Grain Legumes and Green Manures for Soil Fertility in ... - cimmyt

Grain Legumes and Green Manures for Soil 

Fertility in Southern Africa: 

Taking Stock of Progress 

Proceedings of a Conference held 8-11 October 2002 

at the Leopard Rock Hotel, Vumba, Zimbabwe 

Edited by 

Stephen R. Waddington 

Maize Programme and Natural Resources Group 

CIMMYT-Zimbabwe 

Co-coordinator, Soil Fert Net 

Website: 

www.soilfertnetsouthemafrica.org 

A Publication of the Soil Fertility Management and Policy Network for 

Maize-Based Cropping Systems in Southern Africa 

IsaN 970-648-113-3 Harare, Zimbabwe, December 2003

Printed in Zimbabwe 

Correct citation: 

Waddington, S.R. (ed.) 2003. Grain Legumes and Green Manures for Soil Fertility in Southern Africa: 

Taking Stock of Progress. Proceedings of a Conference held 8-11 October 2002 at the Leopard Rock 

Hotel, Vumba, Zimbabwe. Soil Fert Net and CIMMYT-Zimbabwe, Harare,.Zimbabwe. 246 pp. 

Grain Legumes and Green Manures for Soil Fertility in Southern Africa: 

Taking Stock of Progress 

Waddington, S.R. (ed.l 

AGROVOC Descriptors 

Soil fertility; Grain legumes; Green manures; Maize; Cowpeas Vigna 

unguiculata; Striga asiatica; Pigeon peas; Beans Phaseolus vulgaris; 

Mucuna; Biomass; Semi arid Zone; Soil chemistry and physics; 

Fertilizer combinations; Soil cultivation; Land productivity; Soil 

management; Cropping systems; Intercropping; Grain yield; Farming 

systems; Small farn:s; Socioeconomic environment; Economic 

analysis; Agricultural research; Southern Africa 

AGRIS Category Codes 

P35 Soil Fertility 

E16 Production Economics 

Dewey Decimal Cla.ssif. 631.422 

ISBN 970-648-113-3 

Layout: Stephen Waddington and Nigist Bekele

Acknowledgments 

• The Leopard Rock Hotel, Vumba for a first rate experience, including excellent 

conference facilities and accommodation. 

• The Rockefeller Foundation for its continued support to Soil Fert Net and its specific 

funding of most of the participants at the conference and much of the research work 

presented. 

• Rudo Shongedza and other staff of CIMMYT-Zimbabwe for secretarial and 

logistical support.

Key Papers 

Contents 

Introduction and conference objectives 

Stephen Waddington and Mulugetta Mekuria ............... ............................................................... 1 

Enhancing the contribution of legumes and biological nitrogen fixation in cropping systems: 

Experiences from West Africa 

Bernard Vanlauwe, Andre Bationo, R.J. Carsky, J. Diels, N. Sanginga and S. Schulz .................,.....3 

Legumes for soil fertility in Southern Africa : Needs, potential and realities 

Ed Rowe and Ken Giller ........ .................. ..... .... ..... ... ............... ............... .... ........... ... ...............15 

Pathways for fitting legumes into East African highland farming systems: A dual approach 

Tilahun Amede .............................. ..................................................... ........... ........................21 

Questions and answers............................................ ................................. ..............................31 

Rhizobium, N Fixation and Microbiology 

Promoting new BNF technologies among smallholder farmers: A success story from Zimbabwe 

Sheunesu Mpepereki and Ishmael Pompi ..................................................................................33 

Response of bean (Phaseolus VUlgaris, L.) cultivars to inoculation and nitrogen fertilizer in Zambia 

Friday Sikombe, Obed I Lungu, Kalaluka Munyinda and Masauso Sakala ...................................... 39 

Role of phosphorus and arbuscular mycorrhizal fungi on nodulation and shoot nitrogen content in 

groundnut and lablab bean 

Ylver L. Besmer, R. T. Koide and S.J. Twomlow .......................... ........... ....................................43 

Soyabean yield response to different rhizobial inoculation rates on selected sandy soils in Zimbabwe 

Ngoni Chirinda, S. Mpepereki, R. Zengeni and K. E. Giller ............................................................47 

Survival and persistence of introduced commercial rhizobia I inoculant strains in selected smallholder 

field environments of Zimbabwe 

Rebecca Zengem~ Sheunesu Mpepereki and Ken E. Giller ........................................................... 53 

Integrating organic resource quality and farmer management practices to sustain soil productivity in 

Zimbabwe 

Florence Mtambanengwe and Paul Mapfumo ......................... .. .. ............... .. ................ ..............57 

Questions and answers................................: ..... .................... ............ .....................................65 

Screening of Annual Legumes for Adaptation and Use 

Adding a new dimension to the improved fallow concept through indigenous herbaceous legumes 

Paul Mapfum0, Florence Mtambanengwe, Sheunesu Mpepereki and Ken Giller .......... .... ................67 

Screening of short duration pigeonpea in Matabeleland 

Bongani Ncube, Tafadzwa Manjala and Steve Twomlow ............................................................ 75 

Risk diversification opportunities through legumes in smallholder farming systems in ~he semi-arid 

areas of Zimbabwe 

Richard Foti, Joseph Rusike and John Dimes ............................................................................79

Evall!ating mucuna green manure technologies in Southern Africa through crop simulation modelling 

Zondai Shamudzarira ... ......... ... .. ......... ... .. .. ........ .......... ... ... ........ .............. .............. ......... ......... 87 

Questions and answers.. ... .................................... .. .......... ............... .............. .. ....... ............... . 93 

Identification of Best Bet Legumes for On-farm Performance as Grain 

Legumes, Intercrops, Rotations, and Green Manures 

Green manure and food legumes research to increase soil fertility and maize yields in Malawi: A 

review 

Webster Sakala and Wezi Mhango .... ... .. .......................... ..................... .. .......... ..... ..................95 

Green manuring in Zimbabwe from 1900 to 2002 

Lucia Muza ... .. .......... .. ...... .... ... ............ .... ....................... .............. ..... .. .... ........... .. .... ......... . 103 

Grain legumes and green manures in East African maize systems - An overview of ECAMAW network 

research 

Dennis K. Friesen, R. Assenga, Tesfa Bogale, T.E. Mmbaga, J. Kikafunda, Wakene Negassa, J. Ojiem, 

and R. Onyango ....... ... ............ ......... .... .. ........... ... ... .... .. ......... .. .... ........ ... ........ .... ........ .. .... ... 113 

The role of cowpea (Vigna unguiculata) and other grain legumes in the management of soil fertility in 

the smallholder farming sector of Zimbabwe 

Nhamo Nhamo, Walter Mupangwa, Shephard Siziba, Tendai Gatsi and Davison Chikazunga .........119 

Biomass production of green manures and grain legumes in soils of different characteristics in Zambia 

and Zimbabwe 

Pauline Chivenge, Moses Mwale and Herbert Murwira ...... .. ............... ........ ................. ........ .... . 129 

Effect of different green manure legumes and their time of planting on maize growth and witchweed 

(Striga asiatica) control: A preliminary evaluation 

Laurence Jasi, Ostin A. Chivinge and Irvine K. Mariga .. ... ............ ... ........ ........................ ... ....... 135 

Leguminous agroforestry options for replenishing soil fertility in Southern Africa 

Paramu L. Mafongoya, E. Kuntashula, F. Kwesiga, T. Chirwa, R. Chintu, G. Silesh/~ J. Matibini .....141 

Pigeonpea/cowpea intercrop + maize + cassava rotations on smallholder farms in the southern 

coastal area of Mozambique 

Candida Cuembelo..... ........ ....... ........ ... ...... .... ............ ... ............ ........... ... ......... .. ...................155 

Questions and answers..... ........ ... ......... .... ........ ... .... ..... .. ................. .................... ... .............. 159 

Legume Benefits on Maize Productivity and Soil Properties 

Mucuna-maize rotations and short fallows to rehabilitate smallholder farms in Malawi 

Webster D. Sakala, Ivy Ugowe and D. Kayira ...... ... .. ........... ......... .. ................................... .. ... 161 

Residual effects of forage legumes on subsequent maize yields and soil fertility in the smallholder 

farming sector of Zimbabwe 

Walter Mupangwa, Happymore Nemasas/~ R. Muchadeyi anti G.J. Manyawu ........ ....... ...............165 

Time of incorporation of different legumes affects soil moisture and yields of the following crop in 

maize based systems of Zimbabwe 

Bonaventure Kayinamura, Herbert K. Murwira and Pauline P. Chivenge... .......... .. ... ....... .. ............ 169 

Soil fertility improvement through the use of green manure in central Zambia 

Moses Mwale, Cassim Mas/~ J. Kabongo and L.K. Phiri .. ..........................................................173

Effect of surface application and incorporation of sunnhemp and velvet bean green manures on the 

production of field crops 

J. Mulambu, K. Munyinda, S. Ngandu and 0.1. Lungu .. ... ........................... ............................. . 179 

Questions and answers........... ..... ........ ...... .... ... .. .... .. ...... , ... ................ ...... .. ..... ... ..... ........ ; ..... 183 

Improving the Productivity of Grain Legumes and Green Manures 

Performance of green manures and grain legumes on severely acidic soils in northern Zambia, and 

their effect on soil fertility improvement 

Costah Malama and Kenneth Kondowe ........ ... ..... ...... .. ........ ............... ...... ............................. 185 

Agronomic effectiveness of phosphate rock products, mono-ammonium phosphate and lime on grain 

legumes in some Zambia soils 

Obed I. Lungu and Kalaluka Munyinda .. ....................... ................... .. ................................... .. 189 

The effect of phosphorus and sulphur on green manure legume biomass and the yield of subsequent 

maize in Northern Malawi 

Atusaye B. Mwalwanda, Spider K. Mughogho, Webster D. Sakala and Alex R. Saka ...... ....... ... ... . 197 

Management of an acid soil using mine tailings as lime for soybean production 

Lackson K. Phiri, Moses Mwale and Mlotha I. Damaseke.... .. .... .. .... .. ...... .. .... .. .......... .. .. .... .. ...... .205 

Questions and answers... .. ........... ....... ...... .. ...... ....... .... .... ... ... ....... .... ...... ... .. ................ .... ...... 209 

Promotion, Economics and Adoption of Annual Legumes 

Evaluation and promotion of various classes of annual legumes with farmers in Chiota, Zimbabwe 

Dorah Mwenye ........ ........................... ............. ....... ...... ............. ........ ....... ....... ... ................. 211 

Financial and risk analysis to assess the potential adoption of green manure technology in Malawi and 

Zimbabwe 

Mulugetta Mekuria and Shephard Siziba.... ...... .... .. ... ....... .............. ...... ...... ... ..... ............. ....... .. 215 

A socio-economic analysis of legume production motives and productivity variations among 

smallholder farmers of Shurugwi Communal Area, Zimbabwe 

Charles Nhemachena, Herbert K. Murwira, Kilian Mutiro and Pauline Chivenge ................... .........223 

Linking technology development and dissemination with market competitiveness: Pigeonpea in the 

semi-arid areas of Malawi and Tanzania 

Joseph Rusike, Gabriele Lo Monaco and Geoff M. Heinrich ....................... ................................227 

Questions and answers.... .. ..... ... ... ............... ...... .. ...... ... .......... .... .... .................. ... ........ .. .... ... 237 

Synthesis and Working Group Reports 

... , .... ..... ....... ...... ............... .. .................. ........ ................ ............. ............. ........ ..... ..... ... ......239

INTRODUCTION AND CONFERENCE OBJECTIVES 

STEPHEN R WADDINGTON and MULUGETTA MEKURIA 

One of the main thrusts of Soil Fert Net and its 

members since the mid 1990s has been to develop 

and test under farmer conditions a wide range of 

annual legume options that smallholder farmers 

will find useful for soil fertility, and for food or sale. 

Research has also been undertaken on the mechanisms 

and processes by which these legumes provide 

their benefits and the magnitude of benefits 

tha t can be realized under ideal conditions and on 

farm. More recently, initiatives have been undertaken 

to promote these technologies with farmers, 

get farmer feedback about which they prefer and 

encourage farmers to experiment with the legumes. 

Most recently, a range of studies on the economics 

and policy implications of these systems have been 

undertaken. 

In recent years then, tremendous progress has been 

made in identifying suitable best bets, in quantifymg 

their performance on farm, in assessing their 

economic potential and their suitability with farmers, 

and in helping farmers to access the technolo 

. gies. Yet, outputs from these many efforts have been 

widely scattered in annual research reports, in papers 

and articles that are often difficult to access, 

and in some cases are still on computer hard disks 

or have yet to be written up. 

This conference on the soil fertility benefits from 

green manures and grain legumes brought together 

56 participants from Malawi, Zambia, Zimbabwe 

and Mozambique, along with some further key presenters 

from Ethiopia, Kenya and the Netherlands. 

The conference objectives were to: 

9) Provide an opportunity for Soil Fert Members 

and other interested persons to present their research 

on grain legumes and green manures for 

soil fertility management, and to learn from others 

9) Document and synthesize the state of the art on 

this important topic in the region 

9) Showcase the benefits that these technologies are 

having with clients, especially smallholder farmers 

9) Identify socio economic, institutional, and policy 

conStraints affecting the promotion and lli/e of 

legumes and green manures 

9) Develop strategies to fill research gaps and maximize 

the benefits from these technologies. 

The conference was divided into several thematic 

sessions where a mixture of invited and offered oral 

and poster papers were presented and discussed. 

Several key papers were given on strategic topics by 

persons from the region, from East Africa and be-· 

yond. The conference emphasized work conducted 

in Malawi, Zimbabwe, Zambia and Mozambique. 

Session themes included: 

9) Introductory Session of Key Themes 

9) Rhizobium, N fixation and Microbiology 

9) Screening of Annual Legumes for Adaptation and 

Use 

9) Identification of Best Bet Legumes for On-farm 

Performance as Grain Legumes, Interc.rops, Rotations, 

Green Manures 

9) Legume Benefits on Maize Productivity and Soil 

Properties 

9) Improving the Productivity of Grain Legumes 

and Green Manures 

9) Targeting, Integration and Promotion of Legumes 

9) Adoption, Economics and Impacts of Annual 

Legumes with Farmers. 

Towards the end of the conference, synthesizers reported 

key findings. Working Grqups met to summarize 

progress and gap~, and examine the way 

forward. A summary of findings from the Working 

Groups is given at the end of these proceedings. 

The meeting was a great success, with' enthusiastic 

participation throughout. The important papers 

presented that document the state of the art with 

green manure and grain legume research and development 

for soil fertility improvement in southern 

Africa are given in these proceedings. 

Grain Legumes and Green Manures for Soil Fertility in Southern Africa

ENHANCING THE CONTRIBUTION OF LEGUMES AND BIOLOGICAL 

NITROGEN FIXATION IN CROPPING SYSTEMS: 

EXPERIENCES FROM WEST AFRICA 

Abstract 

BERNARD VANLAUWE, ANDRE BATIQNO, 

Tropical Soil Biology and Fertility Institute of CIA T, PO Box '30677, Nairobi, Kenya 

RJ CARSKY, J DIELS, N SANGINGA, and S SCHULZ 

/ITA Nigeria, c/o Lambourn, 26 Dingwall Road, CroydonCR9 3££, UK 

The lack of adoption of improved soil fertility management options to counteract soil fertility decline has led to major 

changes in the research and development paradigm, leading to the currently widely adapted concept of Integrated Soil 

Fertility Management. Legumes in general and biological N fixation specifically have a potentially important role to 

play in ISFM strategies. Four examples are summarized of attempts to enhance the soil fertility status using legumes in 

various agro-ecozones of the West African savanna. These case studies cover the technical aspects of the various legumebased 

systems but also focus equally on the evaluation, adaptation, and adoption processes. 

Alley cropping with leguminous hedgerows is a first example. The technology was proven to be technically sound and 

generated a lot of process work in agroforestry systems. Especially important to note is that the impact assessment phase 

was not in synchrony with the technology development phase, which excluded any useful feedback and delayed the identification 

of the appropriate niches for this system. These were unfortunately found to be geographically limited. The 

Mucuna cover cropping system is a second example. As with alley cropping systems, the inclusion of Mucuna in a 

cropping system was observed to significantly enhance crop yield. Contrary to alley cropping, however, impact assessment 

was implemented during the testing and evaluation phase and useful feedback loops led to clearer insights about 

the specific role Mucuna could play in farmers' fields . This role was more associated with its ability to suppress Imper 

·ata cylindrica weeds than with improving the soil fertility status, thereby also limiting its niche for adoption. As a 

third example and a reaction to the lack of widespread adoption of the former two technologies, dual purpose grain legumes 

- cereal rotations are evaluated. Such systems, using improved legume germplasm that provides net N benefits to 

the cropping system besides grains, significantly enhance cereal yield and supply the farmer with immediate products 

that can be consumed or sold. This technology shows a lot ofpromise and.is currently spreading in the Northern Guinea 

savanna zone of Nigeria, but required the creation of local processing skills and/Or markets for th~ grains and intensive 

interaction between breeders, soil fertility specialists, and farmers. A last example deals with the role ofcowpea in rotations 

in the dry savannas. As with soybean, improved germ plasm of cowpea can also be used to enhance the soil fertility 

status while yielding immediate benefits to farmers. 

In conclusion, several aspects are highlighted that need to be considered when aiming at enhancing the contribution of 

legumes and biological N fixation to cropping systems. These include the need for immediate benefits and the role of 

multipurpose germplasm in providing these, the need to identify niches and the role of markets, and the need for multidisciplinarity 

and full participation of all stakeholders. Finally, some potential routes for future research are indicated. 

Key words: Grain legume, green manure legume, forage legume, dual or multi-purpose legume, biological N fixation, 

cropping system, impacts, west Africa 

Introduction 

The soil fertility status of the soils in sub-Saharan 

Africa (SSA) is generally believed to be poor due to 

poor inherent soil quality and inappropriate soil 

management practices. Such statements are usually 

backed-up by a demonstration of highly negative 

nutrient balances for the major plant nutrients and 

the existence of wide gaps between yields obtained 

under well-managed compared to on-farm conditions 

for the major crops. These facts are generally 

also applicable to the West African savanna ,a.groecozone. 

Although soil fertility replenishment has 

been on the research and development agenda for 

several decades in SSA as this is believed to have 

substantial impacts on the livelihoods of the rural 

population, relatively little has been achieved. The 

reasons for this are plenty and beyond the scope of 

this paper but importantly, the paradigms underlying 

the soil fertility research and development 

agenda have continuously changed to attempt to 

deal with the issue· of non-adoption of improved 

Grain Legumes and Green Manures for Soil Fertility in Southern Africa 3

Table 1. Paradigm shifts underlying the tropical soil fertility research and development agenda and accompanying changes in the framework for 

interactions between the various stakeholders. 

Period 

70's 

Mid-80's 

Mid-90's 

Today 

Soil Fert~ity research and development paradigms 

Nutrient replenishment paradigm: 'Overcome soil constraints to fit plant requirements 

through inputs' (Sanchez, 1994); Green Revolution paradigm 

Focus on biological management of soil fertility; development of the term 'low input 

sustainable agriculture' (LISA) 

Second paradigm (Sanchez, 1994): 'Overcome soil constraints by relying on biological 

processes by adapting germplasm to adverse soil conditions, enhancing soil biological 

activity, and optimizing nutrient cycling to minimize external inputs and maximize their 

use efficiency'; still little or no mention of human and social factors 

Integrated Soil Fertility Management: 'Develop adoptable and sustainable soil 

management practices that integrate the biological, chemical, physical. social, cultural 

and economic processes that regulate soil fertility'; full recognition of the equal role 

of biophysical and social sciences in developing and disseminating improved soil 

management interventions 

Interactions between the various stakeholders 

Top-down, little understanding of the various 

stakeholders in the development process 

Technology transfer; technologies move from research to 

the extension services to the farmer with little feedback 

Participatory approaches; feedback is sought mainly 

from the farmer community regarding improved soil 

management options 

Integrated Natural Resource Management; recognition 

that all stakeholders need to dialogue with each other at 

all stages in the research-to-development continuum 

soil management interventions (Table 1). The 

changes in paradigm with time also lead to the establishment 

of platforms for increasingly more intensive 

interactions between all stakeholders involved 

in improving the status of the soil resource 

(Table 1). 

Currently, the Integrated Soil Fertility Management 

(lSFM) paradigm is widely adhered to. Maybe except 

for the Nutrient Replenishment paradigm, organic 

resources ' and consequently legumes, have 

played a major role in improved soil management 

strategies. This is obviously related to their capacity 

for biological N fixation (BNF) and other positive 

rotational effects. In the Integrated Soil Fertility 

Management paradigm, which advocates the most 

efficient use of all sources of nutrients (organic, 

mineral, soil organic matter-related) and the potential 

interactions between each of these for the provision 

of goods and services, legumes can be hypothesized 

to contribute to crop growth and soil fertility 

improvement in many ~ays (Figure 1). 

Biomass: 

- OM production 

- N contributions 

OlJ!.aniclmineral interactions: 

- pest/disease dynamics 

- soil physical properties 

- soil P availability 

Efficient germplasm: 

- promiscuity 

- adapted to low P 

- adapted to drought 

Figure 1. legumes can potenti.ally contribute to the generation of a 

wide set of properties and functions required in an Integrated Soil 

Fertility Management framework. 

This paper aims- at (i) illustrating relevant experiences 

with enhancing the contribution of legumes 

and BNF to cropping systems in the West-African 

savanna, (ii) evaluating the efficiency of the research 

and development process in relation to the paradigm 

underlying this process, and (iii) highlighting. 

the current line of thought about improved soil 

management through the integration of legumes. 

The paper does not intend to cover all progress 

made with legumes in West Africa, but to foster the 

often lacking exchange of relevant experiences and 

information with other regions in SSA, that are 

more often than not facing similar soil-based constraintsto 

improved crop production. 

The West African Savanna Agroecozone: 

A Biophysically and Socio-economically 

Diverse Environment 

Broadly, rainfall decreases from South to North and 

the rainfall pattern changes from clearly bimodal to 

unimodal (Table 2). Agro-ecozones within the region 

are usually defined in terms of length of growing 

period (Japtap et al., 1995) as highlands are virtually 

absent. Livestock densities also increase from 

South to North due to diminishing disease pressure 

(Mohamed-Saleem and Fitzhugh, 1995). 

Human population density varies widely and is less 

directly linked to latitude, as in each agroecozone 

centres can be identified with large population densities 

(e.g., Squthern Benin in the derived savanna, 

Zaria/Kaduna in the Northern Guinea savanna, 

etc). This is obviously related to pressure on land 

and land use intensification. Smith and Weber 

(1994) postulated that the determinants of intensification 

are either population density or access to 

markets. Within each path of the evolutionary proc- 

4 

Grain legumes and Green Manures for Soil Fertility in Southern Africa

Table 2. Agro·ecozories, rainfall distribution, presence of livestock, potentiallegume·basedtechnologies and potential problems encountered 

with the latter for the West African savanna lone. Source agroecolone definition: Jagtap et aI., 1995. 

Agro·ecozone Rainfall distribution Presence of Potential legume-based technologies Potential problems encountered with 

(length of growing period) 

livestock 

legume technologies 

Deri~ed Savanna Bi·modal Small ruminants Grain legume cereal rotations within one Lack of land in densely populated 

(211·270 days) year; herbaceous cover crops during tlie 

second short season; alley farming with tree 

legumes 

areas; 

Southern Guinea Savanna Bi· to uni·modal Few cattle, small Grain legume cereal rotations within the lack of land in densely populated 

(181 ·210 days) ruminants same year; herbaceous cover crops; alley 

farming with tree legumes 

areas; 

Northern Guinea Savanna Uni·modal Cattle. small Grain legume cereal rotations or 

(151 ·180 days) ruminants intercrops; herbaceous cover crops; fodder 

banks; parkland trees 

Sudano·Guinean Uni·modal Cattle. small Early grain legume cereal rotations or 

(101 ·150 days) ruminants intercrops; parkland trees· 

Sudano·Sahelian Uni·modal Cattle. small Extra early grain legume - cereal rotations 

(61·100 days) ruminants or intercrops; parkland trees 

" 

ess, Manyong, et al. (1996) made a distinction between 

an expansion and an intensification phase. In 

popula.tion-driven expanding . farming systems, increased 

human population results in the opening of 

new land. Fallow periods are still long . enough to 

maintain soil fertility. As new land becomes scarcer, 

land use intensifies with little increase in purchased 

inputs, leading to a progressive decline in productivity 

of labour and land and eventually the abandonment 

of farming. Market-driven systems are 

g~nerated through exogenous factors such as the 

introduction of cash crops. In the expansion phase, 

purchase of inputs is still moderate, while in the intensification 

phase, credit is usually available to increase 

the level of purchased inputs and hired labour. 

Market driven systems require a good transport 

system that provides access to markets. In the 

subhumid zones, 66% of the agricultural systems 

are in the population-driven phase, while 34% in 

the market-driven phase (Manyong et al., 1996). 

Each of the above pathways has implications for options 

available to the farmer to manage soil fertility 

in general, for the best-bet legumes to be integrated 

in existing cropping systems, and for problems related 

to specific legume technologies (Table 2). 

In what follows, specific legume-based technologies 

will be evaluated in terms of their agronomic benefits, 

niche identification, impact assessment, and the 

efficiency of the research and development process 

that brought those tecm-ologies to the farmer. 

Alley Cropping: From a Panacea to a 

Technology with a Very Specific Niche 

The first papers on alley cropping (sometimes called 

alley farming or hedgerow intercropping) were 

published in the early eighties by Kang (e.g., Kang, 

lack of land in densely populated 

areas; disappearance of legume' 

biomass during the dry season; free· 

{lrazing livestock 

Short cropping season excludes long 

duration legumes; disappearance of 

legume biomass during the dry season; 

free·grazing livestock 

Very short cropping season limits 

choice of1egumes; disappearance of 

legume biomass during the dry season; 

free·grazing livestock 

1985). They showed that short term yields of maize 

were substantially enhanced when applying the 

prunings of the hedgerows to the maize, once the 

trees were ready for pruning, usually varying from 

1 to 2 yrs after planting. Legume trees were primarily 

targeted as hedgerow species, mainly because of 

their BNF capacity but also because of their -relatively 

rapid growth and potential source of fodder. 

The great potential demonstrated by the initial published 

results led to a substantial amount of -effort to 

understand and fine-tune the technology and its 

management. Sanginga et al. (2001) reports that certain 

hedgerow trees could fix between 100 and 300 

kg N ha·l yrl while other species fixed less than 20 

kg N ha·l yrl. Substantial differences between 

provenances from the same species were also observed. 

Because the recovery of applied pruning-N 

was often observed to be very low and hardly exceeding 

20% (Vanlauwe et al., 1998a), efforts were 

made to quantify the fate of N not taken up by a 

maize crop using isotopes (Vanlauwe et al., 1998a, 

1998b). Initial observations using litterbags to assess 

pruning-N release .and the N difference method to 

calculate pruning-N recovery, showed poor synchrony 

between N availability and demand by the 

crop. Studies with isotopes, however, cpuld also 

quantify the fate of applied pruning-N as it moved 

through other pools of the alley cropping system 

and consequently the system was observed to be 

tighter in terms of N cycling as compared to earlier 

estimates (Figure 2). Most of the initial testing of the 

resource quality - decomposition hypotheses formulated 

by Swift etal. (1979), was also implemented 

using hedgerow species (e.g., Tiah et al., 

1993). . 

Stimulated by these promising results, the Alley 

Farming Network for Tropical Africa (AFNETA) 


100 

90 

80 

- - 0 - - Uptake by ihe maize crop 

- - iI- - Release from Ihe litler layer 

______ _______ 6 

t>,-- -- -- -- ----- -- -- -- 

Z 70 -e---- Uptake by Ihe syslem (maize. hedgerow. slable SOM pools) 

Gl 

:::::I 60 

"lJ 

'iii 

Gl 

~-50 

40 

0 

0 

~ 30 

20 

10 

0 

-J!r- Release from the litter and particulate organic matter 

__ .. ... _. . .. -. . ..0 ......... . -.... ....... _. .. . -··-0 

0 30 60 90 120 

Time (days after residue application) 

Figure 2. N released from 15N labelled Leucaena /eucocepha/a 

residues and recovered by the various components of an alley 

cropping system during a 120-day maize cropping season in 

Southwestern Nigeria. Synchrony is looked at using the 

'traditional' maize and litter decomposition data and following a 

complete system focus. The particulate organic matter is assumed 

to be on the supply side of synchrony due to their high turnover 

(within one maize season) while the other soil organic matter 

fractions are assumed to be on the demand side. Source: Vanlauwe 

et aI., 1998a. 

was initiated in 1989 and various alley cropping trials 

were established in countries in SSA to test the 

performance of alley cropping systems under a 

wide range of biophysical environments. Most of 

the initial work was carried out on-station at UTA, 

Ibadan, Nigeria. Woomer et al. (1995) summarized 

the data obtained within the AFNETA framework 

and concluded that the system works well with 

maize but not with cassava, cowpea or cotton. He 

also observed that the ratio intercrop:monocrop 

yield was positively correlated with the soil extractable 

P level and negatively with the total N content, 

indicating that the system works best on N deficient 

soils with a relatively high P status. Aihou et al. 

(1999) and Tossah et al. (1999) observed that soils 

with a relatively fertile subsoil led to g~eater biomass 

production and N accumulation than other 

soils. Meanwhile, trials initiated during the earlier 

years of alley cropping research showed that in the 

long term, alley cropping systems were sustainable 

and yields were less variable in the presence of 

minimal amounts of fertilizer N, provided the trees 

were regularly replanted. Vanlauwe et al. 

(unpublished results), for instance, observed maize 

grain yields varying between 2500 and 4000 kg ha- 1 

(average of 2890 + / - 470 kg ha- I ) in a IS-year old alley 

cropping trial with Senna siamea, supplemented 

with 60 kg N ha- I compared to sole fertilizer grain 

yields varying between 600 and 3300 kg ha- 1 

(average of 2080 + / - 910 kg ha- 1 ). 

Whittome (1995) attempted to outline the regions in 

West Africa where alley cropping would potentially 

thrive, using the following criteria: maize-based systems, 

with rainfall> 1200 mm yrl, on non-acid soils 

and with a human population density of > 30 person 

km- I . The outcome of this evaluation was a 

range of limited areas where alley cropping had potential. 

Most of these sites were restricted to Nigeria, 

obviously becaus~ of the high population density 

in that country. In 1996, Dvorak (1996) published 

a first report on the adoption potential for 

alley cropping. She concluded that the potential for 

adoption of alley cropping wa~ limited to areas with 

baseline yields below 2 t ha- 1 , but where soils are 

still of good enough quality to respond to application 

of N, and whose farmers have a flexible demand 

for labour. Drawbacks when evaluating alley 

cropping systems on farm were: (i) hedgerow biomass 

production and/or yield gains were usually 

far below results reported on-station, (ii) the cost of 

establishment is high, (iii) there is a time lag to realization 

of benefits, (iv) and the cropping system is 

inflexible and 'unforgiving' as the penalties for not 

managing the hedges properly can be high. 

While it is beyond doubt that the alley cropping 

concept has generated an enormous amount of 

process work on N cycling and use, organic resource 

decomposition dynamics, soil organic matter 

dynamics, and related topics, impact at tne farm 

level is required before a technology can be called 

successful. In the late nineties, Adesina et al. (1999) 

went back to the sites in Nigeria, Benin, and Cameroon 

where attempts to disseminate the technology 

had been implemented and concluded that despite 

the earlier skepticism about the adoption potential 

of alley farming, the actual rates of adoption were 

encouraging for the complex technology. In Nigeria, 

of the sample of 223 farmers, 93% had heard of the 

technology, 64% had adopted and 53% retained the 

technology. They observed that the technology was 

being adopted in sites with high pressure on land, 

soil fertility decline, erosion problems and fuel 

wood and fodder scarcity. Constraints to adoption 

were mainly technical and management related and 

included too many volunteer seeds (45% of the 

farmers), especially for Leucaena, high labour demand 

(40%), non-adaptability of trees (37%), and 

lack of knowledge (34%). Also impor:tant to note is 

that the technology underwent major changes by 

farmers to suit their circumstances and cropping 

systems (e.g., inclusion of a fallow phase, greater 

height of pruning, wider tree spacing, etc). 

Summarizing the alley cropping story we conclude 

that (i) alley cropping is a technically sound cropping 

system under certain conditions related to soil 

fertility starns, annual rainfall, and target crop; (ii) 

there are a wide range of socio-economic constraints 

to the adoption of alley cropping, (iii) alley crop- 

6 


ping systems are currently utilized but in a modified 

for,m for a variety of reasons, and (iv) the development 

and evaluation of the system following the 

technology transfer paradigm took about 15 years 

(Figure 3). Especially important to note is that the 

impact assessment phase was not in synchrony with 

the phase during which the technology was evaluated 

with farmers. This excluded any useful feedback 

be~een both. 

Mucuna Cover Cropping: Need for Benefits 

Beyond Soil Fertility Replenishment 

When it finally became clear the alley cropping systems 

have limitations with adoptability, another initiative 

had started looking at a basket of options to 

improve soil fertility in the south of the Benin Republic. 

This basket included alley cropping, pigeon 

pea intercropping, mineral fertilizer, and second 

season Mucuna cover cropping (Versteeg et al., 

1998). The Mucuna technology was not new to West 

Africa, as already in the 1930s, work with Mucuna 

was implemented around Ibadan, Nigeria (Vine, 

1953). Mucuna was also observed to create large rotational 

benefits, e.g., increasing maize yields following 

Mucuna by up to 200% on poor soils 

(yielding less than 0.5 t ha·1 and even up to 50% on 

soils yielding over 1.5 t ha·1 maize in the control 

treatments (Figure 4). 

Based on this initial success, screening of various 

species and accessions was implemented in all man~ 

date agro-ecozones of IIT A and usually Mucuna appeared 

as a best bet legume in most zones because 

of its consistently high proportion of N fixed (e.g., 

91%) and total amount of N fixed (e.g., 242 kg , N 

ha·1) (Sanginga et al., 2001). 

During the evaluation of the Mucuna technology in 

southern Benin, farmers observed that the legume 

was very effective at suppressing one of their most 

serious weeds, Imperata cylindrica. lmperata takes 

hold as the length of fallow increases and soil fertility 

declines and requires a substantial ilmount of 

labour to deal with, often forcing farmers to abandon 

their fields. This was obviously a serious constraint 

to crop production in a densely populated 

area as is southern Benin. Evaluation with farmers 

also revealed their reluctance to lose a second season 

food crop because Mucuna did not yield a marketable 

or consumable product (Manyong et al., 

1999). Farmer~ were calculating that the immediate 

opportunity cost of the lost crop "was higher than 

the future benefits of a Mucuna cover crop. Efforts 

to deal with that constraint focussed arOlUld the 

creation of markets for Mucuna seeds, using its residues 

as fodder for livestock, and enhancing the edibility 

of Mucuna seeds for humans and livestock by 

removing toxic L-Dopa (Carsky et al., 2001a). 

The adoption rate of Mucuna in southern Benin tripled 

to 8% of the farmers (amounting to 14000 farmers) 

between 1994 and 1996 (Figure 5), largely 

driven by the intensified effort of programs such as 

Sasakawa Global 2000 (SG2000) who bought about 

15 t of seed in 1995 (Manyong et al., 1999). The decline 

observed after 1996 is likely related to the reduced 

effort of SG2000 to buy seeds and a' collapse 

in the market (Doughtwaite, 2002). The major drivers 

for adoptio~ were need for weeding (39% predicted 

probability) and cash income (41%) 

(Manyong et al., 1999), the l~tter likely largely 

driven by the market created by SG2000. Other factors 

were degraded fields (25%), access to extension 

services (24%), and land tenure security (22%). Although 

the rate of adoption is promising, many con 

Dual pu~ grain legumes 

Mucuna cover crop 

gerrrplasm'derrand 

irllJact assesrrent 

tedln. development/evaluation 

irllJact assesrrent 

tedln .. development/evaluation 

Alley fanning irllJact ~t 

tedlnology evaluation 

tedlnology developrrent 

1978 1982 1986 1990 1994 1998 2002 

Figure~. The various research and development strategies followed 

by the International Institute of Tropical Agriculture while testing 

various systems aiming at improving the soil fertility status. 

400 

~- ClIO 

=:~ 350 

Ill .... 

 

c: 

-0_u 0 300 

._ CII" CII 

>-J: 29) 

CD .... 

NO 

.- .... 200 

IIlCII 

E> 

~; 1S> 

_Ill 

 

CII'Gi 100 

411~ 

111111 

 

CIIc: S> 

~::l 

Uu 

":::l 0 

~:::E 

0 1000 

Control rmize yield (kg'ha) 

Figure 4. Proportional increase in maize grain yield after aMucuna 

crop relative' to the yields in the continuous maize control plots as 

related to the yields in the control plots. Data are asummary of 

various trials in the West African moistsavanna lone. Source: 

Vanlauwe et ai., 2001. 

Grain legumes and Green Manures for Soil Fertility in Southern Africa 

7

- 

6 

~ 

9 

8 

7 

0- 5 

CḎ 

ns 4 

0::: 

3 

2 

-9-1>dopters 

..... ascontirued use 

O__--~~~_.----~----_r----._--__, 

1991 1992 1993 1994 1995 1996 1997 

Figure 5. Dynamics of Mucuna fallow adoption in southern Benin 

(1991·1997). Source: Manyong at ai, 1999. 

straints are likely to halt further adoption. These include, 

loss of a second season (42% of farmers), insecure 

land property rights (19%), unavailability of 

seed (16%), and lack of information (12%) 

(Manyong et aI., 1999). 

In conclusion, (i) the Mucuna cover cropping system 

is a technically sound system under most biophysical 

conditions, (ii) as with alley cropping systems, 

its specific niches are determined by socia-economic 

considerations rather than biophysical ones (e.g., 

not too much pressure on land, high demand for 

labour due to the presence of Imperata cylindrica 

weeds), (iii) in terms of adoption, the Mucuna technology 

is known by nearly all farmers as a tool to 

suppress Imperata, but its use for soil fertility purposes 

is low, and (iv) it took about 10 years to conclude 

the above, as the Mucuna technology was one 

of the first technologies to be evaluated using 

farmer-participatory approaches (Figure 3). The 

identification of an alternative niche for this technology 

(weed suppression) was the result of the 

participatory approach followed a,nd acomplete focus 

on farmers' needs (Houndekon and Gogan, 

1996). Impact assessment was implemented during 

the testing and evaluation phase and useful feedback 

loops led to clearer insights about the potential 

of Mucuna in the target agroecozones. 

Dual Purpose Grain Legume _. Cereal Rotations: 

Multipurpose Options for Redressing 

Soil Fertility Decline 

Whiie alley cropping and Mucuna systems were 

found to have specific and geographically limited 

niches, grain legumes such as cowpea form traditionally 

part of the cropping systems in most of the 

West African agroecozones. Also soybean (Glycine 

max) had become a major grain legume in certain 

areas in Nigeria mainly due to the development of 

local processing techniques and the creation of markets 

(Osho and Dashiell, 1998). Soybean production 

in Nigeria has been estimated at 405,000 t in 1999 

compared with less than 60,000 tin 1984 (www.fao. 

QIg) and this value is expected to increase further 

during coming years. Sanginga et a1. (1999) observed 

that the adoption rates of soybean varieties 

developed at lIT A were over 70% for male and over 

60% for female farmers in Berue State, Nigeria, in a 

period of 10 years. In that same area, soybean was 

for 45% of the farmers the most important source of 

income, leaving the second crop, rice, far behind 

(20%). Although improved varieties of these grain 

legumes have a great potential to be adopted by the 

farmers, the earlier-developed germplasm contributed 

little to improving the soH fertility status. 

These legumes were bred for promiscuity - or the 

ability to establish symbiosis with the native Bradyrhizobia 

- but their N harvest index was usually larger 

than the proportion of N fixed from the atmosphere, 

leading to net negative contributions to ·the 

soil N balance. Through interactions between the 

soybean breeders and soil management staff at 

lITA, breeders were open to develop germplasm 

that produced a lot of leafy biomass without giving 

up on high grain yields (Sanginga et al., 2001). Such 

varieties usually fixed more N than was exported 

with the grains and left a significant amount of N in 

the soil to be potentially taken up by a following 

cereal. Such a variety is, e.g., TGX-1448-2E that produced 

between 470 and 2080 kg of grain ha- 1 

(average of 1290 +/- 500 kg ha- 1 ), between 1000 and 

5340 kg biomass ha- 1 at peak biomass (average of 

2~1@ +/- 1050 kg ha-1), :and fixed between 78 and 

92% of its N (average of 84 +/- 4%) (Iwuafor et aI., 

unpublished data) when grown on 27 farmers' 

fields in Northern Nigeria. Not surprisingly, maize 

growing after these improved soybean varieties had 

significantly higher grain yield (1.2 - 2.3-fold increase) 

compared to a maize control (Sanginga et al., 

2003). In farmer~managed demonstration trials in 

northern Nigeria, initiated in collaboration with the 

non-governmental organization Sasakawa Global 

2000 (SG2000), this variety yielded around 3000 kg 

ha-1 of grain (Iwuafor et al., unpublished data). 

These trials also successfully demonstrated that a 

maize crop grown after soybean can produce a good 

yield with a reduced quantity of N fertilizer compared 

to maize grown after maize. Maize following 

soybean and receiving 85 kg N ha-1 as urea yielded 

slightly more than maize after maize receiving 135 

kg N ha- 1 as urea. While rotational benefits may not 

be as high as those observed after, for example, Mucuna, 

dual purpose soybeans are to be seen as a 

component of an ISFM technology (Table 1) that 

also involved the application of sufficient fertilizer 

8 


N. Applying a limited amoW1t of fertilizer N after a 

dual purpose soybean potentially leads to a more 

efficient use of the fertilizer N compared to a maizemaize 

system. This is because the soybean phase 

may alleviate various constraints to maize production, 

thus increasing the demand for N by a following 

maize crop (Vanlauwe et aI., 2001). This phenomenon 

is usually referred to as positive interaction 

between organic resources and mineral inputs. 

The multi-purpose nature of the above varieties are 

not only related to their capacity to produce a large 

amount of grain and contribute to the N balance of 

cropping systems but also to their ability to suppress 

the parasitic weed Striga hermonthica that affects 

cereal production in Africa, at an estimated 

value of $4S0 million per year for six West African 

countries (Sauerborn, 1991). Soybean can bring 

Striga seeds to suicidal germination and thus reduces 

the pressure on the following maize crop. 

Schulz et al. (2003) reported that the number of 

emerged Striga plants at 12 weeks after planting decreased 

from 0.43 to 0.14 maize- 1 when comparing a 

maize mono-crop with an integrated system that 

included a soybean cropping phase. Large variation 

among soybean cultivars has been found for suicidal 

Striga germination capacity. 

The dual-purpose soybean varieties have only recently 

been introduced to various farmer communities 

in Northern Nigeria, so exact information about 

their spread is not available at this moment. Farmers 

~ the target areas exposed to these varieties, 

however, are excited, not only because of their high 

yields, but also because of the other traits mentioned 

above. Farmer-to-farmer. seed diffusion is 

taking place and old varieties are being abandoned. 

They also observe the improved yields of following 

sorghum or maize crops. In farmer-managed trials 

in Northern Nigeria, that ran for 2 seasons, the 

highest net benefits were obtained with the rotation 

of TGX-144S-2E (1450 US$), followed by the local 

variety Samsoy 2 (1000 US$). The lowest net benefits 

(600 US$) were obtcrined with l..tIblab purpureus 

(Sanginga et aI., 2001). Following these promising 

results and positive initial farmers' reactions, 

SG2000 has started testing soybean rotations in six 

states in Northern Nigeria (lwuafor et al., 2002). 

In conclusion, using resilient, multipurpose and 

adoptable germplasm as an entry point to curb the 

downward spiral of soil fertility decline has proven 

to be a very promising strategy with a potentially 

high impact on farmers' livelihoods. This is mainly 

driven by the fact that there is no time-lag between 

farmers' investments in terms of capital and labour 

and returns, a substantial worry about alley cropping 

and Mucuna rotations that was commonly ex- 

pressed by farmers. Two factors were essential in 

creating the high potential of cropping systems built 

aroW1d dual purpose soybean: (i) the creation.of the 

knowledge for local processing and consumption, 

going hand in hand with the creation of markets for 

soybean and soybean products and (ii) intensive interaction 

between sOYQean breeders, soil fertility 

management specialists, and farmers. 

Cowpea in the West African Dry 

Savannas 

In contrast with the technologies we have discus~d 

above, cowpea (Vigna unguiculata) has been cultivated 

in West Africa since ancient times and appears 

to be a crop native to Africa (Purseglove, 

1991). This is best illustrated by the fact that in 1999, 

cowpea was cultivated on about 7 nilllion ha in 

West Africa, compared with less than 5 million ha 

for groW1dnut and below 1 million ha for soybean 

and bambara bean (Schulz et aI, 2001). In lots of areas 

in Northern Nigeria, cowpea is the first crop to 

harvest after the rains have established. The crop is 

grown from the derived savanna to the sahel for 

food and fodder, although the pressure of pests and 

diseases usually decreases with latitude. 

As with soybean, improvement of the soil fertility 

status in cropping systems with cowpea as a com:" 

ponent can potentially be targeted through the introduction 

of more resilient and multipurpose 

germplasm. There are, however,ftmdamental differences 

between soybean and cowpea-based systems 

that need to be taken into accoW1t when devising 

such a strategy. Because cowpea is a traditional 

crop in West Africa, there is no need to create local 

processing skills or markets for the grains, but it 

also implies that seed traits such as colour, taste, or 

texture become an issue. Secondly, the growth cycle 

of cowpea is shorter than soybean, although early, 

medium, and late varieties are available. This gives 

cowpea another biophysical niche than soybean. 

Thirdly, cowpea is more susceptible to a wide range 

of pests than soybean, and requires chemical or biological 

control. Incorporation of traits related to 

multiple resistance are to be considered when 

breeding for dual purpose germplasm. Fourthly, 

cowpea nodulates in most cases promiscuously, 

while this trait had to be incorporated in improved 

soybean varieties. 

Estimates for N fertilizer replacement values for 

cowpea range from 10 to SO kg N ha·J (Cafsky et aI., 

2003). In a six-month growing season Carsky et al 

(2001b) found that cowpea during the first 2 months 

replaced 30 kg N ha·J as fertilizer to maize during 

the last three months. These values are usually at 


9

Table 3. Millet grain and total dry matter yield at harvest as influenced by millet/cowpea marketable cowpea that will exploit 

crllpping system at Sadore (Niger). Source: Bationo and Ntare, 2000. 

marketing opportunities to improve 

Grain yield (kg hall Total dry matter yield (kg ha·l ) household food security and open 

Cropping system 1996 1997 1998 1996 1997 1998 opportunities for sustainable im 

Continuous millet 937 321 1557 4227 2219 6992 provement in income. 

Millet after cowpea 1255 340 1904 5785 2832 8613 

P > F < 0.001 0.344 < 0.001 < 0.001 < 0.001 < 0.001 

the higher range when residues are incorporated 

and at the lower range after a long dry season. Like 

soybean (Carsky et al., 1997), benefits can be expected 

to be higher after longer duration varieties. 

Schulz et al. (2001) reported that N harvest indices 

were lower and total biomass yields larger in absence 

of chemical treatments, potentially leading to 

higher N contributions to a subsequent cereal crop. 

Cereal-legume rotation effects on cereal yields have 

been reported for the semi-arid tropics (Table 3) 

(Bakayoko et al. 2000; Bationo et al. 1998; Bationo 

and Ntare 2000). In all these studies, the yield of the 

preceding cereal was significantly higher than in 

monocropping treatments. The beneficial effect of 

legumes on succeeding crops is normally exclusively 

attributed to the increased soil N fertility as a 

result of N2-fixation. However, cowpea yield significantly 

responded to rotations suggesting that factors 

other than N alone contributed to the yield increases 

in the cereal-legume rotations. 

Currently, efforts are underway to improve the 

germplasm of cowpea to make it more resistant to 

pests and diseases and adverse environmental c.onditions 

such as drought. Existing germplasm is also 

currently being screened for its ability to access soil 

P that is not readily accessible (Lyasse et al., 2002) 

and to trigger suicidal germination of Striga. Carsky 

et al (2000a) reported that applying P fertilizer to 

soybean tended to increased soybean root density 

and to strengthen its effect on Striga reduction. 

The presence of cowpeas in the cropping systems in 

West Africa is not likely to decrease because of its 

high value in the region. Cowpea grain contains 

about 22% protein, and it constitutes a major source 

of protein for the resource poor farmer. Its fodder 

also provides an important supplement to ruminant 

diets. Cowpea is often the only crop that survives 

severe drought. Cowpea is grown primarily to supply 

farm household food needs but some farmers 

produce surpluses for sale to national and regional 

markets and there is high demand for cowpea in the 

coastal countries such as Nigeria, Togo, Benin and 

Ivory Coast. Thus despite the substantial potential 

that exists for commercialization of cowpea, the opportunities 

are not fully being exploited because of 

weak linkages between farmers and traders. The 

challenge is to help smallholder farmers to move 

rapidly beyond their subsistence needs to produce 

Conclusions and Looking Ahead 

The following conclusions can be drawn from what 

we presented above: 

(i) Improved soil management interventions need to 

generate immediate benefits to the farmer beyond 

an improved soil fertility status, especially in areas 

where land is scarce. Such interventions need to address 

farmers' immediate and longer-term needs. 

(ii) Improved germplasm of commonly-grown 

crops, that addresses various constraints to higher 

yields, is a valid entry point for targeting soil fertility 

depletion. Germplasm that generates multiple 

benefits is likely to be adopted more easily and potentially 

tackles several constraints simultaneously. 

(iii) The role of markets in creating added value to 

certain crops is essential. This was demonstrated 

above for Mucuna in the southern Benin Republic, 

where an artificial market was created, and for soybean 

in Northern Nigeria. 

(iv) Inclusion of improved germplasm alone will 

not lead to sustainable agriculture. Mineral inputs 

are required, very often even to allow the legume to 

grow properly, but also to optimally exploit the 

multiple benefits created by the legumes. In the 

same context, rotational benefits are more often 

than not more than just contributions of extra N. 

(v) There are no panaceas. It is very important to 

identify the appropriate niches for specific classes of 

legumes, both at the macro agro-ecozone scale and 

the farm scale. 

(vi) Regarding the research and development process 

itself, evaluation and impact assessment are essential 

components of the process. Applying participatory 

approaches during the earlier phases of the 

alley cropping story may have led to a much earlier 

recognition of the limitations of this technology. 

(vii) Intense contact between crop breeders, soil 

management specialists and farmers is essential for 

the development of improved germplasm adapted 

to the biophysical and socia-economic conditions 

targeted. Very often, in the International as well as 

National Research Centres, crop improvement ac- 

10 


tivities have been separated programmatically from 

program~ dealing with natural resource management, 

resulting in products that are not adoptable 

by farming communities. 

Future emphasis in legume research and development 

activities could be put on: 

(i) It is important to increase the area cropped with 

legumes i-!1 order to enhance the contribution of 

BNF to agriculture (Giller, 2001). In this context, 

various classes of legumes likely occupy various 

niches within a farm and agro-ecozone. This is a 

very important consideration when targeting specific 

legumes for specific purposes. 

(ii) Dual purpose grain and herbaceous legumes, 

when managed properly, do contribute N to a following 

cereal, although recoveries are often very 

low. It is important to understand the fate of the 

fixed N not recovered by a subsequent crop and, if 

lost, to understand the major loss mechanisms. All 

this information is required to improve the management 

of biologically fixed N. 

(iii) Emphasizing the improvement of accepted 

grain legumes is likely to result in faster pay-offs 

compared with trying to enhance the utility of herbaceous 

legumes. Vast gene banks exist for the major 

grain legumes and these could be exploited for 

specific traits. Biotechnological approaches may 

make this easier in the near future.. 

(iv) A considerable amount of information is available 

related to the performance and rotational benefits 

of all classes of legumes in a wide range of biophysical 

envirorunents. There is an urgent need to 

synthesize this information and avoid unnecessary 

legume screening or related activities. A platform 

such as the Legume Expert System (LEXSYS) 

(Carsky et al., 2000b) could be used as a framework 

for synthesizing this information. 

References 

Adesina AA, Coulibaly 0, Manyong VM, Sanginga 

pc, Mbila 0, Chia~u J ~nd Kamleu DG 1999. 

Policy shifts and adoption of alley farming in 

West and Central Africa. IMPACT, IITA, Ibadan, 

Nigeria, pp. 20. 

Aihou K, Sanginga N, Vanlauwe B, Lyasse 0, Diels 

J and Merckx R 1999. Alley cropping in the moist 

savanna of West-Africa: I. Restoration and maintenance 

of soil fertility on 'terre de barre' soils in 

Benin Republic. Agroforestry Systems 42:213-227. 

Bagayoko M, Buerkert A, Lung G, Bationo A and 

Grain legumes and Green Manures for Sail Fertility in Southern Africa 

Rornheld V 2000. Cereal/legume rotation effects 

on cereal growth in Sudano-Sahelian West Africa: 

soil mineral nitrogen, mycorrhizae and 

nematodes. Plant and Soil 218:103-116. 

Bationo A, Lompo F, Koala S 1998. Research on nutrient 

flows and balances in West Africa: Stateof-the-art. 

Agriculture, Ecosystems and Environment 

71:19-36. 

Bationo A and Ntare BR 2000. Rotation and nitrogen 

fertilizer effects on pearl millet, cowpea and 

groundnut yield and soil chemical properties in 

a sandy soil in the semi arid tropics, West Africa. 

Journal of Agricultural Science, Cambridge 134:277 

284. 

Carsky RJ, Abaidoo R, Dashiell K and Sanginga N 

1997. Effect of soybean on subsequent maize 

grain yield in the Guinea savanna zone of West 

Africa. African Crop Science Journal 5:31-38. 

Carsky RJ, Berner OK, Oyewole BD, Dashiell K and 

Schulz S 2000a. Reduction of Striga hennonthica 

parasitism on maize using soybean ro·tation. International 

Journal of Pest Management 46:115-120. 

Carsky RJ, Weber G and Robert ABC 2000b. LEX 

SYS: a computerized decision-support tool for 

selecting herbaceous. legumes for improved 

tropical farming systems. In Cover Crops for Integrated 

Natural Resource Management in West Africa 

(Eds R.J. Carsky, J.D.H. Keatinge, V.M. 

Manyong, and' A.c. Eteka). Proceedings of a Regional 

Workshop, October, 1999, Cotonou, Republic 

of Benin. lIT A, Ibadan, Nigeria. pp. 201 

208. 

Carsky RJ, Becker M and Hauser S 200la. Mucuna 

cover crop fallow systems: potential and limitations. 

In: Sustaining Soil Fertility in West-Africa 

(Eds G Tian, F Ishida and J 0 H Keatinge), SSSA 

Special Publication Number 58, Madison, USA. 

pp.111-136. 

Carsky RJ, Singh BB and Oyewole B 2001b. Contribution 

of early season cowpea to late season 

maize in the savanna zone of West Africa. Biological 

Agriculture and Horticulture 18:303-315. 

Carsky RJ, Vanlauwe Band Lyasse 0 2003. Cowpea 

rotation as a resource management technology 

for cereal-based systems in the savannas of West 

Africa. Third World Cowpea Research Conference 

held September, 2000, Ibadan, Nigeria. 

Douthwaite B, Manyong VM, Keatinge JDH and 

Chianu J 2002. The adoption of alley farming 

and Mucuna: lessons for research, development 

and extension. Agroforestry Systems., In Press. 

11

Dvorak KA 1996. Adoption potential of alley cropping. 

Resource and Crop Management Research 

Monographs Nr 23, International Institute of 

Tropical Agriculture, Ibadan, ~igeria, pp. 70. 

Giller KE, 2001. Nitrogen Fixation in Tropical Cropping 

Systems. Second Edition. CAB!, Wallingford, 

UK, 423 pp. 

Houndekon V A and Gogan AC 1996. Adoption 

.d'une technoloige nouvelle de gestion des re-o 

sources naturelles: Cas du Mucuna dans Ie sudouest 

du Benin. Recherche Appliquee en Milieu 

Reel. Ministerer due Developpement Rural, Benin 

Republic, 65 pp. 

Iwuafor ENO, Aihou K, Vanlauwe B, Diels J, Sangiriga 

N, Lyasse 0, Deckers J and Merckx R 2002. 

On-farm evaluation of the contribution of sole 

and mixed applications of organic matter and 

urea to maize grain production in the savanna. 

In: Integrated Plant Nutrient Management in sub 

Saharan Africa: From Concept to Practice (Eds B 

Vanlauwe, J Diels, N Sanginga and R Merckx). 

CABI, Wallingford, UK, pp. 185-197. 

Jagtap SS 1995. Environmental characterization of 

the moist lowland savanna of Africa. In: Moist 

Savannas of Africa. Potentials and Constraints for 

Crop Production (Eds BT Kang, 10 Akobundu, 

VM Manyong, RJ Carsky, N Sanginga and EA 

Kueneman) International Institute of Tropical 

Agriculture, Ibadan, Nigeria, pp. 9-30. 

Kang BT, Grimme H and Lawson TL 1985. Alley 

cropping sequentially cropped maize and cowpea 

with leucaena on a sandy soil in Southern 

Nigeria. Plant and Soil 85: 267-277. 

Lyasse 0, Tossah BK, Vanlauwe,B, Diels J, Sanginga 

Nand Merckx R 2002. Options for increasing P 

availability from low reactive Rock' Phosphate. 

In: Integrated Plant' Nutrient Management in sub 

Saharan Africa: From Concept to Practice (Eds B 

Vanlauwe, J Diels, N Sanginga and R Merckx). 

CABI, Wallingford, UK, pp. 225-237. 

Manyong VM, Smith J, Weber GK, Jagtap SS and 

Oyewole B 1996. Macrocharacterization of agricultural 

systems in West Africa: an overview. 

Resource and Crop Management Research 

Monographs Nr 21, International Institute of 

Tropical Agriculture, Ibadan, Nigeria, 66 pp. 

Manyong VM, Houndekon VA, Sanginga Pc, Vissoh 

P and Honlonkou AN 1999. Mucuna fallow 

diffusion in. southern Benin. IMPACT, lITA, 

Ibadan, Nigeria, 21'pp. 

Mohamed-Saleem MA and Fitzhugh HA 1995. An 

12 

overview of demographic an environmental 'issues 

in sustainable agriculture in sub-Saharan 

Africa. In: Livestock and Sustainable Nutrient Cycling 

in Mixed Farming Systems of sub-Saharan Africa. 

Volume II. Technical Papers (Eds JM Powell, 

S Fernandez-Rivera, TO Williams and C 

Renard). ILCA, Addis Ababa, Ethiopia, pp.3-20. 

Ogoke 1], Carsky RJ, Togun AO and Dashiell K 

2002. Effect of P fertilizer application on N balance 

of soybean crop in the guinea savanna of 

Nigeria. Agriculture, Ecosystems and Environment, 

In Press. 

Osho 0 and Dashiell KE 1998. Expanding soybean 

production, processing and utilisation in Nigeria. 

In: Postharvest Technology and Commodity 

Marketing (Ed. RSB Ferris). lIT A, Ibadan, Nigeria, 

pp. 151-156. ' 

Purseglove JW 1991. Tropical Crops. Dicotyledons. 

Longman Scientific and Technical, Harlow, UK. 

Sanginga N, Carsky RJ and Dashiell K 1999. Arbuscular 

mycorrhizal fungi respond to rhizobial inoculation 

and cropping systems in farmers' field 

in the Guinea savanna. Biology and Fertility of 

Soils 30:179-186. 

Sanginga N, Dashiell K, Diels J, Vanlauwe B, Lyasse 

0, Carsky RJ, Tarawali S, Asafo-Adjei B, Menkir 

A, Schulz S, Sing BB, Chikoye 0, Keatinge JDH 

and Ortiz R 2003. Sustainable resource management 

coupled to resilient germplasm to provide 

new intensive cereal-grain legume-livestock 

systems in the dry savanna. Agriculture, Ecosystems 

and Environment, In Press. 

Sanginga N, Okogun JA, Vanlauwe B, Diels J, Carsky 

RJ and Dashiell K 2001. Nitrogen contribution 

of promiscuous soybeans in maize-based 

cropping systems. In: Sustaining Soil Fertility in 

West-Africa (Eds G Tian, F Ishida and J 0 H Keatinge), 

SSSA Special Publication Number 58, 

Madison, USA, pp. 157-178. 

Sanginga PC, Adesina AA, Manyong VM, Otite 0 

and Dashiell KE 1999. Social impact of soybean 

in Nigeria's southern Guinea savanna. IMPACT, 

IITA, Ibadan, Nigeria, 32 pp. 

Sauerborn J 1991. The economic importance of the 

phytoparasites Orobanche and Striga. Proceedings 

of the 5 th International Symposium of Parasitic 

Weeds, Nairobi, Kenya, 1991, pp. 137-143. 

Schulz, S, Carsky RJ and Tarawali SA 2001. Herbaceous 

legumes: the panacea for West African soil 

fertility problems? In: Sustaining Soil Fertility in 

West-Africa (Eds G Tian, F Ishida and J 0 H Keat- 


inge), SSSA Special Publication Number 58, 

Madison, USA, pp. 157-178. 

Srhulz S, Hussaini MA, Kling JG, Berner OK and 

Ikie FO 2003. Evaluation of integrated Striga hermonthica 

control technologies under farmer management. 

Experimental Agriculture 39:1-10. 

Swift MJ, Heal OW and Anderson JM 1979. Decomposition 

in Terrestrial Ecosystems, Studies in 

Ecology Volume 5, Blackwell Scientific Publications, 

Oxford, UK. 

Tian, G, Kang BT and Brussaard L 1993. Mulching 

effect of plant residues with chemically contrasting 

compositions on maize growth and nutrients 

accumulation. Plant and Soil 153:179-187. 

Tossah BK, Zamba OK, Vanlauwe B, Sanginga N, 

Lyasse 0, Diels J and Merckx R 1999. Alley cropping 

in the moist savanna of West-Africa: II. Impact 

on soil productivity in a North-to-South 

transect in Togo. Agroforestry Systems 42:229-244. 

Vanlauwe B, Sanginga N and Merckx R 1998a: Recovery 

of Leucaena and Dactyladenia residue 15N 

in alley cropping systems. Soil Science Society of 

America Journal 62:454:460. 

Vanlauwe B, Sanginga N and Merckx R 1998b. Soil 

organic matter dynamics after addition 9f 15N 

labeled Leucaena and' Dactyladenia residues in alley 

cropping systems. Soil Science Society ofAmerica 

Journal 62:461-466. 

Vanlauwe B, Wendt J and Diels J 2001. Combined 

application of organic matter and fertilizer. In: 

Sustaining Soil Fertility in West-Africa (Eds G 

Tian, F.Ishida ~nd J 0 H Keatinge), SSSA Special 

Publication Number 58, Madison, USA, pp. 247 

280. 

Versteeg MN, Amadji F, Eteka A, Gogan A and 

Koudokpon V 1998. Farmers' adoptability of Mucuna 

fallowing and agroforestry technologies in 

the coastal savanna of Benin. Agricultural Systems 

56:269-287. 

" 


LEGUMES FOR SOIL FERTILITY IN SOUTHERN AFRICA: 

NEEDS, POTENTIAL AND REALITIES 

ED ROWE and KEN GILLER 

Plant Production Systems, Plant Sciences, 'Wageningen University, 

P.O. Box 430, 6700 AK Wageningen, The Netherlands 

Email: ed.rowe@wur.nl.ken.giller@wur.nl 

Abstract 

Legumes have great potential for improving and maintaining soil fertility, through mechanisms such as Nrfixation, 

nutrient release from plant residues, and maintenance of soil organic matter contents. However, this potential remains 

largely unfulfilled. Legumes have a comparatively small role in existing cropping systems, and are often used on poor 

soils where a restricted supply of resources such as phosphorus means that legumes fix N2 at far less than their potential 

rate. Filling gaps in our mechanistic understanding of these processes will allow the design of better systems, but it is 

essential also to consider how new technologies fit into the wholefarming system. New technologies will not succeed 

when productivity gains are small in relation to the amount of time or other resources they require. NUANCES is a 

new framework for analysing trade-offs around soil fertility management in smallholder farming systems. The 

framework is being developed into a modular quantitative dynamic model capabje of integrating crop', livestock, soil and 

bioeconomic models. Flows of carbon and nutrients are considered within imd between heterogeneous farms, and 

influences of and on labour and financial budgets are explicitly included. The analysis will allow the evaluation of 

technologies according to criteria such as agronomic yield, nutrient use efficiency, labour productivity and contribution 

to soil fertility, and the assessment of tradeoffs of investment in different farming strategies and their short and long 

term benefits for improving soil fertility. 

Key words: NUANCES, simulation modeling, annual 

integration, nutrient flows 

legumes, farming system, southern Africa,· technology 

I ntrod uction 

Poor soil fertility remains a major constraint to food 

production in sub-saharan Africa, and thus has 

adverse consequences for food security and the 

sustainability of livelihoods. Amounts of plant 

nutrients in soil are generally small, and nutrient 

balances often negative. In the absence of thriving 

markets for agricultural produce, there are few 

means for purchasing mineral fertilisers to address 

crop demand for plant nutrients. Efforts have 

therefore focused on finding low-cost solutions, 

which include making efficient use of available nutrient 

resources (cattle manure, fertilizers, legumes) 

. Promising 'best-bet' technologies include 

use of grain legumes such as soya bean and groundnut, 

green manures, fodder legumes and improved 

man\.lre storage (Waddington et al., 1998). Creating 

better connections to markets for high-value products 

will also benefit soil fertility by allowing the 

purchase of more inpu ts. 

Nitrogen fixing legumes have great potential for 

inclusion in African farming systems, whether for 

grain or fodder, or in fallow periods either as a 

green manure or a tree. fallow. Trees will rarely be 

worth including simply for their effect on soil fertility, 

but this may be an important additional benefit 

where trees are used for other or multiple functions, 

such as fruit, fuelwood, timber or stakes. Large 

amounts of nitrogen can potentially be introduced 

into the system through fixation by legumes (Table 

1). 

In reality however, legumes contribute far less than 

these potential amounts of fixed N2 (Table 2.). Small 

areas are planted to legumes, and often these are 

obtaining only 25-50% of their N from N2-fixation, 

resulting in overall fixation rates of 5 kg N ha·1 y.l or 

less on farms in Zimbabwe. This under utilization of 

legumes may be due to poor market development 

Table 1. Potential contribution of fixed Nz by legumes in 

different systems. (Summarized from Giller, 2001) 

legume system %Nz fixation Nz fixed Time (days) 

(kg N ha') 

Grain legumes 60·100 105·206 60·120 

Pasture legumes 45·98 115·.280 120·365 

Green manures 50·90 11.0·280 45·200 

Trees .56·89 162·1063 180 

Grain. legumes and Green Manures for Soil Fertility in Southern Africa 15

Table 2. N2 fixed on smallholder farms in Zimbabwe. 

'Average farm size w 3 ha, (Summarized from Chikowo et al. 

2000) 

Legume N2 fixed Area Total fixed 

(kg Nhal) (ha I farm) (kg NI farm) 

Bambara nut 52 0.08 4.2 

Cowpea 47 0.03 1.4 

Peanut 33 0.22 7.3 

Pigeon pea 39 0.34 13.3 

for legume products, or because legumes are grown 

on poor soils where growth and N2-fixation are limited 

and perceived benefits are small. The aim of 

this paper is to examine the benefits of legume technologies 

for soil fertility, and to discuss how these 

benefits can be assessed within an integrated farming 

system. 

Roles of Plant Residues 

Plant residues from crops, mulches or green manures 

have two distinct roles that are in the main 

mutually exclusive, In the short term organic materials 

can, through mineralization, supply nutrients 

to crop plants. In the medium to long term, the accumulation 

of partially decomposed organic matter 

gives the soil better structure, improving aggregation 

and infiltration ~d decreasing run-off. Accumulated 

organic matter also improves the chemical 

properties of the soil, increasing cation exchange 

capacity and slowing down leaching. Medium- and 

100{g-term effects on soil nutrient supply are also 

possible. After the first season the rate of nutrient 

release from a single application of organic matter is 

likely to be small, but with the accumulation of organic 

matter the total nutrient release from slowlyturning 

over materials can become significant. 

As an example, senescing pigeonpea leaves .contain 

substantial amounts of N. Up to 90 kg N ha·J tan be 

added in fallen leaves (Sakala et aI., 2002). However, 

pigeonpea leaves immobilize N for 2 months due to 

low N (1.8%N) and high lignin content (16%) 

(Sakala et aI., 2000) so that no N is contributed for 

companion intercropped maize. After the initial N 

immobilisation phase, net mineralization means 

that substantial amounts of N are made available 

for subsequent crops. Following extensive review of 

literature data, Palm et al. (1997) developed a 

decision tree for allocating organic matter resources 

based on their nitrogen and lignin contents; 

Materials with high N and "low polyphenol / lignin 

concentrations are likely to decompose rapidly and 

thus are suitable for direct applieation for short 

term nutrient supPly. ' Materials with high 

polyphenol / lignin and low N concentrations 

decompose slowly without releasing N and are only 

suitable for surface application as mulch. Materials 

with high polyphenol / lignin and high N contents, 

or with low contents of both, may cause 

immobilization of available soil nutrients and 

should be mixed with fertilizer or higher-quality 

organic material before applying. This decision tree 

was adapted for use in discussion with farmers by 

Giller (2000) by translating plant quality criteria into 

characteristics observable in the field - green colour 

indicating high nitrogen content, ability to crush 

easily indicating low lignin content, and astringent 

taste indicating high tannin content. 

Knowledge Gaps 

Reasearch into the potential roles of legumes in 

African farming systems has given much insight 

into the biophysical processes in which they are 

involved and which affect them. In some cases this 

has been enough to develop technologies with a 

large benefit, and uptake by farmers has been 

substantial. As an example, cropping of soyabean 

on sandy soils in Zimbabwe has been highly 

productive when soil pH and phosphorus status are 

corrected (Mpepereki and Pompi, this volume). 

However, gaps still exist in our understan~ing of 

N2-fixation under field conditions, and of the effects 

of legume residues on the nutrition of subsequent 

crops and on the amount and effects of organic 

matter accumulation (Table 3). 

The Biggest Gap - Integration 

If all the gaps in our understanding of biophysical 

processes were filled, systems could perhaps be 

designed with an optimal productivity and resource 

use efficiency. Ideally, legumes need to be targeted 

in space and time to contribute large amounts of N 

Table 3. Knowledge gaps related to the use of legumes for 

improving soil fertility 

Role 

Knowledge gaps 

N2·fixation • Agroecological ,adaptation of legumes to soils and 

climate 

• Amounts of N2·fixed in different systems and 

agroecologies 

Residue 

contribution to 

crop nutrition 

Soil organic 

matter 

accumulation 

Evidence for improved synchrony of nutrient release 

• 

and plant uptake in the field 

.Measurements of N leac~ing and gaseous losses 

• Provision of nutrients other than N in organic reo 

sources (e.g. cations, S) 

Effects of organic matter quality on long·term build 

• 

up of soil organic matter 

Trade·offs between short and long term benefits of 

• organic resources 

• 

Benefits of enhanced soil organic matter on water 

balances, soil erosion, nutrient USB efficiency etc. 

16 


to the system in such a way that this contributes to 

both short- and long-term soil fertility. There is 

great potential for integrating livestock and 

residue / manure management technologies to 

maintain soil fertility, and in particular to make 

optimum use of mineral fertilizers. Often it is the 

effect of a technology on weeds, pests, diseases or 

vermin which makes or breaks ' it. Problems with 

rats and snakes make many farmers in Lampung, 

Indonesia reluctant to use mulches, and 

agroforestry legume fallows at Domboshawa were 

found to increase cutworm populations, resulting in 

almost complete loss of yield of a subsequent maize 

crop..Conversely, Mucuna pruriens can be highly 

effective at suppressing weeds in some 

environments and this has aided its adoption as a 

nitrogen fixing green manure (see Giller, 2001). 

Creating a farm system model integrating processes 

such as crop rotation, livestock production and 

residue management is conceptually simple. 

Outputs from a crop model can provide inputs of 

crop residue to a decomposition model or stover as 

forage to a livestock production model and vice 

versa. Budgets can be calculated for calories, carbon, 

nutrients and money. Historically, agronomic 

model~ have been developed by a research group 

extendmg an existing model within the original 

software framework. However, developments in 

software technology suggest an alternative 

approach that allows existing models to 

communicate with each other and be linked as 

submodels (Muetzelfeldt, 1995). In such a linked 

model, each part of the system such as a crop field 

or a dairy unit can be simulated by a submodel of 

any level of complexity, provided that standard 

inputs are required and standard outputs produced. 

Quantitative models of pests or weeds could also be 

linked. Linking of all of the various components of 

the farm system would thus allow the exploration 

of opportunities for combing different types of soil 

fertili.ty technologies to understand how they can 

contnbute to overall improvement of productivity 

of the farm as a whole. Optimal farm systems could 

then be designed using techniques such as multiple 

goal linear programming. 

Such biophysically optimal systems might however 

remain unadopted if they were poorly adapted to 

the specific needs and resources of farmers and in 

particular the timing of labour availability. Labour 

requirements are notoriously difficult to assess, 

particularly for new technologies, and 

quantific~t~on of labour supply is complicated by 

opportumtles for alternative employment off-farm 

and hired labour whether paid or unpaid. External 

and internal value judgements about the amount of 

time farmers spend working in the fields also make 

assessment difficult. Labour constraints are 

generally not included in crop models, and labour is 

thus effectively and naively seen as a free resource. 

T~e nee~ for agro~cological models to be integrated 

w~th ~oclOeconomlc models has been identified by 

sCientists from both disciplines. 

NUANCES (Nutrient Use in ANimal and 

Cropping systems - Efficiency and 

Scales) . 

NUANCES (Nutrient Use in ANimal and Cropping 

systems - Efficiency and Scales) is a conceptual 

framework for analysis of trade-offs in African 

smallholder 'farming systems. Heterogeneity is a 

key feature of most farms, as farmers tend to 

concentrate resources in small areas where soil 

fertility is maintained while the majority of their 

fields are effecffvely mined of nutrients. The 

efficiency with which nutrient resources are utilized 

for crop production is likely to vary strongly 

between land of different quality, as will the 

potential growth of different crops or indeed of the 

potentially soil-improving legumes. Documenting 

the extent of variable land qualities within farms is 

therefore an important step in understanding the 

potential impact of different technologies for soil 

fertility improvement. The wealth or resource 

endowment of farming households also determines 

their capacity to invest labour and other resources 

in agriculture as, for example, livestock ownership 

IS often regarded as a key indicator of wealth in 

rural Africa. Poorer farmers are often only able to 

earn income off-farm by selling their labour to the 

wealthier farmers which then restricts the labour 

they can invest in improving productivity of their 

own farms. Farm types will also be identified, 

which might correspond to different wealth classes 

or production systems, to capture the resource 

flows between farms (Figure 1). Resource flows are 

often mediated by livestock, and the framework 

thus includes livestock productivity and manure 

management. 

Resource flow mapping approaches have provided 

valuable insights into the allocation of crops · and 

nutrient resources at various scales, from fields to 

farms, from regions to continents. Assembling static 

balances for nutrients across different land units 

does not however allow for testing of future 

scenarios of how farms could be developed in 

future. Flows which are difficult to measure, such as 

leaching, are generally estimated using simple 

transfer functions, but these functions may not give 

an appropriate response to changing conditions. 

Biophysical models of various degrees of 

Grain Legumes and Green Manures for Soil Fertility in Southern Africa 

17

management operations, such as 

crop rotation, mulching, livestock 

:,; movements or manure transfer. 

Transfers will be possible between 

fields within one farm, between 

farms, and between the farms and 

rangeland. Different methods for 

combining labour and crop / 

livestock models are being 

considered; one idea is for 

management activities to be made 

contingent on there being sufficient 

available labour. The modular 

structure will allow different 

versions of submodels to be 

swapped in or out depending on 

the detail required. 

Figure 1. Resource flows within and between heterogeneous farms 

complexity are however available for different 

flows, and ideally an appropriately complex model 

can be chosen for each part of the system and linked 

together where necessary to allow an integral 

analysis. 

A software framework (Figure 2) is being developed 

to integrate existing crop, soiC livestock and 

bioeconomic models into a model of the whole 

system. This integrated model will be used to 

explore nutrient use efficiency and labour 

productivity, and tradeoffs between short and long 

term contributions to soil fertility, in Afric\U1 

farming systems. 

The NUANCES framework will allow the 

simulation of spatially and temporally complex 

The aim is to create a model that can 

assist in integrating the expert 

knowledge of farmers and of scientists from 

different disciplines. The modular structure will 

allow the principle of 'just-sufficient-complexity' to 

be upheld, since simple models can be used for 

whichever processes are peripheral to the parts of 

the system being considered. As well as providing a 

framework for iterative experimentation and 

modelling, the model will shed light on which 

processes most affect costs and benefits and thus 

allow recommendations to be better targeted. 

NUANCES will lead to the iterative and 

collaborative design of more productive, sustainable 

systems, linking directly to policy at local artd 

greater scales, and will facilitate the development of 

further rules-of-thumb for farmers. 

Farm (x J ",~arm rtln~') 

LV Uves(Qck 

(x 3 sizes· callie, 

sh""p+go.~, 

poUltry) 

P'L finance 

and Labour 

L T Land TyP

Muetzelfeldt, R.1. 1995. A framework for a modular 

modelling approach for agroforestry. Agroforestry 

Systems 30:223-234. 

Palm, C. A., Nandwa, S. and Myers, R.J. 1997. Combined 

use of organic and inorganic nutrient 

sources for soil fertility maintenance and nutrient 

replenishment. In: Buresh, R.J. and Sanchez, 

P.A. (eds.) Replenishing Soil Fertility in Africa. Vol. 

ASSA, CSSA, SSSA, Madison, Wisconsin, USA. 

pp.193-217. 

Sakala, W.O., Cadisch, G. and Giller, K.E. 2000. Interactions 

between residues of maize and pigeonpea 

and mineral N fertilizers during decomposition 

and N mineralization. Soil Biology and 

Biochemistry 32:679-688. 

Sakala, W.o., Cadisch, G. and Giller, K.E. 2002. 

Intercropping of maize and pigeon pea in Malawi: 

grain and biomass yields, N2-fixation in 

pigeonpea, N · balances and residual effects on 

succeeding crops. Plant and Soil, in press. 

Waddington, S.R., Gilbert, R. and Giller, K.E. 1998. 

"Best Bet" technologies for increasing nutrient 

supply for maize on smallholder farms. In: Waddington, 

S.R., Murwira, H.K., Kumwenda, J.D.T., 

Hikwa, D. and Tagwira, F. (eds.) Soil Fertility Research 

for Maize-based Farming Systems in Malawi 

and Zimbabwe. SoilFertNet/CIMMYT-Zimbabwe, 

Harare, Zimbabwe,pp. 245-250: 

Grain legumes and Green Manures for Soil Fertility in Southern Africa 19

PATHWAYS FOR FITTING LEGUMES INTO EAST AFRICAN HIGHLAND 

FARMING SYSTEMS: A DUAL APPROACH 

TILAHUN AMEDE 

African Highlands Initiative (AHI) / Tropical Soils Biology and Fertility Institute of CIA T, 

Code 1110, P. O. Box 1412, Addis Ababa, Ethiopia, T.Amede@CGIAR.ORG 

Abstract 

Food legumes have remained important components of various farming systems in Eastern Africa, but attempts to integrate 

fodder legumes and legume cover crops (Lees) have been unsuccessful. Despite recognising their benefits as soil 

fertility restorers and providers of high quality fodder, farmers remained reluctant to integrate legumes mainly due to 

communitylfanner- specific socio-economic determinants. This paper is based on the experiences of the African Highlands 

Initiative that has worked to integrate legumes in Areka in "the Ethiopian Highlands. The work has tried to understand 

the processes of integration of legumes that have different uses, through participatory research. Areka has an elevation 

of 1990 masl, and an annual rainfall of 1300 mm. The area is characterised by mixed subsistence fanning systems, 

poor access to resources, intensive cropping, land shortage and soil degradation . A participatory evaluation of the 

agronomic performance and adaptability of eight legumes was conducted for three consecutive years during the main 

and short growing seasons, accompanied by extensive data collection on socio-economic determinants. Participatory experiel1ces 

showed that the selection criterion of fanners went far beyond biomass production. The major biophysical 

traits are performance of the species under a specific agroecology (characterised by yield, disease and pest resistance), 

effect on soil fertility and the succeeding crop and its compatibility within the existing cropping system. Specifically, 

farmers identified a firm root system, early soil cover, biomass yield, decomposition rate, soil moisture conservation, 

drought resistance and feed value as important criteria. The total sum of farmers' biophysical criteria showed that Mucuna 

followed by Crotalaria should be the best fitting species, but fanners finally decided on Vetch, the low yielder, due 

to its fast growth and high feed value. The fanners' priority was for livestock feed rather than soil fertility. The final decision 

of the farmers on whether and where to integrate a food legume into their temporal and spatial niches in the system 

is dictated by their food habits, while for a non-food legume it depended on land productivity, fann size, land ownership, 

access to markets and a need for livestock feed. The potential adopters of Lees and forage legumes were less than 

7% of the fanners, while 91 % of the fanners integrated new cultivars of food legumes. A strategic combination of biophysical 

and socio-economic determinants in the form of decision guides was suggested to facilitate the integration of 

legumes into farming communities, and help development agencies and researchers to easily identify potential adopters, 

learn about thl.:; criteria of choice and suggest an improved system of management. It may also help them to identify 

niches or create niches, modify the existing systems and promote the technology for wider use. 

Key words: Legumes, subsistence fanners, selection criteria, integration, decision guides 

Introduction 

Legumes playa pivotal role in nutrient cycling and 

nutrient enrichment in many subsistence-farming 

systems in Africa. They are considered drivers of 

sustainable farming because they intensify the productivity 

and interaction of soit crop, livestock, 

people and other components. In most parts of Africa, 

where livestock products are unaffordable, legumes 

(especially bean, cowpea, pea, chickpea and 

faba bean) are the major sources of protein. The 

maize-based, banana-based and enset-based systems 

are supported mainly by bean and cowpea as 

major protein sources. Legume fodder, as crop residues 

or hay, is also a high value feed for milking 

cows, calves and draught oxen, especially during 

the dry season and in times of high energy demand. 

Legumes increase soil fertility through various 

mechanisms. High quality legume fodder produces 

a high quality manure that could improve soil fertility. 

Legumes can also boost the nitrogen stock in the 

soil through nitrogen fixation and nutrient release 

from their organic residues. Some legumes also release 

root exudates that may increase the availability 

of unavailable/fixed nutrients, e.g. phosphorus, 

through changing the rhizosphere pH and increased 

activity by rhizosphere biota. 

The increasing interest with organic farming in the 

developed world and the challenge to decrease 

costs of inorganic inputs to maintain soil fertility in 

the developing world has focussed the attention of 

researchers and policy makers towards legume 

technology. Organic inputs from legumes could increase 

crop yield through improved nutrient supply/ 

availability and / or improved soil-water hold- 

Grain Legumes and Green Manures tor Soil Fllrtility in Southern Africa 21

ing capacity. Legumes offer other benefits that include 

providing cover to reduce soil erosion, maintenance 

and improvement of soil physical properties, 

increasing soil organic matter/ cation exchange 

capacity, microbial activity and reduction. of soil 

temperature (Tarawali et al. 1987; Abayomi et al. 

2001) and weed suppression (Versteeg et al. 1998). 

There are several studies in Africa that showed 

positive effects of Legume Cover Crops (LCCs) on 

subsequent crops (Abayomi et al. 2001; Fishier and 

Wortmann, 1999; Gachene et al. 1999; Wortmann et 

al. 1994). Studies in Uganda with Crotalarin 

(Wortmann et al. 1994; Fishier and Wortmann, 

1999), and in Benin with Mucuna (Versteeg et al. 

1998) showed that maize grown following LCCs 

produced significantly higher yield than those without 

green manures, mainly through benefits of 

higher amounts of Nand P and partly through nutrient 

pumping from deeper horizons. LCCs could 

also decrease nutrient losses by trapping a huge 

amount of nitrate that could be lost by leaching or 

denitrification if heavy pre-season rainstorms occur 

(Giller, 2001). However, the benefits vary with the 

legume species, their management, soil fertility 

status, the climate and the market value of the preceding 

crop. In some cases, integration of legumes 

for green manuring was not profitable when used 

just to fertilize cereals. Participatory experiments on 

Crotalaria in Uganda showed that a green manure 

did not compensate for the time it occupied in the 

field, although there was an increase in maize yield 

as an after effect (Fishier and Wortmann, 1999). In 

general, the type of LCC species desirable for green 

manuring depends on the assigned use. For weed 

suppression or erosion control, a species capable of 

rapid development of a dense soil cover is required, 

but if the major aim is to intercrop with a cereal, 

then species that grow slowly and erect are more 

suitable (Giller, 2001). 

Despite these positive benefits, there has been relatively 

little success in achieving effective adoption 

of soil-improving cover and forage legumes in Subsaharan 

Africa (Sumberg, 2002, Giller, 2001, Thomas 

and Sumberg, 1995). This could be partly because of 

the absence of methodologies and tools that extensionists 

and community mobilizers can use to facilitate 

the integration of legumes. Information on legume 

technology is diverse and it is accumulated in 

patches. There is, therefore, a need to assemble and 

organise the available information to identify gaps 

and synthesize the data to develop a decision support 

system for farmers, researchers and policy 

makers to select options, niches and systems. 

The objective of this paper is to explore experiences 

with the integration of legumes in subsistence farming 

systems of the East African Highlands, identify 

the biophysical and socio-economic determinants 

affecting their adoption and suggest how those vari-. 

ous determinants could be strategically combined, 

processed and used to develop decision guides. 

Legumes in Various Farming Systems 

Although legumes are important components of 

various farming systems and farmers acknowledge 

the positive contributions of legumes, the amount of 

land allocated to grow them as food, fodder or 

cover crops is relatively small. In the upper highlands 

of Eastern Africa above 2700 masI, including 

the Ethiopian highlands, there are few legumes in 

most farming systems. Lentils are found as a food 

legume, and natural medics and trifolium as feed 

legumes, in proportions of < 2%. In the midhighlands 

of East Africa (1000-2200 masl), both in 

the cereal-based and perennial-based systems, the 

proportion of legumes is higher (about 20-25 %), 

grown as intercrops, intermediate and break crops. 

Without the contribution of legumes in restoring 

soil fertility and breaking pest incidence cycles for 

hundreds of years in this intensively cropped agroecology, 

the production systems may have collapsed 

long ago. The proportion of the legumes decreases 

in the low elevations to less than 10%, because 

those regions are commonly too droughtprone 

to grow most of the traditional legume species. 

In the perennial-based farming systems of Eastern 

Africa, the only dominant legume in the cropping 

system is common bean, intercropped with maize or 

grown sole as a second crop. However, the cultivation 

of beans may contribu te little to soil fertili ty improvement 

(Eyasu, 2002) mainly because 1) the crop 

is harvested by uprooting the whole plant as it 

needs to be stored by hanging bundles on a trellis 

and kept indoors to avoid sprouting; 2) no residue 

is returned to the soil as pods and tops are fed to 

livestock while the stalk is used as feed or cooking 

fuel and 3) beans have the least N-fixing potential, 

particularly in low pH soil with low P availability. 

Why is Adoption of Legume Technology. 

so Slow? 

Thus the proportion of legumes, be it food, feed or 

cover crops, to the various systems is very low. 

There are multiple factors that have affected the 

adoption and dissemination of legumes, which can 

be nested within and defined by three contextual 

factors i) socio-cultural, economic and political ii) 

agroecological and iii) management at the farm 

level (Sumberg, 2002). 

22 


From the food legumes perspective, three factors 

dictate the decision of farmers to grow or not grow 

legumes. First, in African subsistence farming, the 

food habit dictates the amount of land allocated for 

various crops and the type and amount of input invested 

per crop. Since the food habit of most of the 

East African Highlands is cereal-dominated, the 

proportion of cereal to legume consumption in the 

households of East Africa is about 10 to 1. For a 

household with five members in Kenya, on average 

about 500 kg of maize and 100 kg of beans is required. 

Similarly, for the same household size in the 

Ethiopian Highlands 600 kg of barley and 70 kg of 

pea or faba bean is required. Secondly, the fertility 

status of the land and the incidence of pests and diseases 

dictate the frequency of legumes in the cropping 

systems. The proportion of legumes usually 

increases with decline in soil productivity and increased 

incidence of pests and diseases. Thirdly, the 

market value of crops may dictate how much land 

is allocated for legumes. In a few cases, such as the 

Rift-valley of Ethiopia with beans, farmers invest 

land and labour to grow legumes for market. They 

grow legumes for the market and buy cereals for 

consumption at home, as the price of legumes is 

relatively higher than that of the cereals. 

The integration of feed legumes into African farming 

systems has also remained low despite continuous 

research efforts since the 1930s. Sumberg (2002) 

identified several major determinants that affected 

integration. There is a limited tradition to grow 

feed legumes in the region, hence the genetic pool 

of legumes available for growers is limited to a few 

types of recently introduced germplasm. There is 

limited knowledge on legume management and the 

processing and utilization of legumes to make market-orientated 

products. As most of those legumes 

originated in the relatively favourable climates of 

the Andes, it became also challenging to identify 

high yielding, drought-resistant species to integrate 

into the drought-prone environments of Africa. 

Most importantly, because legume technology was 

considered gender-neutral and wealth-neutral, 

socio-economic dimensions were not considered 

during research and extension. 

In recent years, there has been increased research 

interest across the region on the integration of legume 

cover crops (Lees) into the farming systems, 

to help improve and sustain soil fertility. Most of 

the legume cover crops are known to be ideal for 

improving soil fertility, as they are commonly fast 

growing, Nitrogen-fixing, efficient in capturing and 

recycling nutrients and decompose easily (Jama et 

al. 1998). The problem of integration, however, is 

even worse for Lees. This is first because the opportunity 

cost is much higher than the immediate 

benefits of Lees. Second, most Lees are sensitive 

to unfavourable environments (water stress and nutrient 

deficiency.), and very few of them grow well 

in degraded corners of the farm where farmers want 

them to grow. Third, farmers would like to integrate 

legumes that have multiple benefits, i.e. for food, 

feed and soil fertility, while the Lees usually address 

one purpose, i.e. soil fertility maintenance/ 

improvement through the incorporation of the 

green manure into the soil. 

Dual Strategies for Integration of 

legumes 

There are two possibilities to facilitate the integration 

of legumes into East African Highland farming 

systems. Designing a new production system with a 

larger legume component is one option. This could 

be an ideal strategy to integrate legumes, as the prod 

uction system will be geared towards the consumption 

of legumes as major production inputs. 

For example, a policy that prohibits free grazing 

and free herd movements in the Ethiopian Highlands, 

where free grazing is currently practised, and 

the introduction of fast growing feed legumes for 

cut and carry, would enhance the consumption of 

legume technology significantly. Promiscuous legumes, 

which are high yielding in both grain and 

straw, are obvious choices if the system should provide 

high quality manure from few animals and increased 

household income and food. The second 

option is to understand the various farming systems, 

identify the existing temporal and spatial 

niches, creating new potential. niches using the existing 

resources (of land, water, nutrients, solar radiation 

and human resources) and then facilitate the 

integration of legumes by delivering options and 

acknowledging diversities. 

Here I present a case study that justifies the second 

strategy; that of fitting the legume into existing systems 

by identifying the spatial and temporal niches 

of the existing system. 

Experiences of AHI with Integration of 

legumes in the Ethiopian Highlands 

Characteristics of the site 

The research was conducted at Areka, 430 km 

south-west of Addis Ababa, about 1950 masl, representative 

of the mid highlands, with an average 

land holding of less than 0.5 ha. The farming system 

is a perennial highly intensive Enset-based system, 

with a possibility of up to three crops per year. Due 

to a very high human population pressure (>450 


23

people/km2), land holdings are smaller, with fewer 

livestock than in the upper highlands. The average 

livestock holding is less than 1.5 cattle per household. 

Only 15% of the farmers own oxen. Sharing or 

hiring of oxen for ploughing and other farm operations 

is a traditional practice. Unlike the upper highlands, 

where communal land natural pasture and 

free grazing area is available, only crop residues, 

weeds and aftermath grazing are the predominant 

available feed sources in Areka. The cropping system 

is highly diversified. Different forage crop.s are 

grown around the home garden in association with 

coffee, Enset (Enset ventricosum) and fruit trees. 

Crop-livestock integration is strong; farmers use 

crop residues as a feed source, but also return the 

manure to the soil, applied mainly around the home 

garden. The farmers divided their land into several 

plots for various purposes. Trees are planted on valley 

bottoms, slopping areas, farm boundaries, in 

front of the house and in gully areas. Grazing land 

(tittering) is found in front of the house. Some plots 

are left for cut and carry for livestock feeding. These 

plots differ in soil fertility status, with a general decline 

in soil fertility for those plots further from 

houses. 

Determinants of Integration of Legumes 

into Systems 

1. Biophysical Factors Dictating Integration of 

Legumes 

Farmers have multiple criteria to decide whether a 

technology would be appropriate for their circumstances, 

and whether to integrate those tedmologies 

into their farming practices. Although farmers were 

keen to learn about legume technologies in a farmers' 

field school and at on-farm testing sites, they 

demanded time to test them not only under optimum 

research conditions, but also under their own 

real sub-optimal conditions. Experiences from this 

site showed that for a legume to be selected by endusers, 

it should possess the following biophysical 

traits (Amede and Kirkby, 2002): 

a) The biological productivity ofa legume in a 

given agroecology is the principal factor for a 

legume to be considered a potential candidate 

to be integrated into the existing system. The 

most favoured candidate is the one with relatively 

high grain and biomass yield under variable 

agro-ecological conditions (of precipitation, 

temperature, soil fertility and variable management). 

The other criterion was that when farmers 

tested legumes for restoration of soil fertility 

they assume that legumes should improve the 

fertility status of the degraded comers of their 

farm. Therefore, for a legume cover crop to be 

selected for a short term fallow at Areka, the 

major biophysical criterion was whether a species 

can produce high biomass on degraded corner 

plots of the farm. Farmers were not interested 

to grow the LCCs in the fertile comers as 

they were allocated for food crops. The land 

they wanted improved are the border strips, the 

abandoned comers, steep slopes and the barren 

land that failed to produce a reasonable crop 

yield. But most of the LCCs with a strong history 

in improving soil fertility need relatively 

fertile soils to establish, produce a large amount 

of biomass and to fix atmospheric nitrogen. 

That is the reason why farmers selected Crotalaria 

for improving degraded farmlands over Mucuna, 

Canavalia, Tephrosia and vetch (Amede and 

Kirkby, 2002). On individual farmer's fields, 

Crotalaria was the best performing species regardless 

of soil fertility. Similar results were reported 

from Uganda (Wortmann et ai. 1994). On 

the other hand, vetch and mucuna were the best 

performing in fertile comers of the farms. This 

did not agree with the findings of Versteeg et a1. 

(1998), which indicated that mucuna performed 

better than other green manures (including crotalaria) 

to help recover completely degraded 

soils. When those seven species (crotalaria, mucuna, 

canavalia, tephrosia, vetch, stylosanthus, 

and trifolium) were planted in the driest part of 

the season, crotalaria followed by mucuna, performed 

best and produced up to 2.9 t ha- 1 of dry 

matter within three months. 

b) The effect of incorporation of LCCs on the grain 

yield of the following crop is one other very important 

criterion. Application of high biomass 

of LCCs did not necessarily guarantee high 

yield of the following food crop, as the quality 

of the organic material dictates whether nutrients 

accumulated in the LCCs could be released 

a t the required time and in the required 

amount. Participatory experiments on the aftereffect 

of LCCs in Uganda recorded good increases 

in crop yields, although the green manure 

did not compensate for the time it occupied 

the land over a three-crop cycle (Fis!"'ler 

and Wortmann, 1999). Moreover, how large the 

benefit a green manure delivers for growth of 

the following crop depends on the initial fertility 

of the soil and the amount of nutrients that 

the LCC contributes (Giller, 2001). In Areka, 

Tephrosia produced about double the dry matter 

of vetch, but maize yield under vetch was significantly 

higher than under Tephrosia (Amede 

and Kirkby, 2002), which could be explained by 

quality differences and synchrony of the demand 

and supply of nutrients. The most important 

organic quality indicators are nutrient content, 

lignin content and polyphenol content of 

24 


the respective organic resources (Palm et al. 

1997). 

c) Since the opportunity cost of growing an Lee at 

a time when other food crops could be grown is 

very high, those fast-growing early maturing legumes 

that can grow using residual moisture 

should be best fitting. In this case, farmers were 

able to integrate them as intercrops, relay crops, 

and short-term fallows once the major crop is harvested. 

d) Those legumes that did not strongly compete 

with the companion food crop for water, nutrients 

and light when grown in combination with 

foo

ments. It would help researchers to identify the major 

factors of non-adoption and prioritise them in 

relation to socio-economic categories. 

2. Socio-economic Factors Dictating Integration of 

Legumes 

After farmers went through participatory research 

processes for many seasons, and tested favourite 

legumes in their own tie Ids, they were asked to suggest 

the most important socio-economic criteria that 

dictated their selection of one or other legume species 

to be integrated into their systems. 

Results from informal monitoring of farmers' activities 

accompanied by structured questions showed 

21 different factors that affect the integration of legumes 

for different purposes. When farmers were 

asked to prioritise the most important factors that 

affect adoption and integration of legumes, farmers 

mentioned a) farm size, b) suitability of the species 

for intercropping with food legumes, c) productivity 

of their land, d) suitability for livestock feed, e) 

marketability of the product, f) toxicity of the pod to 

children and animals, g) who manages the farm 

(self or share cropping), h) length of time needed to 

grow the species, and i) risk associated with growing 

LCCs -- particularly the introduction of pests 

and diseases. None of the farmers mentioned labour 

demand as an important criterion. Earlier work also 

suggested that farm size and land ownership affect 

the integration of LCCs into smallholder farms 

(Wortmann and Kirungu, 1999). After comparing 

those factors in a pair-wise analysis, five major indicators 

of different hierarchy were identified. 

1) Degree of land productivity: farmers in 

Gummo associated land productivity mainly 

with the fertility status of the soil and distance 

of the plot from the homestead. The homestead 

field is commonly fertile due to a continual supply 

of organic resources. Farmers did not apply 

inorganic fertiliser in this part of the" farm. They 

remained reluctant to_ allocate a portion of that 

land to grow LCCs for biomass transfer or otherwise, 

but grow food legumes (mainly beans), 

as intercrops in the coffee and enset fields. The 

potential niche that farmers were willing to allocate 

for LCes is the outermost field. 

2) Farm size: Despite very high interest by farmers 

to get alternatives to inorganic fertilisers, the 

probability that farmers may allocate land for 

growing LCCs depended on the size of their 

land holdings. For Areka, a farm size of 0.75 ha 

is considered large. Therefore, farmers with 

very small land holdings did not grow legumes 

as sole crops, but integrate them as intercrops or 

relay crops. Therefore, the potential niches for 

LCCs are partly occupied unless their farm is 

highly depleted. 

3) Ownership of the farm: Whether a legume 

(mainly LCCs) could be grown by farmers or 

not depended on the authority of the person to 

decide on the existing land resources, which is 

linked to land ownership. Those farmers with 

insufficient farm inputs (seed, fertilizer, labour 

and/or oxen) are obliged to give their land for 

share cropping. In this type of arrangement, the 

probability of growing LCCs on that farm js 

minimal. Instead, farmers who contracted the 

land preferred to grow high yielding cereals 

(maize and wheat) or root crops (sweet potato). 

As share cropping is an exhaustive profitmaking 

arrangement, the chance of growing 

LCCs in such contracts was almost nil. Without 

ownership or security of tenure, farmers are 

unlikely to invest in new soil fertility amendment 

technology (Thomas and Sumberg, 1995). 

4) Livestock fced: In the mixed farming systems of 

Ethiopia, livestock is a very important enterprise. 

Farmers select crop species/ varieties not 

only based on grain yield but also straw yield. 

Similarly, legumes with multiple use were accepted 

by the community better than legumes 

solely for green manuring. 

5) Market value: For a legume technology to be 

appraised by farmer end-users, the legume 

should bring an immediate and visible benefit, 

~ither direct through the generation of food or 

cash or indirect by making a significant and 

visible contribution to a secondary high value 

product. 

The Decision Guides 

In this paper, we present two guidelines for integra 

tion of legumes into multiple cropping, perennial 

based farming systems. The decision trees were de 

veloped based on the following background infor 

mation from the site. 

1) Farmers prefer food legumes over non-food leg 

umes regardless of the soil fertility status of 

their farm. 

2) The above-ground biomass of food legumes 

(grain and stover) is exported to the homestead 

for feed and food while the below-ground biomass 

from food legumes is too small to affect 

soil fertility. The probability that the manure 

will be returned to the same plot is less as farmers 

prefer to apply manure to their perennial 

crops (Enset and Coffee) growing near the 

homestead. 

3) The tested legumes may fix nitrogen to fulfil 

their partial demand (we have observed nodules 

in all, although we did not quantify N 

fixation), but in conditions where the biomass is 

exported -- like with vetch for feed -- most of 

26 


the nutrient stock would be exported. Therefore, 

we did not expect a significant effect on 

soil fertili ty. 

4} Lees produce much more biomass when 

planted as relay crops in the middle of the 

growing season than when planted late as 

short-term fallows due to possible effects of 

end-of season drought on growth. 

5} The homestead field is much more fertile than 

the outfield; hence those species sensitive to water 

and nutrients will do better near the homestead 

than in the outfield. 

The first guide (Figure 2) is developed based on the 

data obtained from the farmers field and on-farm 

experiments, verified by on-station experiments. 

The overall idea is that not all Lees fit everywhere. 

Some are very sensitive to the availability of nutrients 

and water, at least during their establishment, 

and others do well across environments. When 

farmers got the option to select among seven commonly 

recommended Lees species (Vetch, Mucuna, 

Crotalaria, Canavalia, Tephrosia, Trifolium, 5tylosanthus), 

to integrate into their systems, farmers in various 

socio-economic categories selected different 

species, planted them on different parts of their 

farm and managed them differently. Researchers 

have monitored how the farmers managed the 

Lees, where they planted them, when did they 

plant, how long they were left to grow, how much 

input they invest, how was the biomass production, 

what benefits they get from them and what are their 

final decisions to integrate them into their systems. 

The guide, synthesised from the participatory research, 

has two major frames, one· for legumes suit 

able for maintaining the fertility status of productive 

land and another suitable for improving the fertility 

status of relatively less fertile cropland. Most 

farmers wanted the Lees to improve the plots that 

are 'addicted' to mineral fertilizers, which refers 

commonly to those less fertile corners of the farm, 

the out-fields. The guide showed that there are limited 

Lee options that could be used to improve degraded 

croplands, as the legumes themselves, except 

Crotalaria, were not able to grow under such 

harsh conditions. There are many more Lee options 

for maintaining the fertility status of the fertile 

corners of the farm. Vetch was suggested to be the 

best fitting legume for a short-term fallow. However, 

the guide left a space for other researchers to 

identify an Lee option that may fit into their specific 

production systems. 

The second guide (Figure 3) is intended to assist 

farmers and researchers to identify potential legumes 

that could be compatible with existing spatial 

and temporal niches. This .guide was developed 

based on the homestead being much more fertile 

than the outfield, and that the outfield is larger than 

the homestead field. The most important criterion at 

the lowest level is the presence or absence of livestock 

followed by who manages the farm, market 

access, the size of the land holding and the land 

quality. The factor that dictates the decision at the 

highest level is land productivity, which 'was governed 

mainly by soil fertility status. Growing of 

food legumes was the priority of every farmer regardless 

of wealth (land size, land quality and number 

of livestock). Farmers with)ivestock integrated 

feed crops regardless of land size, land productivity 

and market access to products. 

However, the size and quality of 

land allocated for growing feed 

< 

legumes depended on market access 

to livestock products (milk, 

butter and meat). Those farmers 

withgood market access are expected 

to invest part of their income 

into external inputs, i.e. inorganic 

fertilisers. Hence, farmers 

in this category did not allocate 

much land for growing Lees, but 

< 

applied inorganic fertilisers. In 

or 

Foio 1lliCi'0000i;lIO!l.lli 

the homestead field, there was no 

AlmIHy.of 

CI!OP InI land allocated for Lees in the 

system, because farmers gave pri 

1,P.I.'!"t 

~talGrlo ority to food legumes there, to 

or " 

take advantage of a relatively fertile 

com.er of the farm. The clearest 

spatial niche for growing 

Figure 2. Decision guide that suggests various legumes for improving degraded crop Lees is .the outermost field, espelands 

or maintaining the fertility status of relatively fertile crop land through a short cially in poor farmers' fields with 

or medium term fallow 

exhausted land and limited mar- 


Own livestock 

Gununo farmers for their direct involvement 

in the research. 

References 

Larg~ farm size 

Good market 

Food & feed 

cover 

'-.J"f!Ulne.~. 

Cover crops 

Figure 3. Guidelines for integration of food, feed legumes and legume cover crops 

into small scale farms, with heterogeneous socio·economic conditions 

Abayomi, Y.A., O. Fadayomi, J.O. 

Babatola, and G. Tian, 2001. Evaluation 

of selected legume cover crops 

for biomass production, dry season 

survival and soil fertility improvement 

in a moist savanna location in 

Nigeria. African Crop Science Journal 

9:615-627. 

Amede, T. and R. Kirkby, 2002. 

Guidelines for Integration of Legumes 

into the Farming Systems of 

'East African Highlands. Afnet Pro 

~eedings . African Academy of Sciences. 

In press. 

ket-driven farm products. Those categories of farmers 

experienced share cropping for some time, and 

as a result their farm was on the verge of going out 

of production due to unsustainable land management 

practices. 

Conclusion 

Integration of legumes into various production systems 

and for various clients is complex and requires 

a participatory approach to address both biophysical 

and socio-economic constraints and opportunities. 

The major biophysical traits that need to be addressed 

are adaptability of the species into that specific 

agroecology (which may include yield, disease 

and pest resistance), the effect on soil fertility and its 

compatibility with the existing cropping system. 

The most determinant socio-economic factors are 

land ownership, market value, farm size and tradeoffs 

for various uses. The strategic combination of 

those biophysical and socio-economic determinants 

in the form of decision guides will help farmers, development 

agencies and researchers to identify potential 

adopters, learn about the criteria of choice, 

and learn about the need for improved management 

of the system. Moreover, it may help them to identify 

niches and crea te niches, modify the existing 

systems and promote the technology for wider use. 

Acknowledgement 

I would like to thank Drs Ann Stroud, Roger Kirkby 

and Rob Delve for their valuable inputs and support 

during the research process, Mr. Wondimu 

Wallelu for his valuable inputs in the fieldwork, and 

28 

Eyasu, E., 2002. Farmers' perceptions 

of soil fertility change and management. 

50s-SAHEL, Institute for Sustainable Development, 

Addis Ababa, Ethiopia. 252 p. 

FishIer, M. and Wortmann, c., 1999. Crotal,aria (c. 

ochroleuca) as a green manure crop in maize-bean 

cropping systems in Uganda. Field Crops Research 

61:97-107. 

Gachene, c.K., C. Palm and J. Mureithi, 1999. Legume 

cover crops for soil fertility improvement in 

the East African Region. Report of an AHI Workshop, 

TSBF~ Nairobi, 18-19 February, 1999. 

Giller K., 2001. Nitrogen Fixation in Tropical Cropping 

Systems. 2nd edition. CAB International, 

Wallingford, UK. 423 p. 

Jama, B., R.J. Buresh, and F.M. Place, 1998. Sesbania 

tree fallows on phosphorus-deficient sites: Maize 

yield and financial benefits. Agronomy Journal 

90:717-726. 

Palm, c., R.J. Myers, and S.M. Nandwa, 1997. Combined 

use of organic and inorganic sources for 

soil fertility maintenance and replenishment. 

SSSA Special Publication No. 51, Madison, Wisconsin, 

USA. pp. 193-218. 

Sumberg, J., 2002. The logic of fodder legumes in 

Africa. Food Policy 27:285-300. 

Tarawa1i, S.A., M. Peters, and R. Schultze-Kraft, 

1987. Forage legumes for sustainable agriculture 

and livestock production in sub-humid West Africa. 

ILRI project report, Nairobi, Kenya. 132 p. 


Thomas, D. and J. Sumberg, 1995. A review of the 

evaluation and use of tropical forage legumes in 

Sub-saharan Africa. Agriculture, Ecosystems and 

Environment 54:151-163. 

Wortmann, c., M. Isabirye and S. Musa, 1994. Crotalaria 

ochroleuca as a green manure crop in 

Uganda. African Crop Science Journal 2:55-61. 

Wortmann, C. and B. Kirungu, 1999. Adoption of 

soil improving and forage legumes by small 

holder farmers in Africa. Conference on Working 

with Farmers: The key to- adoption of forage 

technologies. Cagayan de Oro, Mindano, The 

·Philippines. 12-15 Oct., 1999. 

Versteeg, MN., F. Amadji, A. Eteka, A. Gogan, and 

V. Koudokpon, 1998. farmers adaptability of 

MUClma fallowing and agroforestry technologies 

in the coastal savanna of Benin. Agricultural Systems 

56 (3):269-287. 


Questions and Answers 

~ey Papers 

To Bernard Vanlauwe, Andre Bationo et al. 

Q: Where do you place improved fallows in the 

systems you described? 

A: It is important to identify proper modes at the 

biophysical and socio-economic level. 

Q: Promotion of mucuna is limited by its lack of 

utilization options due to anti nutritional factors (Ldopa). 

What use did Sasakawa Global 2000 put 

mucuna to in order to increase the adoption by 

farmers? 

A: SG 2000 bought up mucuna seeds for further 

distribution to interested farmers. 

Q: Markets appear to be critical for the adoption of 

legume technologies, as shown by the mucuna case 

where numbers increased from 20 to 14 000 in 10 

years when Sasakawa was buying seeds. What role 

have markets played in the increased adoption of 

soyabean in Nigeria? 

A: Two routes were followed to create demand for 

soyabean; local processing and development of 

private-sector-driven markets. Both were successful 

in creating a demand although the proportion of 

both mechanisms would need to be verified 

through liT A. 

Q: It is important that we recognize the value of the 

word 'adaptation' in terms of developing 

dissemination messages. Adaptation reflects how 

farmers overcome complexities or constraints in the 

system. 

A: Agricultural adaptation is limited to the 

complexity of the interventions aimed at. 

To Ed Rowe and Ken Giller 

Q: What are the incentives for organizations and 

individuals to share information to develop 

simulation tools beyond the conceptual framework 

of NUANCES? 

A: Many people are interested to look at the broade'r 

costs and benefits of the technology or intervention 

that they are researching, to see whether it really is 

viable for the farmer. We have already seen a great 

willingness to share data, and models, which 

suggests that NUANCES is seen as useful and 

timely. 

Q: You have indicated that soyabean leaves little 

residual N, not enough to support the next cereal to 

maturity. Where then does the observed residual 

effect of soyabean on maize come from? Farmers 

have observed and 'harvested' maize grown on the 

residual effect. 

A: This observation, coming directly from farmers' 

experience, shows that the prototype soil-crop 

model is not predicting soya residue effects very 

well. This is great; the intention was just to 

illustrate the kind of predictions and analyses which 

NUANCES will provide, and the comment 

demonstrates the value of having better and faster 

feedback between model predictions and farmers' 

practice, so both can be improved. 

To Tilahun Amede 

Q: Faba beans were described as high N-fixation. 

What are the attributes that may account for that 

characteristic? 

A: Firstly the seeds of faba bean are large, with a 

considerable nutrient concentration, good enough 

to support nutrient demand during the early stages 

of growth. This leads to a vigorous start with 

prolific leaves able to synthesize enough 

carbohydrate to support N -fixation processes. 

Moreover, faba bean has prolific roots that may 

explore nutrients like P from a wider soil space. 

Q: The recommendation for outfields in the absence 

of livestock is to grow LCe. But earlier you 

presented that farmers found these were not 

growing well in their outfields? 

A: As farmers grow mainly maize and potato in the 

outer fields, and apply some inorganics like DAP, 

the residual nutrients could be enough to support 

the initial start of LCCs. 

Q: To what extent have the decision guides been 

able to predict the growing of a particular grain 

legume in the research areas? Were you able to 

quantify the decision guides, e.g. size of land 

holding and a specific pH that could suit each grain 

legume? . 

A: The simple guide that indiCates which legume 

could be used for what purpose can be used across 

the board as they are developed based on . 

biophysical traits. The other guides may require 

characterization of the socio-economic components. 


Abstract 

PROMOTING NEW BNF TECHNOLOGIES AMONG SMALLHOLDER 

FARMERS: A SUCCESS STORY FROM ZIMBABWE 

SHEUNESU MPEPEREKI 1 and ISHMAEL POMPI 2 

1Department of Soil Science and Agricultural Engineering, University of Zimbabwe, 

PO Box MP167, Mount Pleasant, Harare and 

2Agronomy Research Institute, Department of Research and Extension, 

Ministry of Agriculture, PO Box CY550, Causeway, Harare, Zimbabwe 

Biological nitrogen fixation (BNF) contributes significant amollnts of N into both managed and natural ecosystems and 

forn:s the basis for the age-old practice of rotating legumes with other crops. Benefits of legume N fixation include protein 

nutrition, soil fertility improvement, savings on fertilizer costs and cash income from sale of crop surpluses. The 

packaging and use of superior N-fixing rhizobial strains as commercial legume inoculants is a relatively cheap costeffective 

technology widely adopted by large-scale but not smallholder farmers in Zimbabwe. We report on a promotion 

program that used soyabean as a vehicle to convey the multiple benefits of BNF technologies to poor smallholder farmers 

through a multi-faceted research-extension-promotion effort. The primary objective was to strengthen rural food security 

of smallholder farmers through exploitation of soyabean BNF for soil fertility improvement against rising input 

costs. The main elements of the promotion strategy included training farmers and extension staff in technology applica 

'tion, demonstration of the tangible multiple benefits and facilitation of input/output marketing, all backed by a parallel 

program of adaptive research. The.basic promotion concept used was that of creating a closed loop with four links: training 

(in BNF technology application), production (of soyabean), processing and marketing (TPPM). Coordination of 

stakeholder activities was and continues to be a critical component of the promotion effort. A conceptual framework linking 

various elements (BNF technology, food security, soil fertility, cash income) was used to guide and focus both the 

promotion and research components. The rate of adoption of soyabean BNF among smallholders has been near exponential 

(from 50 farmers in 1996 to more than 10,000 in 2000). This paper outlines the conceptual framework and mechanisms 

used in the promotion of soyabean technologies, the responses of smallholder farmers and the prospects for wider 

scaling up. 

Key words: Soya bean, smallholder farmers, BNF, soil fertility improvement 

Introduction 

Nitrogen deficiency is the main limiting factor for 

high cereal yields iIi sub-Saharan Africa and'yet the 

majority of smallholder farmers use very little mineral 

N fertilizer. Biological nitrogen fixation (BNF) 

contributes significant quantities of nitrogen (N) to 

both natural and managed ecosystems and offers a 

relatively cheap alternative source of N for resource-poor 

farmers. Exploitation of BNF technologies 

in African farming systems requires the identification 

of appropriate N-fixing legumes that have 

multiple benefits to ensure adoption by risk-averse 

rural communities. There is need to develop a research 

agenda that identifies appropriate BNF technologies 

(e.g. effective legume-rhizobium combinations) 

that can be readily adopted by farmers with 

immediate demonstrable benefits to ensure adoption. 

Such research efforts will need to be linked to 

appropriate extension programs that ensure that 

target commlmities benefit in tangible ways. 

Traditional legumes such as groundnut (Arachis hy- 

pogaeae), cowpea (Vigna unguiculata) and bambara 

nut (Vigna jubterranea) that rely on BNF and contribute 

residual fertility to soils are low-yielding and 

are often viewed as minor crops. Yields of these legumes 

have failed to respond consistently to inoculation 

with commercial rhizobiaI strains. Soyabean, a 

relatively new legume in Africa, responds well to 

rhizobiaI inoculation and fixes large amounts of N 

even in marginal soils (Kasasa, 2000; Musiyiwa, 

2001). The multiple benefits of soyabean include soil 

fertility improvement, protein nutrition for humans 

and livestock and cash income from sales of grain 

and processed products. Soyabean is now grown in 

several parts of sub-Saharan Africa including Malawi, 

Nigeria, Zambia and Zimbabwe where it is 

making significant contributions to rural livelihoods. 

Due to limited inoculant production capacity 

in most African countries, promiscuous soyabean 

varieties that effectively nodulate with indigeno\.!s 

rhizobia have been successfully grown without inoculants 

demonstrating their potential for conveying 

the benefits of BNF to poor and marginalized 

communities (Mpepereki et al. 2000). 

Grain Legumes an,d Green Manures for SO'il Fertility in Southern Africa 

3~

Historical Perspective on Soyabean in 

Zimbabwe 

Soyabean was introduced into Zimbabwe (then 

Southern Rhodesia) in the 1930s as a green manure 

crop and later for forage. Large-scale commercial 

production started in the 1960s when a breeding 

program and a Rhizobium inoculant factory were 

established (Corby, 1967). The crop was not promoted 

among smallholder black farmers, most of 

whom had been relocated onto marginal, often 

sandy, soils in low rainfall areas unsuitable for soyabean 

production. Apart from the real agro~ 

ecological limitations, soyabean production, with its 

requirement for rhizobium inoculants that need refrigeration, 

was considered too sophisticated for African 

peasant farmers who had no knowledge on 

how to process it for food. 

After political independence in 1980, government 

adopted a policy of encouraging smallholder farmers 

to increase crop production through various inputs 

and marketing support programs. By the 

1990s, smallholder farmers were contributing over 

70% of national maize and cotton production. A 

soyabean BNF promotion program targeted at Hurungwe 

West district in northern Zimbabwe in the 

late 1980s boosted farmer interest, production and 

consumption of soyabean which all declined when 

project support ended in 1989 (Whingwiri, 1996; 

Mudimu, 1998). Smallholder farm communities 

however continued to face limited dietary protein 

sources, general declining soil fertility and poor 

household incomes against a background of increasing 

mineral N fertilizer prices, following World 

Bank/IMF-induced removal of government subsidies. 

A two-day stakeholders' workshop that was 

held at the University of Zimbabwe in February 

1996 to examine the potential for promiscuous soyabean 

for smallholder farmers recommended two 

major activities. First it resolved that research be initiated 

to characterize indigenous soyabean rhizobia, 

the potential for promiscuous soyabean and to 

quantify the amounts of N fixed and the residual 

fertility benefits for maize grown in rotation. Secondly, 

it was resolved to extend soyabean technologies 

(the use of rhizobial inoculants and promiscuous 

varieties for BNF, production, processing, utilization 

and later input/output marketing) to smallholder 

farmers. A National Soyabean Promotion 

Task Force with representation from farmer organizations, 

private industry, NGOs and public institutions 

(research, extension, university) was formed. 

The Task Force was to be convened by AGRlTEX 

with overall coordination by the University of Zimbabwe 

Faculty of Agriculture. The Task Force objectives 

included promotion of soyabean through appropriate 

research, training farmers in production 

and processing and coordinating the activities of 

various stakeholders. 

This paper outlines the conceptual framework and 

mechanisms used to promote soyabean technologies, 

the scale of operations, feedback from farmers, 

constraints and opportunities and the potential for 

scaling up. 

Conceptual Framework 

The context was that of smallholder cropping systems 

characterized by low productivity due to low 

soil fertility, with N as a major limiting nutrient. 

Biological N fixation (BNF) was identified as a potential 

tool to address N deficiency in these systems. 

Soyabean was chosen as the candidate legume because 

of its high N-fixing potential and soil improving 

properties, food value as a protein and vegetable 

oil source, relatively low production costs and 

high market value. The place of soyabean BNF in 

the total food production system of a typical smallholder 

farm was identified. This was an essential 

step to ensure that the technology would address 

real food security concerns of farmers, a critical element 

for successful adoption. The conceptual framework 

illustrated below (Figure 1) shows the main 

linkage loops and benefits from soyabean BNF in an 

integrated maize-based crop-livestock system. 

Strategies Translating the Concept into 

an Operational Model 

For research, a proposal was written up, funds 

sourced and graduate students engaged to conduct 

research to quantify N inputs from promiscuous 

and commercially inoculated soyabean into the 

cropping system and to measUre and demonstrate 

the residual soil fertility benefits for maize in subsequent 

seasons. Research was conducted to establish 

the prevalence and symbiotic effectiveness of indigenous 

rhizobia on both promiscuous and specific 

soyabean varieties and the adaptability of the latter 

to the more agro-ecologically marginal smallholder 

areas. Experiments were conducted both on-station 

and on-farm under researcher and farmer-extension 

~ L:'::':~::d~ 

So,_be,," BNF~ s,,;. F",~';" ~

management respectively. This meant that researcher-managed 

detailed replicated field experiments 

were placed on a few farms selected for their 

representative soil types, while a larger number of 

simple plus/minus treatment trials were run under 

farmer management with extension officers monitoring 

them. Rhizobial inoculation, liming and basal 

compound fertilizers and promiscuous versus specific 

nodulating soyabean varieties were tested. 

Both farmers and extension personnel helped to set 

up and monitor experiments and gained valuable 

experience and confidence in managing a soya bean 

crop. Scientific data obtained was used to 

strengthen the extension messages that had hitherto 

been extrapolated from work done in large-scale 

commercial production under somewhat different 

agro-climatic conditions. 

For the promotion aspect, the main strategies were: 

training of both farmers and extension staff on how 

t9 apply rhizobial inoculants, how to grow, weed 

and harvest soya bean; facilitating access to inputs, 

setting up technology transfer demonstrations that 

involved farmers and extension staff, regular follow-ups 

and communication in local languages at 

all times. Train-the- trainer workshops targeted extension 

staff in AGRITEX, NGO personnel and 

farmer leaders identified by their organizations and 

employed a hands-on practical approach. Topics 

included how to store and apply rhizobial inoculants, 

use of promiscuous varieties where inoculants 

are unavailable, how to check if nodules are effective, 

identification of pests and diseases and their 

control. 

Training was consolidated by a vigorous pr

ganic amendment for resource-poor farmers who 

cannot afford adequate mineral fertilizers. For 

many African farmers, livestock represent a critical 

investment or "money in the bank", as they can be 

sold to meet food and other budgetary needs of the 

family. 

The lowest loop on our conceptual model (Figure 1) 

emphasizes the link between soyabean BNF and 

cash income. Each soyabean harvest provides food, 

seed and surplus for sale. In the Zimbabwean 

model, the Task Force working with farmer's organizations, 

commodity brokers and processors put 

in place marketing arrangements to ensure that 

farmers received fair prices for their soyabean grain. 

The key to success has been effective load consolidation, 

identification of lucrative markets and affordable 

transport. Initially volumes were small and 

marketing costs very high, but as more farmers took 

up the crop, volumes increased allowing for economies 

of scale. A comprehensive study to analyze the 

economic potential of soyabean showed that there 

are'...potential benefits ... for smallholder farmers, 

particularly the poorer smallholders ... ' in Zimbabwe 

(Rusike et al. 2000). 

For the adoption rate to be sustained, there was 

need to coordinate the efforts of many stakeholders 

that are involved. Figure 2 illustrates the range of 

possible linkages that are involved in the soyabean 

BNF research -extension program in Zimbabwe. To 

facilitate coordination, a unit for that purpose was 

established in 2000 under the Promotion Task Force. 

Its major function was to provide technical backup 

and training to various groups engaged in soyabean 

production and to mobilize stakeholders. Currently 

stakeholders are setting up a soya bean development 

trust to take over coordination of all stakeholder activities 

in research production, processing, marketing 

and training in the whole country. 

FMmmg communilU~S 

Farmer 

~ Ore3msatlOl1S 

! ~ ? \ 

~ten"~ '"1' '"Or'" 

MIcro- 

Seed, 

• • credi1 tOoc ulants 

('oordiutiul Uuit 

UIl.:lnuity of Zimb .. 1I ..... 

Figure 2. Coordination linkages of stakeholders in the soyabean 

promotion program in Zimbabwe. Key activities in the links include 

training, information exchange, adaptive research and movement of 

inputs, outputs and cash. 

36 

Outcomes 

In general the research-extension program has successfully 

introduced and brought benefits of soyabean 

BNF to thousands of smallholder families in 

Zimbabwe. Promiscuous soyabean has enabled 

farmers with no access to commercial inoculants 

also to adopt soyabean. Up to 50% of soyabean produced 

in Hurungwe district in northern Zimbabwe 

in the last four seasons (1998 -2001) was promiscuous, 

while in Zambia and Malawi promiscuous Magoye 

still forms the backbone of smallholder soyabean 

production (Javaheri, 1996). Promiscuous soyabean 

forms the bulk of varieties planted in Nigeria. 

Below we summarize results from various research 

initiatives undertaken within the conceptual framework 

described to illustrate the kinds of information 

being generated. 

Quantities of N fixed by promiscuous arid specific 

soyabean varieties under field conditions were 

measured (Table 1). In demonstrating the residual 

soil fertility benefits of rotating maize with soyabean, 

yields of both grain and stover were quantified. 

Yields of maize following soyabean were significantly 

higher than those of maize after maize, demonstrating 

significant residual fertility effects of soyabean 

(Table 2). Residual effects ensure sustainable 

food production in a soyabean maize rotation. Soya- 

Table 1. Nitrogen yields from promiscuous and specific (improved) 

soyabean varieties at Hotera smallholder farm, Hurungwe, 

Zimbabwe, 1997 

Soybean variety %Nderived from fixation Fixed N(kglhal 

. Inoculation + Inoculation . Inoculation + Inoculation 

Magoye 91 90 73 58 

Local 90 90 57 58 

Roan 91 88 63 66 

Nyala 92 82 46 58 

s.e.d 3.8 15.8 

'Magoye' and 'local' are promiscuous; 'Roan' and 'Nyala' are specific commercial 

~arieties. (Adapted from Kasasa et al. 19981. 

Table 2. Maize yields over two seasons following soyabean 

in asandy loam soil in asmallholder farm, Hurungwe, 

Zimbabwe, 1998·99 

Soyabean variety Soyabean biomass Maize yields (tlhal 

(961971 incorporated Itlhal 

97198 98/99 

Magoye (proml 5.4 2.3 1.2 

Local (prom.l 4.9 2.1 1.4 

Roan (spec.I 3.2 1.8 0.9 

Nyala (spec.l 2.8 1.4 0.8 

Maize control Nil 0.19 0.2 

Prom - promiscuous; Spec. - Specific 


ean residual fertility effects on maize have been 

demonstrated under farmer management, boosting 

adoption of soyabean BNF by smallholder farmers. 

Mineral fertilizer inputs (e.g, Cu, Mg, P, K) will continue 

to be required to prevent nutrient mining of 

soils. Extension messages must continue to emphasize 

the critical importance of inorganic fertilizer 

amendments. 

An important benefit of soya bean BNF has been the 

boost in household incomes from grain sales by 

farmers (Table 3). A critical element in the promotion 

program was the consolidation of loads so that 

economies of scale have enabled the relatively small 

production by each farmer to be sold on the lucrative 

commodity exchange as part of a large batch. 

Thus the conceptual model for promoting BNF includes 

produce marketing as a key element. 

A study of the economic potential of soyabean 

showed that the crop was most profitable for the 

poorest farmers as it had lower input costs but gave 

the highest return on investment (Rusike et al. 

2000). Poor farmers who adopted soyabean for the 

first time between 1997 and 2001 have testified that 

they earned more money from soya bean sales than 

from any other crop that they have ever grown 

(Table 4). The significant boost in family dietary 

protein availability (Table 4) is a critical element of 

household food security, a key benefit of BNF 

among rural communities where poor nutrition 

among the HIV-infected is contributing to the high 

death toll from AIDS related illnesses. 

Table 3. Soyabean grain sales by smallholder farmers from 

four locations over 4 marketing seasons in Zimbabwe 

. location Amounts sold (metric tl 

96/97 97/98 98/99 99/2000 

Guruve 6.2 53 153 210 

Kazangarare 58 280 475 580 

Sadza 0.5 3.5 7 10.2 

Senge 0.2 6 11 18.1 

Total sold 64.9 342.5 646 818.3 

Only sales facilitated by the Soyabean Promotion Task Force are reflected; 

farmers also used other marketing outlets. 

Table 4. Grain, protein and cash returns from soyabean for Tapera 

smallhold farm in Zimbabwe (1998) 

Soyabean Total grain yield Protein from 150/0 Cash from 70% 

variety (kg/hal seed retained grain sold 

(kg/hal 

(US$ equivl 

Magoye 2100 126 471 

local 1900 114 302 

Roan L800 168 496 

Nyala 3100 186 560 

Average smallholder planting: 0.4 ha; average yield: 0.8 t/ha; average price; US 

$360/t (2001). 

Conclusions 

Our experiences with developing and implementing 

a research-extension model for promoting BNF 

technology among peasant farmers in Zimbabwe 

offers lessons for similar initiatives in developing 

countries. Previous experiences of promoting promiscuous 

soyabean in Nigeria (N. Sanginga, pers. 

comm.), Malawi and Zambia (Mpepereki et al. 2000) 

also point to the need for integrated approaches that 

address both the scientific-technological and socioeconomic 

aspects in a holistic way (closing the 

loop). Demonstration of multiple benefits of N 

fi xing soyabean, use of promiscuous varieties, training 

women in home processing, adapting soya bean 

to local diets and facilitating input/output marketing 

(all carried out in the context of a clear conceptual 

framework with stakeholder participation), 

have resulted in rapid adoption of soyabean by 

thousands of smallholder farmers, thereby strengthening 

their food security in a sustainable way. An 

integrated program of adaptive and applied research 

ro support the soyabean BNF promotion initiative 

has provided a scientific basis for a technical 

backup service to adopting farmers. The success of 

such a promotion program depends on the number 

of actual and demonstrable benefits to the smallholders 

and the commitment of all stakeholders to 

implement its various facets in a coordinated way. 

Marketing, both in terms of inputs and outputs, is a 

key driving force for soya bean BNF technology 

adoption. More BNF grant funds must go into activities 

that directly benefit farm families than project 

personnel salaries and per diems. Legume BNF 

can make a difference to rural livelihoods. 

Acknowledgements 

We thank the Rockefeller Foundation for funding 

our soyabean BNF research and extension work in 

Zimbabwe. 

References 

Javaheri, F. 1981. Release of four new soya bean varieties. 

Mimeo Government of Zambia, Lusaka. 

Kasasa, P., Mpepereki, S. and Giller, K.E. 1998. 

Nodulation and yield of promiscuous soyabean 

(Glycine max L. Merr.) varieties under field conditions. 

In: Waddington, S.R., Murwira H.K., 

Kumwenda J.D.T. Hikwa D. and Tagwira, F. 

(eds). Soil Fertility Research for Maize-based Farming 

Systems in Malawi and Zimbabwe. SoilFertNet/ 

CIMMYT, Harare, Zimbabwe. pp. 93-103. 

Kasasa, P. 1999. Quantification of nitrogen fi xation 

by_promiscuous soya bean in Zimbabwean soils. 


37

MPhil. Thesis. University of Zimbabwe, Harare, 

Zimbabwe. 

Mpepereki, 5., Javaheri, F., Davis, P. and Giller, K.E. 

2000. Soyabeans and sustainable agriculture: 

promiscuous soyabean in southern Africa. Field 

Crops Research 65:137-149. 

Musiyiwa, K. 2001. MPhil. Thesis. University of 

Zimbabwe, Harare, Zimbabwe. 

Rusike, J., Sukume, C. Dorward, A., Mpepereki, S. 

and Giller, K.E. 2000. The economic potential of 

smallholder soyabean production in Zimbabwe. 

Soil Fert Net Special Publication. CIMMYT, Harare, 

Zimbabwe. 

38 


RESPONSE OF BEAN (PHASEOLUS VULGARIS, L.) CUL TIVARS TO 

INOCULATION AND NITROGEN FERTILIZER IN ZAMBIA 

FRIDAY SIKOMBE, OBED I LUNGU, KALALUKA MUNYINDA and MASAUSO SAKALA 

Abstract 

University of Zambia, Department of Soil Science, Lusaka, 

and Mount Makulu Research Station, Chilanga, Zambia 

Bean is an important component of the diet of people of Zambia, and many farm households grow it for subsistence and 

barter in their communities. However, grain yields are low, typically 500 to 700 kg ha- 1 with local cultivars and without 

supplemental nitrogen fertilizer application. Many soil and plant factors have been investigated to explain these low 

yields, but there is still limited information on the contribution of soil fertility, variety and inoculation to improvement 

in bean Yi.elds. A field study was conducted to evaluate the response of bean cultivars to applied nitrogen fertilizer and 

to inoculation with native and introduced rhizobium strains. The experiment was set up as a 5 x 5 factorial design comprising 

25 treatment combinations of five common bean cultivars and five nitrogen sources (3 strains and 2 nitrogen 

fertilizer levels). The treatments were replicated four times and arranged in a randomized complete block design at 

Mount Makulu Central Research Station, Chilanga, Zambia . The data collected included nodule count, dry nodule 

weight, dYlJ shoot weight, total nitrogen content in shoots and grain weight. The amount of nitrogen fixed by the inoculated 

crop was estimated by the difference method, using wheat as the non-fixing control crop. The results show that a 

combination of some strains with some cultivars tested is as effective as applying nitrogen fertilizer to the crop. An effective 

strain such as T AL1383 increased grain yield by 38.2% with some cultivars compared to the average grain yield 

of the other four strains with other cultivars. The local strain isolated from nodules of common beans grown locally was 

comparable to introduced strains in Mbala and Lundazi cultivars. The reduction in biological nitrogen fixation (BNF) 

by inorganic nitrogen application was more with the Lundazi cultivar than other cultivars. The native rhizobia strains 

at the trial site were as effective as the introduced strains. This study has shown that optimization of the effect of inoculation 

lies in identifying and matching bean cultivar to Rhizobium strain. Therefore, because of strain/cultivar specificity, 

it may be advisable to develop a broad-spectrum inoculum for use with bean cultivars in Zambia . 

Key words: Nz-fixation, N fertilization, rhizobium strains, common bean 

Introduction 

Beans are produced for both domestic consumption 

and sale in Zambia. Some bean leaves are consumed 

as a vegetable, and only cultivars with palatable 

leaves are consumed; other cultivars have tough 

textured leaves. The major production areas in this 

country are the high rainfall areas of Northern, 

Northwestern, Luapliia Provinces and medium to 

high rainfall areas of Eastern and Central Provinces. 

In other provinces, production of beans is on a small 

scale. 

Most farmers prefer to grow local cultivars for their 

colour and taste. However, average grain yields of 

local cultivars are exceptionally low (500 -700 kg 

ha- I ) even under commercial production (Annual 

Report, 1978). Beans experience a deficient in nitrogen, 

which results in poor yields (Lupwayi and 

Mkandawire, 1996). 

To improve bean yields, in the absence of effective 

rhizobia, it is recommended to apply nitrogen fertilizer. 

However, most resource-poor small-scale 

farmers are unable to afford N fertilizers. The 

cheaper option, therefore, is to exploit Biological 

Nitrogen Fixation (BNF) through inoculation with 

Rhizobia, and use bean genotypes that respond well 

to inoculation. Llipwayi and Mkandawire (1996) 

made similar observations to other researchers, that 

inoculation with some strains of rhizobia increased 

yield in common beans. 

Some factors may cause failure of applied inoculum 

to increase grain yield. According to Weiser et al. 

(1985), soil pH, low phosphorus, high levels of exchangeable 

aluminium and manganese, poor nutritional 

status, and water stress may limit nodulation 

and nitrogen fixation. Further, high levels of applied 

N or soil N can inhibit nodulation (Munyinda, 

personal communication). Nodulation and N2 fixation 

is also influenced by climatic factors such as 

light (Antoninew and Sprent, 1978), temperature 

(Rennie and Kemp, 1981) and cultural aspects such 

as planting density (Graham and Rosas, 1978). 

In Zambia, very little research work has been conducted 

on the inoculation of common bean. A great 


deal of research work has been biased towards soyabean. 

Although there are numerous bean cultivars 

and native rhizobia strains, these have not been 

identified and exploited. Therefore, extensive 

screening of bean land races for N2 fixation is required 

throughout Zambia. 

The objectives of the study were a) to evaluate the 

response of some cultivars of common bean to inoculation 

using native and introduced rhizobia 

strains, b) to identify effective strain/cultivar combination 

for optimal N2 fixation and c) to evaluate 

the response of common bean to nitrogen fe~filizer 

application. 

Materials and Methods 

The field experiment was conducted during the 

2001/2002 cropping season at Mt. Makulu Central 

Research Station located 15 0 32'S 20 0 15'E near to 

Lusaka, Zambia. The station received 617.5 mm annual 

rainfall, though normally the area receives between 

800 and 1000 mm rainfall annually. 

The land used for the trial had not been previously 

planted to any legume. Sorghum was grown during 

the immediate previous cropping season 

(2000/2001). Before planting, soil samples were 

taken at random for analysis to establish the initial 

fertility status of the site. The initial physical and 

chemical characteristics of the soil at the trial site are 

given in Table 1. 

The experiment was set up as a 5 x 5 factorial design 

comprising 25 treatment combinations of five common 

bean cultivars/varieties and five nitrogen 

sources (3 inoculum strains and 2 nitrogen fertilizer 

levels). The treatments were replicated four times 

and arranged in a randomized complete block design. 

Each treatment plot measured 3 m x 1.5 m, with 

four crop rows spaced at 50 cm apart. The harvest 

40 

Table 1. Initial physical and chemical characteristics 

of the soil used 

Texture 

pH CaCb 7.2 

Organic Carbon 1.37% 

Total Nitrogen 0.09% 

Available Phosphorus 

Potassium 

Calcium 

Magnesium 

Zinc 

Iron 

Manganese 

Sand Clay l,oam 

13mgkg' 

0.84 cmol (+) kg' 

42.8 cmol (+) kg' 

2.1 cmol (+) kg" 

11.0 mg kg' 

98 mg kg' 

456 mg kg" 

area for grain yield was a sub-plot of 2 m x 1 m, and 

the two outer rows were used for sampling. Each 

plot was isolated from the adjacent plot by a border 

of 0.5 m. 

Five bean cultivars were evaluated; three land races 

(Mba la, Solwezi and Lundazi) and two improved 

varieties (Carioca and Pembela). An improved 

wheat variety (Coucal) was used as a reference crop 

for N2 fixation. Two exotic rhizobia strains (CIAT 

899 and TAL 1383), one local isolate and native 

rhizobia at the experimental site used as a control 

were evaluated. Nitrogen was applied at two rates 

of 0 and 100 kg N ha- 1 • The nitrogen was applied as 

a split, 20 kg N ha- 1 at planting as compound 0 (N P 

K S: 10:20:10:10) and 80 kg N ha- 1 as urea at 25 Days 

After Sowing (DAS). 

The ~eans were planted in rows 50 cm apart and 10 

cm; within the row. The planting depth was approximately 

4-5 tm deep, and one seed per station. 

Wheat was grown in adjacent plots to the beans and 

it was drilled in rows 50 cm apart. Weeding was 

carried out by hand hoeing in all the plots at 17, 28 

and 45 DAS. 

Five plants from the discard rows of each plot were 

randomly sampled at 50% flowering or 5-6 weeks 

after sowing. After thorough washing the nodules 

were detached, counted and then dried in an oven 

at 65°C for 48 hours to obtain nodule dry weight. 

The shoots were dried in the oven at 70"C for 48 

hours to obtain shoot dry weight. 

The Nitrogen Difference Method by Hansen (1994) 

was used to determine the amount of nitrogen 

fixed. The fixed N was calCulated from: 

N2 fixed = NI - Nnf 

where: 

NI is the N accumulated by the fixing legume and 

Nnf is the N taken up by the rE:ference crop. 

The data were analyzed using the Genstat statistical 

package. The means were separated using the Duncans 

multiple range test. 

Results and Discussion 

Response to Inoculation 

The results show that nitrogen sources had a signifi 

cant (P

4.5 .. - ---.••....---.-.-- .-----.•...-.-.----.-....--..---- ..•. - - .. ---- -.•.....•...--.-....... - 

4-!---------------1' 

3.5 -!------==-----.---I 

I!! 

~ 3 

'" 2.5 i--- 

z 

OJ 2 

:; 

-g 1.5 

z 

0.5 

O+-J-~-r_~-L_.~-~_r~-L-,-~~~ 

Carioca Mbala Solwezi Lu ndazi Pem be la 

Cultivars 

Figure 1. Nodule number for bean cultivars following inoculation 

and fertilization 

Results of percent nitrogen content are presented in 

Figure 2. There was a significant interaction between 

nitrogen source and variety / cultivar. The 

highest percent nitrogen content was obtained from 

the inorganic nitrogen source followed by CIAT 

899. Alt the cultivars responded to inorganic nitrogen 

application except LLLndazi . All the varieties 

also responded to CIAT 899 except Mbala, but the 

response was greater with Carioca and Lundazi 

than wi th the other three cultivars. 

Results of nitrogen fixed by cultivars are presented 

in Figure 3. There was a significant difference (P< 

0.05) between cultivars and nitrogen sources. CIAT 

899 was more effective than other strains across all 

the varieties except Mbala, but it was most effective 

with Carioca and Lundazi. Pembela only responded 

to TAL 1383. 

The native rhizobia at the trial site were as effective 

as OAT 899 with Solwezi and Pembela. Carioca 

was less sensitive to the reduction of BNF by inorganic 

nitrogen. The reduction in BNF by inorganic 

nitrogen was more in Lundazi than other cultivars. 

Deibert et al (1978) reported that nitrogen levels 

above 45 kg ha·1 inhibited nitrogen fixation in soybeans, 

and the trend was the same in Lundazi. This 

result suggests that if Lundazi is to be grown with 

inoculum, the levels of inorganic N in the soil 

should not be excessive (greater than 100 kg N ha·1). 

Effect of inoculation on yield 

Results of the shoot dry weight and grain yield are 

presented in Figures 4 and 5. There was a response 

of shoot dry matter to nitrogen across the varieties. 

CIAl' 899 was more effective than other strains in 

increasing dry matter yields across varieties. The 

effect was more for Carioca, Solwezi and Lundazi, 

than in Mbala and Pembela. This effect was comparable 

to the application of inorganic nitrogen. The 

local isolate was as effective as the introduced 

strains in Mbala and Lundazi, and in Mbala it was 

even more effective than CIAT 899. The native 

0.35 

0.3 

ImCIAT899 -TAL1383 0LOCAL o CONTROL -N FERT. I 

H" 

; 0.25 

Cl 

o 

.~ 0.2 t- 

z 

Ii 

~ 0.15 - t- t- t- . 

:§ 

~ 0 .1 r-- t- t- t 

0 .05 

r-- 

t- t 

o 

t-t 

Carioca Mbala Solwezi Lundazi Pembela 

Cultivars 

Figure 2. Effect of Nsource on %N of bean cultivars 

80 

70 

~60 

..c: 

I:llJ CIAT899 .TAL1383 0 LO CAL DCONTROL.N FERT. I 

~50 '" " r- r 

"t:J 

: 40 i 

Ii: 

! 

~ 30 t- t-- l- e-' 

.., '" ~ 20 l- t- t- r 

z 10 - r Ir I- l I- 

c 

Carioc a Mt·ala Solwezi Lundazi Pembela 

Cultivars 

Figure 3. Nitrogen fixed by bean cultivars following inoculation and 

fertilization 

-&:: 

10 

ra 

9 

Q. 

:§ 

8 

7 

.r:: - 

0) 6 t- - - ' 

'0; 

'--:: 

5 t- .:=-. - 

~ 

4 - r - 

- 

~ 3 - t- - 

'0 

 

2 

- 

- r- - r 

0 

0 1 - r- - 

.r:: 

(/) a 

IEl CI AT899 -TA11383 OLO CAL O CONTROL -N FERT. I 

..................... .. . ............. ............... .. .................................................. .. 

Carioca Mbala Solwezi Lundazi Pembela 

Cultivars 

Figure 4. Effect of N source on shoot dry weight of bean cultivars 

rhizobia strains at the trial site were as effective as 

the introduced strains in Mbala, Solwezi and Lundazi. 

TAL 1383 was specifically selective to Pembela. 

There was a significant yield increase (p < 0.05), 

with application of inoculum and inorganic nitrogen 

fertilizer. The increase was greatest in Lundazi 

with inorganic nitrogen application and least in 

Pembela. Overall TAL 1383 was more effective than 

other strains in increasing grain yield across the cultivars, 

and it was even more effective in Lundazi 

J 

i 

Grain legumes and Green Manures for Soil ~ertility in Southern Africa 

41

1600 

'7 

ns 1400 

~ 

1200 

Cl 

~ 

1000 

"C 800 

.~ 

>. 600 

c:: 400 

'n; 

... 200 

C) 

0 

nI 

ROLE OF PHOSPHORUS AND ARBUSCULAR MYCORRHIZAL FUNGI ON 

NODULATION AND SHOOT NITROGEN CONTENT IN 

GROUNDNUT AND LABLAB BEAN 

YLVER L. BESMER 1, R.T. KOIDE 1 and S.J. TWOMLOW 2 

Abstract 

1Pennsylvania State University, University Park, PA, USA 

21CRISA T, Matopos Research Station, Bulawayo, Zimbabwe 

Rotations with legumes have been suggested as a means to increase cereal production in low-input agriculture in Zimbabwe, 

as cereal yields are currently limited by nitrogen (N). However, NJixation by legumes is often phosphorus (P) 

limited. Soil availahle P explained 67% of the variation in nodule numbers when groundnut was grown on a wide 

range of soils collected from subsistence farm er's fields in southern Zimbabwe. P applications on a luvisol and vertisol 

in Tsholotsho, south -western Zimbabwe, can significantly increase nodule mass, aboveground biomass and total N in 

residues of groundnut (Arachis hypogaea L), lablab bean (Lablab purpureus) and pigeonpea (Cajanus cajan (L) 

Millsp.) . However, P fertilizers are often beyond the economic means of subsistence farmers. The success of the legumes 

therefore, will strongly depend on their ability to utilize the P already l/1 the soil. Arbusclilar mycorrhizal fungi (AMF) 

are components of most natural ecosystems and form a symbiosis, arbuscular mycorrhiza, with approximately 80 percent 

of all terrestrial plants. The fungi can increase plant P uptake by increasing the surface uptake area. We have 

shown in a pot trial that by enhancing the AM colonization through an inoculation with AMF in a luvisol from 

Tsholotsho, nodule number and N content of the shoot significantly increased in groundnut and lablab bean. This study 

indicates that by exploring the biology of the agro-ecosystem , beneficial effects could be obtained by optimizing the mlltualistic 

interactions between the plant, bacteria and fungi. Ways to enhance AMF inoculum potential in the field s of 

subsistence farmers are currently being tested and are disCllssed. 

Key words: Arbuscular mycorrhiza, rhizobia, phosphorus, groundnut 

Introduction 

A majority of subsistence farms in Zimbabwe occur 

on communal land. Maize (Zea mays) is grown as a 

staple, often on nutrient depleted sandy soils iow in 

organic matter (Grant 1967, 1981, 1985). Inorganic 

fertilizers, once subsidized by the government, are 

scarce in rural areas and often beyond the economic 

means of subsistence farmers (Mapfumo and Giller, 

2001). With declining maize yields due to nitrogen 

(N) limitations (Snapp, 1998), there has been renewed 

interest by researchers in using N2 fixing legumes 

in intercropping and rotational cropping systems 

to increase soil fertility (Snapp et al., 2002; Ma 

et aI, 1998), a common practice in much of southern 

Africa prior to the introduction of mineral fertilizers 

(Howard reprinted in Small Farmer's Journal, 1999). 

Groundnur (Arachis hypogaea L), cowpea (Vigna llndiculata 

(L) Walp.) and bambara groundnut (Vigna 

subterranean (L) Thou.) are currently grown for huan 

consumption and animal feed in Zimbabwe. However, 

their capacity to improve soil fertility might be 

limited for at least two reasons. First, they are all 

grain legumes where much of the N is translocated 

to the seeds and removed from the field at harvest. 

Second, optimal N2 fixation might be limited by 

phosphorus (P), as soil P availabilities are generally 

low in the old, highly weathered soils of sub 

Saharan Africa (Warren, 1992; Buresh et a L, 1997; 

Giller, 2001). 

In a previous experiment we have shown that when 

groundnut was grown on a wide range of soils collected 

from subsistence farmer's fields in southern 

Zimbabwe, nodule numbers differed by an order of 

magnitude. Further, soil available P explained 67% 

of this variation (Besmer et al., unpublished), suggesting 

the importance of this element for nodulation 

and N 2 fi xa tion. Applications of P (40 kg P20 s/ 

ha) to two soils from Tsholotsho, a luvisol and vertisol 

(Moyo, 2001), both low in P, significantly 

(p

Arbuscular mycorrhizal fungi (AMF) colonize roots 

of about 80% of terrestrial plant species and can increase 

plant P uptake by increasing the surface uptake 

area (Koide 1991). Synergistic effects on legumes 

are frequently seen when both symbionts (the 

rhizobia and the fungi) are present (Goss and de 

Varennes, 2002; Sanginga et ai., 1999; Fitter and Garbaye, 

1994). Both nodule number and dry weight 

usually increase after mycorrhizal colonization 

(Reddy and Bagyraj, 1991), which is often explained' 

by increased P uptake by the fungi. While mycorrhizal 

fungi are components of most natural ecosystems, 

their abundance and efficacy can be severely 

retarded by common agricultural practices such as 

fallowing, soil disturbance through tilling and weed 

management, and prolonged cultivation of non-host 

plants (Boswell et al. 1998; Kabir et al., 1997; Douds 

et al. 1995; Harinikumar and Bagyraj, 1989). 

The objective of our work was to understand the 

role of AMF for legume performance in subsistence 

farmers' fields, and to determine if an enhanced 

AMF abundance can promote nodulation and N2 

fixation. In this paper we discuss the results of two 

pot experiments. In Experiment 1, the effect of an 

altered AMF abundance on nodulation and shoot N 

content on groundnut was determined by enhancing 

the AMF abundance through an inoculation 

with a common AMF, or by reducing the indigenous 

AMF abundance through a fungicide application. 

In Experiment 2, the abundance of indigenous 

fungi was enhanced and the effects on nodule number, 

nodule mass and shoot N content were determined 

on lab lab bean. Groundnut was chosen since 

it is a common legume grown by the subsistence 

farmer in Zimbabwe, and lablab bean because it is a 

green manure crop and therefore has a higher potential 

to improve soil fertility. 

Material and methods 

General 

The soil used in both Experiment 1 and Experiment 

2 was a Tsholotsho luvisol (from Simeon Moyo's 

farm) where legume P limitations had been demonstrated 

previously. The pH of the soil was 6.2 (1:2 V 

soil: V water), and available P was 1.2 ppm (Olsen). 

Soil was collected to a depth of 15 cm in December 

1999, for Experiment 1, in April 2001 for the inoculum 

production part of Experiment 2, and in November 

2001, for the inoculation part of Experiment 

2. For Experiment 1 the soil was collected randomly 

from the field where plants were currently grown 

and no fertilizers had been added, and for Experiment 

2 from areas where maize had been grown the 

previous year without fertilizer additions. 

Experiment 1 

Groundnut (var. Falcon) was planted on December 

28 1999, in 1.6 L pots and grown for 6 weeks in soil 

amended with P [2 g single superphosphate pot- l 

(19% P20S)], AMF (2000 spores pot l of Glomus intraradices, 

Schenk and Smith), a fungicide Benomyl 

(200 mL of 0.1% solution added one day prior to 

planting), or control consisting of non-sterile soil. 

No additional fertilizers were added and the plants 

were watered as needed. At harvest, nodule numbers 

were determined along with shoot Nand P 

concentrations and AM colonization using standard 

procedures (Brundrett et ai., 1996; Watanabe and 

Olsen, 1965; Jensen 1962). 

Data were analyzed using a one-way ANOV A and 

transformed where appropriate. When transformation 

failed to generate data that fulfilled the underlying 

assumptions, data were analyzed using nonparametric 

tests. 

Experiment 2 

Production ofAMF and control inoculum 

Maize was planted in the soil in July 2001, and 

grown for three months to enhance the abundance 

of indigenous AMF. Control pots consisted of 

maize grown in sterile soil that had been given 

spore washings containing soil bacteria and fungi 

but lacking AMF. The plants were fertilized six 

times with Peters fertilizer 15-0-15NK plus micronutrients 

(at a N concentration of 100 ppm) and 

amended with 0.15 giL MgS04 and 5 mg PIL as 

KH2P04. After three months the maize plants were 

allowed to wilt, and dry soil and cut root pieces 

served as a source of inoculum, which consisted of 

AMF spores, external hyphae and colonized root 

pieces. The control roots were non-mycorrhizal. 

Inoculation experiment 

Lablab bean (var. Rongai) was planted on 3 November 

2002, in pots amended with either 200 mL control 

inoculum or 200 mL of AMF inoculum, and 

grown for 7 weeks and watered as needed. At harvest, 

nodule number and weight were recorded, 

AM colonization determined and shoot Nand P 

concentration measured (Brundrett et ai., 1996; Watanabe 

and Olsen, 1965; Jensen 1962). Data were 

analyzed using a one tailed paired t-test where 

+AMF and control pairs shared the site origin from 

the field. 

Results 

Experiment 1 

Enhancing AMF in the test luvisol significantly increased 

AM colonization compared to the control 

(Table 1). This resulted in a doubling in nodule 

numbers and a significantly higher N content in the 

shoot. There was a strong beneficial effect of P on 

the number of nodules, which resulted in almost a 

doubling in N content in the shoot. However, inter- 

44 


estingly, P was more efficient than a fungicide in 

lowering the AM colonization. Fungicide applica 

tions did not differ significantly from the control in 

any of the variables measured. 

Experiment 2 

Enhancing the indigenous AMF abundance resulted 

in a significantly higher AM colonization, nodule 

mass and N concentration in the shoot (Table 2). 

There were no significant differences between con 

trol and +AMF in shoot weight and shoot P concen 

tration. 

Discussion 

Many factors need to be considered when trying to 

optimize the soil fertility benefits of legumes. Differences 

in residue quantity and quality among legumes 

grown under various conditions need to be 

documented, and factors limiting legume performance 

need to be established. We have shown here 

by' exploring the biology of the agro-ecosystem that 

beneficial effects could be obtained by optimizing 

the mutualistic interactions between the plant, bacteria 

and fungi . Nodule number and mass in 

groundnut and lab lab were significantly enhanced 

by a higher abundance of AMF when the legumes 

were grown in a low P luvisol. This resulted in 

more N in the shoot tissue, a key element for optimizing 

maize production. 

Beneficial effects of AMF on nodulation have been 

documented before (Goss and de Varennes, 2002; 

Ahiabor and Hirata, 1994; Reddy and Bagyraj, 

1991) and have often been linked to the increased P 

uptake provided by the fungi. However, even 

though AM colonization levels were higher in the 

Table 1. Effect of P, enhanced AMF and a fungicide on groundnut 

grown in aluvisol soil collected from a subsistence farmer's field in 

Tshlotshlo. Different superscripts indicate a significant (p < 0.05) 

difference between means, n= 5. 

Treatment Shoot OW Nodule AM Pcontent Ncontent 

(g) (per plant) (%) (mg shoot 1 ) (mg shoot 1 ) 

Control 1.0 be 75 ' 25 b 1.1 b 27 ' 

AMF 1.201> 144 b 57 ' 1.9 b 36 b 

Fungicide OJ' 43 ' 13 be 0.9 b 20 ' 

Phosphorus 1.3' 290 ' 1 ' 6.9 • 46 ' 

Table 2. Effect of enhanced indigenous AMF on lablab bean. 

Control consisted of non·sterile soil collected from a subsistence 

farmer's field in Tsholotsho. Different superscripts indicate a 

significant (p < 0.05) difference between means, n~ 1O. 

I 

Variable Control + AMF 

(non·sterile soil) (non·sterile soil + AMF) 

I Shoot OW (g) 3.3' 3.6' 

Nodule mass (mg plant 1 ) 4S.6 b 202.1' 

Shoot Nconcentration ('Yo) 1.Sb 2.2' 

Shoot Pconcentration (%) 0.11' 0.11' 

AM colonization (0/,) S4.0 b 74.5' 


+AMF treatment in our experiments, shoot P levels 

were not significantly increased. If the AMF effect is 

P mediated, should this not be reflected in an increased 

shoot P content? One possible answer to 

this is that the amount of P needed for optimal 

nodulation is an order of magnitude lower than for 

optimal growth of plants (Gates and Wilson, 1974). 

Since available P in the luvisol soil in this experiment 

was very low, it is possible that the small 

amount of additional P taken up in the +AMF treatments 

remained in the roots and promoted nodulation. 

In that case, differences would not be detected 

in the shoot tissue. Analyses of nodule P contents in 

lablab are currently underway to address this issue. 

Further, unlike many previous experiments, the 

control plants in our experiments were mycorrhizal, 

so differences between +AMF and control plants 

were likely to be smaller than what is normally presented 

when the control plants are non-mycorrhizal. 

We are not proposing large-scale inoculation projects 

as an outcome of these results. Rather, the 

abundance of indigenous fungi should be increased. 

In temperate agro-ecosystems, it has been shown 

that fungal abundance is affected by tillage and fallow 

practices (Boswell et al. 1998; Kabir et a!., 1997; 

Douds et al. 1995; Harinikumar and Bagyraj, 1989), 

but little is currently known about the effects of 

common management practices by subsistence 

farmers in the semi-arid tropiCS. Based on this, 

AMF responses to tillage timing and fallow period 

are currently being investigated on subsistence 

farmers' fields. It is important to remember though, 

that even if AMF abundance is increased through a 

change in current management practices and beneficial 

effects on legume performance observed, it is 

not a sustainable substitute for P fertilizers. Whatever 

P is brought from the soil in harvestable products 

or animal feed need to be replenished. Nevertheless 

AMF can contribute to a better utilization of 

the P applied to the system, thereby reducing the 

amounts and frequency of P application in a sustainable 

agro-ecosystem. 


This project could not have been conducted had it 

not been for our funding sources, the National Geographic 

Society and the Root Biology Program of 

the Pennsylvania State University, which is funded 

by the US National Science Foundation. Our special 

thanks go to the collaborating farmers in Tsholotsho, 

Gwanda and Masvingo. 

References 

Ahiabor BD, Hirata H. 1994. Characteristic responses 

of three legumes to the inoculation of 

two species of V AM fungi in Andosol soils with 

different fertilities. Mycorrhiza 5(1):63-70. 

45

Boswell EP, Koide RT, Shumway DL, Addy HD. 

1998. Winter wheat cover cropping, VA mycorrhizal 

fungi and maize growth and yield. Agriclllture, 

Ecosystems and Environment 67:55-65. 

Brundrett M, Bougher N, Dell B, Grove T, 

Malajczuk N. 1996. Working with mycorrhizas 

in forestry and agriculture. ACIAR Monograph 

32, Canberra ACT, Australia. 374 pp. 

Buresh RJ, Sanchez PA, Calhoun F, (Eds). 1997. Replenishing 

Soil Fertility in Africa. SSSA Sepcial 

Publication No. 51. SSSA, Madison, WI., USA. 

Douds DO, Galvez L, Janke RR, Wagoner P. 1995. 

Effects of tillage and farming system upon populations 

and distributions of vesicular-arbuscular 

mycorrhizal fungi. Agricllltllre Ecosystems and Environment 

52:111-118. 

Fitter AH, Garbaye J. 1994. Interactions between 

mycorrhizal fungi and other soil organisms. 

Plant and Soil 159:123-132. 

Gates CT, Wilson JR. 1974. The interaction of nitrogen 

and phosphorus on the growth, nutrient 

status and nodulation of Stylosanthes humilis H.B. 

K. (Townsville Stylo). Plant and Soil 41:325-333. 

Giller K. 2001. Nitrogen fixation in tropical cropping 

systems. 2nd Edition. CABl Publishing, Wallingford, 

UK. 423 pp. 

Goss MJ, de Varennes A. 2002. Soil disturbance reduces 

the efficacy of mycorrhizal associations for 

early soybean growth and N-2 fixation. Soil Biology 

and Biochemistnj 34:1167-1173. 

Grant PM. 1967. The fertility of sandveld soil under 

continuous cultivation. Part 1. The effect of manure 

and nitrogen fertilizer on the nitrogen 

status of soil. Rhodesia Zambia Malawi Journal of 

Agricultural Research 5: 71-79. 

Grant PM. 1981. The fertilization of sandy soils in 

peasant agriculture in Zimbabwe. Zimbabwe Agricultural 

Journal 78:169-175. 

Grant PM. 1995. Soil fertility and organic matter 

management. In: Twomlow S, Ellis-Jones J, Hagmann 

J, Loos H (Eds) Soil and water conservation 

for smallholder farmers in semi-arid Zimbabwe 

- transfers between research and extension. 

Proceedings from a National Technical 

Workshop, 3-7 April 1995 in Masvingo, Zimbabwe. 

Silsoe Research Institute, Report 

00/95/16.164-171. 

Harinikumar KM, Bagyraj OJ. 1989. Effect of cropping 

sequence, fertilizer and farmyard maure on 

vesicular-arbuscular mycorrhizal fungi in different 

crops over three consecutive seasons. Biology 

and Fertility of Soils 7: 173-175. 

Howard H, 1941. Soil and Health. Reprinted by 

Small Farmer's Journal in 1999, vol 22. 

Jensen WA (1962) Botanical histochemistry. Freeman, 

San Francisco, CA., USA. 

Koide RT. 1991. Nutrient supply, nutrient demand 

plant response to mycorrhizal infection. New 

Phytologist 117: 365-386. 

Kabir Z, O'Halloran IP, Fyles JW, Hamel C. 1997. 

Seasonal changes of arbuscular mycorrhizal 

fungi as affected by tillage practices and fertilization: 

Hyphal density and mycorrhizal root colonization. 


Map;umo P, and Giller K. 2001. Soil fertility management 

strategies and practices by smallholder 

farmers in semi-arid areas of Zimbabwe. Bulawayo, 

Zimbabwe: International Crops Research 

Institute for the Semi-Arid Tropics. 60 pp. 

Moyo M. 2001. Representative soil profiles of ICRl 

SAT research sites. Chemistry and Soil Research 

Institute, Soils Report No. A666. pp 97. 

Muza L, Pashapa L, Feresu S. 1998. Need to revive 

green manuring in soil fertility management in 

Zimbabwe. Zimbabwe Science News 32:51-57. 

Reddy BMS, Bagyaraj OJ. 1991. The symbiotic efficiency 

of pigeonpea to VA mycorrhizal inoculation 

in an alfisol and vertisol. Biological Agriculture 

and Horticulture 8:177-182. 

Sanginga N, Carsky RJ, Dashiell K. 1999. Arbuscular 

mycorrhizal fungi respond to rhizobial inoculation 

and cropping systems in farmers' fields in 

the Guinea savanna. Biology and Fertility of Soils 

30:179-188. 

Snapp SS.1998. Soil Nutrient Status of Smallholder 

Farms in Malawi. Commun. Soil Sci. Plant Anal. 

29:2171-2588. 

Snapp SS, Rohrbach DO, Simtowe F, Freeman liA. 

2002. Sustainable soil management options for 

Malawi: can smallholder farmers grow more legumes? 

Agriculture Ecosystems and Environment 

91:159-174. 

Warren GP. 1992. Fertilizer Phosphorus: sorption 

and residual value in tropical African soils. NRl 

Bulletin 37, Chatham, UK: Natural Resources 

Institute. 91 pp. 

Watanabe FS, Olsen SR 1965. Test of an ascorbic 

acid method for determining phosphorus in water 

and NaHC03 extracts from soil. Proceedings 

Soil Science Society of America 26:677-678. 

46 


SOYABEAN YIELD RESPONSE TO DIFFERENT RHIZOBIAL INOCULATION 

RATES ON SELECTED SANDY SOILS IN ZIMBABWE 

NGONI CHIRINDA 1 ', S. MPEPEREKI', R. ZENGEI\J1 1 and K.E. GILLER 2 

1 Department of Soil Science and Agricultural Engineering, University of Zimbabwe 

Box MP 167 Mt Pleasant, Harare, Zimbabwe 

2 Plant Production systems, Department of Plant Sciences, Wageningen University, 

P. O. Box 430 6700 AK Wageningen, The Netherlands 

*Corresponding author email: (ntchirinda@yahoo.com) 

Abstract 

The recommended rhizobial inoculation rate for soyabean (Glycine max (L) seed in Zimbabwe of 80 g (10 8 cells g-l) 

inoculant for 100 kg seed has generally resulted in unreliable nodulation on the sandy soils that predominate in most 

smallholder areas. In this study, it was hypothesized that a higher seed inoculation rate would improve soyabean grain 

yield hI the smallholder sector. The trial was set up during the 2001/2002 season in two districts, Guruve and Goromonzi, 

to determine appropriate inoculation rates for maximum soyabean grain yield on sandy soils of the smallholder 

sector in Zimbabwe. Two soya bean varieties, Solitaire and Storm, were inoculated with a commercial rhizobial inoculant 

containing Bradyrhizhobium japonicum (strain MAR 1491) at the following rates: 80 g inoculant per 100 kg ofseed 

(the recommended rate), and 2, 3, 5 and 10 times that recommended rate. Seeds were sown at the rate of 100 kg ha- J in 5 

m x 3.6 m field plots arranged in a randomized block design, with each inoculation rate replicated four times. Eight 

weeks after planting, whole plant samples were dug up from the guard rows of each plot and checked for nodulation. 

Plants were harvested from 3 m x 3 m net plots at maturity to determine seed and stover yields. Results showed a significant 

increase in grain and stover yields as well as primary root nodulation with increased inoculation rates for both 

varieties (P


The trial was set up during the 2001/2002 cropping 

season in Goromonzi District of agro-ecological region 

(AER) 2 and Guruve District of AER 3. Average 

rainfall in AER 2 (70S-1000 mm) and 3 (6S0-800 

mm) is suited for soyabean production. Two weeks 

before planting, soils were sampled (from 0-30 em 

depth) in Guruve (Mrs. Kanonama's field) and 

Goromonzi (Mr. Majuru's field). Both fields were 

last planted with soyabean in the 1998/1999 cropping 

season. Maize (Zea mays) was grown at both 

sites during the 2000/2001 season. Cotton 

(Gossypium hirsutum) and common beans (Phaseolus 

vulgaris (L) .) were grown at Majuru and Kanonama 

respectively during the 1999/2000 season. Soil texture 

was determined using the hydrometer method 

and organic matter content (%C) by the Walkley 

Black method (Nelson and Sommers, 1996). Total 

soil nitrogen was estimated using the Kjeldahl 

method (Bremner, 1996). Exchangeable bases were 

determined using ammonium acetate as the extracting 

agent (Summer and Miller, 1996). The most 

probable number (MPN) technique (Vincent, 1970) 

was used to quantify indigenous rhizobial populations 

in the soils. The density of Bradyrhizobium japonicum 

(strain MAR 1491) in the commercial inoculants 

was estimated by plate counts on Yeast Extract 

Mannitol agar. 

Five inoculation rates (the recommended inoculation 

rate (0.8 g kg-I seed), 2, 3, S and 10 times that 

rate) were tested in this experiment using two soyabean 

varieties, Storm (determinate) and Solitaire 

(indeterminate), that nodulate with specific rhizobial 

strains. Seeds were sown at 100 kg seed per hectare 

in S m x 3.6 m plots that were arranged in a randomized 

complete block design with four replicates. 

Basal fertilizer, Compound L (S% N, 18% P20S, 

10% K20 and 0.2S% Boron) and lime were applied at 

rates of lS0 kg ha-I and SOO kg ha-I respectively. The 

crop was weeded at 2 and 6 weeks after sowing. 

The uninoculated plots were weeded first, then 

other plots were weeded, taking precautions to 

avoid cross-contamination by wiping feet, hands 

and hoes with commercial methylated spirit (10% 

methanol) before weeding a different plot. Whole 

plant samples (12) from the guard rows of each plot 

were dug up and checked for nodulation 8 weeks 

after sowing. The number of nodules, nodule colour 

and nodule dry mass (70OC for 24 h) were deter~ 

mined. Plants were harvested at ma 

turity from 3 m x 3 m net plots. Pod 

number, grain and stover yield, and 

expected value (EV) concept (Singleton et al., 1992) 

was used to express yield increases in monetary 

terms. [EV = Y (increase in yield resulting from inoculation) 

x Pr (price of the crop) x P (the probability 

of obtaining a yield increase under the defined 

conditions)]. 

Results 

Rainfall amount during the 200l/2002-season rainfall 

was low at both the Kanonama (416 mm) and 

Majuru (307 mm) site. Rainfall distribution was 

poorer at Majuru. Soils at both sites were sandy 

«7% clay) and acidic (pH

A 

B 

7 10 

6 

• 

8 

....... , 5 "7 

C C 6 

ell ell 

4 

0... 0... 

Vl 

0 

Vl 3 4 

0.9 

0.8 

~ 0.7 Rec 

'-' 0.6 

Vl 

::9 0.5 

esponse is site specific (Singleton and Tavares, 

1986). The nodulating pattern of the two varieties 

tends to suggest that variety Solitaire is less specific 

than variety Storm as it had a relatively high nodule 

count in the absence of inoculation. Because Solitaire 

has an indeterminate growth habit, like the 

popular promiscuous variety Magoye, it could be 

that the indeterminate growth habit is related to 

promiscuity. At the sites used in this study, the indigenous 

rhizobia population was ineffective in fixing 

N as evidenced by the poor grain yield in the 

uninoculated control. However, in the presence of 

effective indigenous rhizobial populations, Solitaire 

could be grown without inoculation. 

Wadisirisuk and Weaver (1985) reported that nodule 

OM is related to N-fixation capacity. Nodules 

formed on Storm increased with inoculation rate 

and had a higher OM than those on Solitaire. This 

could have partly contributed to increased nitrogen 

fixation and grain yield for Storm. 

Pod count was significantly related to grain yields 

(P=0.03) and increased with inoculation rates. This 

result is in agreement with Jayapaul and Ganesaraja 

(1990) who stated that an increase in plant nitrogen 

increases pod count. The seed weight of 100 seeds 

did not differ for the different inoculation rates, possibly 

due to the poor rainfall that could have affected 

seed setting. 

In the SH sector, pieces of land allocated to soyabean 

are small (0.1 ha), with seed requirements of 

about 10 kg. Available i~oculant sachet sizes (80 g) 

result in farmers inoculating their soyabean at rates 

almost ten times higher than recommended. This 

could be the reason why dramatic yields in response 

to inoculation were reported in the first 

phase of soya bean promotion in Zimbabwe 

(Mpepereki et al., 2002). Results from this study 

suggest that the recent introductions of inoculant 

sachets allowing inoculation of small quantities of 

seed at the recommended rate will result in a yield 

decrease. 

As farmers become more confident in growing soyabeans 

and realizing they can make profits, they are 

increasing its area planted. Of the 5 million ha 

opened for resettlement, about30% are virgin sandy 

soils. If 150 VOO ha (10%) of.the sandy soils are used 

for soya bean production, a five times increase in the 

rate of inoculation will increase the demand for inoculants 

above what the factory can supply. The inoculant 

factory in Zimbabwe currently produces 

120000 sachets in a year which are for the 75 000 ha 

of soya bean grown nati0nwide, 10 000 ha of which 

was in the SH sector. 

Conclusions 

This study has shown that the recommended inoculation 

rate of 0.8 g inoculant kg· l seed is insufficient 

for maximum nodulation, soyabean seed yields, 

seed nitrogen and stover yields on sandy soils in 

Zimbabwe. A rate of 8 g kg· l seed, currently being 

used by farmers is uneconomic for soyabean production 

on sandy soils. Generally, the inoculation 

rates of 4 g kg-l seed were observed to result in 

maximum seed yield. 

Recommendations 

Judging from its current capacity (120 000 sachets 

per year), the inoculant factory in Zimbabwe would 

not meet the increased inoculant demand if the inoculation 

rates were to be increased five times. Assuring 

a higher seed inoculation rate by increasing 

the number of viable cells per sachet could be an 

option. While this could increase the number of cells 

in an inoculant sachet, intensifying competition for 

available nutrients, increasing cell mortality, there is 

need for further research to ascertain this. Use of 

granular inoculants could also be considered for the 

harsh SH cropping environments. Improving the 

soil environment for better rhizobia survival by reducing 

soil acidity and increasing soil organic matter 

could enhance rhizobial survival in the soil, 

eliminating the need for high inoculation rates. This 

could allow farmers to benefit from high yields at 

reduced inoculation rates. 

References 

Andrade, OS. and Hungria, M. 2002. Maximizing 

the contribution of biological nitrogen fixation in 

tropical legume crops. In: Finan, T.M., O'Brian, 

M.R., Layzell, O.B., Vessey, J.K. and Newton, W. 

(eds) . Nitrogen Fixation Global Perspectives. Proceedings 

of the 13 lh International Congress on 

Nitrogen Fixation. Hamilton, Canada 2-7 July 

2001. CABI publishing, Wallingford, UK. pp. 

341-345. 

Beck, O.P and Munns, O.N. 1984. Phosphate nutrition 

of Rhizobium spp. Applied and Environmental 

Microbiology 47:278-282. 

Boonkerd, N. and Weaver, R.W. 1982. Survival of 

cowpea rhizobia in soils as affected by soil temperature 

and moisture. Applied Environmental 

Microbiology 43:585-589 . 

Bremmer, J.M. 1996. Nitrogen-Total. In: Methods of 

Soil Analysis. Part 3. chemical methods. SSSA Book 


51

Series no 5. Soil Science SOciety of America and 

American Society of Agronomy, Madison, Wisconsin, 

USA. pp. 1085-1121. 

Davis, P.E. 1994. Seed inoculation, are we recommending 

enough? In: Harnessing Biological Nitrogen 

Fixation in African Agriculture. Challenges and 

Opportunities. Mpepereki, Sand F. Makonese 

(eds). University of Zimbabwe Publications. 

Harare, Zimbabwe. pp. 165-172. 

De Mallaro, M.s. and Izaguirre, M.L. 1994. Seasoflal 

dynamics, host range and symbiotic efficiencyof 

native rhizobial populations in three soil horizons 

of four contrasting savanna sites. Symbiosis 

4:99-105. 

Dudeja, S.s. and Khurana, A.L. 1989. Persistence of 

Bmdyrhizobium sp. (Cajanus) in a sandy loam. Soil 

Biology and Biochemistry 21:709-713. 

FAa. 1991 Expert Consultation on Legume Inoculant 

Production and Quality Control. 19-21 

March 1991. (ed.) J.A. Thompson FAa, Rome. 

Gardner, F. P., Pearce, R.B. and Mitchell, R.L. 1985. 

Physiology of Crop Plants. Iowa State University 

Press, USA. 

Giller K.E. 2001. Nitrogen Fixation in Tropical Cropping 

Systems 2nd Edition, CABI publishing, Wallingford, 

UK. 423 pp. 

Jayapaul, P. and Ganesaraja, V. 1990. Studies on response 

of soybean varieties to Nand P. Indian J. 

Agron. 35 (3):329-330. 

Mpepereki, S. and Makonese, F. 1998. Seasonal 

rhizobial population fluctuations under field 

conditions in Zimbabwean soils. In: Harnessing 

Biological Nitrogen Fixation in African agriculture. 

Challenges and Opportunities. Mpepereki, S and F. 

Makonese (eds). University of Zimbabwe Publications, 

Harare, Zimbabwe. pp. 109-115. 

Mpepereki, S., Makonese, F. and Giller, K.E., 2002. 

Soyabean N2 fixation and food security for 

smallholder farmers: A research-extension 

model for sub-Saharan Africa. h1: Finan, T.M., 

O'Brian, M.R., Layzell, D.B., Vessey, J.K. and 

Newton, W. (eds). Nitrogen Fixation Global Perspectives. 

Proceedings of the 13 lh International 

congress on Nitrogen Fixation. Hamilton, Canada 

2-7 July 2001. CABI publishing, Wallingford, 

UK. pp. 346-351. 

Nelson, D.W and Sommers, L.E. 1996. Total carbon, 

organic carbon, and organic matter. In: Methods 

of soil analysis. Part 3. chemical methods-SSSA 

Book Series no 5 Soil Science Society of American 

Society of Agronomy, Madison, Wisconsin, 

USA. pp. 961-1010. 

O'Hara, G.W. 2001. Nutritional constraints on root 

nodule bacteria affecting symbiotic nitrogen fixation. 

Australian Journal of Experimental Agriculture 

41. 

Sanginga, N., Abaidoo, R., Deshiell, K., Carsky, R.J. 

and Okogun, A. 1996. Persistence and effectiveness 

of rhizobia nodulating promiscuous soyabeans 

in moist savanna zones of Nigeria. Applied 

Soil Ecology 3:215-224. 

Singleton, P.W. and Tavares, J. W., 1986. Inoculation 

response of legumes in relation to the number 

and effectiveness of indigenous rhizobium 

populations. Applied Environmental Microbiology. 

51:1013-1018. 

Summer. M.E. and W.P. Miller. 1996. Cation exchange 

capacity and exchange coefficients. In: 

Methods of Soil Analysis. Part 3. Chemical Methods-SSSA 

Book Series no 5. Soil Science Society 

of America and American Society of Agronomy, 

Madison, Wisconsin, USA. pp. 1201-1229. 

Thompson J.G. and W.O. Purves 1981. A guide to 

the soils of Zimbabwe. Zimbabwe Agricultural 

Journal, Technical Handbook No.3. Chemistry 

and Soil Research Institute. Department of Research 

and Specialist Services. Harare, Zimbabwe. 

43 pp. 

Vincent, J.M. 1970: A Manual for the Practical Study 

of Root Nodule Bacteria. Blackwell Scientific 

Publications Ltd., Oxford, UK. 

Wadisirisuk, P. and R.W. Weaver. 1985. Importance 

of bateroid number in nodules and effective nodule 

mass to dinitrogen fixation by cowpeas. Plant 

and Soil 87:223-231. 

Watkin, E.L.J., O'Hara, G.W. and Glenn, A.R. 1997. 

Calcium and acid stress interact to affect the 

growth of Rhizobium leguminosarum biovar trifoli. 

Soil Biology and Biochemistry 29:1427-1432. 

Zapata F., A.s.K Danso, G. Hardarson and Fried, M. 

1987. Time course of nitrogen fixation in field 

grown soyabeans using 15N methodology. 

Agronomy Journal 79: 172-176. 

52 


SURVIVAL AND PERSISTENCE OF INTRODUCED COMIVIERCIAL 

RHIZOBIAL INOCULANT STRAINS IN SELECTED SMALLHOLDER FIELD 

ENVIRONMENTS OF ZIMBABWE 

Abstract 

REBECCA ZENGENI', SHEUNESU MPEPEREKI' and KEN E. GILLER 2 

1 Soil Science Department, University of Zimbabwe, P. O. Box MP 167, 

Mount Pleasant, Harare, Zimbabwe 

2 Department of Plant Sciences, Wageningen University, P. O. Box 430, 

6700 AK Wageningen, The Netherlands 

The persistence of an introduced rhizobial inoculant strain in smallholder field environments of Zimbabwe was 

determined at two sites with no history of sdyabean production in Goromonzi district. An inoculated soyabean crop was 

introduced in the first season using Magoye, Solitaire and Viking soyabean varieties. There was a positive response to 

inoculation in the first season (2000/2001) with inoculated plants having higher nodule numbers, grain yields and total 

nitrogen contents than the uninoculated plants (p < 0.05). Re-inoculation of previol/sly inoculated plots in the following 

season (2001/2002) however did not result in an increase in nodulation or yields of soyabean indicating that the 

rhizobial strain introduced in the first season persisted into the second season. In a separate experiment to determine the 

persistence of an inoculant strain over many seasons, soils that had been last inoculated in the years 1996, 1998, 1999 

and 2000 were sampled from three districts and assessed for rhizobia I population sizes in the greenhouse. Rhizobial 

numbers were correlated with soil properties and the occupying strains identified. 

Results indicated that rhizobial numbers were positively correlated with soil pH, clay percent and organic carbon, with 

numbers significantly increasing at pHs above 5.5. Rhizobial numbers decreased with year since last inoculation, with 

populations as high as 10 3 cells /g of soil being obtained in fields last inoculated in the year 2000 and less thnn 30 cells /g 

of soil in fields last inoculated in 1996. The rhizobial strain MAR 1491 (USDA 110) was obtained in most fields with a 

history of rhizobial inoculation. 

Key words: Rhizobia, persistence, soyabean, inoculation 

Introduction 

Soyabean, a crop previously restricted to the.largescale 

farming sector of Zimbabwe, has been 

promoted to smallholder farmers through first the 

Soyabean Promotion Programme from 1986-89 then 

more recently through the Soyabean Promotion 

Task Force from 1996 to the present (Rusike et al., 

2000). Soyabean production requires the use of 

rhizobial inoculants to attain optimum nodulation 

and high yields. Commercial rhizobial inoculants 

are however not readily available for use by 

smallholder farmers. This is because marketing 

channels for rhizobial inoculants to smallholders 

were not sufficiently developed to match their 

demand. 

A solution to the problem of limited inoculant 

availabilily could be the use of promiscuous 

soyabean varieties such as 'Magoye' that readily 

nodulate with indigenous rhizobia and hence do 

not require rhizobiaI inoculation (Mpepereki et al., 

1999). These varieties are however not available on 

the local market, they produce lower grain yields 

and their pods readily shatter compared to specific 

soyabean varieties (Kasasa, 1999). The locally 

available commercial specific soyabean varieties on 

the other hand require rhizobial inoculation to 

obtain high yields. Since the use of rhizobia I 

inoculants in smallholder field environments is a 

relatively new technology, little information is 

availaole on the survival and persistence of an 

introduced inoculant strain in field soils. Persistence 

of an inoculant strain would obviate the need for 

inoculation each time soyabean is cropped thereby 

allowing for sustained productivity. The objective of 

this study was therefore to assess survival and 

persistence of introduced rhizobiaI inoculant strains 

in field soils. It was hypothesized that inoculant 

strains survive and persist poorly in smallholder 

field environments due to unfavourable soil 

condi.tions of low soil pH, poor clay and low 

organic matter amounts. 



Persistence of an introduced rhizobial inoculant 

strain over one season was assessed by setting up 

an experiment on two field sites in Goromonzi 

district. The first site Mudzivare, is a sandy soil with 

a low soil pH of 5 and poor nutrient status. The 

second site, Majuru, has a contrasting clay soil with 

a slightly higher pH of 5.8 and favourable nutrient 

amounts. Three soyabean varieties, namely Magoye 

(promiscuous variety), Solitaire and Viking (specific 

varieties) were planted on 5 x 5 m plots set up in a 

completely randomized design replicated four 

times. Treatments included inoculated and 

uninoculated field plots in the first season 

(2000/2001). In the second season (2001 / 2002), half 

of the previously inoculated plot was reinoculated 

while the other half remained uninoculated. The 

control plots that were not inoculated in the first 

season remained uninoculated in the second season. 

Data on nodulation was collected at eight weeks 

after planting (W AP) while grain and total dry 

matter yield in the different inoculation treatments 

was assessed at physiological maturity. An analysis 

of variance of treatment means was determined 

using GENST AT. 

Persistence of an inoculant strain over many 

seasons was assessed under greenhouse conditions. 

Soil samples with a previous history of rhizobial 

inoculation were collected up to a 20 cm depth in 

October 2001 from Guruve district. The fields had 

las t been inocu la ted in the 1996, 1998, 1999 and 2000 

growing seasons. Populations of rhizobia in each 

soil were quantified using the most probable 

number method with the variety Solitaire as the 

trap host. Serial dilutions of each soil were done 

using a base dilution of 10. Three plants were 

planted per pot and inoculated with 1ml of soil 

inoculum. Plants were scored for nodulation at 6 

WAP. Data on nodulation, inoculation volume and 

number of replicates used was fed into the MPNES 

computer programme and populations of rhizobia 

in the different soils estimated. The rhizobial strain 

occupying nodule sites was identified with the 

enzyme-linked immuno sorbent assay. Population 

sizes of rhizobia were then compared with soil 

properties using regression analysis. 

Results 

Number of nodules in different inoculation 

treatments at 8 WAP for three soyabean varieties. 

Although inoculation of the specific varieties 

Solitaire and Viking resulted in increased nodule 

numbers, yields and total N amounts in the first 

season, reinoculation of the same varieties in the 

second season did not result in increased 

nodulation at both sites (Table 1). No significant 

differences in nodule numbers in the different 

inoculation treatments were recorded for the 

promiscuous variety Magoye. More nodules were 

also obtained from the clayey Majuru site than from 

the sandy Mudzivare (p < 0.001). 

Grain _and total dry matter yield in different 

inocuIation treatments 

Re-inoculation of the different soyabean varieties in 

the second season also did not result in increased 

grain and total dry matter yields at both sites 

(Figure 1). An exception was Viking, which gave a 

positive response to re-inoculation for total dry 

matter yield at Mudzivare. The grain yield obtained 

at this site was however very low when compared 

with the total dry matter yield. 

Changes in rhizobial populations in inoculated 

fields over time 

An assessment of persistence of rhizobia over many 

seasons showed that the rhizobial strain used in 

inoculant production persists in smallholder fields 

since a good rhizobial count was obtained from 

fields last inoculated as far back as 1998 (Table 2). 

Rhizobial populations were positively correlated 

with soil pH, clay and organic carbon contents. The 

year since last inoculation also strongly influenced 

rhizobial numbers, with numbers decreasing with 

increasing year since last inoculation (p < 0.001). As 

many as 10 3 rhizobial cells / g of soil were obtained 

in fields last inoculated in the year 2000 while less 

than 30 cells / g of soil were found in fields last 

inoculated in 1996. The strain MAR 1491 was 

detected in most previously inoculated fields and in 

one instance both MAR 1491 and 1495 were 

Table 1. Number of nodules at eight WAP in different inoculation 

treatments at two sites in Zimbabwe 

Variety Majuru Mudzivare 

-inoc inoc SI inoc S2 -inoc inoc Sl inoc S2 

Magoye 7 7 8 3 1 2 

Solitaire 2 14 7 0 2 

Viking 3 8 5 0 

sed 0.639 

inoc - no inoculation 

Inoc 51 - inoculation in season 1 only 

inoc 51 - inoculation in both seasons 

sed - standard error of differences of means 

Table 2. Changes in rhizobial population in inoculated fields from 

Guruve district over time 

....................................... - ........... ........ 

Site Year last Rhizobia cellsl Soil pH % %C rhizobia I 

No. inoculated gsoil x 10 3 in water Clay strain 

1491 1495 

1 2000 4.5 5.9 32 2.3 + 

2 1999 3.0 6.1 29 0.9 + + 

3 1998 0.44 6 36 0.79 + 

4 1996 0.029 6.6 24 1.3 + 

5 Control 0.01 5.5 23 0.8 

54 


1.5 1.5 ~-----------, 

Grain yield In a clayey soil (Majuru) 

Grain yield in a sandy soil (Mudzivare) 

Ii 

S -1.0 1.0 

"C 

-a; 

'>' 

c 

~ 0.5 0.5 I 

None - uninoculated plots 

Season 1 - inoculated in first 

season only 

Both seasons - inoculated in 

both seasons 

0.0 o.0 ...L.L.J::+,ll!a.....----L-l4"~-.........+""""---_1 

5 5 -r------------, 

Dry matter in sandy soil (Mudzivare) 

4 4 

Ii 

oE 

:::.. 

... 

3 3 

~ 

E "' 2 2 

... > 

a 

I 

o 

o 

Magoye Solitaire Viking Magoye Solitaire Viking 

Variety 

Variety 

Figure 1. Grain and total dry rr.atter yield at Majuru and Mudzivare in the second season 

identified. Fields with no history of rhizobial 

inoculation (controls) had none of these strains. 

Discussion 

Although soyabean responded to inoculation in the 

first season of cropping through increased nodule 

number and yields, re-inoculation in the second 

season did not result in a similar increase. This 

indicates that rhizobial strains introduced in the 

first season persisted into the second season. 

Mapfumo (2000) noted that legumes often 

responded to inoculation during the first year of 

introduction into new areas but not in subsequent 

years on the same piece of land. The initially low 

population of indigenous rhizobia may necessitate 

the use of commercial inoculants but as high 

populations of effective rhizobia build up in the 

soil, the need for inoculation may be obviated in 

subsequent years. 

The response to rhizobial inoculation through 

increased nodule numbers and yield for the specific 

varieties Solitaire and Viking and not the 

promiscuous Magoye correspond with results 

obtained by Kasasa (1999) where significantly 

higher nodule numbers were obtained after 

inoculating specific soyabean varieties. Studies by 

Mpepereki et al. (1999) revealed that promiscuously 

nodulating soya bean varieties such as Hernon 147 

and Magoye nodulate and fix nitrogen well in fields 

with no history of rhizobial inoculation, hence no 

significant changes were observed after inoculating 

Magoye in either season. More nodules at the clayey 

Majuru site than the sandier Mudzivare can be 

explained by the observation that a high clay soil 

gives rise to many small nodules because of a better 

moisture retention capacity while sandier soils 

result in larger but fewer nodules due to their poor 

water retention capacity (Mapfumo, 2000). The very 

low amount of grain produced at Mudzivare in 

comparison with its total dry matter yield is a result 

of mid season dry spells experienced at flowering 

resulting in reduced grain production. 

Results from the greenhouse experiment showed 

that rhizobial strains persist in smallholder fields 

and that inoculation history and pH strongly 

influence the populations. This trend is consistent 

with work covered by Mpepereki and Makonese 

(1995) where soyabean rhizobia were not detected 

in fields with no history of legume cultivation. They 

similarly observed that cowpea rhizobia were 

lowest in communal areas that were generally 

acidic, sandy and with low nutrients. Acidic soils of 

pHs below 5 are known to be detrimental to 

rhizobial survival and are unfavourable for 

soyabean-rhizobia symbiosis (Tattersfield, 1996). In 

this experiment, a rise in soil clay and carbon 


55

content resulted in increased rhizobial populations. 

Chatel and Parker (1972) noted that heavy textured 

soils formed micro aggregates, which afforded some 

protection to rhizobia against high temperatures. 

Rhizobia also survive saprophytically in the absence 

of a legume host, therefore organic carbon is an 

essential source of energy required for their growth 

and survival. The rhizobial strain MAR 1491 was 

found in most inoculated field soils since it is the 

most widely used strain in soyabean inoculant 

production in Zimbabwe. In some soils, the strain 

MAR 1495 was also found because it was once used 

in combination with MAR 1491 during inoculant 

production. The uninoculated field controls had 

none of the tested strains because they have no 

history of rhizobial inoculation. 

Conclusion and Recommendations 

Rhizobial inoculant strains survive in smallholder 

field environments for at least three seasons, so reinoculation 

of previously inoculated fields is not 

beneficial to the farmer. Inoculation when 

introducing a soyabean crop in a new area for the 

first time is essential. Thereafter, a second soyabean 

crop grown after rotation with a cereal does not 

require inoculation since a significant population of 

rhizobia will still be present in the soil. Rhizobial 

survival can be further enhanced by raising soil pH 

and organic matter content and in soils with high 

clay amount. Re-inoculation may therefore need to 

be more frequent in sandy soils. 

Acknowledgements 

The Rockefeller Foundation for funding the project. 

Also the farmers, project assistants and the Soil 

Science Department of the University of Zimbabwe. 

References 

Chatel, D.L and Parker, c.A. 1972. Survival of field 

grown rhizobia over the dry summer period in 

western Australia. Soil Biol. Biochem. 5:415-423. 

Kasasa P., 1999. Biological nitrogen fixation by 

promiscuous nodulating soyabean Glycine max 

[L] Merr) varieties in communal soils of 

Zimbabwe. MPhii Thesis. University of 

Zimbabwe, Harare, Zimbabwe, pp 37-41. 

Mapfumo, P. 2000. Potential contributions of 

legumes to soil fertility management in 

sIPallholder systems of Zimbabwe: the case of 

pigeon pea (Cajam!s cajan [L] Millsp.). DPhii 

Thesis. University of Zimbabwe, Harare. 

Zimbabwe, pp 62-64. 

Mpepereki, Sand Makonese F. 1995. Prevalence of 

cowpea and soyabean rhizobia in field soils of 

Zimbabwe. Zimbabwe Journal of Agriculture 

33:191-205. 

Mpepereki S., Javaheri F., Davis P. and Giller K.E. 

1999. Soyabean and sustainable agriculture. 

Promiscuous soyabeans in southern Africa. Field 

Crop Research 65:137-149. 

Rusike J., Sukume c., Dorward A., Mpepereki S. 

and Giller K. E. 2000. The economic potential of 

soyabean production in Zimbabwe. Soil Fert Net 

Special Publication. Harare, Zimbabwe, pp 1-3. 

Tattersfield, J.R. 1996. Soyabean Production and 

Research in Zimbabwe. Seed Co-op Company of 

Zimbabwe, Harare, Zimbabwe. 

56 


INTEGRATING ORGANIC RESOURCE QUALITY AND FARMER MANAGE 

MENT PRACTICES TO SUSTAIN SOIL PRODUCTIVITY IN ZIMBABWE 

Abstract 

FLORENCE MTAMBANENGWE and PAUL MAPFUMO 

Department of Soil Science and Agricultural Engineering, University of Zimbabwe, 

P. O. Box MP 167 Mt Pleasant, Harare, Zimbabwe 

Maintenance of soil fertility on smallholder farms in Southern Africa is almost entirely dependant on locally available 

organic resources, with mineral fertilizers only dominating in exceptional cases for relatively wealthy fc:nners. The exploitation 

of nitrogen-fixing legumes for soil fertility purposes, be it in the form of residues from grain or green manure 

legumes, is not widespread in Zimbabwe smallholder farming systems .. In a newly initiated study in three agroecological 

regions in Zimbabwe, we focus on the differential effects oforganic resource quality and quantity on the manifestation 

of soil fertility gradients and subsequently on crop yields as influenced by fanner management practices. Both 

Master/Innovator farmers and poor farmers are targeted in the exploration of the within1arm soil fertility gradients 

(usually giving rise to high yielding 'rich' and low yielding 'poor' fields) observed under different management regimes. 

Preliminary results showed that mineral fertilizer was the most common nutrient source used by over 85% of the fanners 

from three identified farmer groups (Class A - Master farmers, Class B -. Innovator fanners and Class C - Resource 

poor farmers). Livestock manure and woodland litter featured as common organic nutrient sources in Zimuto, while in 

Chinyika and Chikwaka, less than 50% of the farmers used these as additional nutrient sources. As a result of frequent 

organic nutri~nt source usage, soil organic carbon contents of the rich and poor fields of Zimuto was significantly 

higher than from the other two agro-ecological regions (p < 0.05) with values reaching 11.5 mg C g-l soil (rich field) and 

8.5 mg C g.l soil (poor field) . The bulk of the tested soil had very little total nitrogen « 1 mg Ng-l soil). Although the 

chances of building soil organic matter (SOM) under granitic sands in Zimbabwe are slim, understanding the influence 

of organic resource management on SOM dynamics, short-term N availability and mineral fertilizer use efficiency is 

critically important. Target farm sites serve to complement long-term experiments in explaining the interaction between 

organic resource management and crop yields. In addition to these preliminary results, the principles and approaches of 

the study are also discussed in this paper. 

Key words: Organic resource quality, soil organiC matte~, mineral fertilizer 

Introduction 

Marked differences in soil fertility levels and nutrient 

balances can often be observed for the different 

fields within one smallholder farm. These withinfarm 

soil fertility gradients have a major impact on 

overall farm productivity, yet their dynamics are 

often poorly understood. Several reasons could be 

attributed to this high variability, although inherent 

soil properties, micro-climatic conditions and 

farmer management practices may be obvious determining 

factors (Mapfumo and Giller; ~001; Sanchez 

and Jama, 2002; ·Smaling et aI., 1997). Farmer 

management of fields varies with time of planting, 

type of crop planted, type and quality of added inputs 

(including access to cash for purchasing inputs) 

and the efficiency with which nutrients are 

used. Most smallholder farmers in Zimbabwe recognize 

the importance of soil fertility and its conservation. 

There is also general knowledge among both 

scientists and farmers that the application of organic 

residues improves the physical conditions of soil, 

although infonnation as to why this happens is still 

often very limited (Sanchez etal., 1989; Palm et aI., 

1998). 

Agricultural activities and maintenance of soil fertility 

in a smallholder farming commUnity is almost 

entirely dependant on locally available resources. 

However, given the present scenario where traditional 

strategies for sustaining soil and crop productivity 

have been outpaced by the growing human 

population, a diminishing natural resources base, 

and the downward spiral of many national economies, 

the question of resource availability has become 

topical. The big question is: what options are 

available to the smallholder fanner for soil amelioration? 

Annual woodland litterfall may be as much as 5 t 

ha- l and measured annual litter collections in Masvingo, 

southern Zimbabwe, by Nyathi and Campbell 

(1993), ranged from 0.2 to 1.2 tonnes per household. 

In reality, most communal areas are in a severe 

~tate of deforestation. Livestock manure, cattle 

manure in particular, is a traditional source of plant 

nutrients and can be one of the cheapest sources of 

organic fertilizer in many smallholder communities 


(Mugwiia and Murwira, 1997). However, use of manure 

is a privilege for owners. Cereal and legume 

residues are available to many but ~heir uses as soil 

ameliorants is not widespread as these are usually 

fed to livestock or even burnt in some cases to prepare 

land for the next growing season. 

. Mineral fertilizers are accepted as a way to increase 

productivity (Piha, 1993), but the extent to which 

farmers can depend on this input is constrained by 

the low purchasing power of the majority of farmers 

(Ashworth, 1990). Mineral fertilizers are often purchased 

in amounts inadequate to replace those nutrients 

lost alU1ually·in harvested produce. Where 

ecological conditions are not particularly favourable 

for farming or the soil has been degraded, using 

mineral fertilizers alone may be insufficient 

(Smaling et al., 1997). The foregoing discussion indicates 

the critical need to increase the efficiency with 

which the little available nutrients should be used. 

Consequently, nutrient outflows from individual 

farm holdings have progressively become much 

higher than inputs supplied in both organic and 

mineral nutrient sources. This paper discusses prelirriinary 

results of a recently initiated study 

"Managing soil organic matter for improved nutrient 

use efficiency on smallholder farms in Zimbabwe 

- the NUESOM Project" which aims at addressing 

the role of organic residue quality and 

quantity on soil organic matter (SOM) dynamics in 

smallholder farmers' fields under different management 

systems in Zimbabwe. The study is part of a 

larger collaborative research effort by the African 

Network (AfNet) members of the Tropical Soil Biol- · 

ogy and Fertility (TSBF - CIAT) on " Soil Organic 

M?tter Dynamics for Sustainable Cropping and Environmental 

Management in Tropical Systems: Effect 

of Organic Resource Quality and Diversity" 

(Mapfumo et al. 2001a). The study aims to answer 

the following questio~,s: 

(i) Can the reason for crop yield success be attributed 

to grain legume rotations, litter and livestock 

manure applications and / or simply efficient 

mineral fertilizer management techniques? 

(ii) Do legumes playa role in the superior fertility 

status or productive capacity of soils often observed 

on Master and/or IlU1ovator farmer's 

fields? 

(iii) What is the comparative advantage of legumes 

in these circumstances given the current state of 

knowledge on legume technologies? 

This paper discusses the preliminary results of this 

research. 


Study sites 

Three communal areaS in different agro-ecological 

.regions of Zimbabwe, namely Chikwaka in Natural 

Region (NR) II (31°30'E and 17°40'S; 80 km northeast 

of Harare), Chinyika in NR III (32°25'E and 18°15'S; 

250 km east of Harare) and Zimuto in NR IV (30° 

52'E and 19°50'S; 320 km southeast of Harare) were 

chosen for the major part of the study. Natural Region 

II is a sub-humid zone that receives summer 

rains of between 700 - 1000 mm of rainfall, NR III 

receives rainfall of between 650 and 750 mm with 

relatively high temperatures and infrequent, heavy 

fall of rain while NR IV is a semi-arid area in which 

alU1ual rainfall is between 450 and 650 mm and is 

subject to frequent seasonal droughts (Vincent and 

Thomas, 1961). Chikwaka and Zimuto have a long 

history of settlement (over 70 years of settlement) 

while Chinyika is a resettlement area first cropped 

by smallholder farmers less than 20 years ago. 

Farming systems in all the three areas are maizebased, 

with strong crop-livestock interactions. These 

sites were found to be representative of dominant 

soil types found in most parts of the country in addition 

to rainfall regimes. Although the overall focus 

of the project is on-farm, replicate on-station experiments 

were established at Domboshawa Training 

Centre (NR II) located about 30 km north of Harare 

(31°19'E and 17°36'S) and Makoholi Research 

Station in NR IV (about 280 km south of Harare (30° 

45'E and 19°47'S) for detailed measurement on the 

influence of C quality on SOM formation. 

Selection of farm sites 

On-farm sites that have been systematically and 

consistently managed by known groups of farmers 

were chosen for field monitoring and experimentation. 

Gathered information based on farmer participation, 

key informant interviews and literature on 

biophysical and socio-economic characteristics of 

the respective farming systems, was used to classify 

farmers according to resource endowment and competence 

in farming. The focus was on the farmers' 

history of organic matter management. Detailed key 

informant interviews conducted in the three study 

sites revealed that farmers basically fell into three 

groups: Class A - Master Farmers; Cla'ss B - IlU1ovator 

Farmers; and Class C - Resource-poor Farmers 

(Table 1). 

Field surveys 

Participatory rural appraisal (PRA) techniques 

helped to identify the range and determine the 

quantities of organic resources available to smallholder 

farmers. Particular attention was paid to C 

inputs in relation to general soil fertility management 

practices. Focused group discussions, priority-ranking, 

transect walks and informal interviews 

constituted the major PRA tools. Six replicate sites 

across farms, two from each class (Table 1) were selected 

in each Natural Region for further investiga- 

58 


Table 1. Classification of smaliscale farmers of Chikwaka, Chinyika and Zimuto according to key Data on maize yield estimates 

informants 

in the riCh and poor fields were 

Class Description Resource endowment 

also collected during informal 

interviews with the selected 

Class A • The Master Farmer class • livestock (at least 2 cattle) 

• Undergo all training as recommended by 

farmers. 

• plough 

AREX 

• scotchcart 

• Training is on· site and involves crop and • adequate accommodation whether Soil sampling and analyses 

animal production including general farm galvanized iron sheets, asbestos or grass. 

management • Small livestock (chickens! goats) The'key question was whether 

the local indices for productiv 

Class B • Group contains the Innovator Farmers The majority has the same resources as 

ity could be given a scientific 

• Group comprises mostly the eager·to· Master farmers. 

learn type farmers The remainder normally have at least: meaning. To ,answer this, soils 

• Maximize production through informal • some livestock were collected for analysis 

consultation with AREX Officers and! or • a plough 

other farmers 

• a scotchcart 

from the top 20 cm of the rich 

• Generally do not attend training sessions • adequate accommodation and poor fields before the onset 

of the 2002-2003 rainy sea 

Class C • The group includes farmers who take Usually those who don't or have little of: 

time to adopt a technology • livestock son. At least ten auger samples 

• The majority of the members are • plough were collected from each field 

resource constrained! resource poor • scotchcart 

site, bulked in a bucket and 

• have little or no ambition to learn or know 

what is happening in the local environment. then thoroughly mixed to give 

• rarely attend training meetings 

one composite sample that was 

sub-sampled for laboratory 

analysis. These soils were analyzed for total organic 

C and N using methods described by to Anderson 

and Ingram, (1993). The C and N data from the rich 

and poor fields from the three Natural Regions 

were subjected to a Two-sample T-Test for mean 

comparisons (Minitab Inc., 2000). 

tions bringing the total number of farm sites to 36. 

The farm owners assisted in identifying appropriate 

field sites with attributes that included: 

• a known history of organic matter applications 

(at least 5 years) 

• fields with no external C application in the past 5 

years 

• type of predominant management systems including 

mineral fertilizer application~, organic 

matter management (cereal or other stover, livestock 

manure, green manures, woodland litter, 

compost or household waste, termitaria) and legume/ 

cereal intercrops or rotations. 

• distinct soil textures. 

The 36 sites reported here are part of a total of 120 

field sites being investigated in the three Natural 

Regions for the same attributes. Transect walks and 

informal interviews were conducted with the selected 

farmers to help identify the different productivity 

capacity of their fields. We specifically identified 

the mdices that the farmers used to classify the 

yield potential of their fields. Some of the productivity 

indices for rich and poor fields included: 

a) high yielding (Rich) fields 

• high crop performance 

• crops respond well to additional nutrient inputs 

• heavy soils usually with a high humus content 

• islands of termitaria present 

• soils do not easily dry out. 

b) low yielding (Poor) fields 

• poor crop pe;-iormance despite external nutrient 

inputs 

• very light (sandy) free draining soils 

• low humus content. 

Results 

Maize yields 

Maize yields were higher in Chinyika compared to 

Zimuto and Chikwaka (Table 2). Yields of up to 7.0 t 

ha·J were realized by Class A farmers in Chinyika. 

A verage yields from rich fields were highest in 

Chinyika (5.1 t ha·1) followed by Chikwaka (3.5 t ha' 

1) and Zimuto had lowest yields of 2.7 t ha· J• In all 

the 'three regions, yields from rich fields were in the 

order of Class A > Class B > Class C. However, differences 

between yields from poor fields of the 

three identified farmer groups within each NR in 

the three study sites were insignificant (p > 0.05). 

Comparing the three agro-ecological regions, average 

yields from the poor fields also followed the 

same trend with Chinyika having the highest yields 

(1.3 t ha·1) followed by Chikwaka (0.9 t ha·J) and Zimuto 

with the lowest yields (0.5 t ha·1). 

Nutrient sources 

All host farmers in Chikwaka applied mineral fertilizer 

to their rich fields with at least half of them also 

applying livestock manure to the' same fields (Table 

2). Two out of the six poor fields did not receive 

mineral fertilizer. In Chinyika, there appeared to be 

a strong dependency on mineral fertilizer by all, including 

Class C farmers. Only one Class C farmer 

failed to apply any external nutrient to his fields, 

although he still managed to achieve relatively good 


59

Table 2. Average maize yield estimates and farmer management of fields perceived as rich and poor fields in the three agroeco·regions in 

Zimbabwe 

Agro· Farmers' name and Field status Average . Farm management 

reoion Class maiz&- yields 

Mineral fertilizer Organic fertilizilr Legume rotations 

(kg hal) 

NR II Mr G2 - Class A Rich 5.0 150 D*; 100 ANI 7.0 t ha" manure Groundnut (4 years) 

Chikwaka Poor 0.9 OD;100AN 0 None 

Mrs K -Class A Rich 4.0 oD; 0 AN 6.0 t ha" manure None 

Poor 0.8 150 D; 100 AN 0 None 

Mr Ml - Clas.s B Rich 3.7 150 D; 100 AN 6.0 t ha l manure Groundnut (3 years) 

Poor 2.0 1500; 150 AN 0 Soyabeanl runnerbean 

Mrs M2 - Class B Rich 2.5 1500; 150 AN 0 None 

Poor 0.2 150D;100AN 0 None 

Mrs C - Class C Rich 1.5 150 D; 100 AN 0 Groundnut (5 years) 

Poor 1.0 150 D; 100 AN 2.0 t ha I manure Groundnut intercrop 

Mr Gl - Class C Rich 4.5 00; 0 AN 0 None 

Poor 0.6 150 D; 100 AN 0 None 

NR III Mr Cl - Class A Rich 7.0 2000; 200 AN 5.0 t ha l manure None 

Chinyika Poor 1.0 2000; 200 AN 0 Groundnut (2 years) 

Mr C2 - Class A Rich 6.0 150 D; 100 AN 4.5 t ha l manure Soyabean (3 years) 

Poor 2.0 150 D; 100 AN 0 None 

Mr Ml - Class B Rich 6.5 200 D; 200 AN 2.5 t ha" groundnut stover Groundnut (2 years) 

Poor 0.7 2000; 200 AN 0 None 

Mr M2 - Class B Rich 4.5 200 D; 200 AN 0 None 

Poor 1.5 2000; 200 AN 0 None 

Mr W - Class C Rich 2.5 00; 0 AN 0 None 

Poor 0.5 oD; 0 AN 0 None 

Mr Z - Class C Rich 4.0 150 D; 100 AN 0 Groundnut/bambara (4 yrs) 

Poor 2.0 1500; 100 AN 4.5 t ha" manure Groundnut/bambara (4 yrs) 

NR IV Mr Ml - Class A Rich 3.7 1000; 100 AN 0 None 

Zimuto Poor 0.2 00; 100 AN 4.5 t hal manure None 

Mrs M2 - Class A Rich 3.0 1000; 100 AN 2.5 t ha" manure None 

Poor 0.6 100 D; lDO AN 2.5 t ha" manure None 

Mrs C - Class B Rich 2.5 00; 200 AN 2.5 t ha'composted litter Groundnut/cowpea/bambara intercrp 

Poor 0.4 oD; 200 AN 4.0 t ha" manure None 

Mr Z - Class B Rich 2.7 100D;100AN 0.4 t ha ' litter/2 t ha" manure Groundnuts (2 years) 

Poor 0.4 OD;100AN 1.0 t ha I manure None 

Mrs N- Class C Rich 2.0 50 D; 100 AN 0.7 t ha I manure None 

Mrs T - Class C 

Poor 0.3 50 D; 100 AN 0 None 

Rich 2.1 oD; 0 AN 2.5 tha I manure Bambara (2 years) 

Poo: 0.8 oD; 0 AN 0.9 t hal manure None 

• 0 - Compound 0 fertilizer (7% N; 14% P205; 7K20); I AN - Ammonium Nitrate (34.5%N 

Grain legume adoption 

Less than half of the 18 interviewed farmers in the 

three study areas grow grain legumes in their fields. 

In Chikwaka, groundnut rotations in rich fields 

range from 1 in 3 to 1 in 5 years. Only one Class B 

farmer had tried to rotate maize with 'soya and run 

ner beans in his poor field. In Chinyika, legume ro 

tations included groundnut, bambara nut and soya 

bean on a two to four year cycle for all the three 

farmer classes, and only one Class B farmer utilized 

groundnut residues for soil fertility purposes. In Zi 

muto, only two farmers (Class Band C) grow leg 

umes in 2-year rotations (Table 2). Another Class B 

farmer intercropped groundnut, cowpea and bambara 

nut with maize in their rich field. 

maize yields (about 2.5 t ha- 1 from the ric.h field) . 

Manure usage in Chinyika was far lower than that 

in Zimuto and Chikwaka. Only one farmer (Class B) 

used legume stover as a soil ameliorant in Chinyika. 

In Zimuto, organic nutrient resource usage was 

more widespread with livestock manure and woodland 

litter being the common sources among the 

three farmer classes (Table 2). Most of the farmers 

who used organic fertilizers did not apply basal 

Compound 0 fertilizer. However, except for one 

resource-poor farmer, farmers in Zimuto applied 

the recommended rates of between 100 and 200 kg 

ha- 1 ammonium-nitrate fertilizer to both their rich 

and poor fields. 

60 


Soil carbon and nitrogen 

Soil organic carbon content of the soils from the rich 

fields in Chinyika ranged from about 4.5 (Classes B 

and C fields) to 8.2 mg C g-l soil (Class A farmer) 

while that from the poor fields ranged from 3.2 

(Class B farmer) to 8.3 mg C g.l soil (Class A 

farmer) . In five out of six cases, organic carbon was 

consistently higher in the rich than the poor fields 

(Figure la). The only exception observed was that of 

a Class C farmer whose poor field had about 2.5 mg 

g -1 soil more C than his rich field. Total soil nitrogen 

was less than 1- mg N g.l ranging from 0.5 to 0.8 

mg N g-l soil in Chinyika. There was little difference 

in the soil nitrogen contents between rich and poor 

fields in Chinyika for all the farmer classes (Figure 

Ib). In Zimuto, soil C contents ranged between 2.0 

(Class B farmer) and 11.5 mg C g.l soil (Class A 

farmer) in rich fields and between 2.0 (Class C 

farmer) and 8.2 mg C g-l soil (Class A farmer) in 

poor fields (Figure 2a). Unlike the Chinyika case, 

soil nitrogen was relatively higher in Zimuto, with 

results ranging from 0.6 to 1.2 mg N g.l soil (Figure 

2b). In all the selected field sites, the nitrogen content 

of the rich fields was higher than that of the 

poor fields regardless of farmer class. 

Discussion 

Ownership of resources was the key attribute differentiating 

farmer classes. When it came to farm management, 

the more resource-endowed Class A farmers 

had more soil fertility options at their disposal. 

The biophysical characterization of the smallholder 

farming systems has shown that nutrient sources 

accessible to farmers in the different agroecosystems 

were highly heterogeneous and varied in quantity. 

There was a general appreciation of the role of organic 

nutrient sources in soil amelioration in the 

three Natural Regions, particularly livestock manure_ 

However, it was Class A farmers who frequently 

used mineral fertilizers for crop production 

although they could afford to use other available 

resources. Although there was widespread use of 

manure among all classes, the survey also showed 

that application of woodland litter, composted 

household waste and crop residues to field crops 

was deemed experimental by the innovator farmers 

(Class B) and was also perceived as an option for 

resource poor farmers. In many instances, manure, 

when available, was preferentially applied to the 

rich fields particularly by the Class A farmers. This 

Chinyika (NR ·111) 

12.---------------------------,-------, 

.RiCh field 

a) {2J Poor field 

I - ·0.02; df - 9; p > 0.05 

Zimuto (NR IV) 

t - 1.89; df - 9; p > 0.05 

.Rich field 

EJPoor field 

0> 

.§. 

'6 1 2 b) 

'" t - 0.71; df - 9; p > 0.05 

; 1 

0> 

E 

:; 0 .8 

::> 

iii 

"iii 

z 0. 6 

u 

C 

~0.4 

o 

MrCl MrC2 MrMI MrM2 MrZ MrW 

Class A Class B Class C 

Farme r' s name and class 

Figure 1. Pre-season soil organic carbon (a) and nitrogen (b) 

contents of rich and poor fields belonging to six three different 

farmer groups in Chinyika, Zimbabwe 

Mr Ml Mrs M2 Mrs C Mr Z Mrs T Clk);rstt 

Class A 

Class B 

Farmer's name and class 

Figure 2. Pre-season soil organic carbon (a) and nitrogen (b) 

contents of rich and poor fields belonging to six three different 

farmer groups in Zimuto, Zimbabwe 


61

implied that most fanners prefer to invest their labour 

inputs where returns are already favourable, 

thus apparently following a nutrient concentration 

strategy - the poorer or less productive the field, the 

less applied. The ameliorative role of manure to soil 

was not a new concept to the smallholder farming 

community where cattle manure was viewed as a 

. common and significant soil fertility input in maize 

systems (Tanner and Mugwira, 1984; Mugwira and 

Murwira, 1997). Livestock manure applications to 

soil have been known to increase soil pH, infiltration 

rates, water holding capacities and decrease 

bulk densities (Grant, 1967; Murwira, 1993). 

Preliminary results from this study have revealed 

that cultivation of grain legumes for food is at a low 

scale in all the three Natural Regions, be it as rotations 

or intercrops while exploitation of legumes for 

soil fertility management is virtually non-existent. 

While information on the role of leguminous plants 

in replenishing soil fertility is available (Sanchez, 

1995; Giller, 1998; Mapfumo, 2000), these results 

suggest that the strategies to translate this valuable 

information to the smallholder farming community 

need diversification. Currently there is a general belief 

among smallholder communities that legumes 

are a women's crop (Mapfumo et al., 200tb) and this 

belief coupled with poor extension methods and 

over emphasis on cereal production, have led to reduced 

legume cultivation. Application of leguminous 

residues in arable farming systems provides a 

ready supply of N to growing crops. While very few 

farmers appreciate the role of grain legume residues 

in soil amelioration, some results have shown that 

in cereal cultivation, N contributions from legumes 

can be as high as 250 kg N ha- 1 yrl (Giller, 2001). 

However, in poor sandy soils, reported values have 

mostly been less than 30 kg N ha- 1 (Mapfumo, 2000). 

It is imperative to note that the impact of legumes 

on soil productivity may not only be restricted to N 

contributions, which has been the major focus of 

previous work on organic input research. The quantity 

and quality of C supplied by many of these organic 

materials may also playa significant role in 

soil productivity. Information of the role of decomposable 

C on nutrient release and soil amelioration 

from high quality organics including legume residues 

is not well documented (Kirchmann and Bergquist, 

1989). This information is essential in guiding 

farmers and land-managers to optimally use 

their organic resources, both in the short and in the 

long term. Legume resid ues are most beneficial in 

providing nutrients in the short-term, an option 

more likely to be appealing to most smallholder 

farmers (Palm et al., 2001). In the wake of diminishing 

resources, the groyving of legumes and utilizing 

their residues may be a realistic way of increasing 

soil available C in sandy soils. This study aims to 

address the practicalities of these issues. Are we as 

'researchers doing enough to promote soil organic 

matter build-up in our inherently poor soils? Participatory 

experiments with fanners in Murewa (NR 

II) suggested that Cajanus cajan (pigeonpea) could 

be successfully grown by farmers yielding very 

high biomass of up to 23 t ha- 1 in 2 years (Mapfumo 

et al., 2001). 

While it is difficult to make conclusive statements 

based on these preliminary results, a few lessons 

can be drawn. In Zimuto, use of organic fertilizers 

in arable farming showed that soil C reserves could 

be improved judging from the relatively high contents 

of soil organic C, compared to the soils in 

Chinyika where mineral fertilizer usage takes precedence. 

Although cultivation in Chinyika is barely 20 

years old, soil C and N are already depressed 

probably stemming from the heavy dependency on 

mineral fertilizer with little or no organic inputs. We 

therefore conclude that there is merit to develop 

strategies for the use of organic inputs, to not only 

improve the soil organic C status, but also crop 

yields through efficient nutrient uptake. The term 

organic fertilizer should be given a new meaning 

for the smallholder environment to not only mean 

manure qut crop residues as well. For the nonowners 

of cattle, the window of opportunity rests 

with the growing of legumes with the N-rich stover 

being retained in the field. 

It is imperative to investigate how nutrient availability 

is related to the quantity and quality of C 

supplied in organic resources used by farmers in the 

medium- to long-term if combined use of organics 

and mineral fertilizer is to be optimized. 

References 

Anderson, }.M. and Ingram, }.S.1. 1993. Tropical Soil 

Biology and Fertility: A Handbook of Methods. Second 

Edition. CAB International. Wallingford, 

UK. 221 pp. 

Ashworth, V.A. 1990. Agricultural Technology and the 

Communal Farm Sector. Main Report. Background 

paper prepared for the Zimbabwe Agricultural 

Sector Memorandum. The World. Bank, Agriculture 

Division, Southern Africa Department, 

Washington DC, USA. 159 pp. 

Giller, K.E. 1998. Tropical legumes: Providers and 

plunderers of nitrogen. In: Carbon and Nutrient 

Dynamics in Natural and Agricultural Tropical Ecosystems. 

L. Bergstrom and H. Kirchmann (Eds.). 

CAB International, Wallingford, UK, pp. 33-46. 

Giller, K.E. 2001. Nitrogen Fixation in Tropical Crop- 

62 


ping Systems. 2nd Edition. CAB International, 

Wallingford, UK. 423 pp. 

Grant, P.M. 1967. The fertility of sandveld soil under 

continuous cultivation. Part I. The effect of 

manure and nitrogen fertilize"r on the" nitrogen 

status of the soil. Rhodesia Zambia Mal'lwi Journal 

of Agricultural Research 5:71-79. 

Kirchmann, H. and Bergquist, R 1989. Carbon and 

nitrogen mineralization of white clover 

(Trifohum repens) of different age during aerobic 

incubation with soil. Zeitschrijt fur Pflanzenernahrung 

und Bodenkunde 152:283-88. 

Ma·pfumo, P. 2000. Potential Contribution of Legumes 

to Soil Fertility Management in Smallholder Farming 

Systems of Zimbabwe: The Case of Pigeonpea 

(Cajanus cajan fL] Millsp.). DPhil Thesis. Department 

of Soil Science and Agricultural Engineering. 

University of Zimbabwe, Harare, Zimbabwe.198pp. 

Mapfumo, P, Campbell, B.M., Mpepcreki, S. and 

Mafongoya, P.L. 2001b. Legumes in soil fertility 

management: The case of pigeon pea in smallholder 

farming systems of Zimbabwe. African 

Crop Science Journal 9:629-644. 

Mapfumo, P. and Giller, K.E. 2001. Soil Fertility 

Management Strategies and Practices by Smallholder 

Farmers in Semi-Arid Areas of Zimbabwe. PO Box 

776, Bulawayo, Zimbabwe: International Crops 

Research Institute for the Semi-Arid Tropics 

(ICRISAT) with permission from the Food and 

Agriculture Organization of the United Nations 

(FAO). 60 pp. 

Mapfumo, P., Mtambanengwe, F., Nandwa, S., Yeboah, 

E. and Vanlauwe B. 2001a. Soil Organic 

Matter Dynamics for Sustainable Cropping and Environmental 

Management in Tropical Systems: Effect 

of Organic Resource Quality and Diversity. TSBF 

AfNet SOM Network Proposal. (Typed script). 

Minitab Inc. 2000. Minitab Statistical Software. Minitab 

Release 13.1. Minitab Inc. 

Mugwira, L.M. and Murwira, H.K. 1997. Use of cattle 

manure to improve soil fertility in Zimbabwe: 

Past and current research and future research 

needs. Soil Fert Net Network Research Results 

Working Paper No.2. Soil Fertility Network for 

Maize-Based Cropping Systems in Malawi and 

Zimbabwe. CIMMYT, Harare. 33 pp. 

Murwira, 1993. Nitrogen dynamics in a Zimbabwean 

granite derived sandy soil under manure 

fertilization. DPhil Thesis, University of Zimbabwe, 

Harare, 194 pp. 


Nyafhi, P. and Campbell, B.M. 1993. The acquisition 

and use of miombo litter by small-scale farmers 

in Masvingo, Zimbabwe. " Agroforestry Systems 

22:43-48. 

Palm, CA., Murwira, H.K. and Carter, S.E. 1998. In: 

Soil Fertility Research for Maize-Based Farming Systems 

in Malawi and Zimbabwe. S.R. Waddington, 

H.K. Murwira, JD.T. Kumwenda, D. Hikwa and 

F. Tagwira (Eds.). Proceedings of the Soil Fert 

Net Results and Planning Workshop, 7-11 July, 

1997. Africa University, Mutare. Soil Fert Net 

and CIMMYT-Zimbabwe, Harare. pp 21-29. 

Palm, CA., Gachengo, CN., Delve, R.J., Cadisch, G. 

and Giller, K.E. 2001. Organic inputs for soil fertility 

management in tropical agroecosystems: 

application of an organic resource database. Agriculture, 

Ecosystems and Environment. 83: 27-42. 

Piha, M.1. 1993. Optimizing fertilizer use and practical 

rainfall capture in a semi-arid environment 

with variable rainfall. Experimental Agriculture 

29:404-15. 

Sanchez, P.A. 1995. The science of agroforestry. 

Agroforestry Systems 30:5-55. 

Sanchez, P.A. and Jama, B.A. 2002. Soil fertility replenishment 

takes off in East and Southern Africa. 

In: Integrated Plant Nutrient Management in 

Sub-Saharan Africa: From Concept to Practice. B. 

Vanlauwe, J. Diels, N. Sanginga and R Merckx 

(Eds.). pp. 23-45. CAB International, Wallingford, 

UK. 

Sanchez, P.A., Palm, CA., Szott, L.T., Cuevas, E. 

and Lal. R 1989. Organic input management in 

tropical agroecosystems. In: Dynamics of Organic 

Matter in Tropical Ecosystems. D.C Coleman, J.M. 

Oades and G. Uehara (Eds.). NifTAL" Project, 

University of Hawaii, Honolulu, Hawaii, USA . 

pp. 125-152. 

Smaling, E.M.A., Nandwa, S.M. and Janssen, B.H. 

1997. Soil fertility in Africa is at stake. In: Replenishing 

Soil Fertility in Africa. Buresh, RJ., Sanchez, 

P.A. and Calhoun, F. (Eds.). Soil Science Society 

of America Special Publication 51. Madison, 

Wisconsin, USA.pp. 47-62. 

Tanner, p.o. and Mugwira, L.M. 1984. Effectiveness 

of communal area manure as sources of nutrients 

for young maize plants. Zimbabwe Agricultural 

Journal 81 :31-35. 

Vincent, V. and Thomas, RG. 1961. An Agroecological 

Survey of Southern Rhodesia: Part 1-Agro 

Ecological Survey. Government Printers, Salisbury, 

Rhodesia. 

63

] 



To Sheunesu Mpepereki and Ishmael Pompi 

Q: How did you handle marketing among 

smallholder farmers? 

A: In Zimbabwe, lucrative markets exist for 

soyabean, e.g. for oil expression and livestock feeds. 

Smallholder farmers come together in groups to 

consolidate their harvest into large enough loads for 

transport to market. (:ontracts have been negotiated 

with buyers to accommodate all smallholder crops. 

The Soyabean Promotion Task Force has played a 

coordination role that has been progressively 

passed on to farmers' own organizations. Private 

buyers have supplied weigh scales. 

Q: Is inoculation a full component of the soyabean 

tedmology or do farmers often grow soya bean 

without inoculation? 

A: Yes inoculation is the key technology being 

promoted. The inputs package contains seed, 

inoculant, lime (for acid soils), base fertilizer and 

fungicides (for rust disease). No farmer will plant 

soyabean without rhizobia inoculants if they can 

help it. Some plant promiscuous varieties. 

Unfortunately, there is no breeding program for 

promiscuous varieties in Zimbabwe. 

Q: Where does the soyabean fit within the whole 

farm system ~iven soil fertility gradients? 

A: In Zimbabwe soyabean is planted in outfields, 

often not the most fertile fields. Farmers are 

encouraged to grow soyabean in the more fertile 

fields to enhance yields and income from sales. 

Q: What is the percentage of smallholder farmers 

adopting soyab.ean production technology in 

Zimbabwe? 

A: Adoption rates have been near experiential. 

Numbers increased from a few hundred to over 10 

000 in three growing seasons (1996 -1999). Area 

planted has increased from about 240 ha (1995) to 44 

000 ha in 2000. In one communal area, Kazangarure, 

with about 3000 families, AGRlTEX estimates over 

98% have adopted soyabean BNF technology over a 

four year period (1997 - 2000). 

Q: To what extent could you have soil residual 

effects of the inoculants in the field? 

A: Residual effects of inoculants depend on the 

survival and persistence of inoculation strains. In 

heavy soils (with high clay and organic matter 

content), rhizobia strains survive and are effective 

for up to three seasons or more if the legume is 

grown in a regular rotation. Survival and 

persistence are poor in sandy soils where the 

legume requires to be inoculated every time it is 

planted. 

To Friday Sikombe, et al. 

Q: What were the optimum levels of nitrogen 

fertilizers and inoculation for bean yields? 

A: The optimum levels of nitrogen recommended 

o 

were 100 kg N ha ] which is called the Lima 

recommendation. For the inoculum, the optimum 

level is two 250g-packets of inoculant per hectare. 

Q: The pH of the soils at your site was 7.2. What 

could have been the effects on N- fixation? You also 

applied N-fertilizers at two rates; 0 and 100 kg N 

hao 

]. Don't you think that 100 kg N hao 

] was rather 

too high and could have suppressed nodulation? 

Do you think we have farmers who can apply 

fertilizers at this rate? 

A: The pH 7.2 had no effect on N- fixation. The 

level of 100 kg N hao is the Lima recommendation. 

This level did not affect nodulation except with the 

cultivar, Lundazi. It is true that small-scale farmers 

are unable to apply fertilizer nitrogen at this rate. 

The option, therefore, is to exploit Biological 

Nitrogen Fixation (BNF) through inoculation with 

RhizobiJ, and the use of bean genotypes that 

respond well to inoculation. 

C: A rate of 100 kg N/ha is certainly too high for a 

legume. Its effect would be to limit nodulation and 

N fixation in the beans. 

To Ylver Besmer, et al. 

Q: Did you quantify the AMF inoculants, e.g. spore 

numbers? And is the intervention one that f.umers 

can introduce and manage? 

Why lab lab? Does it h,l\'e any utility Y,11ue for 

farmers or a chance of being integrated into the 

cropping system? 

Quantities of N from groundnut appear e:dremel~' 

low, contrary to common knowledge that the 

residues of groundnut have high amounts of N. 

How do you explain this? 

How did yOll account for Iitterfall by pigeonpe'l in 

calculating N input? 

How was the control for trapping the Aiv1F treated? 


65

General Discussion 

C: Recovery of legume N may be low in a 

subsequent cereal crop but a substantial part often 

remains in the SOM pool, which may build soil 

fertility and benefit future crops. 

C: The N recovery of leguminous tree rotations with 

maize is 10-20% and subsequent recovery is 3-5%. 

However we need fertilizer equivalencies of organic 

based technologies and these should be presented in 

extension manuals. 

C: In relation to the recovery of N from soyabean 

residues by a subsequent maize crop. Results from 

work carried out at CIAT (1993-98) on the Colombia 

Savannas on an Oxisol in the humid tropical 

lowlands (150 masl, >2600 mm rainfall) using ISN 

techniques found 10-20% recovery of N from 

soyabean residue by a subsequent maize crop. A 

substantial amount was lost by leaching (>50%). N 

recovery was

ADDING A NEW DIMENSION TO THE IMPROVED FALLOW CONCEPT 

THROUGH INDIGENOUS HERBACEOUS LEGUMES 

PAUL MAPFUMO', FLORENCE MTAMBANEI\JGWE', 

SHEUNESU MPEPEREKI' and KEN GILLER 2 

1Department of Soil Science and Agricultural Engineering, University of Zimbabwe, 

P. O. Box MP 167 Mt Pleasant, Harare, Zimbabwe 2Department of Plant Sciences, 

Plant Produc'tion Systems, Wageningen University, Wageningen, The Netherlands 

Abstract 

Opportunities for harnessing biological nitrogen fixation of non-cultivated herbaceous legllmes in order to improve soil 

productivity on smallholder farms in Zimbabwe were explored in a study initiated in Decem ber 2001. Over 30 indigenous 

legume species .were jointly identified with farm ers across three agro-ecological regions, ranging from sub-humid 

(800 mm annually) to semi-arid «650 mm). A ewezu Smell Technique, based on the odour released from freshly harvested 

legume roots, greatly enhanced the capacity of farmers to participate in the identification process, The legume diversity 

was highest in Chinyika Resettlement Area where cropping had been going jllst over 20 years. This was contran) 

to the loss of diversity in old Communal Areas where dominance of clean weeding practices for over 70 years has led to 

depletion of the weed seed bank. Legume contribution to the total above grollnd biomass ranged from 3% in 1-year fallows 

under semi-arid conditions to 88% in afield abandoned soon after crop establishment llnder sub-humid conditions. 

The latter case indicated an opportunity for manipulating legume popl/lations to increase their contribution, Overall, 

total fallow productivity did not exceed 3.2 t ha- J dl/e to extreme conditions of poor soil fertility, where soils had generally 

logical processes in the given agro-ecosystems. Despite 

the advocacy for integrated nutrient management 

and ecological approaches to agriculture (e.g. 

Giller and Cadisch, 1995i Breman, 1998), little or no 

research work in Zimbabwe and other parts of 

Southern Africa has focused on natural weeds as an 

organic nutrient resource that can be exploited by 

smallholder farmers for their management of soil 

. fertility. This study, therefore, focused on selfregenerating 

N2-fixing indigenous legumes. These 

legumes are considered an under-utilized component 

of an organic resource pool that may be readily 

available to smallholder farmers in many parts of 

Africa. Assessing and manipulating the diverse N2 

fixing herbaceous legumes in local agro-ecosystems 

provide a good starting point to meet these challenges. 

Based on results of a study initiated in December 2001, 

this paper explores the concept and scope for indigenous 

fallows (Indifallows). The general objective was 

to make an appraisal on the potential for indigenous 

legumes to contribute towards combating the problem 

of poor soil fertility which underlies rural poverty in 

Zimbabwe and other parts of Africa. 

The Indifallow concept 

The Indifallow concept is based on harnessing biological 

nitrogen fixation (BNF) of herbaceous annual 

legumes native to or naturalized under given agroecological 

environments in order to improve the N 

economy of natural fallows at minimal establishment 

and management costs. While it is traditional 

practice to fallow unproductive land, effectiveness 

of fallows as a means for soil fertility restoration has 

often been compromised by the reduction in fallow 

periods as land becomes limiting and also the poor 

quality of the plant biomass generated in the fallowing 

phase. Efforts to improve the quality of faHows 

through agroforestry tree crops such as Leucaena, 

Sesbania and Acacia spp. have often been hindered 

by high establishment costs and lack of immediate 

benefits to the farmer (Cook and Gmt, 1993; Kwesiga 

and Coe, 1994; Snapp et al. 1998). Through use 

of self-regenerating and well-adapted indigenous 

annual legumes, constraints related to seed costs 

and availability, nursery management and biomass 

management are minimized. Thus a focus on BNF 

of indigenous herbaceous legumes will not only 

help to provide a basis for integrating weeds as a 

potential source of N and soil organic matter in 

cropping systems, but also to improve and maintain 

the biodiversity in smallholder agro-ecosystems. 

Most studies on weed management in smallholder 

fanning systems have been concerned with the adverse 

effects of weed competition on moisture and 

nutrient uptake by crops, and the labour costs involved 

in managing the weeds. As a result, the 

strategy has been to eradicate weeds, mainly by depleting 

the seed bank (weed plants are killed before 

flowering). Clean weeding, which involves maintenance 

of weed-free fields until the end of the cropping 

season, is a common practice among smallholder 

farmers in Zimbabwe. This is still driven by 

extension recommendations based on 'green revolution' 

technologies. Although this weed management 

approach has become a tradition, it may compromise 

tl}e long-term sustainability of these cropping 

systems· thro_ugh loss of bio-diversity and reduced 

organic matte'i- inputs. It also leads to the dominance 

of pernicious weeds that are by definition difficult 

to control by hand weeding and cultivation. 

An analysis of 'green revolution' technologies in 

sub-Saharan Africa has shown that they are largely 

incompatible with the socio-economic environment 

on smallholder farms (Quinones et al. 1997). It is imperative 

that the current weed management regimes 

be revised to match the demands for integrated nutrient 

management and reduce labour requirements. 

This may reduce the burden on women and children 

who usually provide labour for key agricultural 

activities such as weeding and seedling establishment 

and transplanting in agroforestry. Technologies 

with minimal labour demands are likely to 

be particularly appropriate for the poorest farmers 

who are often women in single-headed households, 

or families where key members providing labour 

have been lost due to AIDS. As we explore the feasibility 

and merit of indifallows from a soil fertility 

perspective, the key question is whether such legumes 

do exist in smallholder farming systems and 

under what soil conditions. . 

Study Sites 

The research was conducted in three communal 

(smallholder) areas found in different eco-zones of 

Zimbabwe, namely Chikwaka (31° 30' Ei 17° 40' S) 

in Natural Region II, Chinyika (32° 25' Ei 18° 15' S) 

in NR III and Zimuto (30 0 

52' E; 19° 50' S) Communal 

Areas in NR IV. Natural Region II receives over 

750 mm of rainfall annually between November and 

March while NR's III and IV receive 650-750 mm 

and 450-650 mm of unimodal rainfall per annum 

respectively. The soils in all sites are granite-derived 

sand to loamy sands, Haplic Lixisol/ Arenosols according 

to the F AO classification. The sites were 

mainly chosen based on their being representative 

of most smallholder farmmg areas. Chikwaka and 

Zimuto are old Communal Areas where cultivation 

by smallholders has been going on for over 70 

years. The a:verage household landholding in Chikwaka 

and Zimuto was 3 ha, while in Chinyika, a 

resettlement area established in 1982, the landholding 

was 6 ha per household. 

68 



The study used farmer participatory approaches, 

complemented with laboratory-based analyses of 

soils and plant materials. At least two participatory 

rural appraisal (PRA) workshops were held at each 

study site to discuss broader issues of soil fertility 

management and local knowledge of leguminous 

plants. Farmer involvement ranged from identification 

of the legumes, and their niches, to seed collection. 

Members of the local community leadership 

that included councilors, headmen and resident national 

extension officers organized and participated 

in transect walks during the initial legume identification 

exercise. 

Transect walks and legume identification using 

the Gwezu smell technique 

Tr~ect walks were conducted in all study sites. 

Based on the physical slope, farmers identified three 

main field positions, namely, topland, midslope and 

the relatively moist bottomland positions. During 

the transect walk, particular attention was paid to 

cropping patterns (including crop types), weed 

status of the fields, and occurrence of naturally 

growing herbaceous legumes. General discussions 

ensued during the course of the walks and details of 

cropping history and predominant weed species 

were specifically discussed with farmers whose 

fields were surveyed. Farmers were generally able 

to distinguish legumes from non-leguminous plants 

by considering mainly fruit morphology and likening 

them to traditionally grown leguminous crops 

such as groundnut (Arachis hypogaea), common bean 

(Phaseolus vulgaris) and cowpea (Vigna unguiculata). 

Because identification was done when most of the 

species were not yet fruiting, applicability of this 

approach was limited. To aid this process, the research 

scientists then came up with an identification 

approach based on the human sense of smell, hereinafter 

called the Gwezu smell technique. The researchers 

discovered that freshly harvested roots of 

all the identified legumes invariably had ,a distinct 

smell characteristic of an immature groundnut pod. 

An immature groundnut pod is known as Gwezu in 

Shona (Karanga dialect), a Zimbabwean vernacular 

language. Plants were also uprooted, and presence 

of root nodules was considered indicative of a legume, 

taking care to differentiate true root nodules 

from the galls caused by root-knot nematodes. The 

field-identified legumes were then taken to the National 

Herbarium laboratory of the Zimbabwe Ministry 

of Lands~ Agriculture and Rural Resettlement, 

for botanic identification. 

Measuring species diversity and abundance 

Transect walks and PRA group discussions resulted 

in three possible scenarios for legume sampling to 

determine the diversity and abundance of the legume 

species: Scenario I - natural grazing areas that 

have not been cultivated for more than five years; 

Scenario II - fields that had not been cropped in the 

current season (first season of fallowing); and Scenario 

III - cultivated fields in which only the first 

weeding had been done. After further consultation 

with farmers, it was decided that measurement of 

species abundance be focused on the latter two scenarios 

since grazing by livestock would affect measurements 

under natural grazing areas. Consequently, 

the emphasis on Scenario I was only on determining 

species diversity. Individual plant samples 

were collected by farmers, field assistants and 

researchers enclosed in polythene bags, and put in 

cooler boxes for transportation to the National Herbarium 

for identification. For Scenario II, only those 

fields that were free from livestock disturbance during 

the cropping season were sampled. The only 

exception to the sampling protocol was at Mr Zindoma's 

farm where a maize field abandoned soon 

after crop emergence due to lack of fertilizer inputs 

was additionally included. 

Sampling and analyses of plants and soils 

Sampling for Scenarios II and III was done by using 

a network of 4 m x 4 m grids that were made out of 

metal pegs and twine. The grid network was spread 

over the desired field area and four replicate grids 

randomly selected for sampling. For each replicate 

sampling grid, all legume plants belonging to the 

same species were uprooted, checked for nodulation 

and put in a khaki sampling bag after cutting 

off the roots from just above the soil line. Nonleguminous 

plants were collectively sampled from 2 

randomly located grids of 0.5 x 0.5 m 2 drawn from 

within the 4 x 4 m 2 grid. In each agro-region the 

process was repeated at each of the selected 10 farm 

sites where fields meeting desired criteria were 

found. All plant samples were oven-dried to constant 

weight at 60°C and then measured for dry 

mass. The dried samples were then ground in a 

Wiley Mill to pass through a 1 mm sieve, Determination 

of N, P and K concentrations was then done 

using the methods given by Anderson and Ingram 

(1993). 

Soils were sampled from the respective field sites by 

collecting 15 sub-samples from the 0 - 20 cm depth 

per field site using a spade. The sub-samples were 

mixed thoroughly in a clean polystyrene bucket, after 

which a 1 kg composite sample was withdrawn 

and put into a polythene bag for laboratory analysis. 

The soils were analyzed for texture, pH, organic 

C and plant available P according to methods by 

Anderson and Ingram (1993). 


69

Seed collection 

Seed collection was considered an important entry 

point for farmer involvement, and would enhance 

farmers' capacity to identify the legumes. No spe 

cific assignment was given to individual farmers. 

All volunteers made collections for more than one 

species, taking into consideration the spatial distri 

bu tion, time differences in reaching maturity and 

the differences in growth patterns among the differ 

ent species. 

Results 

Species diversity and abundance 

Thirty-three indigenous herbaceous legume species, 

mainly of the genera Crotalaria, Indigofera and 

Tephrosia, were identified among the three agrozones. 

Using the Gwezu smell technique, participating 

farmers and field assistants were able, not only 

to identify the legumes, but also to collect seed. The 

highest number of 28 legume species (Table 1) was 

recorded in Chinyika resettlement area. Ten of the 

total of 33 species were only present in Chinyika 

and not in the other two Communal Areas, while a 

further 10 species were commonly found in all areas. 

There was a critical lack of information on indigenous 

names for the diverse legumes identified. 

Farmers attributed this to fact that these species 

were generally not rated as problem weeds, neither 

were they used as a source of food at household 

level. They were therefore unlikely to be given specific 

names because of their little economic importance. 

Seed collection by farmers was feasible for all 

species except Alysicarpus ovalifolius. 

Because of a severe drought that started mid-way 

through the growing season, the determination of 

species abundance under cropped areas was rendered 

impossible. Weeds failed to germinate after 

the first weeding due to lack of moisture. Results on 

legumes species abundance were therefore only 

available from fallowed areas (Scenario II). Rothia 

hirsuta was the most dominant species in Chinyika 

where it constituted 71% of total legume biomass, 

while Indigofera astragalina was predominant in 

Chikwaka contributing 47% (Figure la and b). Crotalaria 

pisicarpa, predominated in semi-arid Zimuto 

where it contributed 39% to the total legume biomass, 

followed by R. hirsuta with 32% (Figure lc). 

.. Apart from R. hirsuta and I. astragalina which featured 

prominently across all agro-regions, there 

were significant differences in the abundance of 

species from one region to another. For instance, 

Crotalaria cylindrostachys only featured prominently 

in Chinyika while the most significant amounts of 

Zornia glochidiata biomass were measured in Zimuto 

(Figure 1). At Mr Zindoma's additional field site 

70 

Table 1. Species of indigenous herbaceous legumes identified 

to grow as weeds on smallholder farms in three agro·ecological 

regions in Zimbabwe 

legume species 

Agro·ecological region (NR) and site 

NR II: NR III: NR IV: 

Chikwaka Chinyika Zimuto 

Alysicarpus ovalifolius NI I NI 

Eflosema ellipticum NI I NI 

Crotalaria cylindrostachys I I I 

C. glauca NI I NI 

C. laburnifolia I NI I 

C. microcarpa I I I 

C. ochreleuca I I NI 

C. pisicarpa NI NI I 

C. rhodesiae I NI NI 

C. sphaerocarpa I NI 

Chamaecrista absus I I 

C. mimosoides I I 

Indigofera antunesiana NI NI 

I. astragalina I I 

I. brachynema NI NI 

I. demisa I NI 

I. fla vicans NI I 

I. nummularifolia NI NI 

I. praticola I NI 

I. vicioides I NI 

I. wildiana I NI NI 

Macrotyloma daltonii NI I I 

Neonotonia wightii I I I 

Rothia hirsuta I I I 

Stylosanthses fruticosa NI I NI 

Tephrosia acaciifolia NI I NI 

T. longipes NI I I 

T. lurida NI I NI 

T. purpurea NI I NI 

T. radicans I I I 

T. reptans I I NI 

Vigna vexillata NI NI I 

Zornia glochidiata I I I 

Total number of species 18 28 15 

identified 

I - identified; NI - not identified in the area; NR II - 750 mm annual rainfall; 

NR III - 650·750 mm; NR IV - 450·650 mm 

(outside the regular sampling domain), there was a 

dense natural stand of legumes dominated by C. 

cylindrostachys with legumes contributed up to 88% 

of the above ground biomass (Figure '2), suggesting 

populations of the existing species are highly dynamic 

within a single growing season depending on 

soil management. 

No observable patterns in species distribution 

across catenary pOSitions were apparent in all areas 

during transect walks. The only notable exception 

was the growth and abundance of Vigna vexillata 

and Zornia glochidiata in bottomland positions 

(seasoIlally waterlogged or dambo fields) where the 

rest of the species were excluded. While the former 

was only restricted to dambo fields, the latter also 

occurred on topland and mid-slope positions down 

the catena. 


a) Chikwaka 

b) Chinyika 

!21 C. absus 

[!] C. cylindrostachys 

• C. microcarpa 

01. astragalina 

[j I. demisa 

12:11. praticola 

.1. wildiana 

47% BZl R. hirsuta 

tSI T. radicans 

Il§I Z. glochidiata 

i] C. cylindrostachys 

• C. microcarpa 

lID C. mimosoides 

3% GI E. ellipticum 

01. astragalina 

2% 

1(9/. praticola 

71% 

0% 

lID I. vicioides 

1151 R. hirsuta 

!lim Z. glochidiata 

c) Zimuto 

range, for Zimbabwean soils, in Chikwaka 

and Zimuto. Overall, legumes contributed 

a maximum 12% of the total biomass 

harvested under high relative rainfall 

conditions (average 800 mm y r· l ) in 

Chikwaka, and as low as 3% in semi-arid 

Zimuto (Figure 3). The total biomass productivity 

was highest in Chinyika Resettlement 

Area, with slightly over 3 t ha- 1 , 

while Zimuto had no more than 0.75 t 

ha- 1 . Plant productivity was evidently reduced 

by drought in the second half of 

the season. 

N, P and K contents of identified species 

There were high variations in the tissue 

N, P and K concentrations of the legumes 

across the three study areas. About 12 out 

of the 35 sample entries across sites had 

more than 2% N (Table 3). In general 

there were more entries with high tissue 

N concentration from the semi-arid area 

than from the other two regions. Surprisingly 

high total N values of 5.02 and 

5.88% were measured for C. inbul'izijolin 

and T. pUrpllrl!n, respectively, both of 

which were sampled in Zimuto. The 

number of samples with relatively high 

concentrations of P and K was generally 

high in Chinyika. 

t;] C. cylindrostachys 

o C. microcarpa 

Discussion 

!ill C. mimosoides The Gwezu Smell Technique for identification 

of legumes by farmers 

13 C. pisicarpa 

39% [] I. astragalina 

Participatory research is often constrained 

by lack of a common tool for as 

32% 

D I. flavicans 

sessing or evaluating technologies. In 

&3 M. daltonii 

several instances, there is a technical language 

barrier between farmers and re 

~R. hirsuta 

!Ill V. vexillata 

searchers. The Gwezu Smell Technique 

IIJ!l Z. glochidiata provides an identification tool that can be 

shared by researchers and farmers. The 

technique is easy and accessible to all 

Figure 1. The relative abundance of legume species, expressed as %of total legume farmers, and could the refore provide an 

biomass per unit area, identified in communal areas falling under different agro· opportunity for them to distinguish and 

ecoregions in Zimbabwe 

utilize legumes in their own environments. 

It could be combined with other 

Productivity of the legumes on abandoned infertile 

soils 

nodulation and nodule colour development used in 

assessment tools such as physical inspections for 

Most of the identified species were adapted to much the field for legume N 2-fixa tion appra!sals . There is, 

depleted coarse sandy soils with organic C averag however, a research challenge to identify the chemi 

ing less than 0.4% and mean available (Olsen) P lev cal substance responsible for giving this characteris 

els below 3 ppm (Table 2). About 89% of all the soils tic smell. 

sampled had less than 10% clay. The pH was 

slightly acidic in Chinyika but in the strongly acidic 


Other weeds 

12% 

C. 

indrostachys 

T. radicans 49% 

22% 

R. hirsuta,__--IOCIf""", 

6% 

Figure 2. Weed composition of a maize 

field abandoned soon after crop 

establishment due to lack of fertilizer 

inputs in Chinyika smallholder farming 

area, Zimbabwe 

Chamaecrista 

absus 

0.4% 

• Crola/aria microcarpa 

fl C. cytindroslachys 

Kl Chamaecrisla absus 

[j /ndigofera aslraga/ina 

1lI/. Demisa 

III Rolhia hirsula 

~ Tephros/a radicans 

• Other IIeeds 

14 T"'......~~-"-"'O:"'""'..................---,......,,-.,. .wOO 

~ 

g 12 3500 ~ 

~ _ 10 3000 "" g 

o~ 

." 

n -; 8 

..... 

2500 e_ 

~ :; 2000 :: ~ 

on E 6 .. "" 

.. 0 

w:E 

~ 4 

" go 

...J 

1500 ~ 

1000 ~ 

2 500 

... ~ 

o 

o 

Chlkwaka Chlnylka Zlmuto 

Site 

1_%LegulTIII -+-Total Productivity I 

Figure 3. Legume biomass contribution to total biomass product 

under a one·year natural fallow in smallholder farming areas across 

three agro·ecologicai regions in Zimbabwe 

Legume diversity and adaptability 

Diversity of legumes was higher in the relatively 

new resettlement area of Chinyika, where smallholder 

cropping has been taking place for about 20 

years, than in the old smallholder farming areas of 

Chikwaka and Zimuto where cultivation had been 

going on for over 70 years. The trend on legume diversity 

contrasted with the pattern in biomass produ£tivity, 

which showed a decline in legume biomass 

from Chikwaka (-12% of total), a high rainfall 

area, to Zimuto (-3%) a semi-arid area. This suggests 

a loss of legume diversity in the old communal 

(smallholder) areas of Chikwaka and Zimuto, 

and we attributed this to continuous depletion of 

the seed bank due to continuous cultivation coupled 

with clean weeding approaches. However, the 88% 

contribution by legumes to total biomass productivity 

in an abandoned field suggests that boosting the 

population density can enhance the legume contri 

Table 2. Major soil characteristics for selected fallowed field sites on 

which indigenous herbaceous legumes were naturally growing in 

smallholder areas of Zimbabwe 

Areas and Farmer's %Clay %Sand pH Olsen P %C 

Agro·region Name (H2O) (ppm) 

Chikwaka Kaseke 6 85 4.8 3 0.40 

NR II Mutawu 6 87 5.5 1 0.33 

Tafirenyika 6 83 5.3 1 0.32 

Chinyika Majaji 9 78 5.8 3 0.41 

NR III Razio 13 75 6.0 1 0.49 

liodoma 7 82 6.0 3 0.52 

Zimuto Madhava 4 81 5.0 1 0.40 

NRIV Makonese 8 81 5.3

Table 3. N, P and Kconcentrations (means of ~ 5 samples) of indigenous legume species sampled identified species by farmers 

from fields fallowed for asingle growing season in three smallholder farming areas across different themselves was fea-sible is 

agro·regions in Zimbabwe 

indicative of the potential to 

Species Nutrient Concentration manipulate legume densities 

in the stands, once the popu 

Chikwaka (NR II) Chinyika (NR III) Zimuto (NR IV) 

lation dynamics are under 

'lioN %P %K %N %P %K %N %P %K stood. We consider Indifal 

Alysicarpus ovalifolius 1.53 0.04 0.73 

lows as a technology for 

Chamaecrista absus 1.19 0.11 0.77 those poor and vulnerable 

C. rotundifolia 2.12 0.08 1.75 farmers for whom current 

C. mimosoides 1.45 0.07 1.44 1.45 0.05 0.48 research and development 

Crotalaria cylindrostachys 1.63 0.09 2.11 2.09 0.09 2.33 1.77 0.06 1.79 initiatives have failed to 

C. laburnifolia 5.02 0.12 1.04 draw their participation. 

C. microcarpa 2.09 0.08 1.71 .1.99 0.18 2.19 2.81 0.11 1.14 

These poor groups often 

C. pisicarpa 2.67 0.08 1.14 

Eriosema ellipticum 1.62 0.05 0.99 

lack minimal cash require 

Indigofera astragalina 1.95 0.08 0.89 1.62 0.14 2.16 1.18 0.04 0.33 ments to invest into cur 

I. demisa 1.81 0.10 1.28 rently available soil fertility 

I. flavicans 1.71 0.05 1.29 technologies. The challenge, 

I. praticola 1.74 0.13 1.14 however, is that of develop 

I. vicioides 1.41 0.07 1.75 2.59 0.11 1.24 ing strategies for integration 

I. wildiana 2.38 0.10 0.84 

of these legumes into exist 

Macrotyloma daltonii 1.66 0.06 1.14 

Rothia hirsuta 1.59 0.09 1.72 1.91 0.10 2.44 1.66 0.06 1.02 ing cropping systems, and 

Tephrosia longipes 1.32 0.05 0.33 

defining the practical do 

T. purpurea 1.76 0.04 0.43 5.88 0.04 0.73 main within which the tech 

T. radicap~ 2.22 0.07 1.19 nology can work. The exis 

Vigna vexillata 1.65 0.06 1.40 tence of regional research 

Zornia glochidiata 2.72 0.09 1.14 1.59 0.17 1.45 2.27 0.08 0.96 networks such as the TSBF 

Other weeds (mostly grasses) 0.75 0.07 0.73 0.63 0.07 1.35 0.69 0.05 0.59 

CIA T African Network 

I Each value·is a mean of four samples; NR - natural (agro·ecologicalJ region; H implies amissing value due to absence 01 (AfNet) and Soil Fertility 

particular species in the sampling framework 

Network for Maize-Based 

1. Abundant seeding to allow ready propagation 

Cropping Systems in Southern Africa (SoiIFertNet) 

and ready seed collection to reinforce populaand 

testing of Indifallow technologies on a regional 

provide an opportunity for a wider development 

tions. 

basis. 

2. A long-lived seed bank. 

3. Rapid establishment and growth. 

4. Adaptation to poor soils with restricted availability 

of phosphorus. 

Conclusions 

5. Good N2-fixing potential in terms of spontaneous 

nodulation with indigenous rhizobia, good Several conclusions were drawn based on this ex 

nodulation potential and high N concentrations ploratory study. Results showed that smallholder 

in the shoots. 

farming systems across different agro-ecological re 

6. Easy to remove by hand pulling or hoeing gions in Zimbabwe contain sufficient diversity of 

should weeding be required. 

indigenous herbaceous legumes to warrant more 

The legumes thatbestfit these characteristics are STrategic research on Indifallows as a component of 

largely annuals, biennials or short-lived perennials. integrated soil fertility management. However, 

there are indications that that current clean weeding 

Opportunities for Indifallows in small!:tolder practices recommended for these farmers may be 

farming systems 

contributing to loss of agro-biodiversity in farming 

The potential for Indifallows lie1' in the existence of systems. Most of the dominant species were evia 

diversity of annual legumes that can grow in their dently tolerant to poor fertility soils with low P and 

mixtures with little demand for management of inon 

pH levels, yet it was apparent that most of the soils 

terspecific competition. Unlike agroforestry improved 

fields fallowed by farmers in Zimbabwe are too 

fallows and annual green manures, which poor to support any meaningful cropping using the 

are often constrained by conditions of poor initial currently available low cost soil fertility technolo 

soil fertility, establishment costs and high labour gies. Although there is still need to establish why 

demands, the most significant cost variables for In farmers have not significantly exploited these re 

difallows are likely to be seed collection and sow sources over time, we advocate for a paradigm shift 

ing. The fact that seed collection for most of the in weed management approaches towards enhance- 


ment of agro-biodiverse smallholder farming systems. 

There is scope for wider networking and collaboration 

on development of Indifallow technologies 

in the sub-regions in Africa. 


This research was funded by the Rockefeller Foundation 

through a grant to the senior author under 

the African Careers Award (ACA) Programme. We 

thank Dr Stephen Waddington, coordinator of the 

Soil Fertility Network for Maize-Based Cropping 

Systems in Southern Africa (SoilFertNet) for support 

and encouragement in the development of this 

paper. 

References 

Anderson, LP., Brinn, P.J., Moyo, M. and Nyamwanza, 

B. 1993. Physical Resource Inventory of the Communal 

Lands of Zimbabwe - An Overview. NRI Bulletin 60. 

Natural Resources Institute, Chatham, UK. 186 pp. 

Breman, H. 1998. Soil fertility improvement in Africa, 

a tool for or a by-product of sustainable production? 

African Fertilizer lvIarket 11:2-10. 

Carter, S.E., Chidiamassamba, A., Jeranyama, P., Mafugidze, 

B., Maleleka, G.P., Mvena, Z., Mudhara, 

M., Nabane, N., van Oosterhout, S.A.M., Price, L., 

Sithole, N. and Dhalimi, R 1993. Soil fertility management 

in Mutoko Communal ArPa, Zimbamue. Report 

of a field exercise, August 12 - September 3 

1992. TSBF, Nairobi, Kenya. 75 pp. 

Cook. c.c. and M. Grut. 1989. Agroforestry In Sub 

Saharan Africa: A Farmer's Perspective. World 

Bank Technical Paper No. 112, Washington D.C., 

USA. 94 pp. 

Giller K.E. and Cadisch, G. 1995. Future benefits from 

biological nitrogen fixation: An ecological approach 

to agriculture. Plant and Soil 174:255-277. 

Giller, K.E. 2001. Nitrogen Fixation in Tropical Cropping 

Systems. 2nd ed. CAB International, Wallingford, 

UK. 458 pp. 

Giller, K.E. and Wilson, K.J. 1991. Nitrogen Fixation in 

Tropical Cropping Systems. CAB International, Wallingford, 

UK. 313 pp. 

Grant, P.M. 1981. The fertilization of sandy soils in 

peasant agriculture. Zimbamue Agricultural Journal 

78:169-175. 

Kwesiga, F., and Coe, R 1994. The effect of short 

rotation Sesbania sesban planted fallows on maize 

yields. Forest Ecology and Management 64:199-208. 

Mapfumo, P. and Giller, K.E. 2001. Soil Fertility 

Management Strategies and Practices by Smallholder 

Farmers in Semi-Arid Areas of Zimbamue. Bulawayo, 

Zimbabwe: International Crops Research 

Institute for the Semi-Arid Tropics (ICRISAT) 

with permission from the Food and Agriculture 

Organization of the United Nations (FAO). 60 

pp. 

Mapfumo, P. Giller, K.E. Mpepereki S. and Mafongoya, 

P.L. 1999b. Dinitrogen fixation by pigeonpea 

of different maturity types on granitic sandy soils 

in Zimbabwe. Symbiosis 27:305-318. 

Mugwira, L.M. and Murwira, H.K. 1997. Use ofCattle 

lvIanure to Improve Soil Fertility in Zimbamue: Past 

and Current Research and Future Research Needs. Network 

Working Paper No. 2. Soil Fertility Network 

for Maize-Based Cropping Systems in Zimbabwe 

and Malawi. CIMMYT, Harare, Zimbabwe. 33 pp. 

Quinones, A.M., Borlaug, N.E. and Dowswell, C.R 

1997. A fertilizer-based green revolution for Africa. 

In: Replenishing Soil Fertility in Africa. RJ. Buresh 

and P.A. Sanchez (Eds.). SSSA Special Publication 

51. SSSA, Madison, WI, USA. pp. 81-96. 

Scoones, L, Chibudu, c., Chikura, S., Jeranyama, P., 

Machaka, D., Machanja, W., Mavedzenge, B., 

Mombeshora, B., Mudhara, M., Mudziwo, c., 

Murimbarimba, F. and Zirereza, B. 1996. Hazards 

and Opportunities.' Farming Livelihoods in Drylanil 

Africa: Lessons from Zimbamue. Zed Books Ltd, 

London and New Jersey, in association with International 

Institute for Environment and Development, 

London, UK. 267 pp. 

Snapp, S.5., Mafongoya, P.L. and Waddington, S. 

1998. Organic rna tter technologies for integra ted 

nutrient management in smallholder cropping 

systems of southern Africa. Agriculture, Ecosystems 

and Environment 71:185-200. 

Waddington, S.R, Murwira, H.K., Kurnwenda, J.D. 

T. Hikwa, D. and Tagwira, F. (Eds.). 1998. Soil 

Fertility Research for Maize-Based Farming Systems 

in Malawi and Zimbamue; Soil Fert Net and CIM 

MYT-Zimbabwe, Harare, Zimbabwe. 312 pp. 

74 


Abstract 

SCREENING OF SHORT DURATION PIGEON PEA IN MATABElElAND 

BONGANI NeuBE, TAFADZWA MANJALA and STEVE TWOMLOW 

feR/SA T-Bu/awayo, P. O. Box 776, Bu/awayo, Zimbabwe 

Pigeonpea is a new crop fer the semi-arid tropics of Zimbabwe that offers substantial benefits to rural households. These 

benefits include soil fertility enhancement, improved household nutrition and income diversification. To identify adaptable 

varieties for farmers, ICRISAT-Bulawayo conducted screening trials for 10 short duration determinate and 12 

short duration indeterminate pigeonpea varieties on heavy clay (Matopos site) and sandy (Lucydale site) soils during 

the 2001/2002 rainy season at Matopos Research Station. 

At each site, pigeonpea varieties were planted in a randomized block design with three replicates. The crops were planted 

on 10 (Matopos) and 11 (Lucydale) December 2001. Agronomic data included date of planting and emergence, days to 

50% flowering and maturity, plant height, grain yield and woody biomass yield. 

Total rainfall for the season (October 2001 to June 2002) was close to the long-term average, with 537 mrr. at the Matopos 

site and 427 mm at Lucydale. However, it was poorly distributed, and crops effectively received 124 mm at Matopos 

and 133 mm at Lucydale from the date they were planted until harvest. Generally, varieties planted on the clay outperformed 

those planted on the sand because of a more favorable soil water balance. Determinate pigeonpea planted at 

Matopos flowered and matured earlier and on average yielded more grain compared to the crop at Lucydale. The earliest 

maturing determinate varieties on both soils (average 114.5 days on sand) were ICPL 86012 and ICPL 87105. However, 

the grain yield for these two varieties (744 kg/ha and 841 kg/ha respectively) was lower than the highest yielding determinate 

varieties, ICEAP 00781 (1058 kg/haY, ICEAP 00535 (902 kg/luz) and ICEAP 00536 (849 kg/ha), which matl/red 

at about 116 days. The earliest maturing indeterminate varieties for both soils were ICPL87091 (114 days) and 

ICEAP00718 (140 days). For the indeterminate types, earliness was associated with highest grain yield ICPL87091 

(829 kg/ha), ICEAP 00721 (682 kg/ha) and ICEAP 00718 (641 kg/ha), but these were much lower yields than from the 

best performing determinate varieties. 

Key words: Short duration pigeonpea, early maturity, high grain yield, screening, benefits 

Introduction and Background 

Pigeonpea (Cajanus cajan (L.) Millsp.) is a multipurpose 

drought tolerant grain legume crop (Kimani et 

al., 1994) that offers substantial benefits to many 

smallholder farming systems that dominate the 

semi-arid tropics (SAT) of Africa and Asia. These 

benefits include improved household nutrition, fodder 

and browse, soil fertility enhancement, and income 

diversification (Holden, 1993; Karachi and 

Zengo, 1998; Mapfumo et ai, 1998). 

Over the last 30 years, the area cultivated to pigeonpea 

has increased substantially, as have the countries 

that now produce it. Traditionally, pigeonpea 

production in the semi-arid tropics was principally 

confined to the Indian subcontinent. However, over 

the last 25 years the efforts of the International 

Crops Research Institute for the Semi-Arid Tropics 

(ICRlSAT) and oLher organizations have ensured 

that the crop is now widely grown in the SAT of 

Asia, Africa and the Caribbean under a wide variety 

of cropping systems (Kollipara et ai, 1994). In Africa, 

considerable amounts of pigeonpea are produced 

in Kenya, Malawi and Zambia, with varieties 

such as ICP 9145 having been cultivated in Malawi 

since 1987 (Reddy et ai, 1993). Pigeon pea is also an 

important grain legume crop for smallholder farmers 

in Tanzania (Mligo et ai, 2001). Despite the increased 

cultivated area in many countries of sub 

Saharan Africa, few studies have been done on the 

adaptability of pigeonpea in Zimbabwe, through 

efforts by the Department of Research and Specialist 

Services in the late 1980s. Those that have been 

done were confined to the higher rainfall areas of 

the country. Dzowela et al (1995, 1997) carried out 

trials on pigeonpea to establish its potential as livestock 

fodder in Domboshava and Makoholi Zimbabwe 

whilst Mapfumo et al (1998) investigated the 

potential contribution of pigeonpea to soil fertility 

in Domboshava and Murewa communal lands during 

the 1996/97 cropping season. Consequently, 

ICRISAT-Bulawayo conducted screening trials for 

10 short duration determinate and 12 short duration 

indeterminate pigeon pea varieties on heavy clay 

(Matopos) and sandy (Lucydale) 'Soils during the 


2001/2002 rainy season. Short duration varieties 

were selected to suit the weather conditions of the 

region that is characterized by low and unpredictable 

rainfall patterns. This parer summarizes results 

for the 2001/2002 season. 


Site Characteristics 

Matopos. The Matopos site is 1344 m above sea 

level, with flat land on the lower slope within the 

Matopos Hills. The soil at the site is classified as the 

Matopos 3E.4 using the Zimbabwe Classification 

system (Pellic-Eutric Vertisol), which describes an 

imperfectly drained vertisol derived from igneous/ 

metamorphic rocks. The soil is deep (l30 cm+) and 

is water-saturated for short periods every year. It is 

mostly made of black clay with a granular structure 

in the top layers. This soil is fertile with 100% base 

saturation and a high cation exchange capacity 

(CEC) (Moyo, 2001). Prior to the 2001/2002 cropping 

season the land had been planted to millet 

breeding trials that had received 300 kg ha- I basal 

Compound D fertilizer. 

Lllcydale. The Lucydale site is l378 m above sea 

level, located on a gently undulating plain. The soil 

type at the site is described as a Eutric Arenosol 

5G.2 (Zimbabwe Classification) that is a typical 

moderately deep to deep well-drai.'1ed fersiallitic 

soil derived from granite. The soil is excessively 

well drained. The tops layers of the soil are coarsegrained 

sand and loamy sand with apedal structure. 

The soil has a slightly acidic pH 5 in the top layers. 

This soil is less fertile compared to the Matopos soil 

and has a base saturation of 60% in the top layer 

(Moyo, 2001). 

Experimental Design 

At each site, 10 determinate and 12 indeterminate 

varieties were planted. The selected fields at both 

Matopos and Lucydale were previously under millet 

breeding trials. Prior to planting the pigeonpea, 

both sites received a basal dressing of 300 kg ha- i 

Comp--ound D and the soil was ploughed using a 

tractor drawn disc plough to a depth of 0.20 m. 

The design was a complete randomized block design 

with three replicates. The determinate pigeonpea 

plot size was 5.0 m x 2.0 m, with a between row 

spacing of 0.5 m and within row spacing of 0.2 m. 

The indeterminate plots were 5.0 m x 3.0 m in size 

with a row spacing of 0.75 m and plant-to-plant 

spacing of 0.3 m. These are recommended plant 

densit!~s for the Matopos elevation. 

The Matopos plots (both determinate and indeterminate) 

were hand planted on 10 December 2001 

and the Lucydale plots were planted on 12 December 

2001. Four seeds were planted per station about 

5 cm deep to !Ilaximize germination. 

Crop Management and Records of Observations 

All plots were thinned to one plant per planting station 

three weeks after planting the two sites. Weeding 

was done using hand hoes, as weed pressure 

dictated. A single spraying of Dimethoate was carried 

out at both- sites on 12 March to destroy leaf 

eaters. 

Observations were done every day and records 

were taken for date to 50% flower, date to 75% maturity, 

plant height, final plant stand, harvest date, 

woody biomass yield and grain yield. 

The crop at Matopos was affeCted by frost and both 

the indeterminate and determinate crops wilted. At 

the Lucydale site the crops received a further 20 mm 

of rainfall in June and all the varieties started flowering 

again. The indeterminate varieties could not 

reach maturity due to destruction by animals at the 

podding stage. 

A one way Analysis of Variance (for randomized 

blocks) was performed on woody biomass and 

grain yield using Genstat for Windows (5 th Edition). 


Rainfall Data 

Total rainfall for the season, October 2001 to June 

2002 (Table I), was close to the long-term average 

with 536.5 mm at the Matopos site and 427 mm at 

Lucydale. However, it was poorly distributed, and 

the determinate varieties received effectively 124 

mm at Matopos and 133 mm at Lucydale. Given the 

moist condition of the seedbeds at planting at both 

sites, germination was good. 

Responses of Determinate Pigeon Pea Varieties 

Short duration determinate pigeonpea varieties 

flowered approximately two weeks earlier at the 

Matopos heavy· clay site, when compared to the 

sandy Lucydale site, despite that planting was only 

one day apart (Table 2). Despite the earliness of 

flowering at Matopos, a similar length of time was 

required for the pigeonpea varieties to reach 

physiological maturity; typically 116 days from 

planting. However, the early flowering at Matopos 

Table 1. Total rainfall (mm) at ICRISAT Research fields from October 

2001 to April 2002. 

Season longterm 

Month OctOl NovOl OecOl Jan02 Feb02 Mar02 Apr02 Total mean 

Matopos 64 108 134 44 12 2 134 536.5 590 

lucydale 41 132 112 41 16 3 46 427 

76 


Table 2. Characteristics of short duration determinate pigeon pea varieties grown during the 2001/2002 For the indeterminate 

season at Matopos and lucydale. 

varieties, early ma turity 

Days to 50% flowering Days to 75% maturity Grain yield kg ha I Biomass kg ha I 

was associated with 

Site/variety Matopos Lucydale Matopos Lucydale Matopos Lucydale Matopos Lucydale 

highest grain yield. Variety 

ICPL87091 (829 kg/ 

Var 1 IGEAP 00360 61 77 118 118 739.3 504.0 7383.9 6329.9 

ha), ICEAP 00721 (682 

Var 2 IGEAP 00394 61 71 118 118 746.7 343.3 6538.9 8521.7 

kg/ha) and ICEAP 00718 

Var 3 IGEAP 00535 61 73 115 114 902.0 614.0 5201 .0 4468 .9 

(641 kg/ha) were the 

Var 4 IGEAP 00536 61 77 115 115 848.7 634.7 5224.4 4555 .6 highest yielding indeter 

Var 5 IGEAP 00753 61 77 116 118 705.3 360.7 6642.9 5785.8 minate varieties, but 

Var 61GEAP 00781 61 75 117 118 1058.0 370.7 5815 .6 5129.0 these gave lower yields 

Var 7 IGPL 86012 61 73 115 114 744.0 458.0 4365.7 4177.3 than the best performing 

Var 81GPL 87091 61 77 115 118 773.3 398.0 6114.3 4909.1 determinate varieties. 

Var 9 IGPL 87105 61 71 115 114 840.7 592.7 4109.6 4668.1 The average for the inde 

Var 10 IGEAP 93027 61 77 118 115 749.3 

Site mean 61 75 116 116 810.7 

SED 0 3 0.901 0.870 172.3 

had a major influence on yields. The pigeonpea 

yields at Mqtopos averaged 810.7 kg ha- I compared 

with 482.3 kg ha- I for Lucydale. There was no statistical 

difference between varieties at both sites. 

Variety ICEAP 00394 yielded significantly (P>O.OOl) 

more woody biomass than other varieties at Lucydale, 

and performed similarly to the other varieties 

at Matopos. However, it had the lowest grain legume 

yield of any variety on the sands at Lucydale 

and gave one of the lowest yields at Matopos. Varieties 

ICEAP 00535 and ICEAP 00536 performed 

above average at Matopos and significantly 

(P

The indeterminate varieties that can be tried are 

rCPL 87091, ICEAP 00721 and ICEAP 00718. The 

indeterminate varieties however need to be planted 

early if they are to reach full maturity under the low 

rainfall conditions in Matabeleland. 

Woody biomass (with leaves dropped) yield can be 

used as a rough indicator of the fodder yield potential 

of the pigeonpea varieties. Promising determinates 

are ICEAP00394 and ICEAP 00360. The indeterminate 

varieties that can be tried include ICEAP 

00728 and ICEAP 00722. These recommendations 

are tentative because several abnormalities occurred 

during the 2001/2002 season at Matopos where the 

trials were carried out. 

The 2001/2002 rainy season had poorly distributed 

rain. Therefore it is difficult to predict what would 

have happened under a normal season. Pigeonpea 

is known to be prone to Fusarium wilt but the disease 

was not observed during the season. Some IC 

RISAT lines are tolerant to the disease, like ICP 9145 

and ICEAP 40 which is commercially grown in Malawi. 

Reco m mendations 

This trial was able to give information about the 

short duration pigeopnpea varieties that can be 

grown in Matabeleland. There is need to determine 

how the crops perform in a more normal season in 

which rainfall is evenly distributed. There is also 

need to further expand the studies to include soil 

fertility benefits, fodder yield and other benefits 

such as the use as firewood. There is also need to 

explore the market for the crop in Matabeleland. 

Tsholotsho farmers were shown the crop and they 

expressed an interest to grow it, but only if a market 

for the crop could be found. The drive to promote 

pigeonpea should also include showing the farmer!, 

how to utilize it as a food crop. 

References 

Ozowela, B., Holden, S.T., 1993. The potential of 

agroforestry in the high rainfall areas of Zambia: 

A peasant programming model approach. Agroforestry 

Systems 24:39-55. 

Kimani, P.M., Benzioni, A., and M. Ventura 1994. 

Genetic variation in pigeonpea (Cajanus cajan 

(L.) Mill sp.) in response to successive cycles of 

water stress. Plant and Soil 158:193-201. 

Karachi, M., and Zengo, M., 1998. Legume forages 

from pigeonpea, leucaena and sesbania as supplements 

to natural pastures for goat production 

in western Tanzania. Agroforestry Systems 39:13 

21. 

Kollipara, K.P., L. Singh and T. Hymowitz 1993. Genetic 

variation and chymotrypsin inhibitors in 

pigeonpea [Cajanus cajan (L.) Millsp.] and its 

wild relatives. Theoretical and Applied Genetics 

88:986-993. 

Mapfumo, P., Mpepereki, S., and Mafongoya, P., 

1998. Pigeonpea in Zimbabwe: A new crop with 

potential. In: Waddington, S.R., H.K., Murwira, 

J.FO.T. Kumwenda, O. Hikwa and F. Tagwira 

(eds). Soil Fertility Research for Maize-Based Farming 

Systems in Zimbabwe. Proceedings of the Soil 

Fert Net Results and Planning Workshop held 

from 7 to 11 July 1997 at Africa University, Mutare, 

Zimbabwe. Soil Fert Net and CIMMYT 

Zimbabwe, Harare, Zimbabwe. pp. 93-98. 

Mligo, J .K., Myaka, F.A., Mbwaga, A., and Mpangala, 

B.A., 2001. On station research, technology 

exchange, and seed systems for pigeonpea in 

Tanzania. In: SHim. S.N., Mergeai, G., and Kimani 

P. (eds). Pigeonpea Status and Potential in 

Eastern and Southern Africa. Gemblox, Belgium: 

Gembloux Agricultural University; and 

Patancheru, AP, India: International Crops Research 

Institute for the Semi-Arid Tropics. 

Moyo, M., 2001. Representative Soil Profiles of ICRI 

SAT Research Sites. Chemistry and Soil Research 

Institute, Zimbabwe, Soils Report No. A666. 

Reddy, M. V. Nene, Y. L., Raju, T. N., Kannaiyan, J., 

Remanandan, P., Mengesha, M. H. and Amin, K. 

S., 1995. Registration of pigeonpea germplasm 

line ICP 9145 resistant to fusarium wilt. Crop Science 

35(4):1231. 

78 


RISK DIVERSIFICATION OPPORTUNITIES THROUGH LEGUMES IN 

SMALLHOLDER FARMING SYSTEMS IN THE 

SEMI-ARID AREAS OF ZIMBABWE 

RICHARD FOTI, JOSEPH RUSIKE and JOHN DIMES 

feR/SA T-Bu/a wa yo, PO Box 776, Matopos Research Station, Bu/awayo, Zimbabwe 

Abstract 

This paper uses a simulation modeling approach to evaluate the long-term diversification gains and risks associated with 

adoption of a range of fertility options, including legumes, manure, and small doses of inorganic fertilizer in semi-arid 

areas. These options were tested by the Department of Agricultural Research and Extension (AREX), the International 

Crops Research Institute for the Semi-Arid Tropics (ICRISAT), the Tropical Soil Biology and Fertility Programme 

(TSBF), the International Maize and Wheat Improvement Center (CIMMYT) and the Smallholder Dry Areas Resource 

Management Project (SDARMP) in farmer participatonj research trials during the 1999/2000 and 2000/2001 cropping 

seasons in pilot areas in three semi-arid regions in Zimbabwe. The study tests the hypothesis that legume-based soil fertility 

technologies wi(l benefit farmers if diversification into legumes complements farmers' current investments compared 

to alternative investments and the demand on resources is within the boundaries of the resource-endowments of 

the fanners. 

Results indicate that maize-cowpea and maize-groundnut rotations and maize-pigeon pea intercrops and rotations are 

good investment opportunties for diversification from the traditional maize and sorghum soil-mining practices currently 

being pursued by the majority of farm households. 

Key words: Legumes, intercropping, rotation, risk simulation, diversification, return on investment 

Introduction 

During the past decade, there has been growing interest 

in the use of legume-based technologies as 

nutrient sources in smallholder farming systems in 

Sub-Saharan Africa because of constraints on expanded 

use of inorganic fertilizers. Historically, 

legumes were grown as intercrops with cereals, especially 

during pre-colonial times. In Zimbabwe, 

agricultural extension has discouraged intercropping 

in the pasf5() years and encouraged farmers to 

grow pure crops targeted at commercial markets. 

Despite this advice, farmers have continued to grow 

legwnes intercropped with cereals albeit in small 

areas. To re-introduce legumes into the system at 

large enough scale to enable farmers to capture potential 

benefits of biological i'h-fixation (BNF), the 

legume technologies need to give a competitive rate 

of return on investment compared to alternative investment 

options available to households, meet 

farmers' requirements for risk, and fit within the 

boundaries of resource er~dowments of smallholders. 

This paper uses a simulation modeling approach to 

evaluate the long-term diversification gains and 

risks associated with adoption of a range of soil fertility 

options, including legumes, animal manure, 

small doses of inorganic fertilizer and water management. 

These were tested by the Department of 

Agricultural Research and Extension (AREX), the 

International Crops Research Institute for the Semi 

Arid Tropics (ICRISAT), the Tropical Soil Biology 

and Fertility Programme (TSBF), the International 

Maize and Wheat Improvement Center (CIMMYT) 

and the Smallholder Dry Areas Resource Management 

Project (SDARMP) in farmer participatory research 

during the 1999/2000 and 2000/2001 cropping 

seasons in pilot areas in the semi-arid regions 

in Zimbabwe. 

Objectives 

The general objective of the study was to assess the 

potential for adoption of legume-based soil fertility 

improvement technologies. The specific objectives 

are to: 

o Estimate expected profitability and riskiness of 

alternative legume technologies 

o Determine the benefits offered by legume-based 

soil fertility management tedmologies through 

diversification to households with different resource 

endowments and risk preferences. 

Research Approach: Theory and Methods 

The conceptual framework used for guiding the 

study is derived from portfolio choice theory. Portfolio 

theory provides analytical tools and methods 

for analyzing farmers' decision-making under risk 


and how best to invest a given bundle of resources 

among several alternatives while minimizing the 

risk of their portfolios. 

Conceptual Framework and Hypotheses 

For farmers to invest resources in new technologies 

such as organic and inorganic fertilizers, the rates of 

return on these investments need to be competitive 

with returns on available alternative investment 

opportunities. Within a single crop enterprise such 

as maize production, the return on investment in 

new fertilizer technologies is compared with the 

return on capital using traditional soil-mining 

methods. Between different crop enterprises, 

returns on investment in new management 

technologies are compared against other 

investments including livestock, off-farm activities, 

temporary and permant labor migration. Because 

of the uncertainty of payoffs to alternative 

investments, new investments must fit farmers' 

objectives and requirements for risk in addition to 

profitability. Because high returns are associated 

with high risks, farmers face trade-offs between 

allocating resources to activities with high profits 

and more desirable but riskier and therefore less 

attractive outcomes compared to activities with low 

profits and yet less riskier, which makes them more 

appealing. 

Moss, Weldon, and Featherstone (1991) have developed 

an approach for evaluating risk-return tradeoffs 

and diversification opportunities of alternative 

farm investments that is based on the portfolio theory 

of financial markets and the capital asset pricing 

model (CAPM). These analysts argue that for a new 

or candidate enterprise to improve the risk-return 

tradeoff provided by any existing group or portfolio 

of enterprises, the following condition must hold 

If the mean return divided by the standard deviation 

of new investment alternative is greater than 

the mean of the whole farm plan's return divided 

by the standard deviation times the correlation between 

investments the new investment will complement 

the current operation from a risk-return perspective. 

Based on this risk-return condition necessary 

for a new enterprise to improve the risk-return 

trade-off provided by the current farm's portfolio, 

one can calculate a risk diversification index (RDI) 

as follows: 

r r 

RDI _i _.-L n 

r'ip 

a: I 

a: I 

If the RDI is greater than zero, the candidate technology 

is a good investment opportunity for diversification. 

But if RDI is less than zero the candidate 

technology offers no gains through diversification 

to the farm household. 

Applying this conceptual framework generates the 

following two hypotheses about relationships between 

the ~'isk-return characteristics of new technologies, 

farmers' risk management strategies, and 

potential for adoption that are tested in the study: 

l. Legume-based soil fertility management technologies 

are attractive if they lie on the riskefficient 

frontier and offer farmers expected returns 

that are high enough to compensate them 

for additional risks; and 

2. Diversification into legume-based soil fertility 

management technologies will benefit farmers if 

the new technologies complement the current 

farm portfolio and better offset the total risk of 

the whole farm compared to allocating resources 

to alternative farm and non-farm investment 

opportunities available to farmers. 

Where 

O'i 

= the meim return of the potential new investment 

alterna tive, 

= the standard deviation of the new alternative, 

= the mean return of the current whole farm 

r" plan, 

CY p 

= the standard deviation of the current whole 

farm plan, and 

= the correlation between the current whole 

farm plan's return and the new enterprise's 

return. 

Methods 

The study uses enterprise and whole-farm budgeting, 

and simtllation modeling with APSIM and 

@RISK to evaluate the hypotheses. Enterprise 

budgets are constructed for alternative investment 

technologies. Different enterprises are defined by 

different outputs such as maize, sorghum, chickens, 

goats and cattle; sole stands and crop mixtures; and 

different crop prod uction technologies, including 

low and high rates of application of kraal manure, 

pit-treated manure, inorganic Nitrogen fertilizer 

and organic and inorganic fertilizer combinations. 

Enterprise budgets are used to compare the profitability 

of maize and sorghum crop production using 

traditional methods and improved soil fertility 

80 


management technologies with different activities aggregating the returns per unit over the number of 

available to farmers and the profitability of different units produced by 'the households. Because there 

~nterprises for households with different resource are expenses and 'revenues that cannot be allocated 

endowments. The budgets are constructed using to particular enterprises, and cases where we do.not 

yield and input-output coefficients data from farm have a linear budget this may under~stimate resurveys, 

Farmer Participatory Research experi turns to some activities. The budgets are used to esments, 

and yields predicted by the Agric:dtural Pro timate the expected annual returns, the degree of 

duction Systems Simulator (APSIM) model. 'D'e risk, and correlation coefficients for different crop 

input-output coefficients are combined with prices and livestock combinations. 

from the Ministry of Agriculture, Zimbabwe Farmerst 

Union and Commercial Farmers' Union to calculate 

gross margins per hectare for crops and per 

breeding animal unit for livestock. Input prices are 


reported at the supply point. Input prices paid by Tables I, 2, and 3 present the annual returns and 

farmers are estimated by adding input prices re risks per hectare above fixed costs for 11 years from 

ported by suppliers and the cost of transport. Out 1990/91 to 2000/01 on alternative maize and sorput 

prices are reported at the marketing point. ghum soil fertility production technologies by farm 

Farm gate prices are estimated conservatively by household typology. The tables also include the andeducting 

cost of transportation from prices at mar nual return and risk of investing funds in a risk-free 

keting points. The opportunity cost of family labor asset, the POSB savings account. For male-headed 

is estimated by multiplying the minimum wage rate households with resident husband and higher labor 

of engaging in urban employment multiplied by the and draft animal resource-endowments, the most 

probability of finding a job. 

profitable production technologies are, indecreasing 

order of importance, maize plus 9 kilograms of 

Because APSIM crop yield predictions are only ritrogen per hectare, maize plus kraal manure plus 

available for 10 year,s from 1990 to 2000 and for a 18 kilograms of nitrogen per hectare, maizefew 

improved technology options this analysis focuses 

on sole maize and sorghum 

Table 1. Expected annual returns (Zimbabwe $/ha) and risk of alternative maize and sorghum 

grown without fertilizer and with 

soil fertility management technologies for male·headed households, Gwanda, Tsholotsho and 

small quantities of Nitrogen fertilizer; 

kraal and pit manure; manure 

Zimuto, 1990/91,2000/01 

Gwanda Tsholo'tsho Zirnuto 

and, Nitrogen fertilizer combinations; 

sole cowpeas and groundnuts; 

Activity 

Return Risk Return Risk Return Risk 

maize and sorghum-cowpea 

and groimdnut intercrops and rotations, 

maize-cowpea and maizegroundnut 

rotations. The budgets 

include only the physical grain output 

of crop enterprises for primary 

and secondary crops valued at 

pose savings account 

Sorghum+kraal manure 

Sorghum +Okraal manure +ON 

Maize+Okraal manure+ON 

Sorgirum+9N 

Maize +groundnut intercrop 

Sorghum+groundnut intercrop 

289 

·15272 

-2440 

3286 

-849 

1738 

-3486 

328 

3948 

4350 

4526 

5856 

6302 

6403 

289 

·13793 

·156 

5046 

1911 

1260 

·43,83 

328 

5310 

6347 

5759 

7642 

8213 

8872 

289 

4249 

·2285 

328 

7688 

5969 

farm gate prices. Farmers frequently 

produce crops in mixtures 

of more than two crops and the 

budget needs to include the whole 

mixture. The values of byproducts 

such as stalks, which have value as 

livestock feed, in construction and 

composts are not included. Livestock 

budgets include the market 

value of the animal, depreciation 

on the value of breeding animals, 

milk, eggs, and draught power. 

Values of animal manure and tradi 

'tiona I religious ceremonies are not 

included. The enterprise budgets 

are used to construct whole farm 

budgets for current farm plans for 

different household categories by 

Sorghum +pit manure -1888 6752 542 8752 

Sorghum·groundnut rotation 554 7033 2914 8054 

Sorghum+cowpea rotation ·319 7422 8769 11024 

Maize+9N 

4286 7450 7188 7160 6860 7913 

Maize groundnut rotation 3888 7535 6800 8902 5764 7862 

Maize +cowpea rotation 3412 7545 12743 10542 8591 10656 

Sorghum +18N -2330 7703 788 9718 

Sorghum+ ~raal +18N ·154 7759 3314 10160 

Maize +kraal manure +9N 2629 8296 6547 7984 6127 8498 

Maize+pit manure 2414 8723 6417 7768 6281 9q2 

Maize+ cowpea intercrop -1179 9006 6875 10783 168 9807 

Maize +kraal manure 3717 9220 5557 9774 4303 9785 

Sorghum +cowpea intercrop ·5177 9306 3498 13359 

Sorghum +kraal manure +9N -450 9765 2711 11107 

Maize+ 18N 1614 9804 5588 8436 6081 10255 

Maize +kraal manure +18N 3894 10286 8293 8874 8951 10480 


estimated for the soil fertility man 

agement investment options for the 

farm households. 

Table 2. Expected ~nnual returns (Zimbabwe $/ha) and risk of alternative maize and sorghum 

soil fertility management technologies for de facto female·headed households, Gwanda, 

Tsholotsho and Zimuto, 1990/91·2000/01 

Gwanda 

Tsholotsho 

Activity Return Risk Return Risk 

POSB savings account 289 328 289 328 

Sorghum +kraal manure ·12801 3654 ·10128 4911 

Sorghum +Okraal manure +ON 62 4366 3177 6334 

Maize+Okraal manure+ON 3378 4422 5042 5965 

Sorghum+9N 2203 6042 4874 7642 

Maize+groundnut intercrop 2007 6352 1417 8479 

Sorghum +pit manure 344 6435 4908 8839 

Sorghum +groundnut intercrop ·1151 6542 ·843 8883 

Maize+9N 5591 6699 6657 7505 

Sorghum +groundnut rotation 1886 7039 5374 8204 

Sorghum +cowpea rotation 1509 7264 11369 11236 

Maize cowpea rotation 4456 7470 13524 10753 

Maize +groundnut rotation 4422 7564 7563 9056 

Maize +kraal +9N 4505 7767 5937 7952 

Sorghum +18N 697 7863 3548 9749 

Sorghum +kraal manure +18N 2873 7911 6205 10293 

Maize +pit manure 3884 8166 5791 8155 

Maize +cowpea intercrop -482 9005 7217 11232 

Sorghum +cowpea intercrop ·1860 9220 8243 13643 

Maize+ 18N 3491 9286 4625 8780 

Maize +kraal manure 4031 9334 4946 9764 

Maize +kraal +18N 5770 9611 7682 8839 


groundnut rotation, and maize cowpea rotations. 

The ranking is similar in the lower rainfall sites 

(Gwanda and Zimuto) but in wetter sites 

(Tsholotsho) groundnut and cowpea rotation give 

the highest expect annual returns. For de facto female-headed 

households with intermediate resource 

endowments but better access to off-farm 

cash, maize plus kraal plus 18 kilograms of nitrogen, 

maize plus 9 kilograms of nitrogen, and maize 

plus kraal plus 9 kilograms of nitrogen give the best 

returns followed by maize cowpea and maizegroundnut 

rotation in the drier sites. But the maizecowpea 

rotation, sorghum-cowpea rotation, and 

sorghum-cowpea intercrop gives the best returns in 

the wetter sites. For the de jure households who are 

most resource-constrained, the most profitable technologies 

are the same as for the de facto households 

for drier and wetter sites although the legume rotations 

and intercrops are more profitable for de jure 

households compared to de facto households across 

all sites. For all household categories, higher ex 

!Jected returns are associated with higher risks and 

lower expected returns with lower risks irrespective 

of site. This shows that the Capital Asset Pricing 

Model approximates the risk-return characteristics 

Zimuto 

Figures I, 2, and 3 show trade-offs 

Return Risk 

between expected returns and risks 

289 328 

of the alternative soil fertility production 

technologies. The dominating 

technologies are reflected by 

3200 7619 the set of points on the frontier. 

Technology investments that lie 

·11012 7138 

inside the frontier can be eliminated 

from further analysis as inferior 

because households can choose 

from among better options on the 

·1802 

6449 

22751 

10686 

frontier that give higher expected 

returns for the same level of risk. 

The efficiency frontier for maleheaded 

households in Gwanda includes, 

in increasing order of risks 

427 8424 and r,eturns, POSB savings account, 

·1924 22736 sorghum plus kraal manure, traditional 

maize production technology 

without manure and nitrogen fertilizer, 

traditional sorghum, maize 

4974 9054 

plus 9 kilograms of nitrogen per 

·896 9912 

hectare, maize plus pit manure, 

maize plus kraal manure plus 9 

·10118 40691 kilograms of nitrogen, maize plus 

3170 9678 18 kilograms of nitrogen, and 

·6049 39594 

maize plus kraal manure plus 18 

kilograms of nitrogen. The frontier 

for male-headed households in 

Tsholotsho is dominated by POSB savings account, 

traditional maize, maize plus 9 kilograms of nitrogen, 

maize-cowpea rotation, maize plus kraal manure 

and maize plus kraal manure plus 18 kilograms 

of nitrogen. Zimuto male-headed households 

have a smaller available set of risk-efficient 

technologies: POSB savings account, traditional 

maize, maize-groundnut rotation, maize plus 9 kilograms, 

and maize plus kraal manure plus 18 kilograms 

of nitrogen. In contrast, the frontier for de 

facto female-headed households in Gwanda includes 

POSB savings accounts, traditional maize 

technology, maize plus 9 kilograms of nitrogen, and 

maize plus kraal manure plus 9 kilograms of nitrogen. 

For Tsholotsho de facto female-headed households, 

the frontier consists of POSB savings account, 

traditional maize, maize plus pit manure, maize 

plus kraal manure plus 18 kilograms nitrogen, 

maize-groundnut, sorghum-cowpea, and maizecowpea 

rotations. The risk-return frontier available 

for de facto households in Zimuto comprises POSB 

savings accounts, traditional maize, maize plus pit 

manure and maize-cowpea rotation. The frontiers 

for de jure female-headed households are similar to 

those for de facto female-headed households across 

82 


Table 3. Expected annual returns (Zimbabwe $/ha) and risk of alternative rnaize and sorghum 

the three sites although cowpea 

soil fertility management technologies for de jure female·headed households, Gwanda, 

dominates the efficient sets for de 

Tsholotsho and Zimuto, 1990/91·2000/01 

jure compared to de facto house 

Gwanda Tsholotsho Zimuto holds. We conclude that legumes 

Activity Return Risk Return Risk Return Risk lie on the frontier and dominating 

POSB savings account 289 328 289 328 289 328 set for especially de facto and de jure 

Sorghum +kraal manure ·13023 3668 ·7821 4410 households in Tsholotsho and 

Sorghum +Okraal manure +ON 1704 8129 5419 6018 

Gwanda and therefore are attrac 

Maize+Okraal manure+ON 3692 4422 5375 

tive to farmas. The mix of legume 

5986 5560 7518 

based technologies in the portfolio 

Sorghum+ 9N 1981 5994 7181 7347 

will depend on the tolerance for 

Maize +groundnut intercrop 1976 6364 4721 8216 ·15949 6092 

risk of different households. Be 

Sorghum +pit manure 122 6398 6405 8531 

cause legume technologies are asso 

Sorghum +groundnut intercrop ·1795 6476 3966 8355 ciated with high risks and high re 

Maize+ 9N 5954 6714 7759 7529 ·1997 22645 turns, especially maize-groundnut 

Sorghum +groundnut rotation 1696 7024 7986 7992 rotations, they are likely to be at 

Sorghum +cowpea rotation 2296 7278 12735 11081 tractive to mostly households with 

Maize cowpea rotation 5534 7482 14288 10698 6349 10647 a high tolerance for risk. 

Maize +groundnut rotation 4524 7571 9573 8990 ·1777 7727 

Maize +kraal +9N 4869 7796 7044 7976 

Because different technology in 

·2111 22647 

vestments are differently correlated 

Sorghum + 18N 475 7814 5855 9464 

with the current farm plan and 


farm households can choose to in 

Maize +pit manure 4247 8185 6893 8175 4779 8970 

vest resources among several in 

Maize +cowpea· intercrop 800 9076 8291 11151 ·1108 9812 vestments in order to reduce risk 

Sorghum +cowpea intercrop ·545 9171 10167 13431 without reducing expected returns, 

Maize+ 18N 3855 9310 5727 8796 2983 9632 we need to consider the effects of 

Maize +kraal manure 4394 9397 6053 9785 ·6237 39510 including a production technology 

Maize+kraal+ 18N 6134 9632 8789 8865 ·10313 40600 on the whole farm portfolio when 

Sorghum +kraal manure+ 9N 2355 9926 G197 8470 

deciding whether or not to include 

it in the current farm plan. Tables 

15000 • Gwanaa 4, 5, and 6 report the risk diversification indices for 

11 Tsholotsho 

10000 . 

the whole farm for different household typologies. 

..Zimuto 

The analysis is extended to include the benefits of 

5000 

·M :~" 

.l: 

• •• a 

• 

• • ~ • 

peas and temporary migration to urban labor marc 

~000".· • 10000 15000 

:; -5000 kets. The risk indices for maize-cowpea and maize 

C; • 

a: groundnut . rotations and maize-pigeon pea inter 

-10000 

crops and rotations are positive for different kinds 

-15000 +--- _.o_---- - ---- of households across sites indicating that these legume-based 

soil fertility production technologies of 

-20000 -'-------------- 

Risk ($/ha) 

fer significant gains through diversification. Therefore 

they are likely to be adopted by farmers. 

Figure 1. Risk-return tradeoffs of alternative maize and sorghum 

soil fertility management technologies for male-headed households, 

Gwanda, Tsholotsho and Zimuto, 1990/91-2000/01 

.GNan:l:l 

iii 

..-. .. .... diversifying to medium and long duration pigeon 

~.--------------- 

.. 

1WOO~-~r------------ 

10000 l s~_________ 

i woo ~~~~'------------- 

~ ~~, ~~ 

o ~·-~~~~--~-~--~-~ 

E 

;;" -woo 

. ~ OO:O 

II:: -100Xl -1--_"_ ________----.._ _ 

·1 WOO +--. c-r .... - - --------- 

-20).)) '------ - ------ ---- 

II Tsrobtsro 

A ZimJD 

Figure 2. Risk-return tradeoffs of alternative maize and sorghum 

soil fertility management technologies for de facto female·headed 

households, Gwanda, Tsholotsho and Zimuto, 1990/91·2000/01 

Figure 3. Risk-return tradeoffs of alternative maize and sorghum 

soil fertility management technologies for de jure fem

Table 4. Risk diversification indices of alternative maize and 

s~ghum soil fertility management technologies for male·headed 

households. Gwanda. Tsholotsho and Zimuto. 1990/91·2000/01 

Activity Gwanda ,.shelotsho Zimuto 

Sorghwn+ kraal manure ·4.00 ·2.30 

Sorghum +groundnut intercrop ·0.92 ·1.10 

Sorghwn +cowpea intercrop ·0.91 ·0.40 

Sorghum +18N ·0.69 ·0.55 

Sorgum +pit manure ·0.67 ·0.52 

Sorghum+9N ·0.53 ·0.31 

Sorghum +kraal manure +18N ·0.41 ·0.31 

Sorghum +cowpea rotation ·0.37 0.09 

Sorghum +kraal +9N ·0.36 ·0.26 

Sorghum +groundnut rotation ·0.25 ·0.21 

Maize +cowpea intercrop ·0.14 ·0.23 0.14 

Maize+18N ·0.13 0.02 0.13 

Maize +groundnut intercrop ·0.09 .·0.33 0.09 

Maize +pit manure 0.02 0.28 ·0.02 

Maize +kraal +9N 0.04 0.29 ·0.04 

maize+kraal+ 18N 0.09 0.34 ·0.09 

Maize +cowpea rotation 0.09 0.55 ·0.09 

Maize +kraal manure 0.16 0.27 ·0.16 

Maize +groundnut rotation 0.24 0.26 ·0.24 

Maize+9N 0.31 0.53 ·0.31 

Maize +long pigeonpea intercrop 0.56 0.19 0.14 

maize +medium pigeonpea intercrop 0.61 0.53 0.13 

POSB savings account 0.65 0.74 0.80 

Maize +long pigeonpea rotation 0.89 0.74 ·0.02 

Maize +medium pigeonpea rotation 1.13 0.95 ·0.04 

Urban labor market 7.96 7.57 7.80 


The paper evaluates the attractiveness of alternative 

soil fertility management technologies for adoption 

by farm households with varying resources, and 

risk preferences. Results indicate that maize-cowpea 

and maize-groundnut rotations and maize-pigeon 

pea intercrops and rotations are good investment 

opportunties for diversification with traditional 

maize and sorghum soil-mining practices currently 

being pursued by the majority of farm households. 

Consequently, these legume-based soil fertility 


sorghum soil fertility management technologies for de facto 

female·headed households. Gwanda. Tsholotsho and Zimuto. 

1990/91·2000/01 

Activity Gwanda Tsholotsho Zimuto 

Sorghum+ kraal manure ·3.64 ·1.52 

Sorghum+ groundnut intercrop ·0.84 ·1.20 

Sorghum +cowpea intercrop ·0.81 ·0.62 

Sorgum +pit manure ·0.60 ·0.48 

Sorghum +18N ·0.58 ·0.82 

Maize +cowpea intercrop ·0.44 ·0.76 ·0.38 

Sorgbum +cowpea rotation ·0.35 ·0.16 

Maize +groundnut intercrop ·0.33 ·0.72 ·2.51 

Sorghum +groundnut rotation ·0.32 ·0-'5 

Sorghum+9N ·0.31 ·0.39 

Sorghum +kraal manure+ 18N ·0.30 ·0.55 

Sorghum +kraal +9~ ·0.28 ·0.72 

Maize+18N ·0.26 ·0.60 ·0.18 

Maize +pit manure ·0.09 ·0.25 ·0.10 

Maize +kraal +9N ·0.05 ·0.19 ·0.03 

ma.ize +kraal +18N ·0.03 ·0.18 0.05 

Maize +cowpea rotation ·0.03 ·0.03 ·0.28 

Maize+kraal manure 0.01 ·0.24 ·0.12 

Maize +groundnut rotalion 0.08 ·0.13 ·1.14 

Maize+9N 0.24 0.05 0.16 

Maize+ long'pigeon pea intercrop 0.49 0.52 ·0.01 

POSB savings account 0.. 50 0.30 ·0.22 

maize +medium pigeonpea intercrop 0.53 0.20 0.68 

Maize +long pigeonpea rotation 0.80 0.45 1.18 

Maize +medium pigeon pea rotation 0.99 0.76 1.34 

Urban labor market 8.13 7.42 8.14 

The risk analysis presented in this paper is a first 

cu~ to evaluate technologies that merit further 

study. In addition to meeting requirements for risk 

and return, new technologies must fit with the re 

source boUndaries of farmers and management capabilities. 

Mathematical optimization provides 

tools for a more detailed analysis of the benefits and 

adoption potential of the technologies under the se 

vere resource and institutional constraints faced by 

households in semi-arid areas. 

production technologies are likely to be adopted by 

farmers. Significant gains can also result from 

diversification into non-farm assets such as POSB 

savings accounts and urban employment. 

However, more detailed analysis using 

mathematical programming is needed to evaluate 

the· feasibilibity and sensitivity of options to 

changes in environmental factors. 

Reference 

Moss B.C., Weldon N.R. and Feartherstone A.M., 

1991. A simple approach to evalauting risk 

diversification opportunties. Journal of American 

Society of Farm Managers and Rural Appraisors 

55:20-24. 

84 Grain legumes and Green Manures for Soil Fertility in Southern Africa


sorghum soil fertility management technologies for de jure female· 

headed households, Gwanda, Tsholotsho and Zimuto, 1990/91· 

2000/01 

Activity 

Sorghum +kraal manure 

Sorghum +groundnut intercrop 

Maize +cowpea intercrop 

Sorghum +cowpea intercrop 

Sorghum +18N 

Sorgum +pit manure 

Sorghum +kraal +9N 

Maize +groundnut intercrop 

Sorghum+9N 

Sorghum +kraal manure +18N 

Sorghum +groundnut rotation 

Maize+ 18N 

Sorghum +cowpea rotation 

Maize +kraal manure 

Maize +pit manure 

Maize+kraal+9N 

maize +kraal +18N 

Maize +groundnut rotation 

Maize +cowpea rotation 

Maize+9N 

POSB savings account 

Maize+long pigeonpea intercrop 

maize +medium pigeonpea intercrop 

Maize +long pigeonpea rotation 

Maize+medium pigeonpea rotation 

Urban labor market 

Gwanda 

·3.55 

·0.79 

·0.58 

·0.54 

·0.49 

·0.44 

·0.33 

·0.23 

·0.22 

·0.20 

·0.17 

·0.11 

·0.07 

·0.05 

0.07 

0.09 

0.12 

0.25 

0.30 

0.43 

0.44 

0.51 

0.56 

0.83 

1.08 

4.02 

Tsholotsho 

·1.81 

·0.82 

·0.70 

·0.35 

·0.59 

·0.55 

0.01 

·0.78 

·0.22 

·0.38 

·0.21 

·0.68 

0.06 

·0.60 

·0.44 

·0.39 

·0.31 

0.02 

0.34 

·0.20 

0.72 

0.24 

0.52 

0.83 

0.89 

8.17 

Zimuto 

·0.61 

·3.11 

0.54 

0.44 

0.51 

0.65 

0.78 

·0.81 

·0.13 

0.88 

0.27 

0.51 

0.70 

1.04 

1.09 

6.98 


EVALUATING MUCUNA GREEN MANURE TECHNOLOGIES IN SOUTHERN 

AFRICA THROUGH CROP SIMULATION MODELLING 

Abstract 

ZONDAI SHAMUDZARIRA 

CIMMYT-Zimbabwe, PO Box MP163, Mt Pleasant, Harare, Zimbabwe 

After an examination of opportunities for maintaining and improving soil fertility under smallholder systems, a case 

study using crop growth simulation in conjunction with historical weather data is presented. The potential for an improved 

resource management system that incorporates mucuna (Mucuna pruriens) to improve yields of the following 

maize crop in the drier areas of Zimbabwe is assessed with a crop growth simulation model. In this paper, APSIM was 

configured to simulate maize yields from crops receiving various amounts of N, or from an unfertilised maize crop following 

a mucuna crop. To take account of different levels of risk aversion, the technologies are examined in probabilistic 

terms based on the cumulative distribution junctions of yield from 46 years of simulations. 

The analysis highlights large potential benefits from the use of mucuna as a green manure crop. The reason frequently 

proposed to explain low uptake rates ofgreen manure technologies by smallholder farmers is the loss in maize yield during 

the year when the green manure crop is in thefield. In environments where continuous maize cropping yields between 

150 and 500 kg grain/ha, rotation with mucuna was predicted to give maize yield increments of over 1000 kg/ha 

when the mucuna is harvested at maturity and between 3000-5000 kg/ha when incorporated at flowering. Measured 

maize grain yield increases after mucuna in Malawi are of the order of 100-200%. These results are proving useful for 

Soil Fert Net members who have carried out a lot of research on velvet bean and other green manures for the region in 

recent years. 

Key words: APSIM model, green manure, mucuna, computer simulation, semi-arid zones 

Introduction 

Maize is the primary food crop in Zimbabwe and 

occupies about one-half of the total agricultural 

cropland. In the smallholder sector, maize yields 

are low and variable primarily due to low inherent 

and declining soil fertility (Grant, 1981) coupled 

with low and erratic rainfall. About 70% of Zimbabwe 

is covered with coarse sandy soil derived 

mostly from granite. These soils are low in N, P and 

S and low in nutrient reserves and exchange capacity 

due to low organic matter and clay content. The 

risks associated with arable crop production in 

these areas limits the potential use of high input 

strategies by smallholder farmers. Average maize 

yields of between 1.0 to 1.5 t/ha are common under 

smallholder conditions as opposed to yields of 

around 5.0 t/ha in the large-scale commercial farming 

sector. Mataruka and Whingwiri (1988) identified 

soil moisture stress, poor soil and fertiliser 

management, low plant populations, late planting, 

poor weeding and labour bottlenecks as some of the 

major factors limiting maize productivity under 

smallholder conditions. 

Researchers in Zimbabwe and Malawi under the 

Rockefeller-funded Soil Fertility Network have 

come up with a selection of soil fertility technologies 

that offer the 'best bets' for maintaining and 

improving the soil fertility of smallholder maize 

systems in a profitable and adoptable way. The 

range of technologies being tried out include: 

• Flexible mineral N management based on ra:nfall 

for maize 

• Soybean for communal areas 

• Liming of granitic sands 

• Ro~ations with grain legumes and maize 

• Sole crop and intercropped legume green manures. 

Background research papers on these technologies 

are presented in Waddington, MlIrwira, KlImwenda, 

Hikwa and Tagwira (1998) . 

Green manure technologies are not new in Zimbabwe, 

as some work was reported as far back as 

the 1920s to 1940s (Metelerkamp, 1988), although 

use of green manures has been mostly on large-scale 

commercial farms with some informal reports of use 

under smallholder conditions (Hikwa, et al. 1998). 

Rattray.and Ellis (1952) noted that maize yield responses 

were larger when maize followed mucuna 

than after any of the other green manures used extensively 

through the 1950s. Recently there has 


een renewed interest in green manure technologies 

cis the price of mmeralfertiliser has increased. 

The Risk Management Project (RMPJ is a project under 

the CIMMYT Natural Resources Management 

Group with a broad objective of improving farm 

incomes and food self-reliance for poor smallholder 

farmers in Zimbabwe and Malawi l;>y addressing 

problems of low soil fertility, climatic variability, 

low and unstable agro ecosystem productivity 

through the use of simulation modelling and farmer 

participatory research. Risk Management Project 

staff have been working in very close collaboration 

with a group of 14 farmers in the Zimuto smallholder 

area of Masvingo since 1999. The work has 

focussed on farmer-led on-farm experimentation 

with several legume-based soil fertility technologies 

coming out of the Soil Fertility Network trials. The 

approach used aims to enable farmers, together 

with researchers, to analyse and understand farmer 

strategies and practices of soil fertility management 

and to identify technologies that both meet farmers' 

needs and are sustainable. Preliminary results from 

the fieldwork and from focussed group discussions 

with the farmers indicate the robustness of mucuna 

under smallholder conditions and great interest in 

the mucuna technology amongst the farmers. 

However, selection of an appropriate technology 

and management options is complicated by climatic 

variability. This means that management options 

must be assessed on a probabilistic basis. Mo.reover, 

the development of appropriate technologies, 

and the testing of the components, is complicated 

by season-to-season variability. Experiments to test 

different technologies must be run over many seasons 

to obtain reliable results. This is expensive 

and, in many cases, impracticaL Simulation models, 

which integrate the major physical and biological 

processes, provide a solution to this problem. 

Simulation Study 

Simulations were carried out using the APSIM 

(Agricultural Production Systems SIMuJator) model 

in conjunction with a 47-year long-term weather 

dataset for Masvingo, which is around 30 km south 

of the study area and about 50 mm/year drier. The 

APSIM software system allows a wide range of configurations 

of crops, sequences, mixtures and management 

practices to be simulated. It provides a 

flexible structure for the simulation of climatic and 

soil management effects on growth of crops in farming 

systems and changes in the resource base. A 

detaileQ>~scription pf APSIM, including its capabilities,_d~~Jgn 

features, structure, user interface and 

the derivation of its main biological and environmental 

modules is provided by McCown, Hammer, 

Hargreaves, Holzworth and Freebaim (1995). 

The simulation set-up consisted of growing either: 

1. A maize crop, receiving various levels of N, 

year after year. The N levels ranged from 0 kg 

N /ha to 100 kg N /ha. 

2. A crop of mucuna from the opening rains of the 

season. The crop of mucuna was managed in 

two ways: 

a) either the crop was grown to maturity 

and harvested on the 1 st of July with 

60% of the residues incorporated on the 

1 st of November just before maize planting 

(Management 1). 

b) or, the mucuna is harvested at the beginning 

of grain fill with 90% of the 

mucuna material incorporated at that 

time (Management 2). 

3. The muctina crop was grown in rotation with 

an unfertilised maize crop (cv. SC501). Two 

cropping systems were simulated for both residue 

management systems with either one maize 

crop after every mucuna crop (mucuna-maize 

rotation) or two maize crops after every mucuna 

crop (mucuna-maize-maize rotatioll). 

Results and Discussions 

Simuiation runs on maize response to different 

amounts of mineral fertiliser and on maize following 

a mucuna crop were done using the long-term 

climatic data from Masvingo. On moderate fertility 

soils typical of most of the topland fields found in 

Zimuto, maize grain yields in the absence of mineral 

fertilisers were simulated to be, on average, 494 

kg/ha. The yields from such unfertilised crops 

range from total crop failure to about 1600 kg/ha. 

These values are similar to values quoted elsewhere 

from on-farm and on-station results (Shamudzarira 

and Robertson, 2002) and are similar to measured 

yields for unfertilised maize in the area. Figure 1 

shows the simulated responses to a range of different 

amounts of mineral fertiliser additions over 

seven seasons on a typica'lly low fertility soil. There 

is enormous variation in maize response to any 

given rate of fertiliser applied, with -the range being 

greater at higher rates of N applied. Smallholder 

farmers in this area normally cite the "risk" associated 

with the wide variations in yield with N application 

(Figure 1) as one of the reasons they use 

small amounts of mineral fertilisers. The simulations 

also show that in 20% of the seasons there is 

no benefit in use of mineral fertilisers in these environments. 

88 


4500 

4000 . 

3500 

to 

'" 3000 - , "-··1992 

. • . 1993 

0> 

0< 

-0 2 500 - .... - . 1994 

1\j --1995 

2000 

--_·01(·····_·1996 

>= 

c 

'm 

i!l 

1500 

1000 

500 

0 

0 20 40 60 80 100 

N rate kg/h. 

- - 1997 

···--+- 1996 

Figure 1. Maize grain yield responses on a low fertility soil to 

varying amounts of applied N(kg/hal simulated using tong-term 

weather data from Masvingo 

Agronomic nitrogen use efficiencies (NUEs) were 

calculated from the simulated maize yields as extra 

kg grain produced divided by extra kg of N applied. 

Averaged over all years, agronomic NUEs 

declined from around 45-56 kg grain per kg of applied 

N at low application rates to about 33-38 at 

high N rates. In Figure 2, the agronomic NUE for a 

crop receiving 10 kg N/ha is plotted against the 

yield for an unfertilised maize crop. The data suggests 

that responses to low rates of N (commonly 

used by smallholder farmers) are generally larger 

on low than with high fertility soils. 

Simulations were also conducted for the same climatic 

record for maize yields following a mucuna 

crop on a moderate fertility sandy soil. The mucuna 

was managed in two ways: - either harvested at maturity 

(1 July) and then incorporated just before 

maize planting (Management 1), or harvested and 

incorporated at flowering (Management 2). For 

each management system, two cycles of simulations 

were done, one in which a single crop of maize was 

grown following mucuna and the other in which 

two maize crops were grown in succession after a 

mucuna crop. 

Despite having fewer seasons in which maize is 

grown in the mucuna-maize and mucuna-maizemaize 

rotations when compared to continuous sole 

maize, maize grain yields averaged over the 47-year 

record are predicted to be 3-5 times higher in rotations 

that include mucuna (Figure 3). With continuous 

sole maize cropping, just over 20 tonnes of 

maize grain is realised over 47 years compared to 

totals of between 80 and 120 tonnes in rotations that 

include mucuna (Figure 4). 

At a 50% probability level, unfertilised maize grain 

yields are 3.5-4.0 t/ha for the mucuna-maize rotations 

and 3.5-4.5 t/ha for the mucuna-maize-maize 

rotations (Figure 5). These yields are far greater 

than the 200-300 kg/ha obtained for continuous sole 

maize cropping at the same probability level. 


. . 

~ ..e ". 

. .' / . '. 

YIeld response 10 10 kg Nnw 

Figure 2. Simulated maize yield response to a 10 kg N/ha 

application for a crop grown on soils of different fertility status 

300• . 0,-______________---, 

025 00.0+_-------------~~ 

i 2(100 .0 

. 

>. I ~OO.O+_------~~ 

., 

• 

i 1 00 0 0 +_------~ 

Figure 3. Mean maize grain yields for the 47·year record for 

different cropping systems 

OCOOOr-----------------------------------~ 

~~------------------~----------~~.~ 

70000+_----------------~----~~~=---~ 

~~+-------------------~~~----~~--~ 

.~ 

~50000~--------------~~_,~~--------~ 

jf~+_-----------~L-~~-------------~ 

i~r-------~#7~------------------~ 

u~+_---~~-------~--~~----------~ 

l0000+_~~~~~------------------------_i 

SoIematze 'MxlJna-rreize1 - Mucuna-maize2 

- 'Muruna-maize-rT'9ize1 .... 'Muruna-rrnlze-malZ.e2 

Figure 4. Cumulative maize grain production over 47 years for the 

different production systems 

The loss in maize production from the piece of land 

where mucuna is grown in the first season, coupled 

with labour constraints, has been given by many 

workers as one of the limitations in the uptake of 

green manure technologies by farmers (Kumwenda, 

Waddington, Snapp, Jones and Blackie, 1997). 

However, in Northern Malawi on fairly good sandy 

soils, a one season sole crop green manure can increase 

maize yields from 200-300 kg/ha to up to 4 

000 kg/ha. The data presented here also suggests at 

the 50% probability level, a yield surplus from an 

,~ . 

89

unfertilised maize crop of over 2000-3000 kg in 

maize output in mucuna-maize rotations and 5000 

8000 kg in mucuna-maize-maize rotations as opposed 

to continuous sole maize cropping in situations 

where no mineral fertilisers are used (Figure 

6). Muza and Mapfumo (1998) reported a trebling 

of maize yields in Chihota after incorporation of 

mucuna compared to an unfertilised maize crop 

that yielded 466 kg/ha. 

Maize grain yields from continuous sole cropping 

without fertilisers show a slight decline with increase 

in amount of seasonal rainfall (Figure 7). In 

the rotation systems including mucuna however, 

there is a strong positive relationship between grain 

yield and seasonal rainfall. In environments where 

seasonal rainfall is below 350 mm, average maize 

grain yields are higher with the mucuna-maizemaize 

rotation than with the mucuna-maize rotation. 

Where seasonal rainfall is greater than 350 

mm, mean yields are higher with mucuna-maize 

rotations than for two maize crops after every mucuna 

crop. 

1.00r-----------------A::::::o-i 

0.80 t--------------::,..=.+~'------1 

~ 

'Zi 

!0.00 . ..• . •..• "_'_' __ ••• __ • _. ____/; _ ~ ,.. " __ __ ___________ ___ ___ _ 

ro 0.40 - I / / 

-s 

E 

ers, particularly P, to the legume to have reasonable 

biomass production for incorporation (Hikwa et al., 

1998). 

Conclusions 

Crop growth simulation coupled to long-term 

weather data can assist in technology selection by 

generating probabilistic estimates of crop yield. The 

potential benefit of green manure technologies 

through field experimentation has been underestimated 

due to the short-term perspective of most 

field trials. 

The foregoing analysis has highlighted large potential 

benefits from use of mucuna as a green manure 

crop. In environments where continuous maize 

cropping yields between 150 and 500 kg/ha, rotation 

with mucuna gives maize yield increments of 

over 1000 kg/ha when the mucuna is harvested at 

maturity and between 3000-5000 kg/ha when incorporated 

at flowering. The reason frequently proposed 

to explain low uptake rates of green manure 

technologies by smallholder farmers is the loss in 

maize yield duringthe year when the green manure 

crop is in the field. 

Future studies need to confirm that the model has 

not overestimated the response to mucuna or that 

tills response is not limited by factors not considered 

in the model (e.g. weeds, pests and diseases, 

other nutrients). Several workers have reported difficulty 

in establishing legumes under smallholder 

conditions citing soil infertility as the main factor. 

The residual effects from mineral fertiliser management 

on cereals on legume productivity have not 

received much attention and warrant investigation. 

Socio-economic research needs to· assess farmer attitudes 

toward risk and examine other constraints to 

mucuna use (e.g. access to capital, access to maize 

grain storage facilities, farmers' understanding of 

the full benefits of legumes in these systems, alternative 

uses of mucuna). 

References 

Chavunduka, O.M. 1978. African attitudes to conservation. 

Rhodesian Agriwltural Journal 75(3):61 

63. 

Grant, P.M. 1967. The fertility of sandveld soil under 

continuous cultivation. Part II: The effect of 

manure and nitrogen fertiliser on the base status 

of the soil. Rhodesia/Zumbia/Malawi Journal ofAgricultural 

Research 5:117-128. 

Grant, P.M. 1981. The fertilisation of sandy soils in 

peasant agriculture. Zimbabwe Agricultural JournaI78(5):169-175. 

Hikwa, 0, M. Murata, F. Tagwira, C. Chiduza, H. 

Murwira, L. Muza and S. Waddington 1998. Performance 

of green manure legumes on exhausted 

soils in northern Zimbabwe: A soil fertility 

network trial. In: S.K Waddington, H.K. 

Murwira, J.DT. Kumwenda, o. Hikwa and F. 

Tagwira (eds), Soil Fertility Research for Maize 

Based Farming Systems in Malawi and Zimbabwe. 

Soil Fert Net and CIMMYT-Zimbabwe, Harare, 

Zimbabwe. pp. 81-84. 

Kumwenda, J.DT., S.K Waddington, S5. Snapp, R. 

B. Jones, and M.J. Blackie 1997. Soil fertility 

management in Southern Africa. In: o. Byerlee 

and C.K. Eicher, eds. Africa's Emerging Maize 

Revolution. Lynne Rienner, Boulder, Colorado, 

USA, pp. 157-172. 

Mataruka, O.F. and E.E. Whingwiri 1988. Maize 

production in communal areas of Zimbabwe and 

the research options for improving productivity. 

In: B. Gelaw, ed. Towards Self Sufficiency. Proceedings 

of the Second Eastern, Central and Southern Africa 

Regional Maize Workshop. CIMMYT Zimbabwe 

and the College Press, Harare, Zimbabwe. 

pp. 332-346. 

McCown, KL., G.L. Hammer, J.N.G. Hargreaves, O. 

P. Holzworth, and O.M. Freebairn 1995. APSIM: 

a novel software system for model development, 

model testing, and simulation in agricultural 

systems research. AgriclIltural Systems 50:255 

271. 

Muza, L. and P. Mapfumo, 1998. Constraints and 

opportunities for legumes in the fertility enhancement 

of sandy soils in Zimbabwe. In: 

Maize Production T!!clll1ology for til!! FlItllr!!: Clrallenges 

and Opportllllitil!S. Proceedings of the Sixth 

Eastern and Southern Africa Regional Maize 

Conference, 21-25 September 1998. CIMMYT and 

EARO, Addis Ababa, Ethiopia pp. 214-217. 

Shamudzarira, Z. and M.J. Robertson, 2002. Simulating 

response of maize to nitrogen fertiliser in 

semi-arid Zimbabwe. Experilllel1tal AgriclIltllre 

38:79-96. 

Waddington, S.R., H.K. Murwira, J.DT. Kumwenda, 

o. Hikwa and F. Tagwira, (Editors) 1998. 

Soil Fertility Research for Maize-Based S.If~tI'II1S ill 

Malawi and Zimbauw!!. Proceedings of the Soil 

Fertility Network Results and Planning Workshop, 

7-11 July, Africa University, Mutare, Zimbabwe. 

Soil Fert Net and CIMMYT-Zimbabwe, 

Harare, Zimbabwe. 312 pp. 


91


Screening of Annual Legumes for Adaptation and Use 

To Paul Mapfumo, et al. 

Q:If the indigenous legumes dominate natural 

fallows, why is maize productivity so low after 1-2 

years of natural fallows? . 

A: The current weed management approach in most 

farming areas is aimed atldepleting the weed 

(including these legumes) seed bank through clean 

weeding. Therefore the current populations are 

probably too low to make an impact in the short 

term. 

Q: If these 'indifallows' are left for two or more 

years, will they contribute more to productivity? 

A: The duration of the fallow does playa big role in 

indigenous legume species diversity and 

abundance. In Chikwaka, legume biomass from 

one-year fallows was far less than that from twoyear 

fallows for both species richness and 

abundance. 

To Bongani Ncube, et al. 

Q: Would it be useful to separate the woody tissue 

from leaves and their separate response to soil 

fertility? Will the woody tissue immobilize N? 

A: Our research shows that it is beneficial to 

inco~porate both leaves and stems soon after 

harvesting the grain. Stems decompose 

substantially within the year of incorporation 

(based on a five-year rainfall period). There seems 

to be better synchrony for maize under this system. 

Q: Pigeon pea has been researched as long ago as 

the early 80's in Zimbabwe. Did you have an 

unsprayed crop and was spraying economic? 

A: This was a nursery trial aimed at evaluating 

varieties, so we did not put up any control plots. 

Our aim was to assess what would grow in the 

semi-arid region of Matabeleland. 

C: There is a lot of information about pigeonpea 

adaptability in Zimbabwe from Matopos, Makaholi, 

Panmure and Mlezu. See Agronomy Institute 

Annual Reports from 1987, 1988 and 1989. 

A: It is very difficult to get access to this type of 

grey literature from the 1980's. None of the current 

literature we have reviewed makes reference to this 

early work. 

Q: Did the research look at "weed suppression" on 

the experiments? In Mozambique it was found that 

just two weedings were needed when pigeonpea 

was intercropped with maize. 

A: Our aim was to keep the crop as clean ·as 

possible; so weeding was done every time weed 

regeneration occurred. 

To Richard Foti, et al. 

Q: 

1. The 18 kg N ha- 1 for maize seems low for this 

heavy feeder crop. Is this practice not leading to 

nutrient mining? 

2. Does the evaluation of the returns to a crop 

include the quantification of some of the more 

indirect crop values such as barter and exchange for 

labour? 

A: 

1. Farmers are already mining soils by not using any 

inorganic fertilizers at all and are therefore 

foregoing extra income that they could generate 

from using low rates of fertilizers, earn more 

income, buy more fertilizers and move upwards. 

Insisting tha t farmers use rates of fertilizer that they 

cannot afford and that fail to generate a competitive 

rate of return on their investment is retrogressive. 

2. Yes the analysis uses the opportunity cost of these 

resources and products. This will vary for and are 

different for different people and areas, but we 

cannot do the analysis for each and every farmer. 

So we need to compromise and do the analysis for 

one set of prices. We then do sensitivity analysis to 

alternative prices. 

C: In response to the first question, low rates of N 

fertilizer is not the problem contributing to soil 

mining. It is lack of investment in soil fertility in 

general by smallholder farmers, be it inorganics, 

legumes or manure. In dry regions, that is the cause 

of low productivity and low soil fertility. Low rates 

of N better suit the investment profile of semi arid 

farmers and therefore ,r~ "lore likely to be adopted 

than the higher "optimal" rates that arestill 

recommended. 

C: Work from southern Zimbabwe by ICRISAT + 

SDARMP have shown that 18 kg N ha- 1 gives the 

most economic response. Above 18 kg N ha- 1 often 

gives no more yield. 

Q : "Results from the @R.isk analysis appear very 

close to what farmers in semi arid areas are actually 

doing anyway, except for cattle manure in the drier 

Grain legumes and Green Manures. for Soil Fertility in Southern Africa 

93

areas, which many cattle owners do not use. Please 

comment. 

A: Farmers believe that use of manure burns their 

crop in drier weather. This is because in the past 

AGRITEX has recommended high rates of 

application and farmers who have tried these have 

had bad experiences with the recommendations. 

Revising the recommendations downwards and 

experimenting with low quantities of manure is 

resulting in farmers revising their assessment of the 

risk associated with manure use and adopting 

manure. There is also the problem of a shortage of 

labour, especially because of high migration to 

South Africa and HIV-Aids which increases the 

opportunity cost of labour (probably to a higher 

level than used in our analysis) and so is making 

use of manure less profitable than the analysis 

suggests. 


C: We do not have to keep doing screening. How 

are we sure that we target the correct accessions 

within the existing gene banks? 

Q: Is there any research done by scientists here that 

has got results on the contribution of the depth of 

pigeonpea roots for bringing up nutrients and to 

organic matter content? 

C: The screening of alternative legumes should be a 

confinuous process against the biodiversity of pests 

and changing environments. 

C: When to stop screening? The search for new 

materials amongst the available genetic resources 

should continue in accordance with projected 

demands - for better traits, crop diversification, etc, 

so you do not stop( 

94 

Grain legumes and Green Manures for Soil Fertility in Southern Atrica

GREEN MANURE AND FOOD LEGUMES RESEARCH TO INCREASE SOIL 

FERTILITY AND MAIZE YIELDS IN MALAWI: A REVIEW 

Summary 

WEBSTER SAKALA1 and WEZI MHANG0 2 

1 Chitedze Agricultural Research Station, P. O. Box 158, Lilongwe, 

2 Bunda College of Agriculture, P. O. Box 219, Lilongwe, Malawi 

This review was conducted in 2002 to document research on the soil fertility effects of green manures and food legumes 

on the dominant maize-based farming systems in Malawi. The findings indicated that green manures and grain legume 

crops have great potential to increase maize production and improve soil fertility. A great deal of work has been done. 

This can be grouped under the following subheadings: the potential of green manures to improve soil fertility and the 

yield of maize; combined inputs from organic and inorganic (mineral) sources; effect of time of residue incorporation on 

maize grain yield; effect of cropping system (intercropping or rotation) on maize yield; effect of method of residue application; 

and the role of applying inorganic fertilizers on the performance ofgreen manures and grain legumes in Malawi 

maize-based systems. Most of the work that we have found and reviewed lacks socio-economic studies and this needs far 

more attention in the future. 

Key words: Green manures, grain legumes, maize, Malawi 

Introduction 

This paper reviews some of the soil fertility research 

activities conducted in Malawi on green manures 

and annual food legumes. The objective was to 

document what has been done in Malawi on these 

legumes for increasing soil fertility in maize-based 

cropping systems. Information was collected from 

the libraries and annual reports of the Department 

of Agricultural Research on·work carried out by scientists 

at research stations,and on farm. 

General Characteristi

organic and inorganic (mineral) fertilizers, and crop 

removal without the return of nutrients. 

Soil Fertility Status of Malawi Soils 

Soil fertility is defined as the ability of the soil to 

supply the nutrients needed by plants (Ahn, 1993). 

According to Young and Brown (1962; 1965), nitrogen 

is the most limiting nutrient element in Malawian 

soils. Sulphur deficiencies are prevalent in 

some areas. In most upland areas, the soils are 

highly leached and as such, they are dominated by 

iron and aluminium oxides that fix phosphorus into 

forms that-are unavailable for plant uptake. Phosphorus 

studies by Mughogho (1975) on some soils 

in Malawi indicated that soils in Mulanje, in the 

southern region of Malawi, fix a lot of phosphorus. 

This is one of the high rainfall areas that receives 

1200-1800 mm of rain annually. 

Soil Fertility Research Reviews in Malawi 

Mughogho (1989) conducted a review of soil fertility 

research in Malawi. The overall objective of that 

work was to document existing information on soil 

fertility research from Malawi and other appropriate 

sources, to be used as a planning tool and database 

for proposed soil fertility ' studies. Findings 

from that study indicated tQat 'in, addition to low 

soil nitrogen, most soils have large quantities of sesquioxides 

that fix phosphorus into unavailable 

forms, and sulphur is deficient in some areas. Mughogho 

(1989) further recommended the need for a 

detailed study on the characterization of soils in 

Malawi to build upon the work by Brown and 

Young (1962; 1965). The potential of sources of 

phosphate rock, to be used on acid soils needs to be 

explored. 

A review report by Gilbert and Kumwenda (2001) 

highlighted some of the best-bet legumes for smallholder 

maize-based systems. For Instance, Mucuna 

pruriens was described as a promising green manure. 

Successful grain legume-maize rotations and 

intercrops of pigeonpea or Tephrosia with maize 

were observed. 

The following sections look in more detail at green 

manures, crop rotations (especially with grain legumes) 

and agroforestry interventions to raise soil 

fertility and maize productivity in Malawi. 

Green Manures 

Follet et al. (1981) defined a green manure crop as 

one that is grown and incorporated into the soil to 

add organic matter and N and subsequently improve 

crop yields. In Malawi, most farmers have 

used weeds as green man~.ue materials. These are 

96 

incorporated at the time of ridging, weeding or 

banding. The benefits from green manures include 

reduction of nutrient.loss through leaching, the accumulation 

and maintenance of soil N, and improvement 

of soil structure. Other species like Mucuna 

pruriens help to reduce weeds (CIMMYT, 

1998), thereby minimizing competition for soil nutrients 

and water. The success of a green manure 

for soil fertility improvement depends on its quality 

(CN ratio), quantity of the material, and management 

(especially the timing and means of biomass 

incorporation). Proper timing allows nutrient release 

in synchrony with crop uptake. High biomass 

production can be attained if all essential soil nutrient 

ekments are available. For instance, Giller and 

Wilson (1991) noted that phosphate fertilizer applications 

are necessary to support the luxurious 

growth of the green manure and hence its potential 

as an organic source of fertilizer. There are some 

leguminous species with higher quality biomass, 

and good ability to fix nitrogen biologically in Malawi. 

Some of these species include Tephrosia vogelii, 

Sunnhemp (Crotalaria juncea), Tithonia diversifolia 

and velvet bean (Mucuna pruriens). Benefits from 

the use of Mucuna pruriens, Tephrosia vogelii, sunnhemp, 

and bulrush millet have been reported 

(Lungu, 1973; Sakal a et al., 2001; and Mwalwanda, 

2002). However, Lungu pointed out that the one 

year lost to a sole crop green manure or improved 

fallow is a cost to a farmer and therefore this may 

reduce farmer interest and adoption. 

The feasibility of improving soil fertility and maize 

yield through intercropping or rotation of maize 

with legumes was investigated at Chitedze Research 

Station in central Malawi (Kumwenda et al. 

2001) from the 1995/96 to 1998/99 crop seasons. 

The treatments were as indicated in Table 2. 

The results in Figure 1 illustrate that intercrops of 

maize with pigeonpea and sunnhemp gave higher 

yields than the maize/Mucuna system. Maize/ 

Table 2. Treatments from maize x green manure intercrop and 

rotation experiments in Malawi from the 1994/95 to 1998/99 crop 

seasons 

1994/95 1995/96 1996/97 1997/98 1998/99 

Intercrop Maize/PP Sarpe Same Same Same 

Maize/ Same Same Same Same 

Mucuna 

Maize/ Same Same Same Same 

sunnhemp 

Sale Pigeon pea Sale maize Sale maize Sale maize Sale maize 

Sunnhemp Sale maize Sale maize Sale maize Sale maize 

Mucuna Sole maize Sole maize Sole maize Sole maize 

Maize Sole maize Sole maize Sole maize Sale maize 

PP - Pigeon pea 

Same - same treatment as in 1994/95 crop season was grown 


7000 

6000 

';' 

€.. 5000 .~MzSole 

::!. 4000 o MlIPP 

~ 

;;: 3000 mMlISunhemp 

.~ 

~ 2000 

1000 

0 

I nte(crop 

Rotation 

Cropping s~tem 

III MlIMucuna 

Figure 1. Maize grain yield (kg hal) from different cropping 

systems, across four crop seasons in Malawi, 1995/96·1998/99. 

Source: Kumwenda et aI., 2001 

Mucuna mtercroppmg yielded the least due to competition 

for growth resources. Rotation trials gave 

higher maize yields than the intercroppmg system 

due to larger biomass production from the legumes. 

However, an economic analysis indicated that continuous 

maize had higher net benefits than the legume-maize 

rotation. Similar studies on ~igeonpea 

by Sakala (1998) have indicated that a farmer is better 

off opting for maize/ pigeon intercroppmg than 

for a pigeonpea-maize rotation or the maize-maize 

cropping system. 

Other studies have looked at the time of incorporation 

of the green manures as a factor that influences 

their potential as soil fertility enhancers. Kumwenda 

et al. (2001) carried out experiments where 

biomass was mcorporated at either the time of 

maximum flowering, at pod initiation (early incorporation) 

or at harvest (late incorporation). Three 

species, M. pruriens, C. juncea and T. vogelii were 

used, along with maize. Higher maize gram yields 

were obtained from early-incorporated residues 

than with late. mcorpora.tion (Table 3) . This was attributed 

to high CN ratios and high lignin contents 

in the late incorporated residues. However, work 

by Sakala et al. (2001) revealed that with late incorporation, 

lower yields are obtained in the first year 

Table 3. Maize grain yield (kg hal) as influenced by legume 

crop residues and time of incorporation at five sites in 

Malawi, 1996/97 cro season 

legume crop/Maize 

Time of incorporation Mean yield (kg ha") 

M. pruriens Early 3392 

late 2223 

C. juncea Early 3218 

late 2692 

I T. vogelii Early 2845 

late 1483 

I Maize·maize 397 

Source: Kumwenda et ai., 2001 

only but m the subsequent years farmers realize 

higher yields, which was attributed to the build-up 

of nutrients. It was therefore recommended that 

farmers who are constrained for labour would still 

chose late incorporation and realize a longer stream 

of higher maize yields. 

A three-year study was carried out m Southern Malawi 

by Kamanga et al. (1999) to examme the feasibility 

of inter-plantmg nitrogen fixmg perennial legumes 

into maize fields as a way to periodically add 

green manures to maize. The treatments were Sesbania 

sesban, Tephrosia vogelii, Pigeonpea (Cajanus 

ca;an) and maize. The legumes were relay mtercropped 

with maize at first w eedmg. It was shown 

that the application of 48 kg N ha- 1 combined with 

residue incorporation increased maize yields by 62 

71 %. Higher maize yields were obtamed from the 

inter-planting of S. sesban (2.9 t ha- 1 ) and T. vogelii 

(2.6 t ha- 1 ) than from pigeonpea (2 .1 t ha- 1 ) and 

maize stover (2.0 t ha- I ). 

Crop Rotation 

Crop rotation refers to the repetitive cultivation of 

an ordered succession of crops on the same land 

(Mloza-Banda, 1994). The aim is to maintam and 

improve soil fertility, includmg both its physical 

and chemical characteristics. It also ensures that the 

carryover of pests and diseases from one season to 

another is minimized. 

The benefits from crop rotations involvmg grain 

legumes (groundnut, bambara nut and soya bean) 

over a continuous maize-maize cropping system 

have been reported in several studies in Malawi 

(Brown, 1958; Lungu, 1973 and MacColl, 1989; 

Kumwenda, 1996; and Mhango, 2002). This has 

been attributed to improved soil fertility through 

biological nitrogen fixation (BNF) and crop residue 

incorporation. However, grain legume-maize rotations 

are not efficient because of the inadequate biomass 

they prod uce and the small amounts of N retained 

in crop residues to meet the N requirements 

of the subsequent maize crop. Giller Jnd Wilson 

(199\) pointed out that most of the N fixed by grain 

legumes is exported away from the field due to high 

nitrogen grain harvest· indices. Other studies have 

looked into the inclusion of pastures in crop rotations 

to enhance maize prod uction. Maceoll ( 1990) 

reported on long term trials whose aim was to determine 

the contribution to maize yield from a previous 

pasture legume crop. The treatments were 

two rates of N (0 and 80 kg N ha- I ) from CAN fertilizer; 

maize, pure silver leaf, pure stylo, silver leaf/ 

rhodes grass, and stylo/rhodes grass. Pastures 

were grown from 1981 to 1984, and then the plots 


97

Table 4. Yield of maize (t ha I) following different 

cropping sequences and grown at two levels of nitrogen 

fertilizer. across 4 years (1984/85··87/88 crop seasons) 

Cropping sequence 

N letel~ (kg 'N hal) 

Zero 80 

Maize 2.55 5~38 

Silver leaf 3.85 5.90 

Stylo 3.58 5.58 

Silverleaf + Rhodes grass 3.05 5.55 

Stylo + Rhodes grass 2.83 5.33 

Source: MacColI, 1990 

were planted to maize for three years (1985-88). 

Yields were higher when maize followed pasture 

legumes with successful establishment, with- ~ilver 

leaf out-yielding the other species (Table 4). The 

application of inorganic N increased maize yield, 

stressing the need for the combined use of organic 

and· inorganic (mineral) fertilizers. 

Use of combined inputs from organic and inorganic 

(mineral) sources appears to be the best approach to 

address soil fertility problems. Organic fertilizers 

improve the soil physical, chemical and biological 

properties. They also help to build up soil orgariic 

matter because nutrients are released slowly after 

mineralization. However, the amount and quality 

of the organic fertilizers is insufficient to provide 

adequate amounts of nutrients for crops, hence the 

need to supplement with inorganic sources. Mwatoet 

al. (1999) conducted a 2-year study on combined 

inputs of crop residues for smallholder maize production 

in Malawi. The overall objective was to·examine 

the effect of applying inorganic fertilizers 

and crop residues to the soil on the subsequent 

maize yield. Crop residues from maize and different 

varieties of soya bean were incorporated into the 

soil. Inorganic N was applied to maize at several 

rates. Maize grain yields were increased from 0.5 t 

to 1.3 t ha- I after the addition of soyabean residues. 

plus inorganiC fertilizer. . 

Agroforestry 

Agroforestry refers to those land use systems in 

which woody perennials are grown in association 

with herbaceous plants (crops, pastures) and/or 

livestock in a spatial arrangement, a rotation, or 

both, and in which there are both ecological ·and 

economic interactions between the tree and non tree 

components of the system (Young, 1989). Alley 

cropping and improved fallows are among the 

agroforestry systems practiced by some farmers in 

Malawi. Choice of a technology depends on the 

problem to be addre$sed, and the availability of resources 

such as land, rainfall and labour. Research 

work in Malawi has revealed the potential of raising 

soil nitrogen and maize yields with agroforestry 

technologies (Kwapata, 1994; Malawi Agroforestry 

Team, 1994; Makumba, 1998; and Phiri, 1999). 

Maize yields after L leucocephala, S.· spectabilis and S. 

sesban were 4.8, 45 and 4.4 t ha- I respectively, compared 

with 3.2 t ha- I produced from sole-crop maize 

plots (Chirwa and Maghembe, 1994). However, 

there are limitations with agroforestry systems. 

These include: 

• . The benefits from agroforestry technologies are 

long term 

• A high labour requirement with some technologies, 

e.g. alley cropping 

• High seed requirement, implying a cost to the 

farmers 

• Problems with seedling establishment 

• Insect pestattack on some tree species that have a 

high biomass potential, such as L. leucocephala, 

that is susceptible to psyllids (Heteropsylla cubana) 

• Some tree species perform poorly on acid soils 

because of low available phosphorus. 

A lot of research has been conducted with agroforestry 

to identify suitable candidate species for a particular 

technology, for biomass production, timing 

and application methods for biomass, and the effect 

of the technology on maize yield. Faidherbia albida is 

one of the tree species used in agroforestry. It is 

found growing naturally in farmers' fields in some 

parts of Malawi. The yield benefits to maize grown 

under F. albida trees has been reported by many investigators, 

including the Malawi Agroforestry 

Team (1994). Inorganic fertilizer supplements significantly 

further increase maize grain yield under 

the trees. Some of the candidate tree species for 

agroforetsry are Cassia siammea, Gliricidia sepium, 

Leucaena leucocephala, Senna spectabilis and Sesbania 

sesban. Chiyenda and Materechera (1987) conducted 

experiments from the 1983/84 to 1985/86 

crop seasons with L. leucocephala, C. siamea and C. 

cajan. The overall goal was to determine the effect 

of incorporating prunings from these species on soil 

fertility and to assess the response of maize grown 

in alleys. The treatments were three rates of N 

(main plot); the three tree species with maize, and 

maize alone (sub plot); and three alley widths with 

three ridges of maize (as sub sub-plots). Phosphate, 

at 22 kg P ha- I , was applied to all plots at planting. 

According to the results, better yields of maize were 

obtained from plots incorporated with the tree 

prunings, although, they were significantly lower 

than treatments that received 100 kg N fertilizer ha- I 

(Table 5). The plant materials could not provide the 

amount of N that would be provided by moderate 

rates of inorganic fertilizers. 

Kwapata (1994) worked on L. leucocephala in an alley 

98 


Table 5. Maize grain yield (kg ha") as affected by incorporation 

of tree prunings and nitrogen fertilizer levels in Malawi 

Crop Crop System Mean yield 

Nrate (kg ha") 

Season 

(kg hal) 

0 50 100 

1984/85 Zmais 704 1954 3564 2074 

l. leucocephala 617 2454 2794 1957 

S. siamea 467 1856 ·3287 1870 

C. cajan 468 1523 3054 1682 

1985/96 Zmais 151 2317 3233 1901 

l. leucocepha/a 107 1665 2935 1411 

C. siamea 57 1049 1638 915 

C. cajan 255 1918 2870 1681 

Source: Chiyenda and Materechera, 1987 

system to determine the optimal rates and method 

of application of the leaf biomass. Residue incorporation 

gave larger maize yields than did surface 

mulching and this was attributed to a faster mineralization 

rate. Ten t ha- I of fresh L. leucocephala was 

as effective as inorganic fertilizer N applied at 100 

kg N ha- I . 

Conclusions 

• Organic soil amendments from green manures 

and annual legumes have potential to enhance 

.soil fertility and increase maize yields for Malawi 

smallholder farmers. They improve the soil 

physical, chemical and biological characteristics. 

They are relatively cheaper than inorganic 

(mineral) fertilizers but the nutrients provided 

are not adequate to meet crop demands and 

farmer needs. T.herefore, the use of combined in- . 

puts from organic and inorganic fertilizers appears 

to be the best approach to address soil fertility 

problems. 

• Crop residues have alternative competing uses 

such as to feed livestock, as in the case of legumes 

such as groundnut. This reduces their role in soil 

fertility. 

• Intercropping of maizelpigeonpea has proved 

successful. 

• For green manures as soil fertility enhancers in 

maize-based systems, .Mucuna pruriens, pigeonpea, 

and Tephrosia vogelii are promising species. The 

key factors for success include the following: 

o Quality of biomass 

o Quantity of biomass 

o Timing and means of incorporation of biomass. 

• Most of the biological studies lack the socioeconomic 

component of the technologies, and 

these need to be developed. 


• Crop rotations involving grain legumes such as 

groundnut, and pasture legumes like stylo can 

boost maize yield. However, the opportunity 

cosUor the farmer to forgo maize in the first year 

should be considered. Thisis a major restraint to 

adoption. 

• Agroforestry technologies such as alley cropping 

and improved fallows have proved to be successful 

in maize based systems. Researchers should 

consider issues related to direct seeding, resistance 

from pests and tolerance to low soil available 

phosphorus. Extension workers should 

carry out awareness campaigns on the long-term 

benefits from agrbforestry systems. 

References 

Ahn, P.M. 1993. Tropical Soils and Fertiliser Use. 

Longman Group, United Kingdom. 263 pp. 

Benson, T.D. 1997. Developing fertilizer recommendations 

for small holder maize production in 

Malawi. In: Waddington et al., eds. Soil Fertility 

Research for Maize-Based Farming Systems in Malawi 

and Zimbabwe. SFNETICIMMYT, Harare, 

Zimbabwe. pp. 275-285 . 

Brown, P and Young, A. 1965. The Physical Environment 

of Central Malawi. Government 

Printer, Zomba. Malawi. 93 pp. 

Brown, P and Young, A. 1962. The Physical Environment 

of Northern Nyasaland. Government 

Printer, Zomba, Malawi. 107 pp. 

Brown, P. 1958. Results of short-term experiments 

in Nyasaland. Rhodesia Agricultural Journal 

55:626-633. 

Chirwa P.W. and Maghembe, J.A. 1994. Evaluation 

of six multipurpose tree species for alley cropping 

with maize at Makoka, Malawi. Progress in 

Agroforestry Research and Development·in Malawi. 

Proceedings of the second national agroforestry 

symposium held in Blantyre, Malawi. pp.56-60. 

Chiyenda, 5.s. and Materechera S.A. 1987. Effect of 

incorporating prunnings of Leucaena leucocephala, 

Cassia siamea and Cajanus cajan on yield of maize 

in alley cropping system. International symposium 

on nutrient management for food crop productionin 

tropical farming systems. Malang, 

Indonesia. 

qMMYT, 1998. Mother-Baby Trials for the 1997/98 

season. Farmer Participatory "Best-Bet" Soil 

Technologies. Workshop Report. Natural Resources 

College. Lilongwe. Malawi. 29 1h -30 th 

September, 1998. 

99

Follet, R.H., Murphy, L.S. and Donahue, R.L. 1981. 

Fertilizers and Soil Amendments. Prentice-H,all, 

USA. 557 pp. 

' 

Gilbert, R.A. and Kumwenda, J.D.T. 2001. Best-bet 

legumes for smallholder maize-based cropping 

systems in Malawi. In: J.D.T. Kumwenda et al., 

eds. Maize Commodity Team Annual Research 

Report for 1998/99 Crop Season. pp. 217-234. 

Giller K.E. and Wilson, K.F. 1991. Nitrogen Fixation 

in Tropical Cropping Systems. Redwood Press, 

United Kingdom. 313 pp. 

Kamanga, B.C.G., Kanyama:'Phiri G.Y. and Minae, 

S. 1999. Intercropping perennial legumes for 

green manure additions to maize in Southern 

Malawi. African Crop Science Journal 7:355-363. 

Kumwenda, J.D.T. 1996. Effects of previous cropping 

systems on maize yield. In: T.D. Benson ed. 

Maize Commodity Team Annual Report for 

1992/93 Crop Season. Chitedze Agricultural Research 

Station, Lilongwe. Malawi. pp. 110-111 . 

Kumwenda, J.D.T., Saka, AR., Kabambe V.H. and 

Ganunga R.P. 2001. Legume manure cropsmaize 

rotation trial for the 1997/98 and 1998/99 

crop seasons. In: J.D.T. Kumwenda et aL, eds. 

Maize Commodity Team Annual Research Report 

for 1998/99 Crop Season. pp 125-146. 

Kumwenda, J.D.T., Saka, AR., Sakala, W.o., Kabambe 

V.H. and Ganunga R.P. 2001. Effects of 

organic legume residues and inorganic fertilizer 

nitrogen on maize yield at Chitedze Research 

Station. In: J.D.T. Kumwenda et aL, eds. Maize 

Commodity Team Annual Research Report for 

1998/99 Crop Season. pp. 110-124. 

Kwapata M.B. 1994. Maize grain yield response to 

Leucaena leucocephala foliage and inorganic fertilizer 

nitrogen application. In: A.R. Saka et al., 

(eds). Progress in Agroforestry Research and Development 

in Malawi. Proceedings of the second 

national Agroforestry Symposium held in Blantyre. 

Malawi. October 24-28. 

Land Resources Appraisal, 1991. Land Resource 

Evaluation Project/Ministry of Agriculture. Department 

of Surveys. Malawi. 

Lungu N.F. 1973. Rotations and Soil Fertility. Article 

based Research Bulletin No. 11/72 on Rotations, 

written by R.AJ Cox. pp. 139-150. 

MacColl, D. 1989. Studies on maize (Zea mays) at 

Bunda, Malawi. II. Yield in rotations with legumes. 


MacColl, D. 1990. Studies on maize (Zea mays) at 

Bunda, Malawi. III. Yield in rotations with pasture 

legumes. Experimental Agriculture 26:263 

271. 

Makumba, W.1. 1998. Effect of contour planted leguminous 

tree species on soil fertility and maize 

yield at Mkawa, Zomba, Malawi. Makoka Agricultural 

Research Station Annual Report for 

1998/99 crop season. 

MalaWi Agroforestry Team. 1994. Annual Report 

for the 1993/94 crop season. Chitedze Agricultural 

Research Station, Lilongwe. Malawi. 

Malawi Government, Department of Surveys. 1983. 

National Atlas of Malawi. 

Malawi Government" 1999. National Research 

Council of Mala\Afi. Agricultural Sciences Committee 

. and Natural Resources Research 

. 

Master 

Plan. 15p. 

Mhango, W.G. 2002. Response of grain legumes to 

phosphorus and sulphur application in maizelegume 

short term rotation system in Malawi. 

MSc Thesis. Soil Science. University of Malawi, 

Bunda College of Agriculture. 

Mloza-Banda, H.R. 1994. Principles and Practices of 

Crop Management. Field Study Guide. Crop 

Science Department. University of Mala'Yi, 

Bunda College ' of Agriculture. Lilongwe. Malawi. 

137 pp. 

Mughogho, S,K. 1975. Organic Phosphorus Studies 

in some soils of Malawi. MSc Thesis. Cornell 

University, Ithaca, USA 

Mughogho, S:K. 1989. Soil Fertility Research in Malawi: 

A Review. University of Malawi. Bunda 

College of Agriculture. Crop Science Department, 

Lilongwe. Malawi. 86 pp. 

Mwalwanda, AB. 2002. The role of phosphorus and 

sulphur application on performance of legume 

green manure crops in maize-legume shortterm 

fallow in northern Malawi. MSc Thesis. 

Soil Science. University of Malawi, Bunda College 

of Agriculture, Malawi. 

Mwato, l.L., Mkandawire AB.C. and Mughogho, S. 

K. 1999. Combined inputs of crop residues and 

fertilizer for smallholder maize production in 

Southern Malawi. African Crop Science Journal 

7:365-373. 

National Statistics Office, Malawi Government, 

1999. Population Housing Census. Zomba. Malawi. 

100 


Phiri, A.D.K. 1999. Effects of Sesbania sesban and 

Tephrosia vogelii on maize yields at three land 

scape positions in Southern Malawi. MSc Thesis. 

Agronomy. University of Malawi, Bunda 

COllege of Agriculture. 

Sakala W.o. 1998. Nitrogen Dynamics in Maize 

(Zea mays) and Pigeonpea (Cajanus cajan) Intercropping 

in Malawi. PhD Thesis. University of 

London, Department of Biological Sciences, 

Wye College. 217 pp. 

Sakala, W.o., Kumwenda, J.D.T., Saka, A.R. and Kabambe, 

V.H. 2001. The potential of green manures 

to increase soil fertility and maize yields 

in Malawi. Soil Fert Net Research Results Working 

Paper Number 7. ClMMYT, Harare, Zimbabwe. 

8 pp. 

Young, A .1989. Agroforestry for Soil Conservation. 

CAB International, Wallingford, UK. 

Grain legumes and Green Manures for Soil Fertility in Southern Atrica 101

GREEN MANURING IN ZIMBABWE FROM 1900 TO' 2002 

LUCIA MUZA 

Agronomy Research Institute, AREX (formerly Departmentof Research and Specialist 

Services), Ministry of Lands qnd Rural Resettlement, 

CY550, Causeway, Harare, Zimbabwe 

Introduction 

The ploughing under of crops for green manuring became popular with farmers in Zimbabwe, then Rhodesia, 

during the 1900s. Green manuring was mainly practiced before planting a maize crop or potatoes, to help 

supply N to that subsequent crop. Mineral fertilizers were not yet widely used and the ratio of maize to legume 

green manure area reached 4:1 (Tattersfield, 1982). Several research questions were raised about greeri 

manuring and a series of 'trials were conducted, mainly at Harare Research Station, to address the questions 

that farmers had from their early experiences in the 1900s. This paper reviews some of those research areas 

and findings on green manuring in the early 1900-1930s and traces the resurgence of recent work during the 

1990s to 2000s in Zimbabwe. 

Green Manuring from 1900 to the 19505 

Screening of suitable green manure crops 

A need to identify suitable green manure species 

was addressed through the screening of nonleguminous 

and leguminous crops. A series of experiments 

was carried out for ten years, to screen 

crops such as niger oil, sunflower, geotani bean, kaffir, 

florida velvet bean, black velvet bean, sunnhemp 

and mixtures of these crops. Arnold (1909 

1930) summarized the ten-vear results in the Annual 

Reports of Salisbury (Harare) Agricultural Experiment 

Station and in a series of journal articles. 

Pure veivet bean, dolichos, sunnhemp and nigers 

oil, resulted in 8; 8.5; 17 and 15 t of above-ground 

biomass respectively at Harare. Sunnhemp was 

found to have the highest N mobilization in the 

above-ground biomass, whilst niger bean mobilized 

the highest amounts of P and K. Arnold noted that 

although sunnhemp produced the most green manure 

biomass, the subsequent maize crop was less 

vigourous compared with that after velvet bean, 

which had lower above-ground biomass. This may 

have been due to a high % N in the sunnhemp 

which results in rapid mineralization of N, not in 

synchrony with the N requirements of a subsequent 

maize crop. The lower N % (slightly less than 2% N) 

with velvet bean, resulted in a slower rate of N re· 

lease, more likely in synchrony with the needs of 

the subsequent maize crop. It was concluded that 

no single green manure species was suit",ble for all 

~oil types. Niger oil and sunflower generally gave 

lower subsequent maize grain yields. 

Legumes were found to be the best crops for green 

manu:ing because of their ability to fix atmospheric 

N for their own requirement and that of a subsequent 

crop. Sunnhemp, velvet bean and dolichos 

bean were identified as the best among a range of 

potential green manure legumes and there was no 

significant difference between these crops. Sunnhemp 

became popular with equally beneficial results 

chiefly because of its hardiness and suitability 

to a wide range of soils, its ability to smother weeds 

and ease of ploughing its residues under when used 

for green manuring. Sunflower, a non-leguminous 

crop, was found to be suitable for green manuring 

purposes as well because of its ability to produce 

high above-ground biomass (above 4 t/ha in most 

cases), although its effect on the subsequent maize 

was often lower than with the legumes (Arnold 

1928). 

Ploughing under green manure crops vs. harvesting 

green manure crops for hay 

Another research question of concern was whether 

it was more ' economical to plough under a green 

manure crop at flowering or leave the crop to mature 

and harvest the haulms and seed for hay or silage. 

A series of trials . was set up to answer this 

question. Livestock owners found it more profitable 

to use their leguminous crop for hay or silage. There 

was a need to apply farmyard manure to maintain 

soil fertility when legumes were harvested as hay. 

Ploughing under of the whole crop increased maize 

grain yield by 1370 kg/ha, compared to harvesting 

the green manure legume as hay. Removing aboveground 

biomass for other purposes like hay reduced 

the subsequent maize grain yield by 6% with 

sunnhemp, 16% with velvet bean, 11% with dolichos 

bean and 13 % with niger oil green manures. 

The difference between ploughing under and not 

ploughing under was smaller with sunnhemp than 

with the other crops. This was due to the great 

Grain Legumes and Green Manure~ for Soil Fertility in Southern Africa 103

amount of roots with sl:nnhemp. It was concluded 

that there was considerable merit in removing 

Zlbove-ground biomass and lIsing it for an alternative 

purpose such as livestock feed since the reduction 

in the subseq~lent maize crop yields when residues 

were removed was so small. 

Ploughing under green manure crops vs. burning 

Comparisons of ploughing-under mature sunnhemp 

or burning sunnhemp on the field or outside 

the field to avoid sterilization of the soil were reported 

by Arnold (1934, 1935, 1937 and 1939). Burning 

was to cater for farmers that had insufficient 

mZlchinery to plough ina green manure crop. There 

were no differences on the subsequent maize grain 

yield between burning and ploughing under of 

sunnhemp green manures, although the burnt plot 

hZld heZilthier crops during the early stages of crop 

growth. 

Application of mineral fertilizers to green manure 

crops vs. application of fertilizers directly to maize 

Another research question was raised by the Maize 

Association on whether it was more profitable to 

apply fertilizer to a green manure crop that is to be 

followed by maize or to apply the fertilizer directly 

to the maize crop after ploughing under an unfertilized 

green manure. The hypothesis was that a larger 

quantity of vegetative matter would be available for 

ploughing under if the fertilizer is applied to the 

green manure crop. The humus and N available to 

the following maize crop might be increased more 

than if the fertilizer is withheld for direct application 

to the maize. 

These trials were conducted on land of both moderate 

and low fertility status. Phosphorous (P) fertilizers 

were added to the green manure crop in the 

form of raw phosphate, bone meal and super phosphate 

(Arnold 1931 and 1933). Application of P to a 

green manure crop increased the above-ground biomass 

of green manures by an average of 3 t/ha. The 

response was larger on soils of low fertility. 

Twenty-three kg/ha of bone meal and super phosphate 

applied to sunnhemp or velvet bean increased 

maize grain yield from 1000 kg to 2500 kg/ha. It 

was found that maize on its own was unable to 

make the fullest use of the phosphate supplied. N 

fixing bacteria increased with the application of P, 

hence sufficient N was fixed to almost double the 

subsequent maize grain yields. The application of 

fertilizers, mostly P, to green manure crops was 

found to be economically justifiable in very exhausted 

lands particularly if a slow acting fertilizer 

such as rock phosphate was used. 

Residual effects of green manuring 

Arnold ' (1927) reported that the beneficial effects 

conferred by the green manure crop are pronounced 

during the first season after its application 

particularly if the season received heavy rainfall. In 

the second season after green manuring, the benefits 

were small. 

Effects of green manuring on microbial population 

of the soil 

Shepherd (1952) reported that green manuring 

stimulated the growth of antibiotic producing organisms 

and this was suspected to reduce pathogenic 

organisms, thus resulting in higher maize 

yields. Actinomycetes and moulds were shown to 

increase whilst bacteria numbers declined, possibly 

due to antagonistic mechanisms. 

Intercropping and relay cropping of green manure 

legumes with maize 

Loss of a cropping season to a green manure crop 

became

which did not thrive when shaded by maize. By 

June, the bean crop had covered the ground but further 

growth was retarded by frost. Ploughing under 

was in September with maize planted in December. 

The presence of a bean crop at the !11aize grain filling 

stage slightly reduced maize grain yield. Dolichos 

bean reduced the yield of the relayed maize 

crop and the subsequent maize crop, whilst white 

jack bean, khaki jack bean and dhal significantly increased 

maize grain yields (Arnold 1926-27). 

Incorporation of immature vs. mature green manure 

crops 

Arnold (1926, 1927 and 1929) reported results from 

a series of experiments that determined the effects 

of leaving green manure crops to mature before incorporation 

compared with ploughing under "the 

green manure crops at first flowering. The other primary 

objective of the trials was to determine 

whether the ploughing in of two consecutive green 

manure crops in the same season would have toxic 

effects on the land or whether the additional organic 

matter would be more beneficial than the ploughing 

under of one mature crop. Sunnhemp, velvet bean 

and dolichos bean were used in the experiment. Incorporation 

of mature crops was 5-6 weeks later 

than the incorporation of immature crops at flowering. 

This work found out that the growing season was 

too short to permit two velvet bean crops to mature 

unless they were ploughed under before podding. 

The biomass of matUre crops was double for velvet 

bean and dolichos bean and four times for sunnhemp, 

compared with two immature crops. Subsequent 

maize yields obtained after immature green 

manures were less than those from mature crops, 

mostly rl.ue to a higher biomass in the mature crops. 

One fully matured green manure crop was better 

than two immature crops. These were also compareg 

with the effect of a reaped mature velvet bean 

on the subsequent maize. 

These experiments also determined whether irrespective 

of mass of green manure per unit land area, 

immature crops ploughed under will ben~fit the 

land as mU(;h as if the crops are fully grown. Mature 

plants provide a higher percent of organic matter 

than immature plants. Fully developed crops had a 

more beneficial effect than a partially developed 

crop. Mature green .manure crops of velvet bean 

and sunnhemp more than trebled maize yield in the 

first season after green manuring whilst immature 

crops doubled maize yields compared with continu 

·ous unfertilized maize. The second season maize 

after both ma'ture and immature green manure 

crops did not benefit from green manuring. 

Timson (1946) also conc1uded that ploughing under 

of a green manure crop before the end of the rainy 

season led to excessive leaching of the nitrogen. 

Ploughing under at the end of April compared to 

February and March resulted in higher grain yields 

of the subsequent maize crop. 

Green manure crops in rotations 

Rotation experiments were carried out and green 

manure legumes were recommended in the differ 

ent rotation systems; in pure crop production sys 

tems as well as for crop and livestock farming sys 

tems. An example of a recommended rotation, de 

signed to meet the needs of a grain farmer whose 

income was solely dependent on maize and ground 

nut, is given below. 

Year 1 = Maize + fertilizer 

Year 2 =Green manure legume 

Year 3 = Maize 

Year 4 =Groundnut 

A supply of humus in the soil' was maintained by 

ploughing under the velvet bean and dolichos bean 

green manures. The green manure crops followed 

immediately after the maize crop that received mineral 

fertilizer, hence the green manure benefited 

from the resid ue of the fertilizer left in the soil. In 

this way, an adequate supply of humus was retained 

in the soil for the maize that followed the 

green manure crop. 

For dairy farmers, succulent legume crops were also 

included in the rotation to provide ample feed for 

livestock, for example. 

Year 1 =Maize plus farmyard manure 

Year 2 = Oats, velvet bean or dolichos bean mixtures 

for hay 

Year 3 = Maize plus mineral fertilizers 

Year 4 = Sweet potatoes (succulent crop for winter 

food for stockfeed) 

The rotational experiments highlighted the drawbacks 

of continuous (year-after-year) maize cropping 

on the same land. Over 13 years, maize yields 

were trebled in planned rotations compared with 

unplanried rotations similar to those found in smallholder 

communal areas, where continuous maize 

cropping is very common. Leaving the fields fallow 

was also · found to be less productive when compared 

to inclusion ofgreen manures for fodder purposes 

or exclusively as green manures. It was concluded 

that when farm stocks and crops are judi 

, ciously combined, the permanent fertility of the soil 

is increased and larger crops are secured. This was 

calculated to be profitable and a form of insurance 

against unfavorable seasons. 

! 


105

It was proved that a mixed farming system, which 

includes the raising of livestock and the growing of 

different kinds of crops, is more stable than one that 

relies on continuous production of the same crop. 

The practice of ploughing under of green manure 

crops to maintain the humus content of the soil became 

routine for many farmers in Zimbabwe in the 

late 1920s but the opportunity cost of committing 

land to a pure green manure crop limits adoption of 

the technology. 

. For the relaying of green manure crops in January 

or February under a maize canopy, it was concluded 

that the practice could not be relied upon to 

give profitable results because success is largely dependent 

on the amount of rainfall that falls during 

February and March. In a below normal season, 

maize yields are significantly reduced through competition 

with the green manure crops, or the green 

manures fail to grow. 

Fertilization of green manure crop 

In the late 1920s with the introduction of chemical 

fertilizers, the questions of whether it was more 

profitable to apply phosphate fertilizers to green 

manures or directly to a maize crop arose. A series 

of experiments were run comparing application of 

rock phosphate, bone and super phosphates to 

green manures or to maize. It was concluded that if 

the fertility of the land has been r,laintained at a 

moderately high level it was less economic to apply 

the fertilizer to a green manure crop and better to 

apply the fertilizer directly to the subsequent maize 

crop. On fields where previous cropping had reduced 

the soil fertility status to a very low level, the 

application of fertilizers to the green manure crop 

was found to be economically justifiable, particularly 

if a slow acting fertilizer such as raw phosphate 

rock was used. On depleted soil, the application 

of rock phosphate to a green manure crop increased 

the subsequent maize grain yield more than 

four times (Arnold 1931 and 1933). 

Saunders (1959) critically reviewed the available 

evidence on the value of green manuring in Zimbabwe 

up to the late 1950s. Work on continuous 

maize vs. maize in rotations with green manures in 

alternate years or with other crops or green manure 

crops removed for hay was reviewed. When alternate 

green manure were used with maize, slightly 

higher C and N levels in the soil were maintained 

compared with continuous maize cropping in the 

same fields. Use of green manure crops as hay was 

also better that continuous maize, althoush the soil 

CN ratio was not affected. The main benefit of legurne 

green manuring in Zimbabwe at that time was 

to augment available soil N, but it had also beneficial 

effects on the uptake of N, P, K and Ca. Average 

green manure crops were shown to contain 56-112 

kg N ha·1 while a very good crop could contain as 

much as 170 kg N ha- 1 • It was difficult to raise maize 

yields in non-green manure rotations to the same 

levels as those achieved after green manuring, but 

the loss of a season to a green manure crop resulted 

in a reduction in annual green manuring in Zimbabwe 

in the 1950s. 

Green manuring in the 19505 to 19805 

Under Zimbabwean conditions where the cooler 

winter months coincide with the dry seasof., green 

manuring involves the elimination of a productive 

cropping season. W,hether the loss of a season is 

compensated for by increased and sustained soil 

fertility benefits and cereal yields became a major 

issue for both farmers and researchers. With the 

widespread introduction of chemical mineral fertilizers 

in the 1950s and 1960s, use of green manuring 

continued to decline and it had almost disappeared 

from the 1960s to the 1980s. 

The major nutrient contributed by green manures 

was nitrogen but the introduction of cheap N fertilizers 

towards the end of the 1950s made the high 

opportunity cost of committing land, inputs and labour 

to green manuring unattractive to commercial 

farmers (Saunder, 1959) and this replaced the legume 

green manure practice (Tattersfield, 1982). 

Most smallholder farmers had no experience of 

growing green manures, but they did readily adopt 

inorganic fertilizers in the 1970s and 1980s when 

access was greatly improved through government 

schemes (Hikwa and Waddington, 1998). 

.Green Manuring in the 19905 to 2002 

Large rises in the prices of mineral fertilizer for 

smallholder farmers and renewed concern over the 

sustainability of current cropping systems dominated 

by continuous maize, led to renewed interest 

by Zimbabwean researchers in green manuring during 

the 1990s (Hikwa and Mukurumbira 1997). A 

feature of much of the new work was a focus on the 

needs of smallholder soil types and management 

systems. Most smallholders farm sandy · alfisols, 

characterized by coarse textured sandy surface horizons 

derived from granite. The soil structure is 

weak and highly susceptible to crusting and compa(:tion. 

Continuous mining of the soil through 

cropping with iittle fertilizer or organic matter input 

has further depleted the soil nutrients (Hikwa and 

Waddington 1998). . 

106 

Grain legumes and Green Manures for Soil Fertility iii Southern Africa

Jeranyama et al (1998 and 2000) reported on relayintercropped 

cowpea (a food crop) and sunnhemp 

green manure legume with maize in experiments on 

a sandy loam soil at Domboshava, Natural Region 

2. Legumes were planted 4 weeks after planting the 

maize. Herbage biomass (averaged over two seasons) 

was 2.3 t/ha for cowpez: and 3.1 t/ha for 

sunnhemp. Total N accumulation in the legume 

biomass was 111 kg N/ha for stumhemp and 59 kg 

N/ha for cowpea. Relay-intercropped maize fertilized 

with 60 kg N/ha had a grain yield equal to or 

better than those of a sole maize crop at the same 

fertilizer rate. However, at the other N rates, maize 

yields were reduced indicating competition between 

the maize and the legume. In the subsequent 

year, maize following relay intercropped legume 

with maize produced 20% more grain yield than the 

sole maize control. The grain N content of a subsequent 

maize crop was improved by 82% relative to 

the sole maize control. The legume contributed up 

to 36 kg N/ha to the subsequent maize crop. Other 

work on relaying maize and green manure legumes 

(Muza, 1998) reported that introdocing the green 

manuring legumes at 4 to 6 weeks after maize crop 

emergence ~as the best time, but that velvet bean 

tended to intertwine with maize. 

Chibudu (1998) reported on five years (1992-1996) 

of green manuring work with sandy low soil fertility 

status soils in Mangwende in Natural Region 2. 

Farmers, researchers and extension officers formulated 

and set up trials to screen legumes that could 

improve soil fertility, reduce Striga infestation and 

improve maize yields. The legumes used were velvet 

bean, sunnhemp, cowpea and dolichos in either 

a rotation or an intercrop with maize. The results 

showed that crops such as velvet bean, sunnhemp 

~nd cowpea could improve soil fertility, reduce 

striga infestation and subsequently increase maize 

yields. Farmers preferred to use velvet bean for improving 

soils in rotation but not intercropped with 

maize because it choked the maize plants making it 

difficult to harvest the maize crop. Cowpea was pre- 

Table 1. Maize grain yield (kg ha- 1 ) after green manuring with 

different legumes at Makoholi and Mlezu in 1990/91 (Agronomy 

Institute Annual Report) 

Preceding Makoholi Mlezu 

green 

manure 

(Inorganic N fertilizer (kg/ha)) 

crop 

0 40·: 80 120 0 40 80 120 

Dolichos 0.24 0.80 0.75 0.72 2.40 2.85 2.98 3.09 

Cowpea 0.19 0.55 0.84 0.93 2.83 3.35 3.40 3.65 

Sunflower 0.31 0.53 0.51 0.49 1.95 2.44 3.04 3.12 

Sunnhemp 0.44 0.57 0.54 1.38 3.15 2.64 2.88 3.20 

Soyabean 0.23 0.57 -0.78 0.95 1.61 2.73 2.10 2.51 

Maize 0.26 0.34 0.74 0.46 2.35 2.51 3.00 2.99 

Adapted from Agronomy Institute Annual Report. 1990-91 


ferred by farmers for striga control and provision of 

grain for food. 

The Agronomy Institute of the Department of Research 

and Specialist Services (now part of AREX) 

evaluated five potential green manuring species. 

Dolichos lablab, sunnhemp, soya bean, cowpea and 

sunflower were tested for their green manuring potential 

and their effect on a following maize test 

crop, on tWo sandy soils at Makoholi Experiment 

Station in Natural Region 4 and at Mlezu Agricultural 

College in Natural Region 3 Cvable 1). At 

Mlezu, biomass production was highest with dolichos 

(7.7 t/ha) and 7.0 t/ha with sunflower. Soyabean, 

sunnhemp and cowpea had 4.8, 2.7 and 1.6 t/ 

ha of above-ground dry biomass, respectively. At 

Makoholi, the biomass was 1.9, 3.3, 1.3, 1.5 and 1.7 

t/ha for dolichos, sunflower, soyabean, sunnhemp 

and cowpea, respectively .. Table 2 shows the biomass 

yield of v~lvet bean, stmnhemp and cowpea at 

three locations in Zimbabwe in 1995/96, reported 

by Muza, Gatsi, Pashapa and Bwakaya in 2000. The 

nitrogen, phosphorus and potassium contents of the 

green manures in those experiments are shown in 

Tables 3-5. 

Table 6 shows biomass production by velvet bean, 

sunn..~emp and fish bean (Tephrosia vogelii) in relation 

to phosphorus application in Soil Fertility Network 

trials in 1996/97. In the 1996/97 season, 

twelve farmers fields were selected, ten in Natural 

Region 2, one in Natural Region 3 and one in Natural 

Region 4. Either the selected fields were abandoned 

fields due to low soil fertility, or fields where 

Table 2. legume above-ground biomass (kg/ha) at different 

planting times (weeks after maize planting) and sites in 1995/96 

Site legume 4 weeks 6 weeks 8 weeks 

Chiwundura Velvet bean 10556 463 249 

Sunnhemp 2256 162 76 

Cowpea 1 081 307 ~01 

Dolichos 860 83 98 

Tephrosia 23 131 488 

Pigeon pea 359 153 79 

Chihota Velvet bean 4473 178 0 

Sunnhemp 2469 1 097 294 

Cowpea 1 039 336 0 

Dolichos 218 0 318 



Mlezu Velvet bean 3148 1 788 445 

After MUla et al 2000 

Sunnhemp 9554 821 805 

Cowpea 4699 2223 1030 

Dolichos 2875 920 160 


Pigeon pea 538 301 159 

107

Table 3. Above·ground biomass (t/ha) and N, Pand K 

contents (kg/ha) in biomass of green manures grown at 

- Makoholi in 1989/90 

Dry biomass N f' K 

(t/ha) 

Dolichos 1.9 46.7 6.1 39.2 

Cowpea 1.7 41.9 5.2 34.9 

Sunflower 3.3 41.2 8.2 94.1 

Sunnhemp 1.5 40.7 5.1 25.5 

Soyabean 1.3 26.7 4.7 18.4 

Table 4. Average %Nitrogen and %Phosphorus in velvet bean, 

sunnhemp and cowpea above·ground biomass and root!; at time of 

incorporating in April 1996 . 

Velvet bean Sunnhemp Cowpea 

Above ground Roots Above ground Roots Above ground Roots 

biomass biomass biomass 

Nitrogen 1.9 1.38 3.00 0.84 2.16 1.52 

Phosphorus 0.13 0.17 0.12 0.04 0.18 0.14 

MUla et al 2000 

maize grain yields in recent years were less than 500 

kg/ha. Soil pH ranged from 4.1 to 4.8 and there 

was no correction for pH. Velvet bean, surmhemp 

and fish bean were planted with 100 kg/ha P20S or 

without phosphorus. 

Biomass production by the three legumes is shown 

in Table 6 and the grain yield of the maize test crop 

after green manuring forChihota and Zvimba 

(where a maize crop was harvested) are in Table 7. 

Velvet bean performed the best, with six fields gen 

erating an above-ground biomass of over 4 t/ha 

when phosphorus was applied whilst four plots 

wi~h no phosphorus also produced a biomass above 

4 t/ha. Surmhemp performance on the degraded 

soil was very variable. Dieback of the plants after 

crop emergence was common at most sites. 

Velvet .bean gave reasonable biomass on extremely 

nutrient depleted and somewhat acidic soils and 

has the potential to rehabilitate degraded fields 

when coupled with lime and phosphorus. Low pH 

and P levels in the soil inhibit legume growth; hence 

P and lime should be added. 

There is still a need to screen more potential green 

manures, to expand the legume base. 

Green manuring extension work in Chihota 

In the 1999/2000 season, four technologies were se 

lected to help smallholder farmers in Chihota com 

munal area to sustainability improve the crop pro 

ductivity of their farms through improverl soil fer 

tility management practices. This pilot project was 

led by the extension sen-ice in Marondera District. 

Green manuring was one of the technologies se 

1able 5. Total nitrogen and phosphorus (kg ha· 1) in ab.ove· 

ground biomass during the 1995/96 season 

Site Velvet bean Sunnhemp Cowpea 

N PzOs N PZOI N PZOI 

Chiwundura 207 14 68 3 23 2 

Chihota 87 6 74 3 22 2 

Mlezu 62 4 287 12 102 8 

MUla et al 2000 

Table 6. Dry biomass production (kg/ha) by three green manure 

legumes, on exhausted sandy soils in nor\hern Zimbabwe, 1996/97 

season 

Communal Area Velvet bean Sunnhemp Fish bean 

+ P . P + P . P + P . P 

Gokwe South (1) 2368 1916 1688 858 0 0 

Gokwe South (2) 1826 1964 809 1000 0 0 

Nyazura (1) 8020 7240 0 0 0 0 

Nyazura (2) 6490 6610 0 0 0 0 

Chiduku (1) 1257 1865 grazed grazed 70 34 

Chiduku (2) 4538 2703 116 13 64 66 

Mangwende (1 ) 318 317 311 290 145 145 

Mangwende (2) 5351 5250 5000 5040 3127 3125 

Zvimha (1) 2410 1260 0 0 0 0 

Zvimba (2) 85U 1620 0 0 0 0 

Chihota (1) 10665 5290 8460 2315 0 0 

Chihota (2) 4275 3405 505 550 0 0 

AIter Hikwa et al 1998 

lected and 411 farmers participated, in farmer 

groups, in demonstrations of green manuring on 

their farms. Generally, it was found that green manuring 

was a new technology to most of the farmers. 

Few had tried it or seen it. Forty percent of the 

experimenting farmers tried green manuring on 

their own fields whilst 83 farmers outside the 

groups also used it (Mwenye and Kuwaza, 2001). 

There is still a great need to expose far more farmers 

to green manuring through working with other extension 

districts in Zimbabwe. 

Current Work and the Future 

Green manuring work in Zimbabwe is still going 

on, with the Agronomy Research Institute looking 

at the possibilities of · combining mulching using 

green manure legumes and minimum tillage. The 

University of Zimbabwe and the Agronomy Institute 

are also experimenting with different green manures 

to control Striga. Agronomy Institute, Crop 

Breeding Institute and ICRAF are researching the 

possibilities of intercropping Sesbania sesban with 

velvet bean and bushy and trailing cowpea. Researchers 

on livestock feeds at Research stations and 

the University of Zimbabwe are also looking at the 

suitability of velvet bean for use in stock feeds. The 

108 


Table 7. Maize grain yield (kg ha 1 )in 1997/98 following sunnhemp and velvet bean green 

manures grown during 1996/97 in Chihota and Zvimba Communal Areas, Zimbabwe 

Treatment Chihota· Chihota· 

Chigora Chimhembeza 

Maize + OP + 60N 180 157 

Maize + OP + ON 357 308 

Maize + lOOP + 60N 1665 223 

Maize + lOOP + ON 325 165 

Velvet bean (incorporated) + lOOP + 45N 2863 556 

Velvet bean (incorporated) + lOOP + ON 4240 308 

Velvet bean (biomass removed) + lOOP + 45N 3982 1688 

Velvet bean (biomass removed) + lOOP + ON 2532 587 

Velvet bean (incorporated) + OP + 45N 1745 1266 

Velvet bean (incorporated) + OP + ON 2407 710 

Velvet bean (biomass removed) + OP + 45N 143 1010 

Velvet bean (biomass removed) + OP + ON 731 664 

Sunnhemp (incorporated) + lOOP + 45N 4726 1387 

Sunnhemp (incorporated) + lOOP + ON 3628 410 

Sunnhemp (biomass removed) + lOOP + 45N 4715 3104 

Sunnhemp (biomass removed) + lOOP + ON 5984 1989 

Sunnhemp (incorporated) + OP + 45N 2661 516 

Sunnhemp (incorporated) + OP + ON 2082 1120 

Sunnhemp (biomass removed) + OP + 45N 144 951 

Sunnhemp (biomass removed) + OP + ON 890 1063 

Adapted from Murata et al. 2000 

Soil Conservation and Tillage Network based at the 

University of Zimbabwe is also working with some 

of the green manuring legumes in soil conservation 

and proposing their us~ as cover crops. 

Arnold, H.C. 1927. Maize following 

green ·manure crops sown under 

Zvimba· Mean maize the previous season. Rhodesia 

Chimedza Agricultural Journal 24:533-534. 

175 171 

699 455 Arnold, H.C. 1927. Green manuring 

with immature versus mature crops. 

123 670 

Rhodesia Agricultural Journal 24:527 

272 254 

529. 

1395 1605 

1794 2114 Arnold, H.C. 1928. Green manuring 

273 1981 with iIl1mature versus mature crops, 

207 1109 Rhodesia Agricultural Journal 25:309 

1639 1550 311. 

1084 

1355 

1030 

1400 

836 

808 

Arnold, H.C. 1928. Maize following 

green manure .crops sown under 

maize the previous season. Rhodesia 

1503 2539 

Agricultural Journal 25:313-314. 

1285 1774 

564 2794 Arnold, H.C. 1929. Maize following 

182 2718 green manure crops sown under 

1021 1399 maize the previous season. Rhodesia 

1214 1472 

1210 768 


Arnold, H.C. 1929. Green manuring 

1047 1000 with immature versus mature crops, 

Rhodesia Agricultural Journal 26:362 

363. 

Arnold, H.C. 1931. The relative value of certain 

green manure crops. Rhodesia Agricultural Journal 

28:541-544. 

Seed availability is one of the major limiting factors 

to green manuring, hence there is a need to establish 

a sustainable source of green manure legume 

seed near the farming communities. In Chihota, a 

school has been asked to bulk velvet bean seed for 

local farmers. 

Bibliography 

Agronomy Institute 1993. Effect of green manure 

crops on the performance of a second season 

maize test crop. Agronomy Institute Allnual Report 

for 1989/90, Departm~nt of Research and 

Specialist Services, Harare, Zimbabwe. pp. 160 

164. 

Arnold, H.C. 1925 -26. Annual Report of Salisbury 

Experiment Station. 

Arnold, H.C. 1926 -27. Annual Report of Salisbury 

.Experiment Station. 

Arnold, H.c. 1926. Green manuring with immature 

versus mature crops. Rhodesia Agricultural Journal 

23:317-318. 

Arnold, H.C. 1931. Fertilizer and green manuring 

trials. Rhodesia Agricultural Journal 28:544-548. 

Arnold, H.C. 1933. Fertilizer and green manuring 

trials. Rhodesia Agricultural Journal 28:483-484. 

Arnold, H.C. 1934: Effects of burning mature sunnhemp 

or ploughing ·under were compared. Rhodesia 

Agricultural Journal 31 :648-650. 

Arnold, H.C. 1935. Effects of burning mature sunnhemp 

or ploughing under were compared. Rhodesia 



or ploughing under were compared. Rhodesia 



or ploughing under were compared. RhO., 

desia Agricultural Journal 35:737-739. 

Backshaw, G.N. 1921. Green manuring and soil 

management. Rhodesia Agricultural Journal 

18:455-460, 645-647. 

Bary, F. 1995. Socio economic !Study 0f agriculture 


and horticulture in Zimbabwe: Case study of 

Wedza and Buhera. Background information for 

biotechnology research and development projects. 

Commissioned by Zimbabwe Biotechnology 

Advisory Committee (ZIMBAC) and DGIS 

Special Programme oiotechnology, the Netherlands. 

Bellon, M.R., P. Gambara, T. Gatsi, T.E. Machemedze, 

O. Maminimini, and S.R. Waddington. 

1999. Farmers' Taxonomies as a Participatonj 

Diagnostic Tool: Soil Fertility Management in Chihota, 

Zimbabwe. CIMMYT Economics Working 

Paper No. 99-13. Mexico, D.F.: CIMMYT. 

Chibudu, C. 1998. Green manuring crops in a maize 

based communal area, Mangwende: Experiences 

using participatory approaches. In: Soil Fertility 

Research for Maize-Based Farming Systems in Malawi 

and Zimbabwe (S.R. Waddington, H. Murwira, J.D. 

T. Kumwcnda, D. Hikwa and F. Tagwira, eds), Soil 

Fert Net and CIMMYT-Zimbabwe, Harare, Zimbabwe. 

pp. 87-90. 

Grant, P.M. 1970. Restoration of production of depleted 

sands. Rhodesia Agricultural Journal 67:134 

137. 

Giller, K.E., Gilbert, R., Mugwira, L.M., Muza, L., 

Patel, B.K. and Waddington, S.R. 1998. Practical 

approaches to soil organic management for smallholder 

maize production in southern Ahica. In: Soil 

Fertility Research for Maize-Based Farming Systems in 

Malawi and Zimbabwe (S.R. Waddington, H. Murwira, 

J.DT. Kurnwenda, D. Hikwa and F. Tagwira, 

eds), Soil Fert Net and CIMMYT-Zimbabwe, Harare, 


Hikwa, D., Murata, M., Tagwira, F., Chiduza, c., 

Murwira, H., Muza, L., and Waddington, S.R 

1998. Performance of green manure legumes on 

exhausted soils in Northern Zimbabwe; A soil 

fertility network trial. In: Soil Fertility Research for 

Maize-Based Farming Systems in Malawi and Zimbabwe 

(S.R. Waddington, H. Murwira, J.D.T. Kumwenda, 

D. Hikwa and F. Tagwira, eds), Soil Fert 

Net and CIMMYT-Zimbabwe, Harare, Zimbabwe. 

pp.81-84. 

Hikwa, D. and S.R. Waddington. 1998. Annual legurnes 

for improving soil fertility in the smallholder 

maize based systems of Zimbabwe. 

Transactions of the Zimbabwe Scientific Association 

72(supplemenf):15-26. 

Hikwa, D. and L. Mukurumbira. 1997. Evolution of 

the Department of Research and Specialist Services 

(DR&SS) proje~t on integrated soil fertility 

management in the small holder sector of Zimbabwe. 

In: Target (The Newsletter of the Soil Fertility 

Research Network for Maize-Based Cropping 

Systems in Countries of Southern Africa) 

9:6-7. 

Jeranyama P, Hesterman O.B. and Waddington S.R. 

1998. The impact of legumes relay intercropped 

into maize at Domboshava, Zimbabwe. In: Soil 

Fertility Research for Maize-Based Farming Systems in 

Malawi and Zimbabwe (S.R. Waddington, H. Murwira, 

J.D.T. Kurnwenda, D. Hikwa and F. Tagwira, 

eds), Soil Fert Net and CIMMYr-Zimbabwe, Harare, 


Jeranyama P., Hesterman O.B ~ , Waddington S.R. 

and Harwood, R.R. 2000. Relay-intercropping of 

sunnhemp and cowpea into a smallholder maize 

system in Zimbabwe. Agronomy Journal 92:239 

244. 

Murata, M., Waddington, S. and Murwira, H. 2000. 

Rehabilitation of degraded sandy soils using annual 

green manures in Zimbabwe, In: Target 

(The Newsletter of the Soil Fertility Research 

Network for Maize-Based Cropping Systems in 

Countries of Southern Africa) 21:4. 

Muza, L. 1996. Green manuring in the communal 

areas of Zimbabwe. In: Research Results and Network 

Outputs in 1994 and 1995. Proceedings of the 

second meeting of the Soil Fertility Network 

Working Group, .held at Kadoma Ranch Motel, 

Zimbabwe, 18-21 July 1995, CIMMYT, Harare, 


Muza, L. 1997. Integrated soil fertility management 

through legume green manuring in combination 

with inorganic fertilizer in communal areas of 

Zimbabwe. Proposal for funding renewal to the 

Rockefeller Foundation, 1997. 

Muza, L. and Mapfumo, P. 1998. Constraints and opportunities 

for legumes in the fertility enhancement 

of sandy soils in Zimbabwe. In: Maize Production 

Technology for the Future: Challenges and Opportunities, 

Proceedings of the Sixth Eastern and 

Southern Africa Regional Maize Conference held 

in Addis Ababa, Ethiopia, 21-25 September 1998. 

CIMMYT and EARO. pp. 214-217. 

Muza, L. 1998'. Selecting green manure legumes for 

relay and intercropping systems with maize on 

sandy soils in Zimbabwe, In: Cover Crops in West 

Africa, Contributing to Sustab1able Agriculture, (D. 

Buckles, A. Eteka, O. Osiname, M. Galiba and Et 

G. Galiano, eds). IORC, Canada. pp. 251-257. 

Muza, L., L. Pashapa and S. Feresu. 1998. Need to revive 

green manuring in soil fertility management 

110 


in Zimbabwe. The Zimbabwe Science News 32:51-53. 

Muza, L., Gasti, T., Pashapa, L. and F. Bwakaya. 

2000. Soil fertility management through legume 

green manuring in combination with inorganic 

fertilizers in communal areas of Zimbabwe. 

(Rockefeller Grant No. 1994-0020-0048, Report 

covering 1994/95 to 1999/2000 seasons, 

submitted to Rockefeller Foundation in August 

2000). 

Mwenye, D. and C. Kuwaza. 2001. Chihota soil fertility 

project, Agritex, Marondera distrkt, Mashonaland 

East Province, Zimbabwe. End of Season 

Report for the Period 1999-2000. 

Rattray, A.G.H. and B.s. Ellis, 1952. Maize green 

manuring in Southern Rhodesia. Rhodesia Agricultural 

Journal 49:188-197. 

Tattersfield. J.R, 1982. The role of research in increasing 

food potential in Zimbabwe. Zimbabwe 

Science News 16:6-11. 

Timson, S. D. 1946. Green manuring; when to 

plough down the crop. Rhodesia Agricultural JournaI43:11-15. 

. 

Saunder, D.H., 1959. What is the value of green manuring 

in Rhodesia? Proceedings of the fourth annual 

conference of the professional officers of the 

Department of Research and Specialist Services, 

held at the University College of Rhodesia and 

Nyasaland, 9-11 June, 1959. 

·Shepherd, c.J. 1952. Effects of green manuring on 

microbial population of the soil. Rhodesia Agricultural 

Journal 49:198-202. 


GRAIN LEGUMES AND GREEN MANURES IN EAST AFRICAN MAIZE 

SYSTEMS - AN OVERVIEW OF ECAMAW NETWORK RESEARCH 

DENNIS K. FRIESEN 1 , R. ASSENGA 2 , TESFA BOGALE 3 , T.E. MMBAGA 4 , 

J. KIKAFUNDA 5 , WAKENE NEGASSA 6 , J. OJIEM 7 and R. ONYANG0 8 

1CIMMYT/IFDC, Nairobi, Kenya; 

2Agricultural Research Institute-Mlingano Tanzania; 

3Jimma Agricultural Research Center, Jimma, Ethiopia; 

4Selian Agricultural Research Institute, Arusha, Tanzania; 

5 Namulonge Agriculture and Animal Production Research Institute, Kampala, Uganda 

6Bako Agricultural Research Center, Bako, Ethiop;a; 

7Kakamega Regional Research Center, KARl, Kakamega, Kenya, and 

8Kitale National Agricultural Research Center, KARl, Kitale, Kenya 

Abstract 

The Eastern and Central Africa Maize and Wheat (ECAMA W) Research Network, established in 1996, is one of 18 networks 

operating under the Association for Strengthening Agricultural Research in Easterna.nd Central Africa 

(ASARECA). ECAMA W addresses constrain:s to maize and wheat production in the ten ASARECA member countries 

where maize is the number one priority crop and soil fertility is ranked as one of the principal constraints to improved 

maize productivity and production. Nitrogen (N) is the most limiting nutrient in the region yet smallholder 

farmers use very little fertilizer inputs due to high cost, poor infrastructure and risk due to climatic uncertainty. Legumes 

in systems with maize are a potential alternative source of N for the maize crop. During the past 5 years, the 

ECAMA W Network has funded 12 small ~rant projects dealing with green manure and grain legumes in systems with 

maize. Network collaborators have implemented some 24 on-station experiments and 195 on-farm trials to evaluate and 

identify suitable adapted grain legume and green manure species, and to quantify their impact on maize production in 

systems including intercrops, relay crops and rotations. Some 12 legume species were evaluated for nodulation, ground 

cover, resistance to pests and diseases, biomass production, seed production, etc. in the moist and dry mid-altitude, and 

lowland ecologies of Ethiopia, Kenya, Tanzania and Uganda. Mucuna pruriens, Canavalia ensiformis, Crotalaria 

ochroleuca and Dalicos lablab were the most widely adapted and most effective N providers, although other species 

were 10caUy more suited. Green manure legumes intercropped with maize had no significant beneficial effects on maize 

grain yields and, depending on their aggressiveness, sometimes significantly reduced maize yields. Green manure biomass 

praduction was reduced in intercrops and more so when relayed into maize. Depending on the degree of growth 

suppression and the duration offollow-on growth permitted after the maize harvest, green manures had either little or 

substantial effects on maize yields in the following season. The effects of green manures rotated with maize. had more 

consistent and substantive effects on subsequent maize yields with increases as much as 385%, or 2.5-3.0 t/ha, on farmers' 

fields. Grain legumes, including soybean, cowpea, green gram and pigeonpea, had little beneficial or negative effect 

on maize productivity whether grown as intercrops or in rotations. Farmers' reactions to green manures was mixed, 

from reluctance to plant a crop which produced no food to appreciation of the weed suppressing effects and soil fertility 

gains they provided. A frequent question regarded the palatability of mucuna and canavalia seed. Despite considerable 

exposure to green manure legumes, farmers have been slow to adopt them into their farming systems. On the other 

hand, grain legumes, which produced a consumable or marketable product, were highly valued by farmers. 

Key words: ECAMA W Network, legume adaptation, intercropping, relay crops, rotations 

Introduction 

Maize is grown on more than 7.6 M hectares in East 

ern and Central Africa with an average yield less 

than 1.3 t/ha (compared to a potential of 4.5-7 t/ha) 

(Pingali,2001). Average per capita consumption of 

maize grain is 50 kg, but it ranges from 12-103 kg 

per person. Given the large area planted, and its 

importance as a food and cash crop, maize was 

id~ntified as the number one priority for regional 

research by the Association for Strengthening Agricultural 

Research in Eastern and Central Africa 

(ASARECA). Low soil fertility, especially nitrogen 

(N), is one of the principal constraints to increased 

maize productivity in the region (ECAMA W, 1999). 

Fertilizer use is less than 10 kg/ha/yr (Bumb and 


Baanante, 1996; Heisey and Mwangi, 1996) due to (i) 

high price and poor infrastructure, (ii) risk due to 

uncertainty in climate and the price of produce, and 

(iii) lack of access to credit for smqll holders. 

The Eastern and Central Africa Maize and Wheat 

(ECAMAW) Research Network is a network of 

maize and wheat scientists from the National Agricultural 

Research Systems of the ten countries in 

Eastern and Central Africa operating under the Sub 

Regional Organization, ASARECA. ECAMAW scientists 

address priority constraints of regiunal importance 

to improved maize and wheat production 

and productivity and operate through a system of 

small project grants overseen by a Steering Committee, 

a Network Coordinator and CIMMYT project 

scientists that fund the small grants program. 

Due to the poor access farmers have to fertilizers, 

ECAMA W scientists have focussed on green manures 

and grain legumes as alternative sources of N 

for maize systems. The potential for legumes to supply 

N to cropping systems is well known, and the 

benefits and constraints were recently reviewed by 

Giller et al. (1997). Legumes in cropping systems 

can be broadly classified as those that produce a 

consumable seed (grain legumes) and those that are 

grown solely for agronomic purposes, such as a 

source of biologically fixed N (green manures), 

weed control. and ground cover. While grain legumes 

can fix substantial amount of N, with few exceptions 

(e.g., groundnut, cowpea, pigeonpea), most 

of the fixed N is harvested with the grain and little 

is left to the soil and subsequent cereal crops. Green 

manures provide considerable N to the soil when 

grown in rotations with crops but also remove land 

hom production to gain that benefit. Both grain legumes 

and green manures grown as intercrops suffer 

from competition from the companion crop, reducing 

biomass accumulation, biological N fixation and 

the potential benefits to the systems. 

During 1997-2002, the ECAMA W Network supported 

12 small grant projects, each spanning periods 

of two or more years and often implemented 

across several sites, to evaluate grain legumes and 

green manures in maize systems. Most of these projects 

were executed on-farm with farmer participation 

at multiple sites. The objectives of this research 

were to: 

• Identify suitable adapted green manure and grain 

legume species for the major ecologies of ECA; 

• Evaluate appropriate management practices for 

them in intercrops, relay crops or rotations with 

maize; 

& Quantify the impact of green manures and grain 

legumes on maize pFOductivity; 

• Determine ihe fertilizer-N equivalence of green 

manures in rotations; and 

• Evaluate green manures and grain legumes in systems 

on-farm with farmers to ascertain farmers' 

perceptions and acceptance. 

This paper summarizes the results of these regional 

network trials and describes on-going and future 

research and dissemination activities of ECAMA W 

network scientists with legumes in maize-based systems. 

Methods 

Evaluations of legumes for adaptation, biomass 

production and N-fixation 

Regional trials were established at Namulonge 

(Uganda), Arusha (Tanzania), Tanga (Tanzania), 

Jimma (Ethiopia) and Kakamega (Kenya) to screen 

green manure and grain legume species for adaptation 

to the local environment. A core set of 12 species 

(Table 1) were generally evaluated at all sites; 

an additional 10 species (Oolichos-Renga, Cajanus cajan, 

Pueraria phaseoloides, Voandzeia subterranea, Crotalaria 

brevidens, Oesmodium intortum, Lablab purpureus, 

Macroptilium atropurpureum, Phaseolus vulgaris 

(cv. Selian wonder) and green gram) were 

evaluated at single selected sites. Species were 

sown in'small plots on station at the onset of the 

rains and were scored at appropriate periods for 

establishment, nodulation, percent ground cover, 

resistance to pests and diseases, seed and biomass 

production among other criteria. N supply capacity 

was estimated from the total biomass production 

and N content of the biomass. 

Effects of green manures and grain legumes in 

maize-legume systems 

Trials were conducted on station and on farm to 

evaluate promising legume species (based on re 

gional screening trials) in systems with maize. Sys 

tems included the following: 

• Rotations within a year (bimodal rainfall) orin alternate 

years (monomodal rainfall distribution); 

• Relays, including the effect of relay date on green 

manure biomass production and sequenced maize 

production; and 

• Intercrops of green manure or grain legume species 

with maize. 

The effect of legume species and system on maize 

productivity (yield of maize grain per hectare) was 

measured. In some cases, the effect of maize on legume 

biomass production was also determined, including 

the N content of the aboveground biomass 

where possible. All results were subjected to analyses 

of variance and means were separated by the 

Duncan's Multiple Range Test where appropriate. 

114 


Results 

Table 1. Adaptation of green manure and grain legume species to the moist and dry mid·altitude and tropical 

lowland ecologies of Eastern Africa" 

Legume adaptation in Legume species Establish· Nodula· Ground Diseases Seeding 

ment 

• 

tion cov·er and pests capacity 

.ECA ecologies 

Table 1 summarizes re Calopogonium mucunoides 2 2 1 

~ 

sults of regional legume Canavafia ensiformis 2 

species evaluation trials Crotalaria ochroleuca 2 

across five sites for 12 Dolichos lablab 7.6 

species according to six Mucuna pruriens (black) 10.6 

criteria using a 5-point Mucuna pruriens (white) 8.3 

scale from very good Sesbania sesban 

through fair to very Glycine max (Soybean-Nyala) 

poor. Species consid- Glycine max (Soybean.SCs) 

ered include green ma- 

Vigna unguiculata (Cowpea) 

nures and grain legumes. 

Visually, the Vicia dyascarpa (=lana vetch) 

most adapted green ma- Vicia vil/osa (=purple vetch) 

nure species appeared to 

be Mucuna pruriens, Ooli 

2 

"Legend for evaluations 

v.good ' good 

•• 

does not include potential further biomass accumu 

cos lablab, Crotalaria 

ochroleuca and Canavalia ensiformis although there 

were regiona1 variations in adaptation. Grain legume 

species tended to be less adapted than gre.en 

manure species but this may be a reflection of the 

evaluation criteria tha t favoured attribu tes for soil 

fertility enhancement and sequenced maize production. 

lation that may occur if they are allowed to continue 

growing on residual moisture subsequent to the 

maize harvest, the possibility of which depends on 

the presence or absence of free-ranging cattle that 

are often allowed to graze crop residues in these 

systems. 

 

The potential contribution of green manure and Effects of legumes in rotations on maize producgrain 

legume species to soil N status is shown in tion 

Table 2, based on the mean biomass production The effects of green manures grown in rotations 

across one or more sites in the region and thei; 

measured or estimated N contents. 

Table 2. Biomass production and estimated nitrogen content of green manure and 

Green manures were grown as sole 

grain legume residues at ECAMAW regional screening sites (1998-99) 

crops and sampled at the end of the sea 

No. of 

Mean N 

son at a· growth stage considered appro Legume species Common name Biomass yield 

sites 

content§ 

priate for incorporation into the soil. min max mean 

With few exceptions, the mean levels of -- _.. (I-OM/ha) · -- - - (kg-N/ha) 

N provided by green manures were well Calopogonium mucunoides Calopo 3 1.4 4.2 3.U 61 

in excess of a maize crop's requirements. Canavalia ensiformis Jackbean 5 2_9 18.2 12.5 316 

Grain legumes such as cowpea and Crotalaria brevidens Sunhemp 3.4 3.4 3.4 85" 

groundnut left suboptimal amounts of Crotalaria ochroleuca 4 2.0 15.0 8.1 267 

N for a subsequent maize crop. 

lablab purpureus Oolicos lablab 5 2_1 16.6 7_6 131 

Macroptylium atropurpureum Siratro 2.0 2.0 2.0 50" 

Mucurla pruriens (black) Velvet bean 5 2.5 20.7 10.6 289 

Shading and competition for water and 

Mucuna pruriens (white) Velvet bean 2 4.5 12.0 8.3 208" 

nutrients reduces the growth of green 

Pueraria phaseoloides Tropical kudzu 2.1 2.1 2.1 33" 

manures sown as intercrops with maize, 

Sesbania sesban Sesban 12.3 12.3 12.3 308" 

and hence reduces the amount of N Vicia dasycarpa Lana vetch 2 0.5 3.0 1.8 45" 

available for subsequent maize crops. Vicia vil/osa Purple vetch 2 0.6 5.0 2.8 70" 

Figure 1 compares biomass production Cajanus cajan Pigeon pea 1 17.1 17.1 17.1 428" 

of mucuna, canavalia and crotalaria Glycine max (Nyala) Soybean 3 0.5 4.7 3_3 83" 

grown as sole crops, or intercropped Glycine max (SCs-l) Soybean 3 0.3 3.7 1.5 117 

with maize 2-3 weeks after maize emer Phaseolus vulgaris Field bean 0.1 0.1 0.1 3" 

.gence, or relay planted into maize at 2 Vigna radiata Green gram 1 2.0 2.0 2.0 50" 

weeks after tasseling at Jimma, Ethiopia, Vigna unguiculata Cowpea 2 1.2 4.5 2.9 70 

and Namulonge, Uganda. Legume bio 

Bambara 

Voandleia subterranea 1.1 1.1 1.1 28" 

groundnut 

mass was measured at the time of har 

§ mean Ncontents calculated from measured N concentrations of biomass except those marked with 

vesting the maize and, consequently, 

* 

·which are estimated Ncontents based on an average concentration of 2.5% N in the biomass. 

Grain Legumes and Green Manures for Soil Fertility in Southern Africa 1/5

10 .-----------------------____________~ with maize in research station trials at Bako and 

Jlmma (2000) Namulonge (1999) 

- Jirnrna (Ethiopia), Mlingano (Tanzania), Kakamega 

(Kenya) and Namulonge (Uganda) are summarized 

.Sole crop. 

in Table 3. These trials generally compared maize 

~ Inler-cro p 

response to the legume sown in the preceding sea 

'C 6 _Relay crop 

CD 

son with response to the recommended level of N 

>. 

fertilizer. Fertilizer N at locally recommended rates 

If! 

II> 4 

IV 

(see Table 3 footnote) generally increased yields by 

E 

o 

46-108% across sites. With the exception of Mlin 

iii 2 

gano, legume rotations consistently produced significantly 

higher maize yields than unfertilized 

o 

maize in monocrop systems and usually as great or 

Mucuna Canavalia Crotolarla Mucuna Canavalia greater yields than fertilized maize in monoculture. 

Yield gains ranged from 1.5-3.5 t/ha or 27-134%. 

Figure 1. Effect of intercropping and relay cropping on biomass The poor response to the legume rotation at Mlin 

production of legumes in maize·legumes systems at Jimma, Ethiopia, gano may be due to soil hydraulic properties and 

and Namulonge, Uganda. (lnter·cropped legumes sown into maize 2 high rainfall leading to leaching of N mineralized 

3 weeks after maize emergence; relayed legumes sown into maize 2 

weeks after maize tasseling.) 

from the resid~es in the intervening dry season. 

_During 1998-2001, on-farm trials to 

Table 3. Response of maize grain yield· (t/ha) to green manures sown in the preceding evaluate and promote green maseason 

(year) and incorporated prior to sowing maize in the current season - on-station nures in rotation with maize were 

trials 

carried out in Ethiopia, Kenya, Tanzania 

and Uganda on more than 70 

Maize·green 

Ethiopia Tanzania Kenya Uganda 

maRure rotation§ 

farms; 42 were successfuUy harvested 

(Table 4). Green manure spe 

------------------ 

Bako Jimma Mlingano Kakamega Namulonge 

1998 2000 2001 2000 1998 2000 20.01 cies included mucuna, crotalaria, 

Dolicos lablab and canavalia_ Re 

Maize - Fert-N 3.12 a 5.00 a 1.95 a 2.05 a 5.43 a 3.11 a 3.93 a 

sponse of maize to the preceding 

Maize + Fert-N# 5.12 b 8.67 b 3:15 be 4.27 b 4.54 be 6.94 e 

season's green manure crop was 

Mucuna 2.92 b 2.50 a 5.01 e 5.65 b 

compared to monocropped maize 

Canavalia 3.85 cd 2.70.a 6.87 b 4.00 b 6.47 be with and without the recommended 

Crota/aria 8.48 b 4.56 d 7.2~ b rate of fertilizer N for the area. Re 

Sesbania 8.18 b - sponses to fertilizer N were gener 

Dolieos lablab 5.20 ~ ally greater on farmers' fields than 

§ Gfeen manures sown in preceding year and incorporated before sowing maize the following year, except on-station, ranging from 1.4-3.3 t/ 

Kak1imega and Namulonge where rainfall is bimodal and green manures were sown in the previous short rainy ha (21-133% increase). Exceptat 

season and incorporated prior to long rainy season_ 

Shoboka, Ethiopia, green manure 

• Grain yields in a column fonowed by the same letter are not significantly different. 

rotations consistently produced as 

# Fertilizer Napplied - 110 (Bakol. 92 (Jimma, 2000), 69 (Jimma. 2001), 50 (Mlingano). 70 (Namulonge, 2000) 

and 120 (Namulonge, 2001) kg-N/ha _ 

much, and occasionally more, 

maize than the fertilized 

monocrops, increasing maize yields 

Table 4. Response of maize grain yield· (t/ha) to green manures sown in the preceding 

season (year) and incorporated prior to sowing maize in the current season - on-farm trials 

by 1.75-4.5 t/ha (34-384% increase). 

Maize-green 

Ethiopia Tanzania Kenya 

manure rotation§ 

Although green manure species can 

potentially provide an excess of N 

Shoboka Walda Mlingano Kakamega Kitale 

to a subsequent maize crop, not all 

1998 1998 2001 1998 1998 1999 2000 

of the N in the green manure resi 

No. of farmers 14 4 10 7 6 dues may be available and consid 

Maize - Fert-N 2.38 a 3.12 a 1.48 a 0.70 a 6.7 a 4.1 a 4.4 ab erable losses via various pathways 

Maize + Fert-N# 5.55 b 6.48 b 2.78 be 8.1 b 6.5 b 7.3 d (leaching, volatilization) may occur 

Mucuna 3.22 e 9.0 be 6.0 b 3.7 a before the maize crop can access it. 

Crotalaria 3.39 b 6.2 a 6.1 b 5.6 be Thus, on-station experiments were 

Dolieos lablab 2.70 a 6.33 b 11 .1 e 6.9 b 6.5 cd conducted at Namulonge, Uganda, 

Canavalia 

2.31 b 

and Jimma, Ethiopia, to estimate 

the N fertilizer equivalence of mu 

§ Green manures managed as described in Table 3. 

• Grain yields in a column followed by' the same letter are not significantly differe~t. 

cuna, canavalia, sesbania and crota 

# Fertilizer N applied - 50 (Mlingano). 60 (Kitalel. and 110 (Bakol kg-N/ha. laria grown in the preceding season. 

116 Grain legumes and Green Manures for Soil Fertility in Southern Africa

Four rates of fertilizer N were applied to maize in 10 

plots previously under maize or the green manure. (a) Namulonge (b) Jlmma, Ethiopia 

10 

At Namulonge, mucuna and canavalia produced 

maize yields equivalent to about 120 kg-Nlha Ii 8 

J: 

while, at Jimma, sesbania and crotalaria green ma =- 8 

~ 

nures were equivalent to >70 kg-N Iha of fertilizer " 

(Figure 2). 

;; 

>. ~ 

c 6 

~ 0 

• 

'" • • 

6 

N 

Effects of intercropped legumes on maize produc "i 

" 

:I 

4 

tion 

• Mliiu-m.lize 

Table 5 summarizes maize response to green ma 

• Selbllnia maize 

... Crotolarla-malze 

nure intercrops at three on-station sites over several 

years. Although maize responded significantly to N o 40 80 120 o 40 80 

fertilizer in 4 out of 6 site-years, the legume inter- Fertilizer N rate (kg/ha) Fertilizer N rate (kg/ha) 

crop significantly increased the subsequent season's 

maize yield in only one instance (crotalaria at Figure 2. Maize response to fertilizer Nrates following maize 

Jimma in 2000). Mucuna, canavalia and crotalaria monocrops or green manure rotations in the preceding season, at (a) 

intercropped or relayed with maize in regional on Namulonge, Uganda, and (b) Jimma, Ethiopia 

station trials at Jimma, Mlingano and Namulonge 

generally had no significant effect on 

maize yields in the subsequent season Table 5. Effect of green manure ilitercrops and relay crops on maize grain yields§ 

(Table 5). This was attributed to low legmanaged 

(t/ha) sown with the green manure or alone in the subsequent season - researcher 

ume biomass production (and hence N 

trials on station 

fixation) under the shady conditions of Maize·Green manure system Jimma Mlingano Namulonge 

the maize canopy. 2000 2001 1999 2000 2001 2000lR 

Sole Maize - fertilizer N 4.89 a 1.95 ab 1.28 2.05 a 3.76 3.11 ab 

Mixed results of legume intercrops and Sole Maize + fertilizer N§ 6.02 ab 3.15 c 1.77 4.27 b 3.89 4.54 c 

relays were obtained in some 56 on-farm Mucuna intercrop' 5.22 ab 1.60 a 2.12 2.74 a 3.49 3.51 b 

trials conducted in Northern and East 

Mucuna relay crop# 4.86 a 2.36 b 2.30 2.28 a 2.84 3.14 ab 

ern Tanzania during 2000 and 2001 

Canavalia intercrop' 5.96 ab 2.25 bc 1.77 2.40 a 2.53 3.94 bc 

(Table 6). In some cases, maize yields 

Canavalia relay crop# 5.62 ab 1.88 ab 2.26 1.71 a 2.28 2.57 a 

were increased by 60-120% (1-2 t/ha) 

while' in others no significant effects Crotalaria intercrop' 6.52 b 2.15 ab 

were obtained. On the positive side, nei Crotalaria relay crop# 5.34 ab 1.96 ab 

ther did intercropped legumes have any Yields within a column followed by the same letter (or no letter) are not 

negative impact on maize yields in these significantly different 

trials, although experience in other trials § N rate (kg/ha) - 69 (Jimmal. 50 (Mlingano), 70 (Namulonge); 

• planted 2 weeks after maize 

has found considerable competition 

# planted 2 weeks after tasselling 

from legumes such as mucuna if not 

properly managed. 

Discussion and Conclusions 

Table 6. Effect of green manure intercrops or relay crops on yield of maize grain§ (tl 

hal sown with the green manure or alone in the subsequent season - on farm trials 

During the past five years, the Maize·Green manure Tropicallowland ecology 

Dry mid·altitude 

ECAMA W research network has identi system (intercrop or relay) Ngomeni Tanganyika Mlingano 

Hai# 

fied and characterized several green ma 2000 2000 2000 2001 2000 2001 

nure and grain legume species that have No. of f~rms 4 4 18 14 8 8 

good biophysical adaptation to the Sole Maize - 0 kg·N/ha 1.82 a 3.30 2.22 1.48 a 0.8 2.2 a 

moist mid-altitude and tropical lowland Sole Maize - 25 kg·N/ha 2.49 b 3.86 

ecologies of Eastern and Central Africa. 

Sole Maize - 50 kg·N/ha 3.28 c 3.19 2.49 2.78 b 

In monoculture situations, most of these 

Maize/mucuna 2.88 bc 3.57 2.32 2.40·b 1.1 4.1 b 

legumes were able to biologically fix N 

Maize/canavalia 2.92 bc 3.71 2.49 2.31 b 0.0 4.3 

much in excess of a maize crop's re 

b 

q~irements. However, in intercropping MaizelDolicos lablab 1.1 4.6 b 

situations, biomass production and biological 

Maize/calopogonium 2.52 b 3.84 

nitrogen fixation was severely # green manures relayed into maize in 2000; maize sown alone in 2001 

limited by competition for light and § yields within a column followed by the same letter (or no letter) are not significantly 

moisture with the maize crop. Further- different 

T 


more, mtercroppmg was found to be highly management 

sensitive, especially with respect to the 

competition that aggressive legumes such as mucuna 

can exert on the maize crop., As a result of 

these effects, mtercropped green manures were 

found to have very mixed effects on a subsequent 

maize crop's yields. In contrast, green manures 

grown m rotation with maize had benefits that are 

more consistent. The results of these studies by the 

ECAMA W network are therefore m substantive 

agreement with similar results obtamed elsewhere 

in the region (Giiler et al., 1997). 

Farmers' perceptions and constraints to adoption of 

green manure/grain legume systems were ass~ssed 

in mformal questionriaires durmg field days organized 

around on-farm trials. While farmers had 

much interest in new and alternative crops, they 

were concerned with the lack of a consumable or 

marketable product for many of the green manure 

species. For many, the spatial requirements for 

green manures grown m rotations may not be acceptable. 

In intercropping situations, competition 

effects, especially for creepers such as mucuna, were 

also very undesirable. Nevertheless, the benefits of 

both green manures and intercropped legume:> in 

weed control and reduced dependence on fertilizer 

inputs were recognized by farmers. 

Based on this experience, current and fu ture 

ECAMA W activities with green manures and grain 

legumes in maize-based systems are focussing on 

the following: 

• Combining legumes with new low-N maize varieties 

that have been developed by breeders work 

.ing with CIMMYT in the region - these varieties 

·will respond to lower levels of available soil Nand 

should increase the potential benefits of legumes 

that, due to their nature or the system in which 

they are grown, provide less than the required N 

into the system. 

• Introducing and evaluating multi-purpose legumes 

that fit into existing maize systems - in response 

to farmers' concerns regarding a marketable/consumable 

product and the lack of space m 

their systems for fallows. 

• Assessing the economic viability of maize-legume 

systems in on-farm trials with farmers. 

References 

Bumb, B.L., and CA. Baanante. 1996. The role offertilizer 

in sustaining food security and protecting the 

environment to 2020. Food, Agriculture and the 

Environment Discussion Paper 17. IFPRl, Washington, 

D.C, USA. 

ECAMAW. 1999. The Eastern and Central Africa 

Maize and Wheat Research Network (ECAMA W) 

Five-year Plan (2000-2004). Entebbe, Uganda: 

ASARECA/ECAMAW. 

Giller, K.E., G. Cadisch, C Ehaliotis, E. Adams, W. 

D. Sakala, and P.L. Mafongoya. 1997. Building 

soil nitrogen capital in Africa. In: R.J. Buresh, P.A. 

Sanchez and· F. Calhoun (eds.), Replenishing Soil 

Fertil~ty in Africa. SSSA Spec. Publ. No. 51, Soil 

Sci. Soc. Am.!Am. Soc. Agron., Madison, Wisc. 

USA. pp. 151-192. 

Heisey, P.W., and W. Mwangi. 1996. Fertilizer use 

and maize production in sub-Saharan Africa. Economics 

Working Paper 96-01, CIMMYT, Mexico 

D.F. 

Pingali, P.L. (ed.). 2001. CIMMYT 1999-2000 World 

Maize Facts and Trends. Meeting World Maize 

Needs: Technological Opportunities and Priorities for 

the Public Sector. CIMMYT, Mexico D.F. 60 pp. 

118 


THE ROLE OF COWPEA (VIGNA UNGUICULATA) AND OTHER GRAIN 

LEGUMES IN THE MANAGEMENT OF SOIL FERTILITY IN THE 

SMALLHOLDER FARMIN(J SECTOR OF ZIMBABWE 

NHAMO NHAMO, WALTER MUPANGWA 

Soil Productivity Research Laboratory, CSRI, p'Bag 3757, Marondera, 

SHEPHARD SIZIBA, CIMMYT-Zimbabwe, PO Box MP163, Mount Pleasant, 

TENDAI GATSI, Agronomy Research Institute, PO Box CY550, Causeway, and 

DAVISON CHIKAZUNGA, Department of Agricultural Economics, University of 

Zimbabwe, PO Box MP167, Mount Pleasant, Harare, Zimbabwe 

Senior author present address: World Agroforestry Centre (ICRAF)-Zimbabwe, c/o Division of Agricultural 

Research and Extension, POBox CY594, Causeway, Harare, Zimbabwe, Tel/Fax: 263-4-728340, 

Email: nnhamo@mweb.co.zwornnsprl@mweb.co.zw 

Abstract 

Cowpea is a widely grown legume in both high and low rainfall smallholder farming areas of Zimbabwe. Its popularity 

with farmers can be attributed to the multiple uses the crop can be put to and adaptability to different environments. It 

is often used for making relish, it is an important source of protein for human beings, as livestock feed and for enhancing 

soil fertility through biological nitrogen fixation (BNF). This paper reports results of a study on the current cowpea production 

practices by farmers in Chihota, Shurugwi and Zimuto. The aims of the study were to assess the constraints, 

opportunities and the justification for wider use of cowpea for improving soil fertility in maize based farming systems as 

well {IS household livelihoods on smallholder farms. 

The area put under legumes in Chiota, Shurugwi and Zimuto ranged from insignificant to small portions of the farm 

(0.2-2.5 'ha in one season). Farmers were aware that there are sOli fertility benefits associated with the legumes they 

grow, with 97% ploughing under the residues and 80% using the residues to make composts. More farmers grew cowpea 

intercropped (>94%) with cereals, mainly on the homestead fields, than in rotations «6%). Cowpea ranked second 

to groundnut in providing biomass on farms for use in soil fertility improvement. No planned fertilizatio!1 practices 

were applied on cowpea but in intercrops it benefits from fertilizers targeted for maize. Seed availability was a problem. 

Most farmers (88.5%) retained' seed or got it locally from fellow farmers . Only 11.5'% obtained seed from commercial 

seed outlets. Women were at the centre of cowpea production. Pests common under the current scale of production could 

be suppressed using simple methods like using ash or "Surf" washing powder in water. Utilization of cowpea products 

was as boiled beans, porridge and livestock feed. 

We concluded that cowpea was perceived to have high potential to increase soil fertility on most farms in high and low 

rainfall zones of the smallholder farming sector. Most farmers grew cowpea intercropped with maize on small portions of 

their homestead fields, However, the biomass produced did not significantly improve the farm level nitrogen budgets. 

Cowpea had an important dietary role in the rural communities. Seed availability, limited product markets and lack of 

proper fertilization (especially P) were the major constraints to cowpea production. The utilization of more products 

(food and non-food) of the right varieties when properly targeted could greatly improve the role of cowpea in the farming 

systems. Pests were not a major co'nstraint to the production of cowpea since farmers used simple measures like sprinkling 

Surf and ash solutions to contain the pests. The economic and financial benefits of cowpea production could also be 

redized through involving male farmers more in the production process and its promotion. Development of the product 

chain is required. 

Key words: Cowpea, smallholder farming, stakeholder consultation, farmer survey 


Introduction 

Legumes playa signifiCant role in the improvement 

of nitrogen budg~ts through biologiCal nitrogen 

fixation (BNF) and cycling of other nutrients on the 

farm (Giller and Wilson, 1991; Giller, 2001) . Cowpea 

(Vigna unguiculata (L). Waip.) can be considered a 

cheap legume to grow in that its fertility and rainfall 

demands are low. Hegewald (1990) found cowpea 

to produce acceptable yields on acidic oxisols. However, 

the economiCs and yield benefits of BNF in 

maize/cowpea rotation have not been fully explored 

(Shumba et al. 1990). Its role in increasing 

and maintaining soil fertility is not clearly defined. 

Often farmers do not have a planned P and N fe-rtilization 

strategy for these rotations. The fertilization 

of cowpea grown in rotation with maize has not 

been studied thoroughly, especially the economics 

of fulfilling the P requirements of the legume. There 

is need to evaluate the different fertilization practices 

in relation to their agro-economic effectiveness 

for different farmer domains. 

Soils in the smallholder farming sector of Zimbabwe 

are inherently poor in fertility. Deficiencies in both 

macro- and micronu trients have been reported in 

these sandy soils (Grant, 1981; Mashiringwani, 1983; 

Nyamapfene, 1991). Farmers use different strategies 

to either add and/or recycle nutrients on their 

farms. Use of cattle manure, composts, mineral fertilizers, 

crop or grass residues, grass fallows, grain 

legumes and green manures in cereal/legume rotations 

are loosely practiCed · by farmers as ways of 

managing soil fertility (Nhamo et al. 2002). Such 

practices on sandy soils are important in the management 

of the most limiting nutrients and soil organiC 

matter and there is potential to improve their 

efficiencies. In all, the practices have been to add 

whatever is available to the soil or nothing at all. 

This results in addition of much less, just enough or 

more than required nutrients for the'field crops. 

With these practices, the use of both organic and 

mineral fertilizers has no scientific basis. This has 

rendered optimum crop production and profit levels 

difficult to attain on sandy soils. 

Besides the soil fertility contribution, cowpea provides 

the needed proteins in rural households 

through both the pea and the leaves that are used as 

relish. Traditionally, cowpea porridge was an important 

and nutritious dish making part of the diet 

for the farming· communities. It is a multiplepurpose 

legume which can be used for human food 

and livestock feed (Johnson, 1970; Rao and 

Mathuva,2000). 

In the smallholder farming systems of Zimbabwe 

the cultivation of tradition legumes, including cowpea, 

is not emphasized. Current uses of cowpea for 

. improving household food security and soil fertility 

vary from area to area. The reasons for this variability 

are not clear. The aims of this study were to determine 

the current cultivation practices, perceptions 

of farmers on the benefits and constraints of 

effectively utilizing cowpea in their farming system, 

and to evaluate the role cowpea could play in improving 

soil fertility and hence household food security. 


A survey was conducted in three communal areas, 

Chihota (Mashonaland East Province), Zimuto 

(Masvingo Province) and Shurugwi (Midlands 

Province) representing natural regions II, III and IV 

of Zimbabwe respectively. Chihota receives annual 

rainfall of between 800 and 1000 mm whereas Shurugwi 

and Zimuto receive 600-800 mm and 450-600 

rum respectively. The rainfall distribution within 

and across season if variable, and in all the areas 

mid-season droughts are a common feature. 

Farmers in Chihota, Shurugwi and Zimuto rely on 

agriculture for food and to generate income to sustain 

their families. Most families have financial constraints 

and limited agricultural inputs are purchased 

for use in the production of both legumes 

and cereals. 

Within each of the· areas, a formal questionnaire was 

administered to collect information on cowpea practices. 

The questionnaire captured information on 

household characteristics, crop production practiCes, 

and cowpea placement in the farming systems, 

'current constraints and opportunities for in-· 

creased productivity. The semi-structured questionnaire 

was administered to a sample of sixty households 

in each of the three communal areas. Welltrained 

enumerators carried out data collection ~ The 

data collected was captured and analyzed using 

SPSS (Statistical Package for Social Sciences). 

Results 

Like most communal areas in Zimbabwe, maize 

dominates other crops and most of the land was put 

under this staple food crop in Ch,ihota, Shurugwi 

and Zimuto. Tables 1 and 2 show the area under 

several non-legume and five legume crops grown in 

the study areas. Minor traditional crops like millets 

(rapoko) were also commonly cultivated on small 

areas in the study areas. A few farmers grew cash 

crops including cotton, tobacco and paprika. 

120 

Grain legumes and Green Manures for Soil Fertility in Southern Atnca

Table 1. Non-legume cropping pattern of farmers from each of the sites 

Zimuto 

%growers Area (ha) Yield %growers 

(kg/ha) 

Maize 100 2.66 327 100 

Rice 67 1.35 382 21 

Rapoko 64 0.57 512 30 

Sunflower 10 4.69 133 8 

Sorghum 0 6 

Paprika ' 2 0.75 27 10 

Millet 0 0 

Tobacco 2 0.50 1000 2 

Cotton 2 0.04 1163 0 

Total cultivated 93 4.35 100 

Total fallow 62 2.13 48 

Chihota 

Shurugwi 

Area (ha) Yield %growers Area (ha) Yield 

(kg/ha) 

(kg/ha) 

2.32 442 100 3.55 323 

0.62 226 48 0.~3 479 

0.58 253 21 0.51 749 

0.81 632 5 0.85 210 

0.89 693 8 0.60 ~45 

0.75 293 2 0.75 80 

3 0.07 3077 

1.00 600 o 

o 

3.43 102 5.00 

2.63 72 2.24 

Table 2. Area iha) put under various legumes in the 2000/2001 sea· 

son in Chihota, Shurugwi and Zimuto 

Zimuto Chihota Shurugwi 

Sole Intercrop Sole Intercrop Sole Intercrop 

Cowpea 0.93 1.68 1.07 1.59 0.53 2.49 

Bambara 0.64 0.43 0.67 0.50 1.49 0.60 

Garden bean 1.37 0.90 0.85 0.49 0.27 0.38 

Groundnut 1.57 0.60 1.31 0.83 0.77 0.67 

Soyabean 0.20 1.37 O . i~ 

In all three communal areas most of the cowpea was 

intercropped, whereas groundnut and bambara are 

consistently grown as sole crops. A few farmers 

grew soyabean and garden beans for household 

consumption Cfable 2). Groundnut was the dominant 

legume grown by the farmers. 

Though most farmers grew cowpea as intercrops, 

the yields reported for the sole cropped cowpea are 

higher than those grown in intercrops. The national 

average grain yield for cowpea of 300 kg ha- J was 

close to the reported yields for sole cropping (Table 

3). Higher yields were obtained from sole stands 

than from intercrops. 

Farmers acknowledged that legumes are important 

in soil fertility and for breaking disease/pest cycles 

on their fields. Groundnut was perceived by farmers 

to be better than cowpea for improving fertility. 

Farmers utilized cowpea residues for soil fertility 

through incorporation by ploughing under (97% of 

the farmers) and use in composts (80% of the farmers). 

Some ofthe residues were however fed to livestock. 

To farmers the soil fertility benefits derived from 

rotations and intercrops are not significantly different 

(Table 6). Farmers perceive that both growing 

the cowpea in rotation with maize and as an inter~ 

crop with maize has positive soil fertility benefits. 

There were no deliberate fertilization practices followed 

by farmers when growing cowpea. A significant 

amount of fertility inputs are applied to plots 

Table 3. The mean grain yields (kg/hal of different legumes obtained 

by farmers in the 2000/2001 season in the three areas 


Sole Intercrop Sole Intercrop Sole Intercrop 

Cowpea 269.5 61.0 278.5 66.4 339.1 62.5 

Bambara 352.2 299.3 1021.0 40.0 330.3 26.0 

Cowpea, groundnut and bambara nut were the legumes 

commonly grown by most farmers in Chihota, 

Groundnut 551.2 147.9 437.6 90.0 262.0 97.8 

Shurgwi and Zimuto. Groundnut and b~mbara nut 

Garden bean 540.0 39.2 790.7 103.7 80.9 35.0 

were grown by the majority of farmers as sole crops, Soya bean 389.4 

while cowpea was intercropped 

(Table 4) . 

Table 4. Numbers of farmers growing different legumes in Zimuto, Chihota and Shurugwi 


Cowpea was mostly grown 

on the homestead fields 

with a reasonable proportion 

on the top land fields, 

while less than 5% of the 

farmers grow it in the vleis 

and gardens (Table 5). 

Sole Intercrop Total Sole Intercrop Total Sole Intercrop Total 

n n n('!o) n n n ('Yo) n n n(%) 

Cowpea 10 51 61 (100) 19 43 62 (98)) 8 51 59 (94) 

Bambara nut 48 6 54 (89) 43 2 45 (71) 49 5 54 (86) 

Soya bean 1 0 1 (2) 4 0 4 (6) 1 0 1 (2) 

Garden bean 6 3 9 (15) 24 4 28 (44) 4 2 6 (10) 

Groundnut 47 7 54 (89) 51 3 54 (86) 53 3 56 (89) 

(% of farmers growing the legumes in brackets) 

I 

I 


Table 5. Types of fields ('Yo) to which farmers in Ghihota, 

Shurugwi and Zimuto grow cowpea 

Zimuto Chihota Shur~gwi Total 

Homestead field 53 57 52 54 

Vlei margin 10 3 4 

Garden 2 

Topland 37 38 48 41 

Total 100 100 100 100 

Table 6. The perceptions of farmers on which cowpea growing 

pattern results in improved soil fertility 

Shu rug wi Zimuto Chihota 

n % n % n % 

Rotation Yes 55 97 52 91 47 86 

No 2 3 5 9 8 14 

Intercrop Yes 53 88 53 88 40 67 

No 7 12 7 12 20 33 

where cowpeas are grown. The common practice 

observed though was that of intercropping maize 

and cowpea. In the intercrop, manure and fertilizers 

applied were targeted to the maize crops and not to 

the cowpea (Table 7; Figure 1). A few farmers apply 

legume inoculant on the cowpea. 

The fertilizer types used on the maize-cowpea intercrops 

are the recommended ones for sole crops of 

maize. Both the basal and the top-dressing fertilizers 

were applied in limited amount, far bwer than 

the standard recommendations. The rates applied 

were low and similar to those reported by Nhamo 

et al. (2002) of less than 50 kg ha- 1 of compound 0 

and 25 kg ha- 1 of ammonium nitrate fertilizer. The 

use of lime was recorded only in Chiliota, where 

some trials on lime use had been conducted by researchers. 

Seed availability was a problem in all the three areas, 

with most families pianting their own retained 

part of their harvest for seed. About 88.5% of the 

households used retained seed, were given seed by 

other farmers or bought it from other farmers in the 

area (Table 8). Farmers also reported the absence of 

an organized market with attractive prices, lack of 

marketing information and low selling prices in the 

local market. 

The majority of farmers grew the spreading cowpea 

varieties (49%), only 5% used the bush type and the 

remainder used both types. Most of the local varieties 

had their names derived mainly fro~ the appearance 

of the plant or the colour of the bean. 

Names such as rutandavare, chigogova, chitumbe, chitonono, 

dzemavara, chena, jerimeni, dahwa, and 

chipichipi were common in the areas studied. Others 

had names linked t~ the source like zvimugabe and 

mharapara. Names like chinyabundi, kaboko, chizhara- 

Table 7. The soil fertilization practices followed on cowpea crops 

(% farmers) 

Shurugwi Zimuto Chihota Total 

Yes No Yes No Yes No Yes No 

Inorganic fertilizer 30 70 39 61 70 30 47 53 

Cattle manure 61 39 90 10 62 38 71 29 

Legume innoculant 100 8 92 5 95 4 96 

en 

~ 100 iii Urea 

Q) 

. ~ 

.AN 

~ 

~ 80 OCampO 

2 

Ol 

C 60 

'>' 

0. 

a. 

ro 40 

en 

~ 

Q) 

E 

~ 

.E 

:::R 

0 

20 

0 

2 3 

Figure 1. The percentage use of different fertilizers on the maize 

cowpea intercrops where; 1 represents Zimuto, 2 represents Ghiota 

and 3 represents Shurugwi 

Table 8. Sources of cowpea seed for the 2000/2001 growing sea· 

son ('Yo farmers) 

Zimuto Chihota Shu rug wi Tot~' 

Own saved 72.9 68.3 66.1 69.1 

Bought from other 5.1 3.3 3.2 3.9 

farmers 

Given by other farmers 15.3 15.0 16.1 15.5 

Bought commercially 6.8 13.3 4.8 8.3 

Project 9.7 3.2 

Total 100 100 100 100 

wanya and chingwa were also recorded. 

The major pest reported by farmers was the aphid. 

A significant number of farmers did nothing about 

the common aphid problems they encounter and 

they observed that once they receiv.e some rains the 

aphids disappear from their crops (Tables 9 and 10). 

To solve this and other problems, some farmers 

used simple methods like sprinkling "Surf" washing 

powder or ash solutions. 

Cowpea grown on small portions of the farm is 

mainly meant for domestic consumption and very 

little is sold. About 65% of the cowpea bean produced 

in the three areas is eaten at home and the 

greater part of the remainder is barter traded with 

other crops. The utilization of cowpea through 

boiled beans both as a side dish and · as relish was 

most common (Table 11). The use of fresh leaves as 

vegetables was also found to be common. 

122 


Table 9. Common pests on cowpeas as mentioned by 

farmers 

Zimuto Chihota Shurugwi 

Aphids 75.9 58.5 77.9 

Beetles 17 10.2 

Worms 8.6 9.4 8.5 

None 3.4 1.7 1.7 

Termites and ants 3.4 7.5 

Stemborers 1.7 1.9 

Weevils 3.4 1.9 

Others 3.4 3.8 1.7 

Table 10. Solutions to some of the pest problems 

suggested by farmers 

Zimuto Chihota Shu rug wi 

Chemicals 3.3 13.8 4.9 

Traditional herbs 16.4 15.5 8.2 

Cultural practices 11.5 6.9 6.6 

00 nothing 68.8 63.8 80.3 

Women made important decisions on the production 

of legumes and minor crops, including cowpea. 

Among the reasons why women farmers took this 

role were; traditionally it's a woman crop, women 

were responsible for ·providing food and relish in 

the household, men perceived that cowpea was not 

an important crop, women realized the importance 

of the crop and that women sometimes made decisions 

on farm operations in the absence of men. 

Discussion 

Area under legumes and cowpea 

Compared to cereals, the area planted with legumes 

in the three communal areas ranges from insignificant 

to small portions of the farm (Tables 1 and 2). 

Most of the cultivated land is put under maize because 

itis the staple crop, followed by cash crops 

like tobacco, paprika and cotton. however, most 

farmers (99%) devote part of the farm to cultivation 

of at least one legume crop among which are 

groundnut, bambara or cowpea. The small area 

planteo determines the modest contribution made 

by legumes to the N budget on the farms. In most 

~ommunal areas, biomass production is key to the 

utilization of these high quality materials. The successful 

use of these legumes in improving soil fertility 

depends on the quantities of organic materials 

available for use on the farm. Work done by Nhamo 

et a1. (2002) has shown the importance of the 

~mounts of organic materials available on the farm 

in the adoption of some of these organic based soil 

fertility technologies. Biomass production following 

the planting of cowpea on these small areas at low 

Table 11. Scoring on the utilization of cowpea by farmers 

for domestic consumption 


Porridge (Rupiza) . 3 3 4 

Relish (Leaves) 2 2 3 

Relish (Grain) 4 4 2 

Boiled beans (Mutakura) 

plant populations is insufficient to make a big impact 

on the fertility status of soils. 

Farmers in the study areas grew several crops including 

small grains to spread the risk of crop failure. 

Under unpredictable climatic conditions, smallholder 

farmers use such strategies to ensure household 

food security. Millets however are also important 

in beer brewing for the traditional rituals. 

Soil fertility benefits of cowpea 

Farmers perceived that there were soil fertility benefits 

and improved yields of maize grown after cowpea 

(Table 6). The soil attributes linked to these improvements 

varied from the observable soil colour 

to the soil water holding capacity. Cowpea ranked 

second after groundnut in residue production and, 

hence soil improvement potential (Table 4). Most 

farmers intercropped cowpea and maize or other 

cereals. Intercropping maize and cowpea has been 

reported to increase yields in some cases 

(Olasantan, 1988; Jeranyama et al. 2000) and even 

better yields have been reported in rotations where 

there are no moisture competition effects (Kouyate 

et al. 2000; Rao and Mathuva, 2000). However, few 

studies have been conducted comparing the two 

farming systems directly. Work done by Hardter et 

a1. (1991) has shown that while mixed maizecowpea 

cropping had lower yields than rotations, 

continuous monocropping had the lowest productivity. 

The reasons why farmers iritercrop are varied. 

With regards to soil fertility, these can be explained 

scientifically by the residual effects on cereals 

following legumes in rotation and by the below 

ground nutrient transfers that occur in the 

rhizosphere in intercrops (Bandyopadhyay and De, 

1986). 

Incorporating legume residues to the soil improves 

its fertility. Work done on legumes has demonstrated 

the usefulness of legumes grown in rotation 

with cereals in general (Giller and Wilson, 1991, 

Giller 2001). For cereal/legume rotations to be successful, 

a reasonable amount of legume non-grain 

residue/biomass has to be produced and its management 

has to be effeLi;ve. Residues generated by 

legumes are in two forms; the roots (below ground) 

and the stems and the leaves (aboveground) (Giller 

and Wilson, 1991). The agronomic contributions of 

the above and below ground portions of the cowpea 


123

have not been studied. Because of the possible conflicting 

uses that legume leaves have on the farm, it 

is important to quantify the economic b~nefits separatelyand 

together. 

Similar to other field crops, grain legumes require 

soil nutrients for them to grow as well as fix N from 

the atmosphere. The growth is a response to the soil 

type, fertilization and soil water availability. On 

sandy soils that are commonly found in the communal 

areas, the nutrition of N fixers that contribute to 

the successful symbiosis has not been . emphasized. 

Higher cowpea yields from homestead fields (Table 

3) are a result of better soil fertility management. 

Studies have sho~ that they respond well to P application 

(Giller, 2001). However, the P, K, micronutrients 

and lime requirements for cowpea in a 

maize/cowpea rotation have not been worked out. 

Fertilization of the legume in a legume/cereal rotation 

is important if productivity is to improve from 

the current low levels. At present, there is scant information 

on the effective and efficient way of using 

organic and inorganic fertilizers on legume-cereal 

rotations (Giller, 2001). The current practice of adding 

mineral N reduces the N-fixing capacity of the 

legumes in these farming systems (Table 7). For the 

different agro-ecological zones, rates of P application 

need to be worked out. The economics of the 

first application .as well as the residual P effects on 

both the cereal and the legume in rotations, as well 

as in intercrops for the different soil types, are required. 

This information will be important and useful 

in targeting legumes properly on the farm. Work 

on row spacing of maize and cowpea show improved 

yields with wider spacing but the wider 

spacing leads to low plant populations. This leads 

to low biomass production and hence less effective 

utilization of the BNF from legumes. 

Soil nutrients interact with the available moisture .. 

As reported by Muza and Mapfumo (1999), soil nutrient 

and water interactions have a large effect on 

the overall biomass production of legumes. Cowpea 

has the advantage of a deep rooting system that 

makes it adaptable to different agro-ecological 

zones. However different varieties have different 

attributes so proper targeting is important for effective 

use of cowpea in farming systems. Grown in 

rotations with maize, cowpea has been reported to 

reduce weed pressure in the residual season 

(Kamau et al. 1999). Similar findings have been reported 

in intercrops, except that the weed suppression 

in intercrops is in the first season. 

Cowpea utilization 

Most cowpea grown. is utilized as boiled beans for 

either direct consumption or as relish. It remains a 

cheap source of protein especially during the dry 

winter season. In Shurugwi a study done at a school 

positively correlated the consumption of cowpea to 

the high tum out and class performance of primary 

school pupils (SDARMP, 1997). Other less commonly 

used dishes include porridge, scones/bread, 

and there is potential for more. For the benefit of the 

communities these other benefits in addition to soil 

fertility technologies have proven to be important in 

technology acceptability. In the case of cowpea, the 

health effects of the dishes have to be considered to 

see how these could be made part of the diet of 

HIV / AIDS affected persons. Cowpeas provide both 

calories and protein (Venter et al. 1997). For food 

security, indigenous and traditional crops need to 

be improved since their important contribution has 

largely been ignored in recent years. 

Constraints in cowpea production 

Fertilization. Constraints on using legumes effectively 

in soil. fertility management in the smallholder 

farms are varied. Low perception about minor 

crops, little biomass from a small area planted, 

seed availability problems, lack of exposure to information 

on their production and little information on 

the potential benefits of using the legume crop in 

maize-based farming systems are some of the reasons 

why legumes are little used in fertility management 

(Rusike et al. 2000). Cowpea grown in intercrops 

benefits from the fertilizer applied on the 

maize. The nitrogen from basal and topdressing 

maize fertilizer reduces the amount of N fixation by 

the legumes. This reduces the benefits from the 

cowpea and the potential N addition to the nutrient 

budget through BNF. Farmers sometimes also complain 

about the higher labour demands with legume 

crops compared to the cereals (Jeranyama et al. 

2000). Soil fertility management based on rotations 

can be used to come up with integrated soil fertility 

management strategies that have the potential to 

improve he livelihoods of people in the smallholder-farming 

sector of Zimbabwe. Use of combinations 

of organic and inorganic nutrient sources 

can produce better crop yields and improve the soil 

organic matter levels in the long term (Murwira et 

al. 2002 ; Nhamo et al. 2001). 

Marketing. The cowpea market is underdeveloped. 

The whole product chain has not .been developed 

and supported enough to benefit the farmers. Seed 

sources identified in the study are mainly local and 

little commercial seed finds its way to the farmers. 

Interested farmers are therefore faced with the uncertainty 

of growing unproved seed. A large proportion 

of the farmers keep some of their harvest for 

seed for the commonly grown legumes. Considering 

that some of the grain is consumed by the family, 

seed availability could be one of the root causes 

of the low areas for cowpea (Rusike et al. 2000). A 

124 

Grain legumes and GreenManures for Soil Fertility in Southern Africa

few of the farmers interviewed acquire cowpea seed 

from approved seed dealers and the local market 

for retained seed is not organized. This leads to reduced 

areas ,under cowpea and other legumes. The 

high percentage of farmers who rely on retained 

seed posses a problem in seed availability and viability. 

The viability of seed depends on the storage 

conditions under which the bean is stored. These 

post harvest storage facilities have not been developed 

in the smallholder farming sector resulting in 

limited storage, fast loss of quality seed and small 

quantities that can be stored at anyone time. The 

use of inferior cowpea varieties could also have 

caused reduced areas under their cultivation. Most 

farmers grew the spreading type of cowpea and had 

retained seed used over long periods. Over time, the 

vigor of the seed could have declined causing reduction 

in the potential yield. As observed by Franzel 

and Scherr (2002), some cropping systems function 

below their potential productivity because of 

using poorly adapted species, varieties and management 

practices. 

The current poor market structures for cowpea do 

not warrant investment in proper fertilization, use 

of pesticides and other planned agronomic practices 

on the crops. The economics of cowpea beyond barter 

trade need to be explored to include organized 

national markets as well as export markets. Such a 

development would enhance the direct and indirect 

financial benefits of cowpea to farmers. Promoting 

other products from cowpea of dietary, direct and 

indirect monetary importance creates a market for 

the legume. 

Pests and diseases. In this study, pest and diseases 

on cowpea were not regarded by farmers as a major 

constraint to production. The suggested solutions to 

pests showed that those that have grown cowpea 

know about them in general and that the occurrences 

have not been large enough to reduce the 

yields by economic margins. Several options followed 

by farmers need to be refined and avoid the 

wait-for-rains strategy which could reduce yiel:is to 

below economic levels. The use of Surf and ash solutions 

has been documented through the experiences 

shared by farmers in Shurugwi. Use of uncertified 

seed produced without inspection could be 

one way in which there has been a build up of diseases 

over the years (Madamba, 2002). The implications 

Of pest build up with increased area under 

cowpea also need to be looked at. Practicing rotation 

can always break the disease cycles. 

Gender in cowpea production. Whilst it is widely 

agreed that women are overall responsible for 

growing cowpea and other legumes for the family, 

they are faced with serious knowledge limitations 

on 'sustainable agronomic practices with these 

crops. Women make decisions on the area to which 

the legumes are cultivated since they are the ones 

who keep and know the quantities of seed available 

for these crops. Very few received training or advice 

on cowpea production from extension agents. Most 

legumes are labeled as women crops in all the communal 

areas visited though labour to work on fields 

with legumes is provided by the whole family. The 

implications of this are that cowpea production becomes 

low priority, is perceived as a non-cash generating 

activity and' hence no fertilizers or fertility 

practices are targeted towards its production. However, 

the farmers who use legumes for consumption 

and local trade ranked them as highly important in 

improving the livelihoods and food security of the 

household at particular times of the year. For the 

effective and wide production of these legumes, the 

myths and beliefs around their production present a 

challenge. Since gender is central to their production, 

there is need for a partieipatory 'degenderization' 

of the commonly grown legumes. 

Research and development of such crops have 

lagged behind too much compared to what are referred 

to as men crops or cash crops like maize, tobacco 

and cotton. 

Conclusion 

The potential of cowpea to improve soil fertility and 

household food security and income was high. Most 

farmers intercropped cowpea with maize. The area 

put under legumes in the three areas ranged from 

insignificant to small portions of the farm. Farmers 

acknowledged the role of cowpea in soil fertility 

used in both rotations and intercrops. However, no 

planrted fertilization practices on cowpea were followed 

by farmers. 

The cowpea product chain was undeveloped in Chihota, 

Shurugwi and Zimuto. The current utilization 

of cowpea was mainly through four simple dishes 

in the form of porridge, relish (bean and leaves) and 

boiled bean (mutakura). Farmers incorporated some 

of the residues while some were fed to livestock. 

There is therefore need for diversification through 

the utilization of more products. Traditional crops 

have been recommended as part of the diet for people 

suffering from HIV / AIDS, and cowpea could 

find a place in some of these diets. Seed availability 

was a major problem to farmers with the majority 

using retained seed. Varieties suited for the different 

agro-ecological zones need to be studied to improve 

grain and non-grain biomass production of 

cowpea. The area under cultivation needs to be 

properly fertilized for both rotations and intercrops. 

Seed availability and markets of the cowpea need to 

Grainlegumes and Green Manures for Soil Fertility in Southern Africa 

125

e orgaruzed, both the local and external channels. 

Women played an important role in the. production 

of cowpea and there is need for sensitization of all 

stakeholders to remove the gender bias of the crop 

so that its economic, financial, nutritive and other 

benefits could be explored to the maximum. Farmers 

need to be empowered with knowledge to enable 

them to appreciate the real economic and financial 

value of cowpea in the household, on agricultural 

markets and in the farming system. 

Pests are not a major constraint under the current 

cowpea production and farming systems. Increasing 

area under cowpea cultivation could lead to demand 

of a more systematic way of dealing with 

pests and diseases. 

Acknowledgements 

The authors are thankful to the Soil Fertility Net 

work for Southern Africa (especially Stephen Wad 

dington, Mulugetta Mekuria and Johannes Karig 

windi) for taking a keen interest in the subject and 

supporting those involved in this study through the 

provision of finance (from the Rockefeller Founda 

tion) and transport to visit the study sites. 

References 

Bandyopandhyay, S.K. and R De. 1986. N relationship 

in a legume - non legume association grown 

in an intercropping system. Fertilizer Research 

10:73-82. 

Brown, S.J.M. Anderson, P.L. Woomeri M.J. Swift 

and E. Barriois. 1994. Soil biological processes in 

tropical ecosystems. In: P.L. Woomer and M.J. 

Swift (eds), The Biological Management of Tropical 

Soil Fertility. John Wiley, UK, pp 47-80. 

Gilbert, A.R. 1998. Undersowing green manures for 

soil fertility enhancement in the maize-based 

cropping systems of Malawi. In: S. Waddington, 

H.K. Murwira, JD.T. Kumwenda, D. Hikwa and 

F. Tagwira (Eds) Soil Fertility Research for Maize 


Proceedings of the Soil Fert Net Results and 

Planning Workshop, Afri;::a University, Mutare, 

Zimbabwe, SoilFertNet and CIMMYT 

Zimbabwe. pp 73-80. 

Giller, K.E. 2001. Targetting management of organic 

resources and mineral fertilizers: Can we match 

scientists' fantasies with farmers' realities? In: B. 

Vanlauwe, N. Sanginga, sJ. Diels and R Merckx 

(Eds) Balanced Nutrient Management Systems for 

Moist Savanna and Humid Forest Zones of Africa. 

CAB International, Wallingford, UK. In press. 

Giller, K.E. and K.J. Wilson. 1991. Nitrogen fixation 

in tropical cropping systems. CAB International, 

Wallingford, UK. 

Grant, P.M. 1981. Fertilization of sandy soils in 

peasant agriculture. Zimbabwe Agricultural Jou r 

naI78:169-175. 

Hardter, R, W.J. Horst, G. Schmidt and E. Frey. 

1991. Yields and land use efficiency of maizecowpea 

crop rotations in comparison to mixed 

and monocropping on an alfisol in Northern 

Ghana. J. Agronomy and Crop Science 166:326-337. 

Jeranyama, P., O.B. Hesterman, S.R Waddington 

and RR. Harwood. 2000. Relay-intercropping of 

sunnhemp and cowpea into a smallholder maize 

system in Zimbabwe. Agronomy Journal 92:239 

244. 

Johnson, D.T. 1970. The cowpea in the African areas 

of Rhodesia. Rhodesia Agricultural Journal 67:61 

64. 

Kamau, G.M., J.K. Ransom and H.M. Saha. 1999. 

Maize-cowpea rotation for weed management 

and improvement of soil fertility on sandy soil in 

coastal Kenya. In:. Maize Production Technology 

for the Future: Challenges and Opportunities. Proceedings 

of the Sixth Eastern and Southern Africa 

Regional Maize Conference. CIMMYT and 

EARO, Addis Ababa, Ethiopia, pp. 223-225. 

Kouyate, Z., K. Franzluebbers, A.s.R Juo and L.R. 

Hossner. 2000. Tillage, crop residue, legume rotation, 

and green manure effect on sorghum and 

millet yields in the semiarid tropics of Mali. 

Plant and Soil 225: 141-151. 

Madamba, R. 2002. Producing cowpea seed/ 

Kurima mbeu yenyemba/ Ukulinywa kwehlanyelo 

yendumba. P.N. Naykanda ~nd D. Hikwa 

(Eds). Department of Research and Extension, 

Harare, Zimbabwe. 

Mashiringwani, A.A. 1983. The present nutrient 

status of the soils in the communal areas of Zimbabwe. 

Zimbabwe Agricultural Journal 80:73-75. 

Murwira, H.K., P. Mutuo, N. Nharr..o, R Rabeson 

and c.A. Palm. 2002. The fertilizer equivalency 

values of orgaruc materials of differing quality. 

In: B. Vanlauwe, J. Diels, N. Sanginga and R 

Merckx (Eds.). Integrated Plant Nutrient Management 

in Sub-Saharan Africa: From Concept to Practice. 

CAB International, Wallingford, UK. pp 113 

122. 

126 


Murwira, H.K., M.J. Swift and P.G.H. Frost. 1995. 

Manure as a key resource in sustainable agriculture. 

In: J.M. Powell, S. Fernandez-Rivera, O.T. 

Williams and C Renard (Eds) Livestock and Sustainable 

Nutrient Cycling in Mixed Farming Systems 

of Sub-Saharan Africa, Volume II, Technical 

Papers, Addis Ababa, Ethiopia, pp. 131-148. 

Muza, L. and P. Mapfumo. 1999. Constraints ·and 


of sandy soils in Zimbabwe. In: Maize 

Production Technology for the Future: Challenges 

and Opportunities. Proceedings of the Sixth Eastern 

and Southern Africa Regionai Maize Conference, 

CIMMYT and EARO, Addis Ababa, Ethiopia, 

pp. 214-217. 

National Academy of Sciences. 1984. Tropical legumes: 

Resources for the future. A report of ad 

hoc advisory panel of the Advisory committee 

on Technology Innovation, Board on Science and 

Technology for International Development, 

Commission on the International Relations - National 

Research Council. National Academy of 

Science, Washington D.C, USA, 1979. pp 272 

278. 

Nhamo, N., H.K. Murwira and K.E. Giller. 2001. 

The effects of combining cattle manure with inorganic 

N fertilizers on maize yield on sandy 

soils in Zimbabwe. In press. 

Nhamo, N., T. Sithole, D.KC Dhliwayo, W. Mupangwa, 

T. Gatsi, D. Chikazunga and H. Nemasasi. 

2002. The prospects of smallholder farmers 

utilizing organic materials as soil fertility 

amendments in Murewa and Mvuma, Zimbabwe. 

Soil Fertility Network Research Results 

Working Paper No.9. 24 pp. 

Nzuma, J.K, H.K. Murwira and S. Mpepereki. 1998. 

Cattle manure management options for reducing 

nutrient losses: Farmer perceptions and solutions 

in Mangwende, Zimbabwe. In: S. Waddington, 

H.K Murwira, JD.T. Kumwenda, D. 

Hikwa and F. Tagwira (Eds) Soil Fertility Research 

for Maize-Based Farming Systems in Malau.d and 

Zimbabwe. Proceedings of the Soil Fert Net Results 

and Planning Workshop, Africa University, 

Mutare, Zimbabwe, Soil Fert Net and CIMMYT 


Nyamapfene, K. 1991. Soils of Zimbabwe. Ne:.. 

handa Publishers, Harare, Zimbabwe. 

Olasantan, F.O. 1998. lntercropping of cassava 

(Manihot esculenta) with maize or cowpea tinder 

different row management. Field Crops Research 

19:41-50. 

Palm, CA., CN. Gachengo, R.J. Delve, G. Cadisch 

and KE. Giller 2001. Organic inputs for soil fertility 

management in tropical agroecosystems: 

application of an organic resource database. Agriculture, 


Palm, CA., R.J.K. Myers and S.M. Nandwa. 1997. 

Combined use of organiC and inorganic nutrient 

sources for soil fertility maintenance. In: R.J. Buresh, 

P.A. Sanchez and F. Calhoun (Eds) Replenishing 

Soil Fertility in Africa. SSSA Special Publication 

51. SSSA, Madison, WI, USA. pp 193-217. 

Rao, M.R. and M.N. Mathuva. 2000. Legumes for 

improving maize yields and income in semi arid 

Kenya. Agriculture, Ecosystems and Environment 

78:123-137. 

Rusike, J.C Sukume, A. Dorward, S. Mpepereki and 

K.E. Giller. 2000. The economic potential for 

smallholder soyabean production. A Soil Fert 

Net Special Publication, CIMMYT, Harare, Zimbabwe. 

64 pp. 

SDARMP Diagnostic Survey. 1997. Smallholder Dry 

Areas Resource Management Programme, Zimbabwe. 

Shumba, E.M., H.H. Dhliwayo, B. Kupfuma and G. 

Gumbie. 1990. Response of maize in rotation 

with cowpea to NPK fertilizers in low rainfall 

area. Zimbabwe Journal Agricultural Research 

28:39-45. 

Van Kessel, C and J.P. Roskoski. 1988. Row spacing 

effects on N2-fixation, N-yield and soil N uptake 

of intercropped cowpea and maize. Plant and Soil 

111:17-23. 

Venter, S., E. Van den Heever, R. Matlala, and H. 

Maphanga. 1997. Cowpeas as a vegetable. Farmers' 

weekly, June 6,1997, pp. 27-31. 

Wade, M.K. and P.A. Sanchez. 1984. Productive potential 

of an annual intercropping scheme in 

Amazon. Field Crops Research 9:253-263. 


BIOMASS PRODUCTION OF GREEN MANURES AND GRAIN LEGUMES iN 

SOILS OF DIFFERENT CHARACTERISTICS IN ZAMBIA AND ZIMBABWE 

PAULINE CHIVENGE', MOSES MWALE 2 and HERBERT MURWIRA' 

1TSBF-CIA T, Box MP 228, Mt. Pleasant, Harare, Zimbabwe 

E-ma;i: tsbfzim@zambezi.net 

2 Mt. Makufu Research Station, P. Bag 7, Chilanga, Zambia 

Abstract 

Several green manure and grain legumes 'have been identified as having potential for use in soil fertility improvement 

and for food in southern Africa, The ecological boundary conditions under which the different legumes perform have, 

however, not been ascertained, A study was carried out in the 2001/02 season to assess the influence of soil characteristics 

on legume establishment, growth and biomass production and grain yield in Zambia and,Zimbabwe. On-farm experiments 

were established in different agro-ecological zones in Zambia and Zimbabwe to capture soils of different texture, 

pH, soil fertility status and CEC. Orought experienced during the season resulted in low yields for all the legumes. 

Of the five legumes planted, the green manures, Crotalaria grahamiana and C. juncea, and Mucuna pruriens, gave 

higher biomass than the grain legumes, Cowpea (Vigna unguiculata), and Soybean (Glycine max). Crotalaria 

juncea produced biomass yields around 2300 kg ha- 1 in Zimbabwe and Crotalaria ochraleuca accumulated up to 

10000 kg ha- 1 . Cowpea had biomass yields as low as 150 kg ha- 1 while soyabean had close to 2000 kg ha- 1 biomass yields 

in Zambia. There were no significant soil textural effects on legume biomass yields in the Zimbabwean sites. In Shurugwi, 

wetland soils had higher biomass yields than dryland soils mainly because of the drought that was experienced 

during the season. In Zambia, Mucuna pruriens had high biomass yields on the loamy sands while Crotalaria 

ochraleuca had the highest yields under the sandy clay loams. There were weak but positive correlations of legume biomass 

yield with clay content, organic C, soil pH and available P. 

Key words: green manure and grain legumes, biomass yields, soil characteristics 

Introduction 

Herbaceous legumes have been shown to have potential 

in soil fertility improvement in many parts of 

Africa. Legumes play a significant role in the improvement 

of nitrogen budgets through biological 

nitrogen fixation and cycling of other nutrients, reducing 

the amount of mineral nitrogen fertilizer required 

(Giller, 2002) . Legumes can be grown either 

as sole 'crops in rotation with cereal crops or intercropped 

with cereal crops, depending on their compatibility. 

Green manure legumes like Mucuna pruriens, Crotalaria 

species and Tephrosia species have been identified 

to have potential to produce high biomass and 

increase soil fertility in tum (Muza, 1997; Gilbert, 

1997). Giller and Wilson (1991) reported that green 

manure legumes have potential to accumulate up to 

250 kg N ha- 1 year). Grain legumes that include Glycine 

max, Vigna unguiculata and Cajanus cajan have 

~een shown to have high potential to yield high 

amounts of grain and some residual leaf litter and 

root biomass that can contribute significantly to soil 

fertility improvement (Mapfumo et al. 2001; Nyakanda 

et al. 1997; Saka et al. 1998; Schulz et al. 2001). 

Growth performance of the different legumes varies 

from site to site, with some legumes being more tolerant 

of low soil fertility conditions than others. It is 

usually essential to add P fertilizer to enhance legume 

growth, especially in the communal area soils, 

which are infertile. Hikwa et al (1998) showed that 

biomass yields of Mucuna pruriens doubled with P 

fertilization while there were no P responses with 

Crotalaria juncea. Soil' moisture affects legume performance 

with both waterlogging and drought conditions 

reducing crop growth of some legumes 

(Muza and Mapfumo, 1998). The biophysical 

boundary conditions under which different legumes 

perform need to be ascertained. 

The objectives of this study were to establish the influence 

of soil biophysical conditions on legume 

biomass and grain yields in Malawi, Zambia and 

Zimbabwe, and to evaluate legume performance in 

different agro-ecological zones: Only data from 

Zambia and Zimbabwe are reported. It was hypothesized 

that legume biomass and grain yields 

would increase with increase in clay content, CEC, 

C content and P content. 



On farm experiments were carried .out on soils with 

different characteristics to cover soils of different 

texture, C content, P content, pH and CEC in Zambia 

and Zimbabwe. The legumes were planted in a 

randomized block design, together with fertilized 

and unfertilized maize as treatments. 

Zambia 

Four legumes, Mucuna, Crotalaria juncea, Glycine max 

and Vigna unguiculata, were planted on six farms in 

Zambia in Choma, Magoye, Kabwe, Muswishi, Kasarna 

and Mungwi. Choma and Magoye are in medium 

rainfall areas with annual rainfall of 600-800 

mm, while Kabwe and Muswishi were in high rainfall 

areas (800-1000 mm annual rainfall), and Kasarna 

and Mungwi receive 1000-1200 mm rainfall 

annually. Kabwe and Misamfu were on sandy 

loamy soils, while Mungwi was on a loamy sand 

and Muswishi on a sandy clay loam (Table 1). All 

the sites had ~ow .contents of available P oH~~ than 

7 ).lg P g-I SOlI WIth the exception of Kabwe which 

had .40 ).lg P g-I soil available P. CEC was less than 

10 cmol+kg-I except for Muswishi, which was on.a 

sandy clay loam and had a CEC of 39 cmol+kg-I 

(Table 1). 

ZImbabwe 

The experiment was carried out at 24 sites in two 

districts, Murewa and Shurugwi. Murewa was a 

high rainfall area, receiving up to 1000 mm rainf~ll 

Table 1. Initial soil characterization of Zambian sites 

Site pH (KCI) %sand %clay %silt Textural class 

Kabwe 6.3 77 9 14 Sandy loam 

Misamfu 4.2 70 15 15 Sandy loam 

Muswishi 5.5 58 22 21 Sandy clay loam 

Mungwi 5.1 82.1 9.8 8.1 loamy sand 

apnually while Shurugwi was in a low rainfall area 

receiving around 650 mm annual rainfall. Six legumes 

were planted; Mucuna, Crotalaria juncea, Crotalaria 

grahamiana, Glycine max, Cajanus cajan and Vigna 

unguiculata. 

The sites in Murewa covered red and black clays, 

loamy sands and sands (Table 2). The coarse textured 

soils had low C contents with most of the sites 

on coarse textures having less than 0.8% C, while 

sites on heavier soils had C cQntents of up to 3% C 

(Table 2). All the sites in Murewa had low contents 

of available P and low. CEC with most of the sites 

having less than 3 ).lg P g-I available P and less than 

6 cmol+kg-I CEC. 

All the sites in Shurugwi were on coarse textured 

soils, all with It;sS than 7% clay content (Table 3). pH 

at the sites in Murewa averaged around 5.5 (Table 

2) and were lower than those in Shurugwi which 

had a mean of 7 (Table 3). The sites in Shurugwi 

were of low soil fertility status than those in 

Murewa and Zambia. Most of the sites in Shurugwi 

had less than 0.6% C, 3 ).lg P g-I available P and less 

than 3 cmol+kg-1 CEC (Table 3). 


_Biomass yield of different legumes in Zambia and 

Zimbabwe 

At all the sites, green manure legumes had larger 

biomass yields compared with the grain legumes, 

%C %N iJg g"P Mgme% Ca me% Na me% Kme% CEC 

1.48 0.11 40 2.3 5 0.9 10.1 

0.86 0.06 7 0.8 0.08 0.36 5.8 

1.44 6.2 2.3 4.15 0.05 0.48 39.1 

0.7 0.03 5 0.7 2.1 0 0.1 5.8 

Table 2. Initial soil characteristics of sites in Murewa, Zimbabwe 

Farmer pH (H2O) %sand %silt %clay Textural class 

Nzvere 5.3 86 6 7 loamy sand 

A.Oarare 5.4 32 15 52 Clay 

Kwari 5.6 30 20 49 Clay loam 

Mutsago 5.4 46 14 37 Sandy clay 

Chokurongerwa 5.3 82 6 11 loamy sand 

Matambanadzo 5.4 82 6 11 Sand 

Mandebvu 5.3 82 6 11 Sand 

Kaitano 5.2 84 6 9 loamy sand 

Musegedi 5.4 86 6 7 loamy sand 

Mugwagwa 5.3 86 6 7 loamy sand 

Takarova 5.5 84 6 9 loamy sand 

Ndoro 5.6 42 16 41 Clay 

B.Oarare 5.5 32 18 49 Clay 

Gwara 5.5 46 20 33 Sandy clay loam 

%C %N iJg g'P Mg me% Ga me% Na me% Kme% CEC 

0.44 0.08 2.12 0 0.1 0.03 ·0 0.13 

1.98 0.17 2.27 0.89 1.35 0.05 0.04 2.33 

2.28 0.21 2.07 0.87 1.53 0.03 0.05 2.48 

1.26 0.11 2.02 0.87 1.35 0.05 0.05 2.32 

0.46 0.08 2.02 0.01 0.09 0.03 . 0 0.13 

0.77 0.11 2.96 0.11 0.37 0.08 0.03 0.59 

0.7 0.1 3.11 0.07 0.26 0.02 0.02 0.37 

0.45 0.07 2.57 0 0.05 0.02 0 0.07 

0.39 0.08 2.22 0.01 0.06 0.03 0 0.1 

0.64 0.14 1.63 0 0.06 0.02 0 0.08 

0.49 0.07 2.81 0.04 0.22 0.04 0 0.3 

3.06 0.27 1.63 2.45 2.92 0.04 0.03 5.44 

2.96 0.26 3.56 2.78 3.61 0.05 0.04 6.48 

2.81 0.38 1.78 1.59 2.29 0.06 0.02 3.96 

130 


Table 3. Initial soil characteristics of sites in Shurugwi. Zimbabwe 

Farmer pH (H2O) %sand %silt %clay Textural class 

Gweru 6.87 83 11 6 Loamy sand 

Munyika 6.86 85 9 6 Loamy sand 

Marime 7.13 87 7 6 Sand 

Manatsa 6.82 85 11 4 Loamy sand 

Masendeke 7.33 87 9 4 Sand 

Chimviri 7.38 95 3 2 Sand 

Majoni 7.15 87 11 2 Sand 

Makovere 6.67 90 8 2 Sand 

Mugwagwa 7.45 89 7 4 Sand 

Ngwalati 7.38 85 11 4 Loamy sand 

Gwatsvaira 6.78 91 7 2 Sand 

%C %N j.Jg g.lp Mgme% Ca me% Na me% K me% CEC 

0.28 0.03 0.9 0.1 0.61 0.08 0 0.79 

0.59 0.04 2.76 0.14 0.95 a.09 0 1.18 

0.4 . 0.02 1.01 0.14 0.93 0.09 0.03 1.19 

0.49 0.03 3.27 0.13 1.21 0.07 0.03 1.44 

0.35 0.02 1.07 0.07 0.52 0.07 . 0.01 0.67 

0.37 0.04 1.18 0.02 0.49 0.07 0 0.58 

0.32 0.03 0.51 0.06 0.71 0.07 0 0.84 

0.46 0.01 2.93 0.02 0.42 0.07 0.04 0.55 

0.24 0.01 2.54 0.07 1.11 0.09 0 1.27 

0.23 0.01 0.56 0.15 2.47 0.08 0 2.7 

0.18 0.02 0.73 0.04 0.41 0.07 0.02 0.54 

with Crotalaria juncea and Crotalaria ochraleuca giving 

the highest biomass yields in Zim!:>abwe and Zambia 

(Table 4). Mucuna had yields around ·2000 kg 

ha·J in Zimbabwe while it yielded more than 7000 

kg ha·J in Zambia (Table 4). The mean yields for 

Crotalaria juncea were 2300 kg ha·J in Zimbabwe and 

10000 kg ha·J for Crotalaria ochraleuca in Zambia 

(Table 4). In western Kenya, Ojiem et al (1998) observed 

higher dry matter accumulations of up to 9 t 

ha- J for the green manure legumes (C ochraleuca, C 

grahamiana, C incana and Mucuna) while soyabean 

accumulated dry matter of about 2 t ha· 1• 1,\ Zambia, 

soyabean accumulated biomass yields close to 2000 

kg ha-] while less than 400 kg ha·J biomass yields 

were obtained in Zimbabwe. Cowpea had up to 800 

kg ha·J biomass yields in Zimbabwe while in Zambia 

it was as low as 150 kg ha·1. 

Schulz et al (2001) reported that biomass yield and 

N contribution potential of the different legumes 

varies and may be ranked in terms of soil fertility 

improvement in the following order: green manure 

crops> forage crops> low harvest index grain legumes> 

high harvest index grain legumes. Higher 

biomass yields were observed in Zambia than in 

Zimbabwe, probably because the sites that were 

sampled in Zambia were not under moisture stress 

while the Zimbabwe sites were affected by drought. 

Table 4. Biomass yield of legumes obtained in the 2001/02 season 

from different agro·ecological zones in Zambia and Zimbabwe 

(Murewa and Shurugwi) 

Treatment Biomass yield (kg ha· 1 ) 

Zambia Zambia Zimbabwe Shurugwi 

(800·1000 (1000·1200 (800·1000 « 650 mm 

mm rainfall) mm rainfall) mm rainfall) rainfall) 

Cowpea 146 835 651 

Crotalaria gnihamiana 1715 2056 

Crotalaria juncea 2312 2316 

Crotalaria ochralueca 12841 10000 

Mucuna 4981 10250 2331 1562 

Soyabean 2482 1062 391 

LSO (0.05) 1083.6 974.3 276.2 457.5 

There were no differences in biomass ootained in 

Murewa and Shurugwi; probably because both areas 

were affected by drought (with annual rainfall 

of 542 mm and 595 mm respectively), reducing the 

potentials for legume growth (Table 4). 

Effect of soil characteristi

18000 

1600) 

14000 a) Zambia 

12000 

m 

-" 

l1cxxx) 

~ BOOO 

~ 

iii 

600) 

4000 

2000 

"'''' 

""" 

",m 

,A=. 

'fu21~·25."il 

R>=.Of11!'1' 

:: '1\'t"·TII\T\) 

·~nI5·~'j"/x~1~., 

..".,.., 

'Y· .....ro:.c·JJ2.45 

rf"O.17ID 

"""" 

Y"fil1(&.7'*ltD 

R'''O.OZIO 

.~ ;(, 

~ -- 

..... 

 

- ........ 

-acltiII1I'I8 

C

pressed growth of most crops. The observations 

made in the 2001/02 season however indicate that 

moisture is essential for legume growth. Legume 

biomass yields increase with increase in clay content, 

pH and soil fertility in gener,al. There are interactions 

of the different soil characteristics such as 

soil moisture, clay content, soil pH, and soil fertility 

on legume biomass production. There is therefore, a 

need to explore the effects of the interactions of tlie 

different soil characteristics on legume establishment, 

growth and biomass yield. Green manure legumes 

outyield grain legumes and all legumes require 

some soil moisture to produce meaningful 

biomass yields that can impact on soil fertility improvement. 

More data analysis is required to discriminate 

the importance of the various parameters 

measured. A spatial analysis of the data could help 

in drawing up recommendation domains for the 

various legumes. 


This work was supported by IFAO through a grant 

to TSBF-CIAT. 

References 

Gilbert, R.A. 1998. Undersowing green manures for 


cropping systems of Malawi. In: S.R. Waddington, 

H.K. Murwira, JD.T. Kumwenda, O. Hikwa 

ar:td F. Tagwira (eds), Soil Fertility Research for 

Maize Based Farming Systems in Malawi and Zimbabwe. 


Harare, Zimbabwe. pp 73-80. 

Giller, KE. 2002. Targeting management of organic 

resources and mineral fertilizers: Can we match 

scientists' fantasies with farmers' realities? In: B. 

Vanlauwe, N. Sanginga, S. J. Oiels and R. Merckx 

(eds), Integrated Plant Nutrient Management in 

Sub-Saharan Africa. CAB International, Wallingford, 

UK. pp 155-172. 

Giller, K.E. and K.J. Wilson. 1991. Nitrogen fixation in 

tropical cropping systems. CAB International, Wallingford, 

UK. 313 pp. 

Hikwa, 0, M. Murata, F. Tagwira, C. Chiduza, H. 

Murwira, L. Muza and S. Waddington. 1998. 

Performance of green manurE legumes on exhausted 

soils in northern Zimbabwe: A soil fertility 

network trial. In: S.R. Waddington, H.K. 

Murwira, JD.T. Kumwenda, O. Hikwa and F. 




Harare, Zimbabwe. pp 81-84. 

Ojiem, J.O., J.K. Ransom, a.M. Odongo and 'E.A. 

Okwuosa. 1999. Agronomic and chemical characterization 

of potential green manure species in 

Western Kenya. In: Maize Production Technology 

for the Future, Proceedings of the Sixth Eastern 

and Southern Africa Regional Maize Conference, 

21 -25 September 1~98, Addis Ababa, Ethiopia, 

CIMMYT Maize Program and Ethiopian Agricultural 

Research Organization. pp 210-213. 

Mapfumo, P., B.M. Campbell, S. Mpepereki and P. 

Mafongoya. 2001: Legumes in soil fertility management: 

The case of pigeonpea in the smallholder 

farming systems of Zimbabwe. African 

Crop Science Journal 9(4):629-644. 

Muza, Land P. Mapfumo. 1999. Constraints and 


of sandy soils in Zimbabwe. In: Maize 

Production Technology for the Future, Proceedings 

of the Sixth Eastern and Southern Africa Regional 

Maize Conference, 21 -25 September 1998, 

Addis Ababa, Ethiopia, CIMMYT Maize Program 

and Ethiopian Agricultural Research Organization. 

pp 214-217. 

Nyakanda, c., I.K Mariga, B.H. Ozowela and H.K. 

Murwira. 1997. Biomass production and maize 

yield under tree-based improved fallow of Sesbania 

and Pigeon pea. In: S.R. Waddington, H.K 

Murwira, JD.T. Kumwenda, O. Hikwa and F. 



Soil Fert Net and CIMMYT-Zimbabwe, Harare, 

Zimbabwe. pp 115-119. 

Saka, A.R., JoO.T. Kumwenda, A.G. Allison, J.B. Kamangira 

and W.T. Bunderson. 1998. Integrating 

pigeon peas into smallholder farming systems to 

improve soil fertility and crop yields in Malawi. 

In: Maize Product-ion Technology for the Future, 

Proceedings of the Sixth Eastern and Southern 

Africa Regional Maize Conference, 21 -25 September 

1998, Addis Ababa, Ethiopia, CIMMYT 

Maize Program and, Ethiopian Agricultural Research 

Organization. pp 218-222. 

Schulz, S., R.J. Carsky and S.A. Tarawali. 2001. Herbaceous 

legumes: The panacea for West African 

soil fertility problems? SSSA Special Publication 

No. 58, Madison, WI, USA. pp. 179-195. 

[ 


133

EFFECT OF DIFFERENT GREEN MANURE LEGUMES AND THEIR TIME 'OF 

PLANTING ON MAIZE GROWTH AND WITCHWEE.D '(STRIGA ASIATICA) 

CONTROL: A PRELIMINARY EVALUATION 

Abstract 

LAURENCE JASI, OSTIN A. CHIVINGE and IRVINE K. MARIGA 

Department of Crop Science, Faculty of Agriculture, University of Zimbabwe, 

P. O. Box MP 167, Mount Pleasant, Harare, Zimbabwe 

A pot experiment was established at Henderson Research Station (30 0 58', 17 0 35') and a field experiment was conducted 

at Mlezu (29 0 30', 19 0 8'), Zimbabwe during the 2001/2002 cropping season to evaluate the effect of green manure 

legumes, and their time of planting when intercropped with maize, on Striga asiatica emergence and maize 

growth and yield. The green manure legumes tested were velvetbean (Mucuna pruriens, fish bean (Tephrosia vogelii), 

sunnhemp (Crotalaria juncea), and dolichos (Lablab purpureus). There was less Striga asiatica incidence 

when legumes were planted at the same time as maize, compared with staggering the planting dates. There were no significant 

differences among the green manure legumes in their ability to suppress Striga. Planting the maize and legumes 

at the same- time increased interspecies competition and reduced maize leaf area and plant height significantly. 

Velvet bean reduced maize leaf area more than the other legumes. Howe-ver, the competitive effect of the legumes did not 

reduce grain yield in the pot experiment. 

Key words: Striga asiatica, Striga suppression, velvet bean, fish bean, sunnhemp, dolichos 

Introduction 

Striga asiatica is one of the biological constraints to 

maize production in the smallholder farming areas 

of Zimbabwe. Striga species are difficult to control 

because they cause damage before the Striga plants 

emerge from the soil after most weeding operations 

have been. completed (Musambasi, 1997). Research 

in Zimbabwe and elsewhere hilS shown that S. asiatica 

can be controlled by several methods. These include 

hand pulling before flowering, hoeing, ridging, 

trap and catch cropping, intercropping with 

legumes such as cowpea (Vigna unguiculata (L.) 

Walp], bambara nut (Vigna subterranea) and soyabean 

(Glycine max), crop rotations with non hosts 

and false hosts, resistant varieties, herbicides (such 

as 2,4-D, dicamba and trifluralin), use of multipurpose 

trees and timely planting (Chivinge et aI, 2001; 

Kasembe, 1999; Musambasi, 1997). However, all of 

these have shortcomings as shown by very little or 

no adoption. Nitrogen (N) has been shown by many 

workers to reduce Striga infestation and improve 

maize grain yield. Smallholder farmers sometimes 

apply N on the soil around the maize plant at about 

four weeks after planting. However, sources of mineral 

N are very expensive for smallholders. Alternative 

methods that can add N to the soil and at the 

same time control Striga are therefore urgently 

needed. Green manure legumes such as velvetbean 

(Mucuna prurie-ns), fish bean (Tephrosia vogelW, sunnhemp 

(Crotalaria juncea), and dolichos (Lablab purpureus) 

are an important source of nutrients 

(particularly biologically fixed N) in Zimbabwe 

(Chibudu, 1998): Green manure legumes have been 

used by Zimbabwe smallholders for soil N amelioration 

in areas such as Chihota, Mangwende and 

Zvimba (Hikwa et al,1998;Chibudu, 1998). However, 

the effects of these green manure legumes on 

Striga asiatica dynamics have not been studied in 

Zimbabwe. In Sudan, L. purpureus planted on the 

same day as sorghum (Sorghum vulgare) reduced the 

Striga plant population density by 48-93%, their dry 

weight by 83-97% and number of capsules by 52 

100% (Babiker, 2000). This present study was conducted 

to investigate the effect of maize/green manure 

legume intercrops and their time of planting 

on Striga asiatica emergence and maize yield components 

in Zimbabwe. 


A pot trial was established in January 2002 at Henderson 

Research Station, just north of Harare. Blac:;: 

polythene bags measuring 30 cm diameter and 40 

cm height were used. The experiment was arranged 

as a completely randomized design and replicated 

four times. Three maize seeds of hybrid SC501 were 

planted in each pot and these were thinned to one 

plant after two weeks. Legumes (Mucuna pruriens, 


Lablab purpureus, Crotalaria juncea, Tephrosia vogelii) 

were planted at the same time as the maize in half 

the pots and then two weeks after planting maize in 

the others. Sale crops were also irfcluded as a control 

treatment. The legumes were thinned to one 

plant per pot at two weeks after planting. Compound 

D (8N: 14P20s: 7K20) was applied to supply 

0.6 g N/maize plant, 1.1 g P20s/maize plant and 0.6 

g K20/maize plant; an equivalent of 300 kg/ha of 

the fertilizer in the field. Ammonium nitrate (34.5% 

N) was applied at 4 and 8 weeks to make up to 0.8 

g/pot total N. Plants were given supplementary irrigation 

to field capacity as necessary. Maize plant 

heights were taken at 4 and 8 weeks, while leaf area, 

shoot dry weight and root dry weight were tak~n at 

4, 6 and 8 weeks after planting. Striga counts per pot 

were recorded at 40 days after planting and weekly 

thereafter. 

A similar trial was conducted at Mlezu Institute of 

Agriculture, also near to Haran~, in a field that had 

been artificially infested with S. asiatica the previous 

year. Maize was planted to achieve a plant population 

density of 37 037 velvet bean and dolichos 

plants/ha, 74 064 sunnhemp plants/ha and 10 000 

fish bean plants/ha. Compound D was applied as 

initial fertilizer at a rate of 300 kg/ha. No topdress 

N was applied because of prolonged dry spells. S. 

asiatica counts were taken from the four centre rows 

every two weeks after planting. Data analysis was 

done using GENST AT 5 Release 3.22. Treatment differences 

were compared using the least significant 

difference (LSD P

.. 

~ 

Table 2. The effect of time of planting green manure 

legume and type of maize/legume intercrop on maize 

leaf area (cm 2 ) 6 WAP. 

Maize leaf area 

Intercrop Planted at same Planted 2 weeks 

time 

after maize 

Maize/dolichos 2077a 2529 

Maize/velvet 1311 b 2790 

Maize/sunnhemp 2269a 2094 

Maize/fishbean 1891a 2925 

LSD (P < 0.05) 753 

SED 365 

Time of planting 1887 2584 

LSD (P < 0.05) 377 

SED 183 

CV% 23 

Means followed by the same letter in acolumn are not significantly 

different IP < 0.05). 

1.6 

1.4 

U 

eo 0.8 

0 

..J 

0.6 

c , 0.4 

e 0.2 

~ 

. ~ 

Vi 

20 40 60 80 100 120 

Time (Days) 

Figure 2. Effect of time of planting green manure legumes on Striga 

of December. This was followed by poor rainfall 

distribution in February where 60 mm of rainfall fell 

on one day during the month. The subsequent 

months also had a very low rainfall frequency. The 

experiment was severely affected. 

Striga asiatica counts. Although not significantly 

different, legumes planted two weeks after maize 

(PLTW) allowed slightly higher Striga asiatica counts 

of 0.25 compared to 0.06 for simultaneous planting 

in the field at Mlezu (80 days after plantirlg). Legume 

intercropsshowed that they do not differ in the 

way they suppress S. asiatica emergence. Time of 

establishment of the legume is more important. 

Discussion 

Table 3. The effect of intercrop type and time of planting 

of green manure legumes on Strig;: asiatica dry weight (gl 

pot) 

I)ntercrop 

S. asiatica dry weight (g/pot) 

Planted Legumes planted 

simultaneously 2 weeks later 

I Maize/dolichos 0.24 0.26 

Maize/velvet bean 0.05 0.13 

Maize/sunnhemp 0.26 0.49 

Maize/fish bean 0.17 0.50 

Significance (P < 0.05). 

(lntercrop·Time) 

NS 

SED 0.20 

Table 4. The effect of intercropping and time of 

planting of green manure legumes on maize yield (g/pot) 

Intercrop 

Maize grain yield (g/pot) 

PLST 

PLTW 

Maize/dolichos 2.1 2.7 

Maize/velvet bean 1.4 11.9 

Maize/sunnhemp 12.4 7.9 

Maize/fish bean 5.2 4.9 

Significance (P < 0.05). 

IIntercrop·Time) 

NS 

SED 7.0 

Grain yield. There were no interaction effects of 

type of intercrops and their time of planting on 

maize yield. Maize/dolichos and maize/velvet 

bean resulted in higher yields when legumes were 

planted two weeks later, although this was not significant. 

The opposite was true for maize/ 

sunnhemp and maize/fish bean (Table 4). 

Field Experiment at Mlezu 

The rainfall distribution at Mlezu is in Figure 3. The 

experiment was established during the second week 

250 

208.6 

The study demonstrated that differences in planting 

date for the component crops influenced competition 

between component crops in the green manure 

legume/maize intercrops (measured as leaf area 

and plant height). Planting legumes and maize at 

the same time increased inter species competition 

for growth limiting factors, resulting in reduced 

maize leaf area and maize plant heights. Where the 

planting of maize and legumes was staggered, interspecies 

competition was reduced and maize attained 

higher leaf areas and height than with simultaneous 

planting. Velvet bean reduced maize leaf 

area. This could be attributed to velvpt bean's ro- 

200 

"§ 

/ill Frequency 

~ '" ~Rainfall 

§ fi 150 

108.5 

=5 

ust growth habit that enhances its competitive 

ability for growth limiting factors. Gilbert (1998) reported 

that Mucuna can be excessively competitive 

with maize because it has an aggressive climbing 

growth habit. When maize is intercropped with velvet 

bean and established at the same time, interspedes 

competition increases, particularly at later 

stages of growth. For instance, there was no difference 

in maize leaf area between the sole maize and 

the intercropped maize at two and four weeks after 

planting. During this period, interspecies competition 

of the intercrop appears not affect maize leaf 

area. Probably plants were too young to interfere 

with each other but as they grow, competition begins 

that reduces leaf area at later stages of growth. 

When legumes were planted two weeks later, S. asiatica 

counts were generally higher than when legumes 

were planted at the same time as maize. This 

trend was also observed in the field experiment. 

Legumes caused suicidal germination of S. asiatica 

when planted at the same time and this could have 

reduced the S. asiatica numbers. When maize and 

legumes were planted two weeks apart, more S. asiatica 

plants emerged as the parasite that germinated 

from the maize stimulant successfully attached to 

the maize roots. Carsky et al (1994) postulated three 

reasons for reduction of Striga when intercropped 

with cowpea. These include suicidal germination of 

Striga, release of nitrogen into the soil and shading 

which consequently lowers soil temperature. These 

reasons were also supported by Musambasi et al 

(2002) wh0 suggested that legumes could provide 

shade which smother and kill S. asiatica. These reasons 

can therefore be used to extrapolate the results 

obtained. Planting legumes and maize at the same 

time, allowed legumes to quickly develop a crop 

canopy that produced a shading effect, lowering the 

soil temperatures that could have affected the emergence 

of S. asiatica. Babiker et al (1993) found thaI, 

the density of Striga hermontheca was reduced in a 

sorghum-dolichos intercrop. Probably maize is a 

better germination stimulant than the legumes 

tested. The S. asiatica that germinates due to the legumes 

is of no significance as compared to S. asiatica 

that maize stimulates and supports. However, the 

time of planting the green manure makes the difference 

in terms of S. asiatica numbers. Similar trends 

were observed with S. asiatica dry weights. 

There was no grain yield from the field experiment 

owing to the poor rainfall distribution. Yield from 

the pot experiment was not ,influenced by the green 

manure legumes or their time of planting. It would 

be interesting to find out how these factors influence 

yield in a normal season under field conditions. 

The competitiv~ effects of the green manure 

legumes experienced during the fourth to the sixth 

week after planting were not enough to significantly 

reduce yield in the pot experiment. 


Green manure legumes intercropped with maize 

should be planted two weeks later to reduce competition 

among component crops. For S. asiatica, the 

green manure legumes should be established at the 

same time with maize in a field heavily infested 

with S. asiatica. A legume that does not compete 

with maize for resources should be planted at the 

same time as maize. The experiment needs to be 

conducted again in another season to get conclusive 

results in the field. 

Acknowledge,ments 

We extend our gratitude to the Rockefeller Founda 

tion Forum on Agricultural Resource Husbandry 

for funding this work. 

References 

Babiker, A.G.T., N.E. Ahmed, A.H. Mohammed, M. 

E. EI Mana, and S.M. EI Tayeb, 1993. Striga hermontheca 

on sorghum: Chemical and cultural 

control. In: British Crop Protection Conference 

Weeds, Brighton, UK. pp. 103-108. 

Babiker, A.G.T. 2000. Striga research in the Sudan: 

Towards an integrated control strategy. In: 

Mgonja, M.A., Chivinge, O.A. and Monyo, E.s. 

(eds). Striga Research in Southern Africa and Strategies 

for Regionalized Control Options: Proceedings 

of the SADC Striga Working Group Workshop 

22-23 May 2000, Dar-es Salaam, Tanzania. 

SADC/ICRISAT Sorghum and Millet Improvement 

Program, ICRISAT-Bulawayo. pp. 55-67. 

Carsky, RJ., L. Singh, and R Ndikawa, 1994. Suppression 

of Striga hermontheca on sorghum using 

a cowpea intercrop. Experimental Agriculture 

30:349-358. 

Chibudu, C. 1998. Green manuring crops in a 

maize-based communal area, Mangwende: Experiences 

using participatory approaches. In: 

Waddington S.R, H.K. Murwira; J.D.T. Kumwenda; 

D. Hikwa and F. Tagwira (eds). Soil Fertility 

Research for Maize Based Farming Systems in 

Malawi and Zimbabwe. Proceedings of the Soil 

Fert Net Results and Planning Workshop held 

from 7 to 11 July 1997 at Africa University, Mutare, 

Zimbabwe. Soil Fert Net and CIMMYf 

Zimbabwe, Harare. pp. 97-90. 

138 


Chivinge, O.A., E. Kasembe, and I.K. Mariga, 200l. 

The effect of cowpea cultivars on witchweed and 

maize yield. Proceedings of the British Crop Protection 

Council Conference-Weeds. Brighton, UK. pp. 

163-168. 

Gilbert, RA., 1998. Undersowing green manures for 


cropping systems of Malawi. In: Waddington S. 

R, H.K. Murwira; J.D.T. Kumwenda; D. Hikwa 

and F. Tagwira (eds). Soil Fertility Research for 

Maize Based Farming Systems in Malawi and Zimbabwe. 

Proceedings of the Soil Fert Net Results 

and Planning Workshop held from 7 to 11 July 

1997 at Africa University, Mutare, Zimbabwe. 

Soil Fert Net and ClMMYT-Zimbabwe, Harare. 

pp.73-79. 

Hikwa, D., M. Murata, F. Tagwira, C. Chiduza, H. 

Murwira, L. Muza, and S.R. Waddington, 1998. 

Performance of green manure legumes on exhausted 

soils in northern Zimbabwe: A Soil Fertility 

Network Trial. In: Waddington S.R., H.K. 

Murwira; J.D.T. Kumwenda; D. Hikwa and F. 

Tagwira (eds). Soil Fertility Research for Maize 

Based Farming · Systems in Malawi and Zimbabwe. 

Proceedings of the Soil Fert Net Results and 

Planning Workshop held from 7 to 11 July 1997 

at Africa 'University, Mutare, Zimbabwe. Soil 

Fert Net and ClMMYT -Zimbabwe, Harare. pp. 

81-84. 

Kasembe, E. 1999. The effect of different cowpea 

[Vigna unguiculata (L.) Walp] cultivars and time 

of ridging on witchweed [Striga asiatica (L.) 

Kuntze] management in the smallholder 'farming 

sector. MPhil thesis. University of Zimbabwe, 

Zimbabwe. 

Musambasi, D. 1997. Maize witchweed (Striga asiatica 

(L.) Kuntze management technologies in the 

smallholder farming sectors of Zimbabwe. MPhil 

thesis, University of Zimbabwe. 

Musambasi, D., O.A. Chivinge and I.K. Mariga. 

2002. Intercropping maize with grain legumes 

for Striga control in Zimbabwe, African Crop Science 

Journal 10(2):161-171. 

GrainLegumes and Green Manures for Soil Fertility in Southern Africa 139

LEGUMINOUS AGROFORESTRY OPTIONS FOR REPLENISHING 

SOIL FERTILITY IN SOUTHERN AFRICA 

PARAMU L. MAFONGOYA 1·, E. KUNTASHULAl, F. KWESIGA 2 , T. CHIRWAl, 

R. CHINTU 1 , G. SILESHl 1 , and J. MATIBINll 

'Zambia ICRAF/Agroforestry Project, P.O. Box 510046, Chipata, Zambia; 

2SADC-ICRAF A grofores try Research Project, P. O. Box MP 163, 

Mount Pleasant, Harare, Zimbabwe 

(* Corresponding author, E-mail: mfongoya@zamnet.zm) 

Abstract 

Nitrogen is the major nutrient limiting maize production in Zambia and Southern Africa. Removal of subsidies on inorganic 

fertilizers made them very expensive and most farmers cannot afford them. Short duration planted fallows using 

a wide range of leguminous trees have been found to replenish soil fertility and increase subsequent maize yields. 

Species such as Sesbania sesban, Tephrosia vogelii and Cajanus cajan have been found excellently suited for planted 

fallow technology. These improved fallow crop rotations are being adopted by small-scale farmers in Eastern Zambia. 

Since the seminal paper of Kwesiga and Coe (1994), research has been done to understand how the planted tree fallows 

replenish soil fertility and improve maize yields. 

A wide range of species has been screened as alternatives to sesbania fallows to overcome some of the limitations of sesbania. 

Species such as Gliricidia sepium, Leucaena leucocephaJa have maintained maize yields of 3 t/ha over 8 years 

of cropping when sesbania fallows yields declined to 1.1 t/ha after 3 years of cropping. The selection criteria for good 

fallow species are high biomass production and litterfall. Maize yields after fallows were highly correlated to biomass 

and litterfall yields. High quality biomass, which is low in lignin, polyphenol and high in N, is needed for higher maize 

yields. Mixing of gliricidia and sesbania fallows resulted in higher maize yields compared to single species fallows (3.0 

vs. 1.8 t/ha). Mechanisms on how mixed fallows work need further investigation. 

Preseason inorganic N (N0 3 +NH4) was highly correlated with maize yield (r 2 = 0.62) and this could be used to select 

fallow species and management practices. Nutrient budgets of N, P and K showed over 8 years that a positive balance of 

Nand P was maintained for coppicing fallows while a negative balance of K started showing from the fourth year onwards 

on fertilized maize, gliricidia, ieucaena and sesbania fallows. This points to the need to .use inorganic fertilizers 

such P and K to supplement N supply from leguminous fallows. Improved fallows increased infiltration, reduced runoff, 

increased water storage, and reduced soil loss. The order was sesbania = tephrosia > natural fallow =maize + fertilizer. 

The biophysical limits of most fallow species and other emerging issues such as pests and diseases, the need to inoculate 

with rhizobium, amount of N fixed by different species and provenances and soil acidification under improved 

fallows are under further research. 

Biomass transfer technology using biomass from leguminous trees was evaluated on maize and vegetable production in 

the dambos (wetlands). Maize and vegetable yields were significantly increased by application of high quality biomass 

from gliricidia and leuceana. However, financial analysis showed that it is not viable to apply biomass on a low value 

crop like maize. Biomass transfer was economically viable on high value crops such a vegetables. 

Key words: Eastern Zambia, improved fallows, soil fertility, nutrient budgets, nitrogen fixation and sustainability 

Introduction 

Soil infertility is now increasingly recognized as the 

fundamental biophysical root cause for declining 

food security is smallholder farmers of sub-Saharan 

Africa (Sanchez et al. 1997). Maize is a staple food 

crop in Southern Africa. Nitrogen is the major nutri- 

ent that limits maize productivity, with phosphorus 

and potassium in limited cases. Although inorganic 

fertilizers are used in the region, the amounts applied 

are normally insufficient to meet .crop de 

mands due to high costs and uncertain availability. 

Mostcciuntries in southern Africa have developed 

fertilizer recommendations for major crops, some- 


times with regionally specific adaptations. However, 

the amount of fertilizer used in southern Africa 

is very small in comparison to other parts of the 

world, with the highest rates found in a country like 

Zimbabwe in the commercial sector. For most 

smallholders, fertilizer use is as low as 5 kg/hal 

year (Gerner and Harris, 1993). While the need for 

increased use of inorganic fertilizers in southern Africa 

is clear, there are problems with an approach 

based exclusively on inorganic fertilizers where water 

supply is limited and variable. In many areas 

outside the higher rainfall zones or away from irrigated 

areas, any sensible farmer will use expensive 

mineral fertilizer with caution and supplement with 

organic sources. In most areas, fertilizer is therefore 

used mainly on home fields, gardens or high value 

crops such as cotton and vegetables. 

While the need for increased fertilizer use in Southern 

Africa is apparent to all, the challenge of achieving 

this is very great. A high external input strategy 

cannot rely on fertilizer-seeds-credit packages, without 

addressing other requirements for successful 

uptake of green revolution technologies, including 

water management, credit systems, infrastructure, 

fertilizer manufacture and supply and access to 

markets. Most African conditions are unlike the 

plains of Asia so that the approaches which produced 

such successes there are not easily transferable 

to the African continent. Given the acute poverty 

and limited access to mineral fertilizers an ecologically 

robust approach of improved fallows is discussed 

in this synthesis. This approach is a prodllct 

of many years of agroforestry research and development 

by ICRAF and its partners in southern Africa. 

Improved Fallows 

Improved fallows and their topology 

Improved fallows are the deliberate planting of fastgrowing 

species, usually wood tree legumes, for 

rapid replenishment of soil fertility. Fallows are as 

old as agriculture in southern Africa. Grass fallows 

are a common feature of the farming systems in the 

sub humid and semiarid zones of the region. Improve 

fallows were not a major area of research during 

the green revolution due to the focus to eliminate 

soil constraints by use of mineral fertilizers. 

With the development of the second soil fertility 

paradigm based on sustainability considerations 

(Sanchez, 1994), the biological dimensions of soil 

fertility began to receive increasing attention and 

research on improved fallows has increased since 

the mid 1980s. Reported work includes Kwesiga 

and Coe (1994), Drechsel et al. (1996), Rao et al. 

(1998) and Snapp et al. (1998). 

Large-scale adoption of short-term improved fallows 

by farmers is now taking place in southern Africa 

and east Africa. The main species used are legumes 

of the genus Sesbania, Tephrosia, Leucaena, 

Gliricidia, Crotalaria and Cajanus. 

Non-coppicing fallows 

Since the seminal work of Kwesiga and Coe (1994) 

on Sesbania fallows, a lot has been learnt about the 

performance of improved fallows. There has been 

extensive testing ·of fallows on farm to determine 

the maize productivity and processes that influence 

fallow performance. The performance of Sesbania 

and Tephrosia in a wide range of biophysical conditions 

is shown on Table 1. Improved fallows of twoyear 

duration with both species significantly increased 

maize yields well above unfertilized maize 

(which is a cpmmon farmer's practice). Fertilized 

maize performed better than improved fallows in 

most cases. It is very clear from these results that 

the residual effects of fallows on maize yield declined 

after the second year of cropping. In a third 

year of cropping, maize yields were similar to unfertilized 

maize. Farmers have asked researchers 

how can they extend the residual effects of fallows 

beyond two years of cropping. Suggestions have 

included applying low rates of inorganic fertilizer in 

the second or third year of cropping to increase residual 

effects. Alternatively, farmers can use species 

of trees which coppice after cutting and use 

coppice regrowth to increase residual effects. 

Coppicing fallows 

Most of the work on improved fallows has concen 

trated on Sesbania sesban, but this species has draw 

backs. When cut at fallow termination, which is 2 

years of growth, it will not resprout or coppice. 

Hence Sesbania fallows are called non-coppicing fal 

lows. Non-coppicing species include Tephrosia vogeUi, 

Tephrosia candida, Cajanus cajan and Crotalaria 

spp. In the case of Sesbania farmers must rely on a 

Table 1. Maize grain yield after Sesbania sesban and Tephrosia 

vogeli; fallows on farmers' fields in eastern Zambia during 1998· 

2000 

Fallow species 

Maize grain yield t ha'\ 

Land use system (LUS) Year 1 Year 2 Year 3 

Farmers testing Sesbania fallow 3.6 2.0 1.6 

Sesbania sesban Fertilized maize 4.0 4.0 2.2 

fallows 

Unfertilized maize 0.8 1.2 0.4 

LSD (0.05) 0.7 0.6 1.1 

Number of farmers 8 6 4 

Farmers testing Tephrosia fallow. 3.1 2.4 1.3 

Tephrosia vogelii Fertilized maize 4.2 3.0 2.8 

fallows 

Unfertilized maize 0.8 0.1 0.5 

LSD (0.05) 0.5 0.6 0.9 

Number of farmers 17 9 5 

142 


fresh supply of seedlings or seed reserves to generate 

their fallows. Trials at Msekera Research Station, 

Zambia have shown that natural regeneration 

of Sesbania fallows through seed reserves is possible, 

but highly erratic. Farmers therefore prefer to reestablish 

fallows from bare-rooted seedlings. 

The residual effects of sesbania fallows on subsequent 

maize yields have been shown to be high for 

two or three seasons, but they will start to decline 

rapidly in the third season (Table 1). This may be 

related to depletion of soil nutrients and deterioration 

in soil chemical and physical properties. It can 

be hypothesized that fallows with coppicing species 

will increase residual effects beyond 2 to 3 years, 

because of the additional organic inputs derived 

each year from coppice regrowth. Coppicing species 

include Gliricidia sepium, Leucaena leucocephala, 

Calliandra calothyrsus, Senna siamea and Flemingia 

macrophylla. An experiment was established in the 

early 1990's at Msekera Research Station to test this 

hypothesis. The species tested were Senna siamea, 

Gliricidia sepium, Leucaena leucocephla and Flemingia 

macrophylla, which were compared with grass fallows, 

and continuous maize with or without recommended 

fertilizer as additional controls. The experiment 

has been cropped for 8 seasons during 

which maize and coppice growth were monitored 

(Figure 1). 

The species showed significant differences in coppicing 

ability and biomass production (Table 2). 

Leucaena, gliricidia and Senna siamea had the highest 

coppicing ability and biomass production while calliandra 

and flemingia performed less. Sesbania, as 

expected, did not coppice. The trends in maize 

yields over the 8 seasons are shown in Figure 1. 

Maize yields were high for the first 3 seasons and 

declined to the same level as control plots for sesbania. 

Flemingia and calliandra showed low maize 

8.0 

7.0 

~ 5.0 

~ 4.0 

.'" 

.lij 3.0 

i3 

2.0 

""""-Gliricidia ~L eucaena ~M+F 

-+- M·F ....- Sesbania --lIE- Nalural fallow 

6.0 

I I I I I I I I 

I = SED 

1.0 

0.0 

1995 1996 1997 1998 1999 2000 2001 2002 

Years/seasons 

Figure 1. Grain yield (t ha l) of maize obtained from various 

fallow species for eight seasons at Msekera, eastern Zambia 

yields over years. There were no significant differences 

in maize grain between gliricidia and leucaena 

fallows over the seasons. 

The effects of different fallow species on maize yield 

can be explained partly by the different amounts of 

biomass added and the quality of the biomass and 

coppice regrowth during the dry season. Species 

such as leucaena and gliricidia have good coppicing 

ability and produce large amounts of high quality 

biomass, with high nitrogen content and low contents 

of lignin and polyphenols. Biomass with low 

lignin and polyphenols and high N release N rapidly, 

resulting in higher maize yields. Although sesbania 

produces high quality biomass, it's inability to 

coppice renders it unable to supply biomass during 

the cropping period leading to less prolonged residual 

effects. Species such as flemingia, calliandra and 

Senna siamea produce low-quality biomass, which is 

high in lignin, polyphenols and low in nitrogen. 

This will lead to N immobilization and reduced 

maize yields. 

We hypothesized that the coppicing of gliricidia can 

utilize the residual soil water after maize harvest 

and recover soil nitrogen below the maize rooting 

depth during the long dry season from April to October. 

We monitored the soil water and nitrogen 

dynamics in. all treatments for two seasons, 1997 to 

1998, to test this hypothesis. This information will 

be used to simulate the long-term trend of maize 

yield, water and nitrogen dynamics using the 

WaNuLCAS model. Theoretical simulations indicated 

that gliricidia coppicing could utilize enough 

residual soil water in an average rainfall year of 980 

mm/year to produce 2-4 t/ha of tree biomass and 

increased maize yield. 

At the end of the dry season, soil moisture profiles 

confirmed that the coppicing gliricidia treatment 

utilized about 40 mm more water, primarily from 

below 75 cm soil depth, than in either the sesbania or 

continuous cropping treatments. This is probably 

an under estimation of the total deep uptake of residual 

water by the coppicing gliricidia since soil wa- 

Table 2. Total seasonal coppice biomass (t ha I) recorded from 

various fallow species at Msekera, eastern Zambia during 1995· 

2002 

1996 1997 1998 1999 2000" 2001 2002 

C. calothyrsus 0.3 0.4 0.2 0.4 0.6 0.4 0.6 

S.siamea 2.8 2.1 1.6 1.7 1.8 1.2 2.2 

F. mycrophylla 0.6 0.6 0.3 0.6 0.7 0.4 0.5 

G. sepium" 1.7 1.5 1.3 1.1 3.1 1.4 1.2 

L. leucocephalla" 3.5 2.6 1.7 2.8 3.4 2.2 1.9 

"Ieucaena has more twig biomass added to the system than Gliricidia which also 

has low survival 

"" Biomass cut in 2 months interval INov. Jan & Mar) normal- Nov. Dec & Jan) 


143

ter content at 180 cm was still well below that of the 

other treatments. Deeper access tubes are required 

to determine the actual depth of water extraction by 

gliricidia roots. Based on the amo_unt of biomass 

produced by gliricidia, we would expect the uptake 

of an additional 40 mm of water, i.e. rooting depth 

would have to go beyond another metre deeper. 

This trend in the soil water profile between the 

three treatments was maintained even after five 

months with a total of 767 mm of rain, indicating 

the maize crop in the gliricidia treatment used more 

water than in the other two treatments. In addition, 

the high soil water content in both the sesbania and 

no fallow treatments indicate that nitrogen leaching 

can be a serious problem during this rainy period in 

both the sesbania and no fallow treatments. Indeed, 

measurements of inorganic nitrogen profiles for all 

three treatments confirmed substantial differences 

in N levels below 75 cm depth, with maximum concentratiof'.s 

in the no fallow treatment, followed by 

the sesbania and gliricidia treatment (Figure 2). 

These findings indicate that coppicing gliricidia provides 

a much more sustainable system than the sesbania 

fallow systems because of its ability to utilize 

residual soil water and to prevent N leaching in 

such environments. 

We have concluded that gliricidia and leucaena have 

potential as coppicing fallows. Cumulative maize 

yield of these fallows is greater than from sesbania 

after 4 years of cropping. This is because of the constant 

nutrient replenishment obtained from harvesting 

coppice regrowth. We will continue the trial for 

another 5 seasons to test the sustainability of coppicing 

fallows in terms of nutrient budgets such as 

NPK. In addition, we have established on farm trials 

to evaluate responses widely and screen more 

coppicing species. 

Mixed fallows 

Improved fallow systems with shrub legume species 

like sesbania have become a key agroforestry 

technology for soil fertility management in southern 

Africa and western Kenya. Large increases in maize 

yields have been reported following short duration 

(9-24 months) fallows with single species. Sesbania 

has been the main focus for improved fallows for its 

ability to add huge amounts of high quality biomass 

and fuelwood provision. The dependence on a few 

successful fallow species has revealed some drawbacks. 

Sesbania is susceptible to root-nematode and 

Mesoplatys beetle. Introduction of new species has 

led to the outbreak of new pests and diseases as observed 

with Crota/aria grahamiana in western Kenya 

(Cadisch et al. 2001). Thus there is an urgent need 

to diversity the specie5 and fallow types to farmers. 

Mixing species with compatible and complementary 

rooting or shoot growth patterns in fallows may 

lead to a more diverse system and maximize above 

and below ground growth resource utilization. Undersowing 

herbaceous legumes under open canopy 

species may use more photosynthesis radiation by 

the whole canopy and increase primary production. 

Mixing shallow-rooted species with deep-rooted 

species can enhance the soil water and nutrient uptake 

zone within the soil profile. More importantly, 

it will enhance utilization of subsoil nutrients, e.g. 

nitrate lost through leaching . . Mixing species in fallows 

may also reduce the risk of failure with fallow 

establishment, in case one species is susceptible to 

water stress, diseases and pests. Multiple products 

. obtained from mixed fallows and increased biodiversity 

of the system are other positive characteristics 

that make the whole system more attractive. 

We tested a variety of mixed fallows of tree legumes 

or tree legumes with herbaceous legumes to test the 

above stated hypotheses. 

Mixing coppicing fallow species such as Gliricidia 

sepium and a non-coppicing species (Sesbania sesban) 

significantly increased maize yields compared to 

single species fallows (':;-igure 3) . However mixtures 

of noncoppicing species did not increase maize 

yield compared to sole species. The mixture of cop 

20 

40 

(.) 

ToUillnorgenlc N (mWkoJ 

0.00 2.00 4.00 6.00 8.00 10.00 12.00 

60 

......... Cajanus cajan 

60 ___ Natural (allow 

_____ Maize with fetilizer 

100 

~ Maize without fetilizer 

120 __ Sesbanla sesban 

140 

160 

160 

200 

200 

(bl 

Inorganic nitrate (rng N kg ,l, 

6 8 10 12 16 

Figure 2. Total inorganic nitrogen (a) and inorganic nitrate (b) (mg I 

kg soil) as affected by two year fallow species and soil depth at 

Msekera, eastern Zambia in February 1998 and February 2001 

144 


picing and noncoppicing species reduces the level 

of subsoil nitrate and controlled mesoplatys beetles. 

However mixing gliricidia or tephrosia or sesbania 

with herbaceous legumes such as mucuna or archer 

dolichos reduced tree growth and hence maize yield. 

These mixtures also lead to the build up of mesop:atys 

beetle, which may have led to a larger attack 

of sesbania by the beetles (Sileshi and Mafongoya, 

2002). 

Prediction of improved fallows performance 

Many studies have shown a 3 to 4-fold increase in 

maize grain yields after two year improved fallows. 

Most of these studies were conducted under research 

station conditions. However, when improved 

fallows are tested in a wide range of environmental 

conditions there is variability in maize 

grain yields. The explanations advanced for this 

variability is based on trial and error. There is need 

for a predictive understanding of how fallows perform 

in different agroecological conditions. 

The work done for many years has shown how organic 

decomposition and nutrient release is affected 

by the levels of polyphenols, lignin and nitrogen 

contents of the organic inputs (Mafongoya et al. 

1998). Recently we have found also that maize 

yields after fallows with various tree legumes were 

negatively related to the L+P: N ratio (Figure 4). 

Fallow species with high N, low lignin and low 

polyphenols such as gliricidia and sesbania gave 

higher maize yields compared to species such as 

flemingia, calliandra and senna. This work has clearly 

shown that it is not the quantity of polyphenols 

which is critically important but also the quality of 

the polyphenols as measured by their protein binding 

capacity (Mafongoya et al. 2000). Legume species 

for improved fallows can be screened for their 

suitability based on the above characteristics. 

Soil inorganic N before a cropping season is an accepted 

test for soil N for soil productivity. Results 

of our studies in Southern Africa show that preseason 

inorganic N can also be an effective indicator 

of plant available N after different improved fal 

48 

lows. Our results indicate that preseason inorganic 

N (N03 + NH4) can be more related to maize yield 

than preseason N03 alone in a tropical soil with a 

pronounced dry season (Figure 4). Large amounts 

of NH4 can accumulate during a dry season, and it 

may not be nitrified when the soil is sampled at the 

beginning of the rainy. season. We concluded that 

preseason inorganic N is a relatively rapid and simple 

index that is related well to maize yield on N 

deficient soils and hence it can be used to screen fallow 

species and management practices. 

Improved fallows of S. sesban tested under a wide 

range of conditions showed a strong linear relationship 

between maize yield and standing biomass at 

fallow clearance (r2=0.50). Preseason inorganic was 

also related to standing biomass (r2= 0.60) and 

standing biomass was related to clay content of the 

sites (r2=0.SO). From these studies the impact of improved 

fallows on maize yield were clearly related 

(a) 

35 y = -0_19x+ 3.24 

1 R2 = 0_83 

30, 

.,-;;; 2_5 

r. 

to amount of biomass, quality of biomass in terms 

of L+P: N ratio, litter fall, and preseason inorganic 

N. These results were in agreement with those reported 

by Mafongoya et al. 1999). 

Based on these results we can safely conclude that 

the main predictors of fallow performance are quantity 

of biomass, quality of the biomass, preseason 

inorganic N and texture on the soil. The relevance of 

these predictors needs to be tested over a wide 

range of conditions and with different fallov: species. 

Soil Chemical Properties 

The major soil chemical changes that take place under 

tree fallows are increases in labile pools of SaM, 

N stocks, exchangeable cations and extractable P 

(Rao et al. 1998). Details of the mechanisms of soil 

improvement by tree fallows were reviewed by 

(Buresh and Tian, 1998). In theory, planted tree fallows 

are expected to improve soils faster than natural 

fallows since the land is completely covered by 

fast growing leguminous trees for 2 to 3 years. 

However the magnitude of these soil improvements 

depends on tree species, length of fallow, soil and 

climatic conditions. In this section, we will concentra 

te on these changes as measured from experiments 

in southern Africa. 

Biological nitrogen fixation and N cycles 

The contribution of leguminous trees through N2 

fixation is well recognized, although not all legumes 

fix N2. Nitrogen fixation in the humid and subhumid 

zones of Africa has been reviewed by Sanginga 

(1995). There has been little work on quantification 

of N2 fixation by trees in southern Africa. This work 

has proved to be difficult due to constraints in the 

methodologies for measurins N2 fixed. A series of 

multi-location trials have been set to measure the 

amount of N2 fixed by different tree genera and 

provenances (Table 3) using the 15N natural abundance 

method. The data on percent Ndfa shows 

high variability among provenances of the same 

species for N derived from atmospheric N2 fixation. 

Sanginga et al. (1990) found that percent Ndfa 

ranged from 37 to 74% for provenances of Leucaena 

leucocephala. The data shown in Table 3 falls within 

the range reported by Sanginga et al. (1990). These 

preliminary data show the huge potential of trees to 

fix N2 and increase N inputs in N deficient soils. 

Our future analysis will focus on factors responsible 

for this variability in N2 fixation across s~tes and 

how to optimize N2 fixation under field conditions. 

Barrios et al (1997) m~asured availability of soil N 

following 2- and 3-year fallows a N- deficient soils 

in eastern Zambia. His results confirmed that tree 

fallows increase N availability compared to continuous 

cropping without fertilization. Subsequent N 

measurements down to 200 cm in the soil profile 

showed significant N inorganic accumulation at 

depth during the cropping phase (Figure 2). 

These results show that improved fallows can create 

a very "leaky" N cycle after fallow clearance. Most 

of the N is leached beyond the rooting depth of 

maize and this N is released from organic inputs 

before peak N demand by maize. Hence there will 

be asynchrony between N release and N demand by 

maize. Consequently there is need to design systems 

which try to minimize N losses and increase N 

use efficiency, and cycling. 

Based on those results we designed mixed fallows 

of coppicing species and noncoppicing species. The 

hypothesis is that the coppicing species will act as a 

permanent "safety net" for N when the noncoppicing 

fallows are cut due to resprout growth and deep 

root system in the soil. Results of gliricidia and sesbania 

mixed fallows have shown higher maize prcr 

ductivity and efficient N cycling compared to single 

species fallows (Figure 3). 

Soil acidification and cations 

There are several reports on soil pH and improved 

fallows. Topsoil pH decreased under fallows of 

Acacia auriculiformis (Drechsel et al. 1996). However 

Oonsson et al. 1996) found no changes in soil pH 

after fallows. Our results over a 10-year period 

showed significant decrease in topsoil soil pH, 0-60 

cm and an increase in soil pH with depth (Figure 5). 

This decrease in topsoil pH and increase in soil pH 

is attributed to leaching of N03 and cations such as 

magnesium as shown in Figure 6. The movements 

of cations from the topsoil were also confirmed by 

low CEC in 0-20 cm (6.25 compared with 9.50) in 

the 20-100 cm soil profile. These pH changes, which 

will take place after fallows, may have little effect 

Table 3. Biological nitrogen fixation (%BNF) of coppicing species/ 

provenances across three sites in eastern Zambia after 1 year of 

growth 

Kalichero Kalunga Masumba 

Treatment %BNF Nkg/ha %BNF Nkg/ha %BNF N kg/ha 

A. angustisma 52.1 210.4 61.8 201.4 54.8 260.8 

C. calothyrsus 48.4 81.4 44.1 214.4 48.7 193. 

G. sepium 79.2 212.4 71.4 408.4 70.8 297.5 

l. collinsii 74.7 303.2 57.2 236.7 102.1 475.9 

l. diversif(Jlia 35/88 77.5 196.8 33.8 88.6 50.0 161.1 

l. diversifolia 53/88 58.4 121.5 14.0 40.5 46.9 112.6 

l. esculenta 52/87 70.9 99.3 . 46.6 110.1 46.7 274.5 

l. esculenta·Machakos 84.7 223.6 35.2 120.2 69.0 538.0 

l. pallida 58.6 87.8 33.7 125.2 44.7 168.1 

146 


pH 

-0.40 -0.20 0.00 0.20 0.40 0.60 0.80 1.00 

20 

40 

60 

K80 

;; 100 

c 

~ 120 

140 

160 

180 

200 

__ dena feb02-Noll97 

-e-dena Noll9B-NoII97 

Figure 5. Changes in soil pH as affected by two-year fallow species 

and soil depth at Msekera, eastern Zambia 

on maize productivity in base rich soils. However 

their long term effects on acidic, low-activity clay 

soils may have a major effect on crop yields and 

threaten the long-term sustainability of improved 

fallows. Management practices such as zero tillage 

after fallows and regular application of lime may 

have to be adopted to deal with such problems of 

decrease in pH and cation leaching. 

Soil carbon and improved fallows 

The debate on carbon and global warning has 

gained momentum. Of late, there has been increased 

scientific interest in measuring carbon sequestration 

in different land use systems to mitigate 

climate change issues. Agroforestry land use systems 

have been cited to sequester the most soil C 

without a lot of scientific evidence. We monitored 

soil C in long-term trials with improved fallows. 

There were significant increases in soil carbon in the 

topsoil as compared to deep horizons (Table 4). 

These results are in agreement with those of Onim 

et al.; (1990) in western Kenya using improved fallows. 

Of scientific and practical importance is how 

is the carbon protected against loss after fallow 

clearance. Our research program is looking how 

different soil aggregates store C and how soil aggregation 

is affected by soil texture and fallow management 

over the long term. This will enable us to 

model carbon dynamics and climate change. 

Table 4. Amount of organic carbon (%) measured in different 

soil depths under a two·year non·coppicing fallow species at 

Msekera, eastern Zambia 

Year 

Soil depth (em) 1997 2002 Percent 

increase 

0·20 0.95 1.i 2 17.89 

20-40 0.78 0.94 20.51 

40·S0 O.Sl 0.77 2S.23 

SO·100 0.51 0.55 7.84 

100·150 0.3S 0.49 3S.11 

150·200 0.28 0.37 32.14 

SED 0.05 O.OS 20.00 

c mol kg" 

0.3 0.5 0.7 0.9 1.1 1.3 1.5 

o+---~--~--~--~--~~~ 

20 

40 

E 60 

.!!. 80 

'-e-Cc 

ị. 100 -+-M+f 

~ 120 -->

with the natural fallow, which did not lose its aggregate 

stability. The decrease in aggregate stability 

was more pronounced under sesbania and maize 

without fertilizer as compared ~ith cajanus and 

maize with fertilizer. Under a sesbania fallow system, 

the improvement in soil structure is more evident 

and this is reflected by results from our time to 

runoff studies. Time to runoff after fallow clearing 

was in the order of: natural fallow> S. sesban > fertilized 

maize. After one season of cropping, time to 

runoff decreased in all treatments except that the 

natural fallow maintained the longer time to runoff, 

reflecting good maintenance of aggregate stability. 

Through rainfall simulation studies we evaluated 

effects of improved fallows on runoff infiltration 

soil and nutrient losses under improved fallows. 

Tree fallows of sesbania, gliricidia mixed with archer 

dolichos increased infiltration rates significantly 

compared with continuously fertilized maize plots 

(Figure 7). Fallows compared to no tree plots also 

significantly reduced soil loss (Table 5). 

Improved fallows improve soil physical properties 

as evidenced by increase in infiltration rates, increased 

infiltration decay coefficients, reduced runoff 

and soil losses. However these benefits are short 

lived and they decline rapidly during the first year 

of cropping. This was supported by increase in soil 

loss in the second year (TableS) and decrease in in 

40 

35 

i )0 

! 2S 

~ 20 

c 

.g 15 

g 

~ 10 

S. sesban T. vogelii N. rallow fl:t1ilizal. maize G. sepium+ A. 

dolichos 

Treatmenls 

I-October 2000 [JOclober2001 I 

Figure 7. Infiltration rate under different fallows measured at 

Msekera (source; Nyamadzowo et a/2002) 

Table 5. Soil loss (g/m2) measured under various fallow species 

and maize at Kalunga Farmers Training Center in eastern Zambia 

Treatment October 2000 October 2001 

Sesbania sesban 0.0 5.0 

Tephrosia voge/ii 4.5 15.8 

Natural fallow 0.0 19.5 

Fully fertilized maize 63.8 ..0.5 

~iratro (Macroptilium atropurpureum) 0.0 0.7 

ILSO 15.3 . 

Source: Nyamadzowo et al2002 

filtration rates as well (Figure 7). However, mixing 

a coppicing species like gliricidia and a herbaceous 

legume like archer dolichos maintained high infiltration 

rates and reduced soil loss over two years of 

cropping. 

Sustainability of Improved Fallows 

Improved fallows with sesbania or tephrosia have 

been shown to give maize grain yields of 3 to 4 t/ha 

without any inorganic fertilizer addition. Palm 

(1995) showed that organic inputs of various tree 

legumes applied at 4 t/ha can supply enough nitrogen 

for maize grain yields of 4 t/ha. However, 

most of these organic inputs could not supply 

enough phosphorus and potassium to support such 

maize yields. 

The question ferr sustainability is: Can improved fallows 

potentially mine P and K over time while 

maintaining a positive N balance? To answer that 

question we conducted nutrient balances on improved 

fallow trials at Msekera Research Station. 

These plots were under fallow-crop rotations for 8 

years. The objectives of these studies on nutrient 

balances addressed the following questions: 

• Can nutrient balances be used as land quality indicators? 

• Can they be used to assess soil fertility status, 

productivity and sustainability? 

• Can they be used as a policy instrument for the 

types of fertilizers to be imported or distributed 

to farmers? 

The nutrient balances considered nutrients added 

through leaves and litter fall, which were incorporated 

after fallows as inputs. The nutrients in maize 

grain harvested, maize stover removed and fuelwood 

taken away at end of the fallow were considered 

as nutrient exports. 

For all the land use systems, there was a positive N 

balance two years of cropping after the fallows 

(Table 6). Fertilized maize had the highest N balance 

due to the annual application of 112 kg N/ha 

for the past 10 years. However, unfertilized maize 

had lower balances due to low maize grain and 

stover yields over time. The tree-based fallows had 

a positive N balance due to BNF and deep capture 

of N from depth. These results are consistent with 

those of Palm (1995) that showed that organic inputs 

could supply enough N to support maize grain 

yields of 3 to 4 t/ha. 

However in the second year of cropping (1999) the 

N balance was very small. This is consistent with 

our earlier results, which showed a decline of maize 

148 


Table 6. Nutrient balance (kglha) under two year non·coppicing 

fallow species at Msekera, eastern Zambia 

Nitrogen Phosphorus· Potassium 

limd use systems 1998 1999 1998 1999 1998 1999 

Cajanus cajan 27 5 21 8 13 ·9 

Sesbania sesban 22 5 39 24 -42 ·32 

Natural fallow 8 11 19 15 ·10 ·4 

Fully fertilized maize 150 103 57 43 ·19 ·17 

Unfertilized maize 31 11 29 20 19 ·1 

yields in the second year of cropping after two-year 

fallows. The huge amount of N supplied by fallows 

could be lost through leaching beyond the rooting 

depth of maize. Our leaching studies have clearly 

shown substantial inorganic N at depth under 

maize after improved fallows. These results imply 

that if cropping goes beyond three years after fallows 

there will be a negative N balance. Thus the 

recommendation of two years of fallows followed 

by two years of cropping is well supported by N 

balances and maize grain yield trends. Most of the 

land use systems showed a positive P balance. This 

can be attributed to low offtake of P in maize grain 

yield and stover. In addition, this site had a high 

phosphorus status. The trees could also have increased 

P availability through secretion of organic 

acids and the increased mycorrhizal population in 

the soil. These issues are under investigation at our 

site. In general, we have observed positive P balances 

over eight years. However this result needs 

to be tested on farm where the soils are inherently 

10winP. 

Most land use systems showed a negative balance 

for K. For tree based systems, sesbania showed a 

higher negative K balance compared to pigeonpea. 

This is attributed to the higher fuelwood yield of 

sesbania with subsequent higher export of K compared 

to pigeonpea. The higher negative K balance 

for fully fertilized maize is due to higher maize and 

stover yield which exports a lot of potassium. This 

implies that the K stocks in the soil are very high 

and that K mining has not reached a point where it 

negatively affects maize productivity. However in 

sites with low stocks of K in the soil, maize productivity 

may be adversely affected. 

Nutrient balances were conducted for coppicing fallows 

using gliricidia compared to non-coppicing fallows 

using sesbania for four cropping seasons after 

fallow clearance. Gliricidia fallows maintained a 

positive N balance. This was attributed to resprout 

growth, which was applied to maize as a source of 

nutrients and deep capture of N from depth by the 

well-established gliricidia rooting system. All land 

use systems showed a positive P balance. However 

from the third season of cropping onwards sesbania 

fallows, fertilized maize and gliricidia fallows had a 

large negative balance. This was attributed to removal 

of nutrients in ·stover maize or leaching of K 

from surface soils .. 

Overall, the tree based fallows maintained a positive 

Nand P balance. However on low· P status, a 

negative P balance would be expected. There was a 

negative K balance with most land use systems. It 

can be hypothesized that as we scale up improved 

fallows on depleted soils on farmer's fields, 1

and K to maize as an equivalent amount of commercial 

NPK fertilizer, and in some caGes maize yields 

were higher with tithonia biomass than commercial 

inorganic fertilizer. Recent work in Malawi 

(Ganunga et al. 1998) and Zimbabwe (Jirl and Waddington, 

1998) have similarly reported tithonia biomass 

to be an effective nutrient source for maize. 

Biomass transfer using leguminous species is a far 

much sustainable means of maintaining nutrient 

balances in maize:based systems as these trees are 

able to fix atmospheric N2. Tithonia is not a legume, 

and it does not biologically fix atmospheric N2. The 

transfer of tithonia biomass to fields, therefore, constitutes 

the cycling of nutrients within the farm and 

landscape rather than a net input of nutrients to the 

system. The continual transfer of nutrients from 

tithonia hedges to crop fields constitutes nutrient 

mining and might not be sustainable for long periods. 

Whereas the application of fertilizers to tithonia 

could ensure sustained production of tithonia, 

this is unlikely to be an option for resource-poor 

farmers. The integration of tithonia with N2-fixing 

legumes may merit investigation. 

Synchrony between nutrient release from tree litter 

and crop uptake can potentially be achieved in a 

biomass transfer system. The management factors 

that can be manipulated to achieve this are litter 

quality, rate of litter application, method and time 

of litter application (Mafongoya et al. 1998; 1999). 

However variability in climatic factors such rainfall 

and temperature makes the concept of synchrony ·an 

elusive goal to achieve in practical terms (Myers et 

al. 1994). 

Although prunings from MPTs increased maize 

yield, cutting transporting and managing prunings 

on crop fields require high labour inputs Oama et al. 

1997; Jama et al. 1998; Mutuo et al. 2000). Where 

family labour is available at no additional cost, the 

technology can be profitable even where land is 

SC(lrce Oama et al. 1997; Mutuo et al. 2000). However, 

considering that farm labour is one of the most 

constraining inputs in smallholder agriculture, the 

associated cost makes this technology unattractive 

and may serve as a disincentive for its adoption by 

farmers. In monetary terms, the higher maize yield 

does not compensate for the high labour cost. In 

promoting this technology, farmers may require to 

be provided with additional resources to invest in 

labour and land. Most economic analyses have 

shown that it is unprofitable to invest in a biomass 

transfer system when labour and land are scarce. 

However, in areas where land is abundant and the 

prunings are applied to high value crops like vegetables, 

the technology·is profitable (ICRAF, 1997). 

Biomass transfer could find a niche for vegetable 

production in dambo areas of southern Africa. A 

dambo is a shallow, seasonally or permanently waterlogged 

depression at or near the head of a natural 

drainage network, or alternatively occurs independently 

of a drainage system (Chenje and Johnson, 

1996; Breen et al. 1997). Dambos cover about 240 million 

hectares in sub-saharan Africa (Andriesse, 

1986). They are some of the most productive natural 

ecosystems in the Southern African region. They 

provide water for domestic use, good soils for agricultural 

production, grazing grounds for livestock, 

fish and support a wide range of wildlife and birds 

(Raussen et al. 1995). Dambos are considered extremely 

vulnerable to poor agricultural practices, 

and hence dambo cultivation was illegal for instance 

in Zimbabwe. However, rising population pressure 

has caused the agricultural use of dambos to become 

increasingly important (Kundhlande et al. 1994). For 

example, vegetable gardens cover 15000-20000 ha 

(Bell et al. 1987) of the estimated 1.28 million ha of 

dambos in Zimbabwe. 

However, without applying fertilizers or cattle manure 

smallholder farmers cannot produce vegetables 

successfully in some vf the dambos (for example 

in eastern Zambia) that are degraded due to continuous 

cultivation for over 25 years (Raussen et al. 

1995). The removal of subsidies and increase in interest 

rates in most of sub Saharan Africa has 

caused decline in inorganic fertilizer use, and this 

decline in the smallholder sector is even greater, 

suggesting that for many farmers the use of fertilizer 

is not a viable option any more. Cattle manure 

use could also become limited since not all farmers 

have animals to produce adequate quantities of manure. 

In addition, transport problems for the large 

quantities of manure needed and the spread of 

weeds due to the manure use may limit its utilization. 

Therefore, the use of biomass transfer in sustaining 

vegetable production in the dambos of southern 

Africa could be a viable option. 

An experiment conducted with 43 farmers by Kuntashula 

et al (2003) showed that Gliricidia biomass 

transfer technologies produced cabbage, onion and 

subsequent maize yields comparable with the full 

fertilizer application (Tables 7 and 8). The biomass 

transfer technologies also recorded higher cabbage, 

onion and maize net incomes than the control, and 

required lower cash inputs than the fully fertilized 

crop (Figures 8 and 9). Like in maize based systeMS, 

net incomes of the biomass treatments in vegetable 

production were substantially reduced by the labour 

costs for pruning and incorporation of the biomass. 

However, in vegetables the high price of 

products more than compensated these costs. The 

study concluded that the use of gliricidia biomass 

150 


Table 7. Mean cabbage and onion yields (fresh weight) in dambos 

using inorganic fertilizers or organic inputs from manure, gliricidia 

and leucaena biomass in Chipata South district, 2001 

Treatments Cabbage yield Onion yield 

(t ha·l ) n 31 (t hal) n - 12 

Manure 10 t ha· l + half fertilizer 66.8 96.0 

Fully fertilised 57.6 57.1 

Gliricidia 12 thaI t 53.6 79.8 

Gliricidia 8t hal 43.1 68.3 

leucaena 12 t ha I 32.6 

Control 17.0 28.1 

S.e.d (p - 0.05) 5.41 11.48 

.. leucaena 12t ha·' was not used in onion trials 

I Biomass treatments are reported on dry weight basis 

N - Number of farmers participating in the experiment 

transfer could be a viable alternative to inorganic 

fertilizer although farmers taking up the technology 

will however need adequate supply of labour. 

The effectiveness of biomass transfer of nutrient 

sources using organic inputs from MPT species depends 

on their chemical composition (Mafongoya 

and Nair, 1997). These systems can meet the N requirement 

of most crops in smallholder farming 

systems. However, they cannot meet the requirement 

of P. There is need to apply inorganic sources 

of P in addition to organic sources. When biomass 

is also valued as fodder there is need to assess the 

trade off of applying it directly to the soil or feeding 

it to livestock and then applying the resultant manure.. 

There is evidence to indicate that depending 

on the quality of the biomass there may be no adliiICabbage 

• Maize + Cabbage 

16000,--------------------------- 

14000 

12000 

10000 

8000 

6000 

4000 

2000 

o 

Table 8. Maize grain yields (t ha· t at 13% moisture content) on 

residual plots in dambos after cabbage and onion production using 

inorganic fertilizers or organic inputs from manure, gliricidia and 

leucaena biomass'in Chipata South district, 2001 

Treatments After cabbage After onion 

(n - 21) (n - 10) 

Manure lOt ha·1 + Y2 fertilizer 4.2 3,1 

Fully fertilized 3.9 2.5 

Gliricidia 12 t ha It 4.9 3.9 

Gliricidia 8 thai 4.3 3.3 

leucaena 12 t ha I 3.2 

Control 2.9 1.7 

S.e.d (p - 0.05) 0.48 0.49 

.. leucaena 12t ha I was not used in onion trials 

I Biomass treatments are reported on dry weight basis 

N - Number of farmers participating in the experiment 

vantage in feeding it to livestock and then applying 

the manure as a source of N to crops (Mafongoya et 

al. 1999). However, in other Instances, it has been 

shown that it is more advantageous to first feed the 

biomass to livestock and then apply the resulting 

manure to crops Gama et al. 1997). 

In summary, the biomass transfer system hasgreatest 

potential when biomass is of high quality and 

rapidly releases nutrients, the opportunity cost of 

labour is low, the value of the crop is high and if the 

biomass does not have other valued uses other than 

as source of nutrients. 

Future Research Needs 

This synthesis has described the progress that has 

been made during the past 10 years in research efmlOnion 

• Onion + maize 

6000 ;------------------~ 

5000 

4000 

3000 

2000 

1000 

• Above nel incomes are reported per hectare basis. Average actual area put to cabbage by a 

• Above net incomes are reported per hectare basis . Average actual area put to onion by a farmer 

Figure 8. Net income US$ hal)" from cabbage and subsequent 

Figure 9. Net income US$ hal)" from onion and subsequent maize 

maize on 31 vegetable gardens in Chipata South District in 2001 on 12 vegetable gardens in Chipata South District in 2001 


forts to understand the mechanisms involved in 

how improved fallows work. A lot of knowledge 

has been generated. However other aspects of improved 

fallows have received littie research. These 

will be highlighted in future research directions. 

Accumulation of litter on the soil surface and micro 

climate changes may lead to increased activity of 

soil macro fauna under tree fallows, particularly in 

subhumid zones of southern Africa. No reports 

have been published on the role of soil fauna, the 

functions of specific groups and the scope of their 

manipulation through quality of biomass produced 

by different species. The increased soil fauna will 

playa significant positive role in litter decomposition, 

nutrient mineralization and improvement of 

soil physical properties. This area deserves further 

research. 

The work on improved fallows has focused on few 

species - such as Sesbania, Tephrosia, Crotalaria and 

Gl~ricidia . With regard to extrapolation, further 

work is needed to identify more species for improved 

fallows. Given a large number of potential 

species, the selection process could be accelerated 

by creation of a database containing fallow performance 

in relation to environmental factors such as 

rainfall, soil type and chemistry and incidence of 

pests and diseases. Our recent trials across sites 

have shown a great potential for Tephrosia candida as 

alternative species to Sesbania and T. vogelli and 

equally Leucaena collinsii and Acacia angustissima as 

alternative coppicing fallow species to G. sepium . . 

The biophysical limits of improved fallows need be 

developed to facilitate scaling up with minimum 

research efforts. Simulation modeling, both as a research 

and extrapolation tool, has a potential for integrating 

research results, identifying key components 

or process that merit greater research attention, 

identifying ecozones where appropriate fallow' 

species and management techniques have a good 

chance of success. 

The debate on global warming and carbon sequestration 

has gained momentum recently. Agroforestry 

land use systems have been reported to have 

huge potential to sequester soil carbon. However 

there are few studies if any in Southern Africa, 

which have measured C sequestration in improved 

fallows. The relationship between soil aggregates 

and carbon storage needs further research. 

As noted earlier, the interaction of pests with soil 

fertility is gaining widespread attention due to 

wider interest in scaling up of improved fallows. So 

far most of the resear.ch efforts have concentrated on 

insects pests and nematodes. Equally important are 

plant diseases and weeds. Little effort has been invested 

in these issues. With scaling up across many 

ecozones, the incidence of new pests and diseases 

will increase. Hence, there will be need to monitor 

pests and diseases with farmers to determine the 

few economic pests to deal with in a concerted research 

programme. Such work is now underway in 

southern Africa. 

Many of the species currently used in improved fallows 

are prolific seed producers. If not managed 

well these species can become invasive weeds and 

become a menace to other ecosystems. To date 

there has been no concerted research effort to determine 

the weediness of introduced fallow species. 

There is urgent need to use current models to predict 

the potential of new species to become invasive 

weeds, study the reproductive biology and design 

management ,practices that will reduce the weediness 

of improved fallow species. 

On nutrient depleted soils, two-year fallows with 

fast growing leguminous trees such as sesbania and 

tephrosia can replenish soil N stocks for the production 

of 3 to 4 tlha of maize grain yield , The residual 

effects of such fallows extend to 2 to 3 years after 

fallow termination. Ho;vever coppicing fallows like 

gliricidia can maintain maize grain yields of 3 tlha 

over an 8-year period after fallow clearance. However, 

where soils are deficient in P, inorganic P 

sources are needed to increase productivity of the 

soil. 

Research during the last decade has established the 

main mechanisms on how improved fallows work. 

Despite significant progress in biophysical research 

in improved fallows in southern Africa, the application 

of that science by small-scale farmers is still 

minimum. The main challenge now is to increase 

the generation of viable and acceptable fallow options 

that can make improved fallows more productive 

to increase the income and food security of 

small-scale farmers. 

Future research issues on biomass transfer will involve 

the residual effect of low and high quality biomass, 

combination of organic and inorganic sources 

of nutrients, effect of biomass banks on nutrient 

mining, agronomic research of biomass transfer of 

different leguminous species, and economic analysis 

of the systems. 


The authors are very grateful to SIDA and CIDA for 

funding this research for more than 10 years. 

152 


References 

Andriesse, 1986. Area and Distribution. The Wetlands 

and Rice in sub-Saharan Africa. Proceedings 

of a Workshop, 4-8 November, 1985, 

Ibadan, Nigeria, pp. 15 - 30. 

Adejuwan, J.O. and Adesina, F.A. 1990. Organic 

matter and nutrient status under cultivated fallows: 

an example of Gliricidia sepium fallows 

from South Western Nigeria. Agroforestry Systems 

10:23-32. 

Barrios, E., Kwesiga, F., Buresh, R.J. and Sprent, J.I. 

1997. Light fraction soil organic matter and 

available nitrogen following trees and maize. 

Soil Science Society of America Journal 61:826-831. 

Bell, M.R, Faulkner, P., Hotchkiss, R, Lambert, R, 

Robert, N. and Wind ram, A. 1987. The use of 

dambos in rural development, with reference to 

Zimbabwe. Final Report aDA Project R 3869. 

Breen, CM., QUllu1, N.W. and Mander, J.J. (eds) 

1997. Wetlands conseroation and management in 

southern Africa: Challenges and opportunities. 

xxviii. IUeN The World Conservation Union. 

164 pp. 

Buresh, RJ. and Tian, G. 1997. Soil improvement by 

trees in sub-Saharan Africa. Agroforestry Systems 

38:51-76. 

Cadisch, G., Gathumbi, S.M., Ndufa, KJ. and Giller, 

K.E. 2001. Resource acquisition of mixed species 

f~lIows - competition or complementarity? In: 

Balanced Nutrient Management Systems for the 

Moist Savanna and Hum:d Forest Zones ofAfrica (B. 

Vanlauwe, N. Sanginga and R Merckx, eds.)., 

Kluwer Academic Publishers, Dordrecht, The 

Netherlands. Volume In press. 

Chenje, M. and Johnson, P. (eds) (1996). Water in 

southern Africa. Maseru/Harare: SADC/IUCN/ 

SARDC 

DrechseL P., Steiner, K.G. and Hagedorn, F. 1996. A 

review on the potential of improved fallows and 

green manure in Rwanda. Agroforestry Systems 

33:109-136. 

Gachengo, CN. 1996. Phosphorus release and availability 

on addition of organic materials to phosphorus 

fixing. soils. MPhil Thesis, Moi University, 

Kenya. 

Ganunga R, Yerokun O. and Kumwenda JD.T. 

1998. Tithonia diversifolia: An organic source of 

nitrogen and phosphorus for maize in Malawi. 

In: Waddington, S.R et al. (Eds), Soil Fertility Research 

for Maize-based Farming Systems in Malawi 

and Zimbabwe., Soil Fert Net and CIMMYT 


Gerner, H. and Harris, G. 1993. The use and supply 

of fertilizers in sub-Saharan Africa. In: Reuler, 

H.V. andPr~, W. (Eds). The Role of Plant Nutrients 

for Sustainable Food Crop Production in sub 

Saharan Africa. Vereniging van Kunstrnest Producenten, 

Leidschendam, The Netherlands. 

Hullugalle, N.R. and ,Kang, B.T. 1990. Effect of 

hedgerow species in alley cropping systems on 

surface soil physical properties of an Oxic 

paleustaJf in: southwestern Nigeria. Journal of Agricultural 

Science, Cambridge 114:301-307. 

ICRAF 1997. Annual Report 1996. International 

Centre for Research in Agroforestry, Nairobi, 

Kenya. 

Jama B.A., Swinkels RA., Buresh RJ. 1997. Agronomic 

and economic evaluation of organic and 

inorganic sources of phosphorus in western 

Kenya. Agronomy Journal 89:597-604. 

Jama B, Buresh RJ., Ndufa J.K. and ShepherdK.D. 

1998. Vertical distribution of roots and soil nitrate: 

Tree species and phosphorus effects. Soil 

Science Society ofAmerica Journal 62:280-286. 

Jama B., Buresh RJ. and Place F.M. 1998. Sesbania 

tree fallow on phosphorus-deficient sites: maize 

yield and financial benefits. Agronomy Journal 

90:717-726. 

Jiri, O. and s. Waddington 1998. Leaf pruhings from 

two species of Tithonia raise maize yield in Zimbabwe, 

but take a lot of labour! Target (Soil Fert 

Net Newsletter) 16:4-5. 

Jonsson, K, Stahl, L. and Hagberg, P. 1996. Tree fallows: 

a comparison between five tropical tree 

species. Biology and Fertility of Soils 23:50-56. 

Kundhlande, G., Govereh, J., Muchena, O. 1995. Socioeconomic 

constraints to increased utilisation 

of dambos in selected communal areas. In: 

Dambo Farming in Zimbabwe: Water Management, 

Cropping and Soil Potentials for Smallholder Farming 

in the Wetlands, (Eds ROwen., K Verbeek., J. 

Jackson., and T. Steenhuis). Harare: CIFAD and 

University of Zimbabwe. pp. 87-96. 

Kuntashuia, E., Mafongoya, P.L., Sileshi, G. and 

Lungu, S. 2003. Potential of biomass transfer 

technologies in sustaining vegetable production 

in the wetlands (dambos) of eastern Zambia. Experimental 

Agriculture In Press. 

Kwesiga, F. and Coe, R 1994 Potential of short rotation 

sesbania fallows in eastern Zambia. Forest 

Ecology and Management 64:161-170. 

Mafongoya P.L. and Nair P.K.R 1997. Multipurpose 

tree prunings as a source of nitrogen to maize 

under semiarid conditions in Zimbabwe. Nitro 


153

gen recovery rates in relation to pruning quality 

and method of application. Agroforestry Systems 

35:47. 

Mafongoya P.L., Giller K.E. and ~alrn CA. 1998. Decomposition 

and nitrogen release patterns of tree 

prunings and litter. Agroforestry Systems 38:77 

97. 

Mafongoya P.L., Barak P. and Reed J.D. 1999. Carbon, 

nitrogen and phosphorus mineralization 

from multipurpose tree leaves and manure from 

goats fed these leaves. Biology and Fertility of 

Soils. In press. 

Mafongoya, P.L., Barak, P. and Reed J.D. 2000. Car 

" bon, nitrogen and phosphorus mineralization of 

tree leaves and manure. Biology and Fertility of 

Soils 30:298-305. 

Mapa, RB. and Gunasena, H.P.M. 1995. Effect of 

alley cropping on soil aggregate stability of a 

tropical Alfiso1. Agroforestry Systems 32:237-245. 

Maroko J.B., Buresh R.J. and Smithson P.C 1998. 

Soil nitrogen availability as affected by fallowmaize 

systems in two soils in Kenya. Biology and 

Fertility of Soils 26:229-234. 

Myers RJ.K., Palm CA, Cuevas E., Gunatilleke I.U. 

N. and Brossard M. 1994. The synchronization of 

nutrient mineralization and plant nutrient demand. 

In: Woomer PL 'and Swift MJ (eds.) The 

Biological Management of Soil Fertility, John Willey 

and Sons, Chichester, UK. pp. 81-116. 

Mutuo P.K., Mukalama J.P., Agunda J., Kobare S., 

Palm CA., Jama B., Kinyangi J., Amadola S. and 

Niang A. 2000. 00- farm testing of organic and 

inorganic phosphorous source on maize in Western 

Kenya. In: The Biology and Fertility of Tropical 

Soils:TSBF Report, 22 pp. 

Nyathi P. and Campbell B.M. 1993. The acquisition 

and use of miombo litter by small-scale farmers 

in Masvingo, Zimbabwe. ,Agroforestry . Systems 

22:43-48. 

Nziguheba G., Palm CA., Buresh RJ. and Smithson 

P.C 1998. Soil phosphorus fractions and adsorption 

as affected by organic and inorganic 

sources. Plant and Soil 198:159-168. 

Nziguheba, G., Merckx, R and Palm, CA. 2002. 

Soil phosphorus dynamics and maize response 

to different rates of phosphorus fertilizer applied 

to an acrisol in western Kenya. Plant and Soil 

243:1-10. 

Onim, J.F.M., Mathuva, M., Otieno, K. and Fitzhugh, 

H .A. 1990. Soil fertility chat.ges and response 

of maize and beans to green manures of 

leucaena, sesbania and pigeonpea. Agroforestry 

Systems 12:197-215. 

Palm, CA. 1995. Contribution of agroforestry trees 

to nutrient requirements of intercropped plants. 


Palm CA., Gachengo CN., Delve RJ., Cadisch G. 

and Giller K.E. 2001. Organic inputs for soil fertility 

management: some rules and tools. Agriculture, 


Rao, M.R, Nair, P.K.R and Ong, K. 1998. Biophysical 

interactions in tropical agroforestry systems. 


Raussen, T., Daka, AE., and Bangwe, L. 1995. Dambas 

in Eastern Province: Their agroecology and use. 

Chipata: MAFF, DepaTtrnent of Agriculture. 

Sanchez, P.A 1994. Tropical soil fertility research, 

towards the second paradigm. Transactions 15 th 

World Congress of Soil Science (Acapulco, Mexico) 

1: 65-88. 

Sanchez, P.A., Shepherd, J.D., Soule, M.J., Place, F. 

M., Buresh, RJ., Izac, AM.N., Mokwunye, AU., 

Kwesiga, F.R., Ndiritu, CG. and Woomer; P.L. 

1997. Soils fertility replenishment in Africa: An 

investment in natural resource capital In: Buresh 

RJ, Sanchez PA and Calhoun F (eds) Replenishing 

Soil Fertility in Africa, Soil Science Society 

of America Special Publication 51. SSSA and 

ASA, Madison, WI, USA. pp 1-46. 

Sanchez, P.A et a1. 1997. Soil fertility replenishment 

in Africa: an investment in natural resource capital. 

In: Buresh, R.J., Sanchez, P.A. and Calhoun, 

F. (Eds). Replenishing Soil Fertility in Africa. SSSA, 

Wisconsin, USA 

Sanginga, N., Vanlauwe, B. and Danso, S.K.A. 1995. 

Management of biological N2-fixation in alley 

cropping systems: Estimation and contribution 

to N balance. Plant and Soil 174:119-141. 

Sanginga, N., Zapata, F., Danso, S.K.A. and Bowen, 

G.D. 1990. Effect of successive cuttings on up~ 

take and partitioning of nitrogen 15 among plant 

parts of Leucaena leucocephala. Biology and Fertility 

of Soils 9:37-42. 

Sileshi, G. and Mafongoya, P.L. 2002. Incidence of 

Mesoplatys ochroptera stal (Coleoptera: Chrysomelidae) 

on Sesbania sesban in pure and mixed 

species fallows in eastern Zambia. Agroforestry 

Systems I In press. 

Snapp, S.$., Mafongoya, P.L. and Waddington, S. 

1998. Organic matter technologies for integrated 

nutrient management in smallholder cropping 

systems of southern Africa. Agriculture, Ecosystems 

and Environment 71:185-200. 

154 

Grain legull'K!s and Green Manures for Soil Fertility in Southern Africa

PIGEON PEA/COWPEA INTERCROP + MAIZE + CASSAVA ROTATIONS 

ON SMALLHOLDER FARMS IN THE SOUTHERN 

COASTAL AREA OF MOZAMBIQUE 

Abstract 

CANDIDA CUEMBELO 

INIA, Caixa Postal 3658, A v. Das FPLM, 2698, 

Mavalane, Maputo 8, Mozambique 

Low soil fertility is a limiting factor for agriculture production in Mozambique that is aggravated by unsustainable 

cropping systems used bysmallholding farmers. They practice rainfed low input agriculture on land areas that range 

from 0.5 to 4 hectares. 

This paper gives the background and brief information on results from a very initial set of on going on-farm rotation 

experiments on nutrient management carried out in Inharrime district within Inhambane province in southern Mozambique. 

The experiments are conducted on farmers' fields located in smallholder farming areas surrounding Nhacoongo 

Research Station and are based on legumes in rotation and intercropped with maize and cassava. Crop residue incorporation 

was considered part of the technology. Of the five sites (one farmer=one site), harvesting of cowpea took place in 

three sites only because of environmental stresses and yields were low.. 

Key words: Legume, maize, cassava, intercrop, rotation, lin farm experiment 

Introduction 

Low soil fertility is one of the most limiting factors 

for agricultural production in Mozambique, plus 

climatic irregularity and adversity and soil erosion, 

aggravated by unsustainable cropping systems used 

by small-holding· farmers. In general, the soils of' 

Mozambique are low to moderate fertility. 

Inadequate soil fertility is one of the major biophy,?ical 

constraints for crop production. Soil fertility de 

pletion is continuous due to nutrient losses caused 

by poor soil husbandry and extractive traditional 

methods of cultivation with no replacement of nu 

trients by smallholder farrriers. They practice rain 

fed agriculture in land areas that range from 0.5 to 4 

hectares. On average, smallholder farmer maize 

yields are very low (about 200-400 kg/ha) and inorganic 

fertilizers are unaffordable, aggravated by 

constrained access to credit systems. 

In addition, Mozambique's agriculture production 

has been greatly affected by the civil war that ended 

in 1992 followed by adverse conditions such as 

droughts and floods resulting in a large food deficit. 

Thus, there is a need to increase crop production, 

particularly at the level of small farmers. Sustain 

able soil fertility management practices have to be 

developed, recommended and adopted if the objectives 

of food security and sustainable natural re 

source management are to be achieved. 

Inharrime district is located on soils that are representative 

of soils in the coastal belt where low fertility 

is the major constraint to crop yields. Sandy soilS 

are predominant that are characterized by low nu 

trient reserves, low organic matter content and low 

cation exchange capacity. Nhacoongo research sta 

tion is located within this district and it was the 

ideal place within Inhambane province for the experiment 

on rotations. 

In Mozambique, about 90 percent of food production 

is from Tainfed systems. The major food crops 

are maize, sorghum, millet, rice, cassava, bean and 

groundnut and other additional crops and fruit 

trees mostly intercropped. The major part of crop 

production is subsistence and low input/low out 

put in nature. Commercial farming contributes only 

about four percent of total production. The highest 

human population density of the agricultural re 

gions is in the coastal belt south of the river Save. 

There is no correlation, in the south, between population 

density and envirorunental conditions, particularly. 

climate and soils. However, there is a 

strong correlation between climate and crop distri 

bution. Cassava and maize are the basic staples of 

this area, and they are known to deplete soil nutrients. 

Mainly because of the environmental conditions, 

maize production has declined in some areas 

and cassava has expanded. For many areas, the introduction 

of new technologies for soil fertility im 

provement may reverse the situation. 


The first steps to improve soil fertility management 

based on legumes has taken place through crop rotation 

studies with green manure cover crops including 

improved fallow (pigeon'p~a, lab-lab and 

cowpea) in Nhacoongo Research Station in lnhambane 

Province. The rotation system consists of three 

main components, namely pigeon pealcowpea 

intercrop (hot season), sole maize (cool season) and 

groundnutlcassava intercrop (in the following hot 

season. 

Methodology 

Site Description 

Inharrime district, located in the south of Inham 

bane Province, lies from latitude 24°10'30" and 

24°37'30" South and longitude 34°30'00" - 35°25'00" 

East. It has 76,518 inhabitants relaying on subsis 

tence agriculture. 

According to the Thornthwaite-modified classifica 

tion (Reddy, 1986), the climate is wet semi-arid 

(Table 1). The rainfall is irregular and erratic due to 

occurrence of low-pressure centers. There are two 

growing seasons (Table 1). 

Inharrime is located along the coastal zone. Most of 

the soils are sandy loams except the low plateau, 

then the middle and the high plateau. The dominant 

soils are arenosols (Table I), used for most of the 

crop production. Locally there are fluvisols and 

soils with hydromorphic properties. 

Material and Methods 

The experiments were planted in lnharrime district 

(after having carried out previous studies at the re 

search station of Nhacoongo) on five smallhold 

farmers in the areas surrounding the research sta 

tion. In consultation with local farmers, the criteria 

were based on farmer's availability and int~rest, 

particularly the ones cultivating the legumes. At the 

trial s,ites, soil samples taken for chemical character 

istics and texture determination showed nutrient 

deficiencies (Table 2). 

Table 1. Experimental site details 

location 

Inharrime district 

Altitude 

43 mabove sea 

level 

Average 'annual rainfall 800·1000 mm 

Annual mean temperature 23·26°C 

Potential evapotranspiration 1275mm 

Growth period 

130·139 days 

Main crop season 

September·March 

~ Crop season April to September 

Dominant soils 

Sandy soils 

Research station soils 


Farmers fields soils 


Table 2. Soil analysis results of sandy soils from a representative 

sample of the smallholders farmer'S field areas 

Farm Ca Mg K Na Bas8$ pH in P % % 

H2O Olsen Organic Total 

Matter ' N 

0.40 0.18 0.08 0.00 0.70 6.1 1.49 0.6 0.06 

2 1.11 0.25 0.08 0.06 1.50 5.8 1.08 0.5 0.09 

3 0.79 0.30 0.16 0.04 1.30 5.9 1.76 0.6 0.07 

4 0.69 0.22 0.12 0.04 1.10 6.0 1.35 0.5 0.07 

5 0.10 0.34 0.02 0.02 0.50 ' 5,5 1.22 0.5 0.07 

The experiment was arranged in a randomized 

complete block design with each site being one replicate 

(equivalent to 1 farmer) with three treatments 

involving maize \Z~'.;' mays), cassava (Maninhot esculenta) 

and legumes such as cowpea (Vigna unguiculata), 

pigeonpea (Cajanus cajan) and groundnut 

(Arachis hypogaea). Maize and cassava spacing was 

0.80"0.40 mana 1"1 m respectively, 0.40"0.40m for 

pigeonpea , 0.80"0.80m for cowpea and 0,30"0.25 for 

groundnut. The plot area covered 25 m 2 and the ~otal 

area was 100m 2 , Other inputs i..cluded nitrogen 

from the legumes and the crop residues. 

The experimental treatments were intercrop legumes, 

cassava and maize in rotation as follows: 

Tl. Maize+Cowpea +Groundnut in the wet season 

followed by maize in the dry season (this in 

one year). 

T2. Cowpea + Pigeonpea in the wet season and 

sole maize in the dry season. 

T3; Cassava+Cowpea and Groundnut, lasting in 

the field with cassava until the end of the dry 

season. This represents the normal farmer cropping 

system. 

40 kglha of P 20s was applied in all treatments since 

the sandy soils are highly phosphorus deficient, expecting 

that the Nitrogen input would corne from 

the legumes . . 


Ants destroyed groundnut because of late sowing 

and the dry season maize was not sown because of 

drought (long dry spell in the beginning of the season). 

With little rain, cowpea did produce some but 

very low yields. As it has been postulated by Parsons 

and Howe, 1984 in Giller and Wilson (1991), 

this grain legume has the ability to maintain lower 

osmotic potential in it's leaves under water stress 

conditions. The best performer among legumes was 

pigeonpea that resisted the environmental stress 

(rain, temperatures, low fertility). 

156 


TlIJil 3. Means ilf legume grain Results were not consisyields 

(kg/hal 

tent because of the adr---~--------------~ 

TreatlThlllt Cowpea Pigeon pea verse environmental. 

l' 390 conditions. Very low 

2 495 720 yields of all involved 

3 250 

crops were obtained 

(Table 3). Legumes are 

affected in several ways under such water deficiency 

conditions. Survival or rate of growth of microorganisms 

or other processes such as plant infection 

or nodule development may be affected, as 

may the fixation of N2 (Giller and Wilson 1991). Furthermore, 

their survival is improved by the presence 

of clay particles and organic matter in soils at 

high temperatures, as happens in sandy soils. 

However despite all this, farmers did want to continue 

with the experiment in the following season 

asking for a reformulation, that is, focusing the rotation 

on maize and legumes excluding cassava. 

Soil chemical analyses (Table 2) showed severe nutrient 

deficiencies, including nitrogen deficiency. 

This showed a need to include a basal fertilization 

with this element in the following seasons and not 

only phosphorus, aiming to guarantee the initial 

plant growth. Measures on soil nutrient status and 

organic matter shall be done during the experiment 

implementation, to verify effects of the new technologies 

on management of sandy soil. 

References 

Westering, R.M. 1997. Evaluation of length of growing 

period and crop growing possibilities in Mozambique. 

Nota tecnica no 76-INIA/DTA. 

Reddy, S.J. 1986. Agro-climate of Mozambique as 

relevant to dry land agriculture. Comunica~ao no 

47 INIA/DTA. 

Carta Nacional de Solos (escala 1: 1.000.000), 1995. 

INIA-DTA-Comunica~ao N° 73. 

INIA- DTA 1996. Resultados da avalia~ao 

generalizada para as para as culturas de milho, 

mapira e mexoeira. 

Pililao, F. 1974-1987. Evolu~ao da toponomia e da 

divisao territorial. 

Giller; K.E. and K.J. Wilson, 1991. Nitrogen Fixation 

in Tropical Cropping Systems, CABI International, 


Grlin Legumes and Green Manures for SOil Fertility in Southern Africa 157

Questions and Answers 

.Identification of Best Bet Legumes for On-farm Performance as Grain Legumes, 

Intercrops, Rotations, and Green Manures 

. 

To Webster Sakala and Wezi Mhango 

Q: In the intercrop and rotation practices on maize 

cultivation, what levels of additional fertilizers were 

added apart from the green manures? 

A: On average, it was half the requirement of the 

maize crop, which was around 40 kg N ha- 1 , but it is 

difficult to give a specific figure because we looked 

at several studies. 

Q: Results of green manures and grain legumes 

appear to be promising in Malawi. What is the 

uptake rate of these technologies among 

smallholder farmers? What were the yields of 

soyabean intercropped with maize? I thought that 

soyabean is very sensitive to shading. 

A: The grain legumes/green manures are currently 

being promoted amongst the smallholder farmers. 

The farmers make choices depending on the 

resources that they have, e.g. land, so that if they 

have less land they go for intercropping, and for 

improved fallows or rotation if land is more 

plentiful. Soya bean is usually grown in pure 

stands. 

Q: As early incorporation of green manures may 

cause an N loss due to early showers (leaching), the 

nutrient supply could have been better if the 

legumes were incorporated late, to synchronize the 

peak N demand of maize with nutrient release. 

What is your comment o,n that? 

A: This is not a problem in Malawi because of the 

nature of our rainfall where after incorporation we 

experience a 6-8 month dry season before the next 

growing season. This may be a problem in countries 

where they receive some heavy showers before the 

next crop season. 

To Dennis Friesen, et al. 

Q: It looks like you do not have a soil fertility 

problem in East Africa because you get up to 6 t/ha 

maize grain yield. At one site you get almost 5 t/ha 

without fertilizer and 6 t/ha with fertilizer, 

displaying a low response to fertilizer. 

A: Generally those high maize yields without 

fertilizer were obtained in trials conducted on 

station where the soils are better and where there 

has been a history of fertilizer use. Yields on farm 

were generally closer to the mean yields reported 

for the region (1-2 t/ha) L with some exceptions in 

high potential areas. 

Q: Did you apply any fertilizer when you 

intercropped and how much? 

A: In general, DAP is applied to the maize adjacent 

to the planting hole at the recommended rate, which 

varies in the region (generally around 46 kg P20S / 

ha). 

Q: Intercrops are notoriously difficult to manage, 

and where they work can depend on various site 

effects. Yet some have been very successful. How 

can we move towards making predictions of where 

they will succeed, and recommendations for their 

management? 

A: Factors that affect the growth of the intercrop 

such as moisture, soil fertility, light penetration of 

the maize canopy, can probably be used to predict 

intercrop growth using a modelling approach. 

However, farmers need rules of thumb to predict 

when to 'plant the intercrop, what maize varieties to 

intercrop into, etc. Perhaps modelling can help to 

develop these guidelines. 

To Nhamo Nhamo, et al. 

Q: Did you check the specific names for the cowpea 

varieties that the farmers received from donors? 

A: Yes but we asked for the local names for these 

varieties, so if it was introduced by the donor and 

the farmers did not remember then we could have 

missed them. 

To Pauline Chivenge, et al. 

Q: It would be useful to establish how the various 

legumes perform under amount of rainfall. Did you 

conduct correlation analysis for rainfall? 

A: No, the drought did not allow that. 

Q: The low available P in your soil is not consistent 

with the high biomass yield. Could this observation 

suggest that the plants are obtaining P from 

insoluble P soil sources? 

A: Quite right, but also the northern part of Zambia 

where these yields were obtained received very 

good rainfall. 


159

To Laurence Jasi, et al. 

C: One should not expect to see a significant effect 

of legumes on Striga infestation after one crop. Our 

experience in Western Kenya is that-the effects 

require long term implementation of rotations since 

the purpose of the legumes are to 1) stimulate 

suicidal Striga germination, and 2) improve soil 

fertility and biological activity to reduce the Striga 

seed bank in soil. This requires several seasons of 

rotation. 

Q: Can crop models be used to predict the result of 

this experiment? 

A: There is some capacity in APSIM to look at crop 

x weed interactions and weed management issues. 

For parasitic weeds however, the model is not 

parameterized (due to limited understanding of th'e 

science) to handle parasitic weeds like Striga. 

Q: Your treatments did not reduce Striga emergence 

below that in the control, but your conclusion is not 

that your hypothesis should be rejected, but that 

more work is needed. Why not simply conclude 

that green manures do not (in this case) usefully 

reduce Striga emergence? 

A: It is too early to make a conclusion. Green 

manures can induce the suicidal germination of 

Striga. With time the Striga seed bank is reduced. In 

the long term, probably positive results may be 

obtained. 

To Paramu Mafongoya, et al. 

Q: When is manure supposed to be called manure? 

At times you have 10 t of material, of which 2 t is 

organic manure and 8 t of sand! 

A: Analyze the manure for sand and other materials 

and then you can correct for sand. 

Q: You have shown that incorporation of 

leguminous tree biomass (e.g. Gliricidia) increased 

pH significantly. There is some work from 

Australia indicating that growing legumes acidifies 

the soil significantly. Do you feel the cation/base 

concentration in the tree biomass justifies the 

increase in pH? 

A: This is explained by leaching of N03 and 

accompanying Mg2+ during the cropping phase 

when there is no tree to recover N. In coppicing 

fallows this explained the addition of cations in tree 

biomass. 

Q: How successful are agroforestry technologie's 

such as improved fallows in improving soil fertility 

in degraded soils like those in Kagoro in eastern 

Zambia where ICRAF is located? 

A: In Kagoro soils, mixtures of Glricidia and Sesbania 

gave maize yields of 3 t/ha compared to 4 t/ha for 

fully fertilized maize. 

Q: 

(1) It is good that now we are beginning to put 

science to the observations of the benefits of 

agroforestry trees that have been demonstrated 

over the years. 

(2) How could soil loss from runoff have been 

measured in October as indicated by the data? 

A: The data were collected by rainfall simulation 

techniques. 

160 


MUCUNA - MAIZE ROTATIONS AND SHORT FALLOWS TO 

REHABILITATE SMALLHOLDER FARMS IN MALAWI 

WEBSTER D. SAKALA, IVY LlGOWE and D. KAYIRA 

Chitedze Agricultural Research Station, P. O. Box 158, Lilongwe, Malawi 

Abstract 

An experiment was initiated in the 1999/2000 season to evaluate four different ways of improving maize yields on degraded 

and abandoned parts of smallholder farms in Lilongwe, Kasungu and Mzuzu Agricultural Development Divisions 

(ADD) in Malawi. Selected sites were farm fields abandoned by farmers due to very low maize yields. In the first 

season (2000/2001), two treatments were planted with maize, which was either fertilized with an area specific fertilizer 

recommendation or not fertilized. The other two treatments were planted either to a one-year improved fallow of mucuna 

or left to a natural fallow. In the second season (2001/2002) maize was planted to all the four plots without fertilizer except 

a control where fertilized maize followed fertilized maize. Average maize yields from the four sites ranged from 0.8 t 

(at Vibangalala) to 2.0 t ha·1 (at Zombwe). Maize grain and stover yields (3.6 t ha- I and 6.8 t ha- I ) were highest and differed 

significantly where maize was fertilized with the hybrid maize. area specific fertilizer recommendation compared 

with other treatments. For the non-fertilized plots, maize following mucuna had the highest grain and stover yields of 

1.5 and 3.5 t ha- I respectively. Total N yield for both grain and stover followed the trend of maize yield. Nitrogen concentration 

in the grain was not significantly different between treatments. These results indicate that resource poor 

farmers with abandoned fields who cannot afford fertilizers would benefit by using green manure improved fallows compared 

to continuous cropping with maize or leaving the field to a natural fallow. 

Key words: Mucuna, maize rotation, short fallow, area specific fertilizer recommendation, smallholder farm, rehabilitation 

Introduction 

Results from initial assessments of the mucuna-maize 

rotation system conducted on-farm and on station in 

Malawi from 1997/98 to 1999/2000 showed that 

maize yields following unfertilized MUClma 

(Kalongonda) were significantly higher than maize 

yields after continuous unfertilized maize (up to 3.5 t 

ha- 1 vs. 1 t ha- 1 ) (Sakala et al. 2000; Gilbert and Kumwenda, 

2001; Sakala et al. 2001; Sakala and Mhango, 

2003). Tlle results showed that Mucuna could be a 

good alternative source of fertilizer for maize production 

in Malawi for farmers who cannot afford fertilizer. 

In the same work, it was clear that the best way to 

manage mucuna in a maize based cropping system is 

through rotation rather than intercropping or relay 

cropping of mucuna with maize. The objectives of the 

new work descr:ibed here were to i) demonstrate to 

more farmers that mucuna can be used for rehabilitating 

smallholder farms, ii) collaborate with Non Governmental 

Organizations (NGO) on scaling up this 

promising technology and iii) measure the yield and 

nutrient benefits of the technology on farms. 

Materials and methods 

The experiment was initiated in the 1999/2000 season 

in Ntheu, Kasungu, and Vibangalala Extension 

Plarming Areas (EPA) in Lilongwe, Kasungu and 

Mzuzu ADDs. The soil characteristics at the sites are 

in Table 1. In the first season (~000/2001), two treatments 

were planted with maize, which was either 

fertilized or not fertilized, and the other two treatments 

were either planted to a one year improved 

fallow of mucuna or left to a natural fallow. In the 

second season (2001/2002), maize without fertilizer 

was planted to all the four plots, except the first 

treatment where fertilized maize followed fertilized 

maize. Crop residues were incorporated at the end 

of the first season. 

The four treatments were arranged in a randomized 

complete block, with farmers as replicates. Each plot 

comprised of 10 rows spaced at 90 cm and 10 m 

long. Maize seed was planted at 37000 plants per ha 

(0.9 m x 0.9 m x 3 plants). The sale crop of maize 

received 35:10:0+2S (N:P20s+S) per hectare from 

23:21:4S as a basal fertilizer, and from urea as atop 

dressing. Maize yield was determined by harvesting 

four middle rows (each 9.1 m long) of each plot, and 

the yield was adjusted to 12.5% moisture content. 

Mucuna was planted at 74407 seeds per hectare (90 

m x 15 m x 1 plant) in 2000/2001. Maize yields were 

analyzed using GENST AT (Payne, 1978). Analysis 

of variance was the main procedure used for testing 

significances of differences between means. 


Table 1. Soil characteristics of some selected parameters across 

four sites at the end of the second season in 2002 

Treatment pH OM (%) NOl (ppm) NH4 (ppm) 

Fertilised maize 5.5 3.0 1.3.8 99.7 

Unfertilised maize 5.5 3.5 14.1 84.2 

Maize after 5.5 3.8 13.3 65.1 

mucuna 

Maize after 5.5 4.1 12.1 66.3 

natural fallow 

SED Prob SED Prob SED Prob SED Proo 

Site 0.13 < 0.001 0.49 < 0.001 2.39 < 0.037 15.4 < 0.001 

Treatment 0.13 NS 0.494 NS 2.39 NS 15.4 NS 

Site xTrt. 0.264 NS 0.988 NS 4.78 NS 30.2 NS 

Results 

Soil chemical characteristics of the treatments 

There were no significant differences at the end of 

the second season on soil chemical characteristics, 

although organic matter tended to be higher following 

the natural fallow and mucuna compared to unfertilized 

and fertilized maize monocultures (Table 

1). 

Effects of mucuna on soil cover compared to natural 

fallow 

No measurements were taken on soil cover in plots 

with the short term fallow of mucuna and the plots left 

to natural fallow. However, farmers in the study sites 

observed that where mucuna was planted there was a 

good soil cover compared with the plots left to natural 

fallows. 

Effect of treatments on maize grain and stover 

yield 

Average maize grain yield was highest where maize 

was fertilized with the area specific fertilizer recommendation. 

For the non-fertilized plots, maize following 

mucuna had the highest grain yield (1.5 t 

ha·1) and highest stover yield (with 3.5 t ha- 1 ) 

(Tables 2a and b). The lowest inaize grain yield (0.4 

t ha- 1 ) was from plots where unfertilized maize followed 

unfertilized maize (Tables 2a and b). Among 

the -twenty-six sites, Vingalala site had the lowest 

average maize yield of 0.8 t ha- 1 but had similar 

maize yield trends to the other sites. 

Figure 1. Grain and stover nitrogen across sites for four maize 

mucuna systems in Malawi 

Effect of treatments on grain and stover N yield 

Nitrogen yield for both grain and stover (Figure 1) 

was highest for the fertilized treatments followed by 

the maize following mucuna treatment, with the 

least grain and stover yield obtained from maize 

that followed the natural fallow. There were significant 

differences in nitrogen concentration in the 

grain due to sites; the lowest nitrogen concentration 

was obtained from Vibangalala and the highest 

from Kasungu. The N.in the maize crop after mucuna 

was almost double that from unfertilized 

maize. 

162 

Effect of sites and treatments on N concentration 

There were significant differences in nitrogen concentration 

in the grain due to sites_ The lowest nitrogen 

concentration was obtained from Vibangala site 

and the highest concentration was from Kasungu 

(Table 3). The N in the maize crop after mucuna was 

almost double that from unfertilized maize. 

Conclusion 

Similar yield trends were obs~rved at all four sites 

and strongly indicate that resource poor farmers 

who cannot afford fertilizers would benefit a lot by 

Table 2a. Maize grain yield (t hal) from 26 farms located at four 

sites in 2001/2002 

Site No. of Fert- Unfert· Maize Maize Mean 

farms ilized ilized after after (t hal) 

Maize Maize Mucuna Fallow 

Ntcheu 6 3.6 1.1 1.6 1.4 1.9 

Kasungu 5 1.7 0.8 1.2 0.9 1.1 

Vangalala 6 1.4 0.4 0.7 0.6 0.8 

Zombwe 5 3.6 0.9 2.4 1.3 2.0 . 

Mean 2.6 0.8 1.5 1.0 1.5 

SED Prob. 

Site 0.248

Table 3. Percent nitrogen (% N) concentration in the grain at 

maize harvest 

No of Fertilised Unfertilised Maize Maize Mean 

farms Maize Maize after Mu· .after 

cuna Fallow 

Ntcheu 6 1.4 1.3 1.6 1.1 1.4 

Kasungu 5 1.7 1.4 1.9 1.5 1.6 

Vangalala 6 1.1 0.95 0.9 0.78 0.9 

Zombwe 5 1.8 1.55 1.5 1.5 1.6 

Mean 1.5 1.3 1.5 1.2 1.4 

SED Prob. 

Site 0.15

RESIDUAL EFFECTS OF FORAGE LEGUMES ON SUBSEQUENT MAIZE 

YIELDS AND SOIL FERTILITY IN THE SMALLHOLDER 

FARMING SECTOR OF ZIMBABWE 

WALTER MUPANGWA 1, HAPPYMORE NEMASASI 2 , R. MUCHADEYI 3 , 

and G.J. MANYAWU 3 

12 Soil Productivity Research Laboratories, P. Bag 3757, Marondera 

(' Present address, International Crops Research Institute for the Semi-Arid Tropics, 

POBox 776, Bulawayo) 

3 Grasslands Research Station, P. Bag 3701, Marondera, Zimbabwe 

Abstract 

The use offorage legumes as an alternative to mineral N sources has shown potential in many integrated livestock/crop 

production systems. The objective of our research was to demonstrate the effect of forage legume residues (litter and 

roots) on maize yield and, soil N, P and organic carbon levels. On-farm trials were established in Wedza (Natural Region 

II and III) and Buhera (Natural Region IV) districts of Zimbabwe. The treatments were: maize only, maize/cowpea 

intercrop, maizelvelvetbean intercrop, maizellablab intercrop, sole lablab, sole cowpea, sole velvetbean and ley in the 

1998/99 and 1999/2000 seasons. In the 2000/2001 season, all plots were planted to a maize crop. 

In Natural Region II the order in which maize grain yield increased in response to the treatments was maize/cowpea > 

maizellablab > velvetbean > maizelvelvetbean > maize> cowpea> ley. In Natural Region III the order of highest subsequent 

maize Ylelds was: cowpea> maizellablab > ley > maize/cowpea z lablab > maizelvelvetbean > maize z lablab. In 

Natural Region IV, maize/cowpea intercropping recorded the highest grain yield (1.75 t/ha). Residual soil mineral N 

content was lowest under continuous maize cropping in all three Natural Regions. In the drier regions, soil N content 

in the ley treatment was comparable to legume treatments, but it was low in NR II. In NR II, maize/cowpea intercropping 

and velvetbean sole cropping had the highest available soil P concentration. The same trend was observed in NR III 

and IV, but the residual P levels were lower in drier regions compared with that of NR II. Rotation of maize with sole 

velvetbean and maize/cowpea intercrop gives the highest maize yield gains. Residual N gains were higher in sole 

cropped legumes than intercrops in high rainfall areas whereas in drier areas ley gave the highest resi~ual N benefit. 

Key words: forage legumes, intercropping, maize yield, mineral nitrogen, organic carbon, residual effect, soil P. 

Introduction 

The farming system in the smallholder-farming 

sector of Zimbabwe is characterized by a close integration 

and complementarity of crop and livestock 

production. Continuous cultivation with no and 

sometimes minimum organic and inorganir: fertilizer 

inputs has led to soil impoverishment which 

has been named as one. of t!:te major causes of declining 

crop yields. Efforts to improve the quality 

and quantity of grazing in the communal rangeland 

through methods such as veld reinforcement and 

good grazing range management practices have 

been defeated by the 'tragedy of · the commons' 

(Scoones, 1994). The widespread use of inorganic 

fertilizers has been stopped by the high costs 

of the fertilizers while the manure is largely of poor 

quality and is often not available in sufficient quantities. 

Integration of legumes into predominantly cereal 

cropping systems is one way of improving soil nitrogen 

(N) status of the soil through biological nitrogen 

fixation. The forage legumes will in turn provide 

additional high quality feed for livestock. In 

addition, through intensive cultivation, arabl~ lands 

can assist in providing supplementary forage and 

other products for livestock feed . This will ease the 

pressure on already degraded less productive 

rangelands. 

Forage legumes are known to improve soil physical 

and chemical properties. Data from Chalk (1996) 

have shown that cereals intercropped with grain 

legumes benefit in terms of increased grain and N 

yields. Literature reports that N is a key factor in the 

response qf cereals following legumes compared 

with cereals following non-legumes. The legume 

may potentially add N to the soil . N pool through 

symbiotic N2 fixation, and it may also remove less 

inorganic N from the soil compared with the cereal. 


~e decomposition of legume residues during the 

post harvest fallow period preceding the sowir,g of 

a cereal may explain differences in !he relative contribution 

of fixed-N to the N economies of intercropped 

and rotation systems (Peoples and Herridge, 

1990). Thus cereals cropped in sequence with 

legumes derive N benefits compared with cereal 

monocul ture. 

Hybrid maize (Zea mays L.) is a crop that requires 

and extracts high amounts of nutrients, which produces 

optimal and/or economic yields in highly fertilized 

soils' or soils of high inherent fertility status, 

provided rainfall is not limiting. Most of Africa is 

struggling with structural adjustment programs that 

have left resource poor farmers in serious economic 

problems. The prices of most agricultural inputs 

have been escalating while the financial resources of 

peasant farmers are dwindling. Resource poor farmers 

can hardly afford to buy mineral fertilizers. The 

effect of intercropping legumes with tropical cereals 

has been reported. Gryseels and Anderson (1983), 

Dzowela (1987), Natarajan and Shumba (1989) and 

Manyawu (1994) have reported the effects of the 

legume component on the cereal crop (maize) in 

intercropping. In Zimbabwe, a nitrogen equivalent 

of 40 to 7S kg N/ha has been realized in legume/ 

maize crop rotations (Mukurumbira, 1985). Biological 

nitrogen fixation (BNF) is an enormous potential 

for the maintenance and improvement of soil fertility 

in the tropics. 

The main objective of this study was to evaluate th~ 

effect of intercropping and rotating forage legumes 

with maize on maize yield and soil fertility. Specific 

objectives were to determine the residual effect of 

forage legumes on (a) maize grain yield (b) aboveground 

biomass and litter yields of the tegumes (c) 

soil mineral N levels (d) soil organic carbon content, 

and (e) soil P content. 


Experimental sites 

On farm trials were established in Natural Regions 

(NR) II, 1II and IV of Wedza (Mashonaland East 

province) and Buhera (Manicaland province) dis 

tricts. Two wards, Chamatendere (NR II) and Ma 

dzimbabwe (NR III), were selected in Wedza dis 

trict. Another ward (Gaza Munyanyi) was selected 

from Buhera district. Three farmers were identified 

in each ward through consultatIon with extension 

officers and farmers. 

Trial establishment 

The treatments were as follows: sole maize, maize/ 

cowpea intercropping, maize/velvet bean inter 

crQPping, maize/lablab intercropping, sole lablab, 

sole cowpea, sole velvet bean and ley. Each farmer 

hosted all the eight treatments. The trials were laid 

out so that each farmer formed the sampling unit. 

There was no blocking at each farmer's field. Each 

farmer in a ward formed the replicate. In the 

1998/99 and 1999/2000 seasons, study sites were 

planted to sole crops and intercrops listed above. In 

the intercrops, the maize and legume were planted 

in alternate rows. Planting of the maize and the legumes 

was done at same time in November 1998. In 

2000/2001, all plots were planted to a maize crop. 

, 

Plots measuring 10m x 10m were used at all sites. 

Maize in monocrops was planted at the recommended 

0.9m x 0.4Sm while legumes in monocrops 

were planted at O.4Sm x O.lSm. Compound D (8% 

N:14% P20S: 7% K20) and calcitic lime (96% neutralizing 

value, 4.5% Mg) were broadcast at 2S0 and 

SOO kg ha- 1 respectively before planting. Maize variety 

SCS01 (medium season variety) was planted in 

Wedza and SC401 (short season variety) was 

planted in Buhera. All the legumes were inoculated 

with the appropriate Rhizobia strains at planting. In 

the 1998/99 and 1999/2000 seasons, the maize was 

topdressed at knee height and tasseling with 60 kg 

N ha- 1 each time. At harvest the legume aboveground 

biomass was taken for livestock feeding, 

leaving litter and belowground biomass contributing 

towards soil fertility. Harvesting was "done in 

Apri12001. 

Measurements 

Soil samples were collected from a depth of up to 30 

cm. Organic carbon was determined by the Walkley-Black 

procedure while P was extracted by the 

bicarbonate method (Wanatabe and Olsen, 1965). 

Mineral N (N0 3--N + N~+-N) was extracted from 

soil by the 1M KCl/0.1M HCl solution and determined 

by the calorimetric procedure. Aboveground 

biomass and litterfall were measured for each legume. 

Data collected were subjected to analysis of 

variance using statistical analysis system (SAS) program 

[SAS, 1990] to evaluate treatment effects. 


The results presented for.2000/01 season are used to 

show the rotation effect of forage legumes to subsequent 

maize grain yield and soil fertility parameters. 

In Natural Region II, Wedza (Chematendere ward), 

maize/cowpea intercropping had a significantly (P 

maize/lablab > velvet bean > maize/ 

166 


Table 1. Effect of forage legumes on subsequent maize grain yields 

(t ha- 1 ) at Wedza and Buhera sites (2000/2001 season) 

Treatment Natural Regionll Natural Region III Natural Region IV 

Maize 2.27 1.50 0.67 

Maize\cowpea 4.64 1.65 1.75 

Maize\lablab 4.42 2.40 0.88 

-Maizelvelvetbean 3.93 1.89 0.83 

Cowpea 2.93 2.60 0.81 

Velvetbean 4.30 1.49 1.03 

Lablab 3.09 1.94 0.80 

Ley 2.39 2.18 0.98 

LSD (P < 0.05) 

Ward xtreatment interaction was significant at P< 0.05 

Isd 0_05 - 2.02 

velvet bean > lablab > cowpea > ley > maize. In 

Natural Region III, Wedza (Madzimbabwe ward), 

neither intercropped nor sole Gopped legumes had 

any significant effect on maize yields in the 

2000/2001 season. In Natural Region IV, Buhera district, 

maize / cowpea intercropping recorded the 

highest grain yield (1.75 t/ha), which was about 

four-fold higher compared with that obtained in 

other treatment combinations (Table 1). The increased 

maize grain yields following forage legumes 

could be due to the N-sparing effects of the 

legumes planted in the previous season. The residual 

effect of the legumes on maize stover yields was 

not significant during the 2000/2001 season. 

All the three legumes have a potential of producing 

high herbage yields noting that they produced more 

than i 500 kg ha- 1 across all regions and when they 

are intercropped (Figure 1). The three legumes chosen 

for the study (cowpea, lab lab and velvet bean) 

have a wide range of attributes and adaptation 

(Skerman et al., 1988). They are widely used for 

intercropping with maize (Almseged et al., 1991). 

Being short-lived perennials, they are easy to manage 

in any of the cropping systems. 

.. 000 

3500 

3000 

2500 

ns 

Table 2. Effect of forage legumes on soil mineral Nat Wedza and 

Buhera sites 

Treatments 

Mineral nitrogen (ppm) 

Natural Region II Natural Region III Natural Region IV 

Maize 2.45 9.67 5.56 

Maize/cowpea 6~85 10.1 4.91 

Maize/velvetbean 4.84 12.2 10.1 

Maize/lablab 3.96 6.84 5.13 

Cowpea 8.05 8.44 2.33 

Velvet bean 6.59 8.58 6.20 

Lablab 6.64 8.48 2.33 

Ley 5.29 8.81 11.6 

Mean 5.58 9.14 6.02 

LSD (P < 0.05) n.s n.s 

n.s. - no significant difference among treatments 

The increases in mineral N were probably due to 

adjusted carbon/nitrogen ratios in legume-cereal 

intercropping systems. From Table 2, it is evident 

that residual soil N content was lowest under continuous 

maize cropping in all three Natural Regions. 

Therefore, incorporation of legumes in cereal 

cropping systems is important. In the drier Regions 

(III and IV), soil N content in the ley treatment was 

comparable to legume treatments, but it was low in 

NR II. This is probably due to more leaching associated 

with high rainfall. 

Forage legumes had no significant effect on soil organic 

C (Table 3). A lack of significant changes in 

percent organic carbon would be expected given the 

short duration of this study. OrganiC carbon is reported 

to take over 10 years to increase by just 2.7% 

(Piha, 1995). Treatment effects also showed no differences 

across the regions. 

Results in Table 4 show that the cropping system of 

a forage legume, such as sole cropping and legumecereal 

intercropping, significantly affected available 

soil phosphorus concentration only in NR II, 

(Chematendere ward, Wedza) where maize/ 

cowpea intercropping and velvetbean 

sole cropping left higher 

available P concentrations than 

the other treatments. Lablab sole 

cropping and maize/velvet bean 

intercropping generally resulted 

in significantly lower available 

soil P concentration. 

1500 

1000 

soo 

c 0 ..... p. II lablllh Mz\ Cowp Mz\lablab M zlvelvel L.y 

legum. '),p. 

Figure 1. Forage and litter production of dual purpose legumes when intercropped with maize in 

Wedza and Buhera, Zimbabwe 

Conclusion 

Forage legumes had a positive 

effect on maize yields and the 

soil fertility parameters measured. 

The residual effects of sole 

velvetbean and maize/cowpea 

intercrop gave the highest maize 

Grain Legumes and Green Manure~ for Soil Fertility in Southern Africa 167

Table 3. Effect of forage legumes on soil organic carbon content (%) 

at Wedza and Buhera 

Treatments Soil organic carbon content (%) 

Natural Region II Natural Region IH 

Natural Region IV 

Maize 0.47 0.38 0.24 

Maize/Cowpea 0.50 0.36 0.38 

Maize/velvet bean 0.52 0.28 0.40 

Maize/Lablab 0.41 0.37 0.25 

Cowpea 0.43 0.28 0.67 

Velvet bean 0.39 0.30 0.29 

Lablab 0.56 0.31 0.42 

Ley 0.48 0.51 0.28 

Mean 0.47 0.35 0.35 

LSD (P < 0.05) n.s n.s ns 

n.s. - no significant differences among treatments. 

yield benefits. Monocropping of maize promotes 

unsustainable crop yields as revealed by lower 

maize yields. Ley gave significant residual N benefits 

especially in lower rainfall zones. Sole cropping 

and intercropping had similar effects on soil organic 

C build up. Residual soil P differed significantly between 

treatments in the higher rainfall region and 

seemed to depend on the demand of the crop combinations 

previously grown. 


The authors are very grateful to the Biotechnology 

Trust of Zimbabwe (BTZ) for funding the research 

work upon which this paper is based on. There is 

also need to mention the co-operation of SPRL technical 

and field staff for the fieldwork a!1d soil analysis. 

We also acknowledge the efforts of extension 

officers and farmers in the participating wards. 

References 

Alemseged, Y.B., King, G.W., Coppock, V.L. and 

Tothill, J.e. 1991. Maize-legume intercropping 

in a Semi-Arid area of Sidamo Region, Ethiopia: 

Maize Response. pp. 77-84. 

Chalk, P. M. 1996. Nitrogen transfer from legumes 

to cereals in intercropping. In: Ito, 0:, Johansen, 

e., Adu-Gyamfi, J. J., Katayama, K., Kumar Rao, 

J. V. D. K. and Tego, T. J. (eds.). Dynamics ofRoots 

and Nitrogen in Cropping Systems of the Semi-arid 

Tropics. ICRISATIJIRCAS, Hyderabad, India. pp. 

351-374. 

Dzowela, B.H. 1987. Maize stover improvement 

with legume forages. In: Kategile, J.A., Said, A. 

N. and Dzowela B.H. (eds.), Animal Feed Resources 

for Small-scale Livestock Producers. Proceedings 

of the second P ANESA workshop held 

at ILARD, Kabete, Nairobi, Kenya. 11-15 November 

1985. IDRC Manuscript Report. pp. 174 

181. 

Table 4. Effect of forage legumes on available soil P205 content at 

Wedza and Buhera sites 

Treatments 

Available soil Pcontent (ppm) 

Natural Region II Natural Region III Natural Region IV 

Maize 7.13 9.14 8.30 

Maize/cowpea 23.9 19.6 10.9 

Maize/velvet bean 6.21 12.2 9.38 

Maize/lablab 7.20 13.3 9.38 

Cowpea 12.8 9.14 8.29 

Velvetbean 20.3 21.2 11.1 

Lablab 8.51 13.6 8.94 

Ley 9.16 , 16.0 14.0 

Mean 11.9 14.3 10.0 

LSD (P < 0.05) ns ns 

n.s. - no sigmficant difference among treatments. 

Gryseels, G. and Anderson, F.M. 1983. Research on 

farm and livestock productivity in the central 

Ethiopia highlands: Initial results, ILCA Research 

Report No.4. 

Manyawu, G.J. 1994. Agronomic evaluation of Lablab 

purpureus accessions intercropped with maize 

(Zea mays) for grain and forage production. Department 

of Research and Specialist Services, Division 

of Livestock and Pasture, Zimbabwe. Annual 

Report. 

Mukurumbira, L. M. 1985. Effects of rate of fertilizer 

nitrogen and previous grain legume on maize 

yields. Zimbabwe Agricultural Journal 82: 177-179. 

Natarajan, M and Shumba, E.M ., 1989. Intercropping 

Research in Zimbabwe: Current status and 

outlook for the future. In: Proceedings of a Workshop 

on Research Methods for Cereal- Legume intercropping 

in Eastern and Southern Africa, Lilongwe, 

23-27 January 1989. CIMMYT and Southern Africa 

on-farm Research Network. Report No. 17: 

190-193. 

Peoples, M. B. and Herridge, D. F. 1990. Nitrogen 

fixation by legumes in tropical and subtropical 

agriculture. Advances in Agronomy 44: 155-223. 

Piha, M. 1995. Soil Fertility Handbook. Department of 

Soil and Agricultural Engineering, University of 

Zimbabwe. 93 pp. 

Scoones, I. 1994. Living with uncertainty. New directions 

in Pastoral development in Africa. pp. 

116-121. 

Skerman, P.I.; Cameron, D.G. and Riveros, F. 1988. 

Tropical Forage Legumes, F.A.o. Plant Production 

and Protection Series, No.2, F.A.O. of 

United Nations, Rome, Italy. 

Wanatabe, F. S. and Olsen, S. R. 1965. Soil Science 

Society ofAmerica Proc. 29: 677-678. 

168 

Grain legumes and Green Manures for Soil Fertility in Southern Africe

TIME OF INCORPORATION OF DIFFERENT LEGUMES AFFECTS SOIL 

MOISTURE AND YIELDS OF THE FOLLOWING CROP 

IN MAIZE BASED SYSTEMS OF ZIMBABWE 

BONAVENTURE KAYII\JAMURA, HERBERT K. MURWIRA and PAULINE P. CHIVENGE 

Abstract 

TSBF-CIA T, University of Zimbabwe, PO Box MP 228, 

Mount Pleasant, Harare, Zim.babwe 

This study reports on an evaluation of the performance of different legumes and their time of inc0T]2oration into soil on 

maize yields in Murewa and Shurugwi communal areas of Zimbabwe. Five legumes, Crotalaria grahamian a, Crotalaria 

juncea (sunnhemp), Mucuna pruriens, Vigna unguiculata (Cowpea IT18) and Glycine max (Magoye) were 

planted in the 2000/01 season followed by maize in the 200l/02 season. The plots were subdivided into two, with legume 

incorporation at flowering in one sub-plot while legumes in the other plot were incorporated at the onset of the following 

season. Mucuna gave the highest biomass yields (4800 kg ha- l ) in Murewa while Crotalaria grahamiana had the highest 

yields (7500 kg ha- l ) in Shurugwi. Higher maize yields were obtained following incorporation of Crotalaria grahamiana 

(2900 kg ha- l ) than Mucuna (2300 kg ha- l ) in Murewa. However Mucuna pruriens had produced higher biomass 

in the previous season. Similar results were obtained in Shurugwi where Crotalaria grahamiana gave higher 

maize yields (1800 kg ha- l ) than Mucuna pruriens (1400 kg ha- l ) . Generally, the early-incorporated legume plots gave 

higher maize yields in the second season than the late incorporated crop, although they were not statistically different. 

At the onset of the second season, soil was sampled from the different plots to analyze for moisture content. Mucuna 

pruriens was shown to conserve higher amounts of mJisture than the other legumes, while late incorporated legumes 

had higher soil moisture content than early-incorporated legumes. It was concluded that Mucuna pruriens and Crotalaria 

grahamiana are potential best-bet legumes in the communal areas of Zimbabwe and in cases where labour is a 

constraint, farmers could incorporate their legumes late. 

Key words: Crotalaria grahamiana, Crotalaria juncea, Mucuna pruriens, Vigna unguic.ulata, Glycine max, incorporation 

time 

Introduction 

Declining soil fertility has undermined crop production 

in Zimbabwe smallholder farming systems. 

With the scarcity and ever-increasing prices of inorganic 

fertilizers, there has been a need to depend 

more on natural processes such as biological nitrogen 

fixation (BNF) in crop production systems. Soil 

improving herbaceous legumes have potential to 

improve soil fertility in various parts of the world 

(Fujita et al., 1992), and can be used as green manure 

and cover crops in areas of different agroecological 

cha~acteristics. 

Green manures have the potential to accumulate up 

to 250 kg N ha- 1 per year (Giller and 'Nilson, 1991), 

and result in subsequent cereal yield increases of 

600 to 4100 kg ha- J (Peoples and Herridge, 1990). 

Evaluation of plants for soil fertility improvement 

remains a priority in this scenario to get the best 

plant species that are suitable for a particular area. 

Some plants have already been identified as best 

bets for green manuring in different situations 

(Buresh et al., 1993). 

Proper management of plant residues for nutrient 

supply requires quantitative knowledge on its nutrient 

release characteristics. The use efficiency of the 

nutrients released by' green manure remains a critical 

point in soil fertility management. Soil water dynamics 

and nutrient management are the main factors 

to consider to achieve sustainable integrated 

cropping systems in a semi-arid environment 

(Biederbeck and Bouman, 1994). 

Incorporated organic materials have several functions 

in the soil other than supplying nutrients. 

They improve soil aggregation (Elliot and Papendick, 

1986), reduce erosion (Young, 1989) and conserve 

moisture. The organic residues from green 

manure help to stabilize the soil structure, increase 

water-holding capacity of the soil, and increase the 

infiltration of moisture into the soil and percolation 

through the soil. Applying crop residues also leads 

to significant N and water interactions (Bolton, 

1981). 

Improvement in scarce available water usually triggers 

an improvement in the use efficiency of scarce 


available nutrients and vice versa, and this leads to 

improved crop production and less movement of 

nutrients into the environment. The OM reduces 

evaporation losses and hence improves N use efficiency, 

and provides nutrients other than N. 

]Jnder field conditions the fluctuations in sDil water 

content affect the release of N from green manure. 

A quantification of this effect is essential for predicting 

the supply of mineral N at a particular time 

(Brar and Sidhu, 1995). There is a progressive decline 

in mineral N· produc tion within the soil wi th 

decrease in soil water level (Brar and Sidhu, 1995). 

In a crop rotation (intercropping or relay) that includes 

growing green manure plants, ~he cereal 

crop and the green manure plants are likely to com 

'pete for nutrients and moisture during the alternate 

season (McGuire et aL, 1998). Green manure crops 

planted during the fallow period may use the moisture 

needed for seed germination at planting, however 

this disadvantage is counter-balanced by other 

benefits of growing green manure during the fallow 

period. 

This study sought to evaluate the performance of est biomass followed by C. juncea (Table 2). The 

Crotalaria grahamiana, Crotalaria juncea, Mucuna prubiomasriens, Vigna unguiculata (Cowpea IT18) and Glycine affected by dry spells that came after crop establish 

production of Mucuna pruriens was more 

max (Magoye) legumes, and the effects of time of ment, while C. grahamiana produced higher biomass 

incorporation of residues on maize yields in in similar conditions. 

Murewa (high rainfall area) and Shurugwi (low 

rainfall). 


The trial was conducted in two consecutive seasons 

(2000/01 and 2001/02). Green manure and grain 

legumes were grown in the first season followed by 

the maize in the second season. The experiment 

comprised of six treatments; three green manure 

crops (Crotalaria grahamiana, Crotalaria juncea and 

Mucuna pruriens), and two gra)n legumes (Vigna unguiculata 

(Cowpea) and Glycine max (Soya bean) and 

a control treatment with maize. The plots were split 

into two subplots for the analysis of the influence of 

time of incorporation of the residues at flowering 

and at the onset of the following season. For the 

maize control the plot was divided into two, one 

was bare (nothing grown in the subplot) and the 

other one had maize crops. Soil samples were taken 

when the plant materiai had been ploughed in the 

soil for the early incorporation subplots, while 

plants were still standing in the other subplots. 

Plant materials for biomass production measurements 

were taken before residue incorporation in 

each subplot. Soil samples for moisture content 

analysis were taken in each plot from 0-10, 10-20, 

20-30 and 30-40 cm depths. They were dried and 

the moisture content determined. The resul ts were 

statistically analyzed using SAS software. 


Biomass production 

Green and grain legumes were grown in the 

2000/01 season for biomass production, and the 

crop residues were incorporated into the soil at dif 

ferent times for a subsequent maize crop in 2001/02. 

Biomass production was higher in Murewa (high 

rainfall, @ 900 mm) than in Shurugwi (low rainfall, 

@450 mm)) in the 2001/02 season. 

In Murewa, Mucuna pruriens produced the highest 

biomass followed by Crotalaria grahamiana, with 

cowpea producing the least biomass (Table 1). The 

N content of the residues was also determined. 

Adding all the' Mucuna pruriens residues harvested 

was equivalent to the addition of 156 kg N per ha. 

In Shurugwi, C. grahamiana outyielded Mucuna pr?lriens, 

with Crotalaria grahamiana producing the high 

Soil moisture content and maize yields 

In the second season (2001/2002) of the study, mois 

ture content was determined just before planting of 

maize, and maize yields were determined at har 

vest. An analysis of variance of moisture content 

measurements showed that soil depth had a signifi 

cant effect on moisture content. The interaction be- 

Table 1 .. Legume biomass yields production for 2000/01 season 

in Murewa 

Treatment Biomass yield {kg/hal Total N{kg/hal 

Cowpea 1442 4B.0 

C. grahamiana 4535 137.0 

Mucuna pruriens 4746 155.6 

C. juncea 4120 llB.5 

Soybeans 2300 77.7 

LSD 1242.3 34.31 

Table 2. Legume biomass yields production for 2000/01 season 

in Shurugwi 

Treatment Biomass yield {kg/hal Total N(kg/hal 

Cowpea lOBO 39.4 

C. grahamiana 7507 24B.3 

Mucuna pruriens 4932 121.4 

C. juncea 5129 144.1 

Maize 

2120 (grain) 

LSD 1290.2 62.4 

170 


tween factors tested in the study did not show any 

significant effect for moisture content (Table 3). The 

mean separations by LSDo.os (least significant differ~ 

ence) of depth (LSDoos = 0.4975) showed that there 

was more moisture at 30-40 cm soil depth than 

other depths (Figure 1). 

The effect of time of incorporation on moisture content 

approached significance (0.0677). The comparison 

of means using LSDo.os (0.35) shows as well that 

there is no significant difference in moisture content 

between late (4.20%) and early (3.87%) incorporation 

of green manure residues. Statistically, the difference 

between times of incorporation of crop residues 

was significant at 10%. The timp. of soil sampling 

might have influenced the difference between 

treatments on moisture conservation. Samples were 

taken when the soil was too dry because of early 

cessation of rain and late onset of rains for the following 

season. However, the numerical difference 

shows that late incorporation of green manure conserved 

more moisture. This might be because in 

early incorporation the soil is exposed to the sun, 

and this increases evaporation, while in late incorporation 

the plants continue to cover the soil, -Nhich 

reduces evaporation. 

Table 3. ANOVA table of moisture content measurements 

Source OF Type III SS Mean Square FVal Pr > F 

Treatment 5 2.01 0.40 0.36 0.8772 

Depth 3 60.34 0.11 17.78 < 0.0001 

Time 3.86 3.86 3.42 0.0677 

Treatment" Depth 15 16.00 1.07 0.94 0.5204 

Treatment "Time 5 3.16 0.63 0.56 0.7309 

Depth"Time 3 3.04 1.01 0.90 0.4460 

Treatment 15 8.42 0.56 0.50 0.9372 

"Depth"Time 

Residual 94 106.22 1.13 

Total 143 210.74 

There was a significant difference wi.thin treatments 

between late and early incorporation of crop residues 

(Figure 2) . . Late incorporation of cowpea, C. 

juncea, C. grahamiana and maize conserved more 

moisture than early incorporation, while incorporation 

time of mucuna and soyabean residues did ~ot 

show any significant ~ffect on moisture conservation. 

A trend of means (not statistically significal'1t) of the 

different treatments shows that soya bean had the 

least moisture content, followed by C. grahamiana, C. 

juncea, cowpea, maize and the highest to conserve 

moisture was Mucuna pruriens (Figure 3). This 

might be due to how these plants cover the ground. 

Mucuna pruriens provides a good cover because of 

its bushy habit that in return reduces evaporation 

from the ground. 

Addition of organic residues improves crop production 

through moisture conservation and nutrient 

supply to crops. Incorporated organic residues augment 

the water retention capacity of the soil by improving 

the structure and physical environment of 

soil. The maximum benefits are achieved by good 

timing of incorporation for growth of the' subsequent 

crop. Consequently, there is need to conserve 

soil moisture to avert moisture deficits at the time of 

sowing, and provide much-needed nutrients at 

early stages of plant growth. 

4 '~------------~========~ 

_ Early(a ll\owa nng) 

_ L

Conclusion and Recommendation 

Mucuna pruriens and Crotalaria grahamiana are potential 

best bets for soil fertility amelioration in communal 

areas of Zimbabwe. High biomass production 

was achieved for the two crops in the areas of 

study. Mucuna pruriens was more susceptible to dry 

spells that occurred in the middle of the growing 

season than C. grahamiana, but both crops need adequate 

moisture at planting for good establishment. 

Higher maize yields were obtained in plots where 

Mucuna and Crotalaria grahamiana residues were incorporated 

compared to other legumes. Time of incorporation 

had an effect on yield of the subsequent 

crop; maize yields in early-incorporated plots were 

higher than in late incorporated plots for all legumes. 

Incorporation of Mucuna residues conserved more 

moisture than other legumes, and moisture content 

was higher (but not significantly so) in late incorporated 

plots. This was probably due to the removal of 

plant cover in early incorporation, and hence high 

evaporation rates before the rains that depleted 

moisture in the soil. . 


The authors greatly acknowledge IFAD for the 

funding of the research, as well as farmers in 

Murewa and Shurugwi for their collaboration in the 

trials. 

References 

Biederbeck, V.O. and O.T. Bouman, 1994. Water use 

by annual green manure legumes in dryland 

cropping systems. Agronomy Journal 86:543-549. 

Bolton, F.E. 1981. Optimising the use of water and 

nitrogen through soil and crop management. 


Brar, D.S. and A.S. Sidhu 1995. Effect of soil water 

on pattern of nitrogen release during decomposition 

of added green mantire residue. Journal of 

Indian Society of Soil Science 43:14-17. 

Buresh, R.J., T.T. Chua, E.G. Castillo, S.P. Liboon 

and D.P. Garrity 1993. Fallow and Sesbania effects 

on soil nitrogen dynamics in lowland ricebased 

cropping systems. Agronomy Journal 

85:316-321. . 

Elliot, L.F and 'R.J. Papendick 1986. Crop residue 

management for improved soil productivity. 

Biology, Agriculture and Horticulture 3:131-142. 

Fugita, K, B. Ofusu and S. Ogata 1992. Biological 

nitrogen fixation in mixed legume-cereal cropping 

systems. Plant and Soil 141:155-175. 

Giller, KE. and KJ. Wilson 1991. Nitrogen fixation 


Wallingford, UK 313 p. 

McGuire, A.M., D.C. Bryant and R.F. Denison 1998. 

Wheat yields, nitrogen uptake, and moisture following 

winter legume cover crop vs. fallow. 

Agronomy Journal 90:404-410. 

Peoples, M.B. and D.F. Herridge 1990. Nitrogen 

fixation by legumes in tropical ~d subtropical 

agriculture. Advances in Agronomy 44:155-223. 

Young, A. 1989. Agroforestry for soil conservation. 

CAB International, Wallingford, UK, 276 p. 

172 

Grain legumes and Green Manures for ~oil Fertility in Southern Africa

SOIL FERTILITY IMPROVEMENT THROUGH THE USE OF GREEN 

MANURE IN CENTRAL ZAMBIA 

MOSES MWALE', CASSIM MASI 2 , J. KABONG0 2 and L. K. PHIRI' 

1Mt. Makulu Central Research Station, PIB 7, Chilanga, genetics@zamnet.zm, 

2World Vision International, P. O. Box 31083, Lusaka, Zambia 

Abstract 

Farmers identify low soil fertility as a major problem affecting crop production in Chibombo, Central Province, Zambia. 

This means fertilizer is a prerequisite to crop production, particularly maize. But the use of fertilizer is not often viable 

due to its high cost and poor availability. To boost crop production, there was need to test alternative cost-effective soil 

fertility improvement techniques. An experiment was therefore conducted to reduce the soil fertility problem using 

green manures. Sunnhemp (Crotalaria juncea) and Velvet beans (Mucuna pruriens) were grown either as sole crops 

or intercropped with maize (in the 1998/99 season). Continuous maize (fertilized and unfertilized) and a natural grass 

fallow were used as controls. Only phosphorus (50 kg PzOs ha·1) was applied to the unfertilized (zero nitrogen) maize 

while compound 0 (10:20:10:8 : N P K S) at 100 kg ha·1 was applied to the other maize treatments. No fertilizers were 

added to the green manure treatments. All maize plots (except unfertilized maize) were top dressed with urea at a rate 

of 23 kg N ha- 1 . The dry matter yield of sunnhemp and velvet bean were determined just before flowering and all the 

above ground biomass was ploughed under. The maize was harvested at physiological maturity and grain weight determined. 

In the 1999/2000 season, maize was planted in all plots and cultural practices were the same as for 1998/99. 

Unfertilized maize had the lowest dry matter and grain yield (less than 1 t ha- 1 ) followed by maize after the grass fallow. 

All fertilized and green manure treatments yielded significantly more grain than unfertilized maize. There were no significant 

differences (p=0.05) between the maize grown after the green manures and that which. received fertilizer. Incorporating 

sunnhemp biomass not only increased the yield of maize but also increased the organic matter content of the 

soil at the experimental sites. 

Key words: Green manure, sunnhemp, velvet bean, profitability, sustainability, Zambia 

Introduction 

Low soil fertility is a major problem affecting crop 

production in Chibombo District of Zambia's Central 

Province. This has been made worse by farmers 

who commonly monocrop with maize year after 

year. This makes fertilizer use a prerequisite to crop 

production, particularly maize. The current prices 

of fertilizers are beyond the reach of most farmers in 

the area due to liberalization of the economy and 

the removal of fertilizer subsidies. The few that can 

afford fertilizers have reduced their application 

rates to far below those recommended, resulting in 

low crop yield!> per unit area cropped . . This has affected 

both food availability and income for many 

people. 

To boo~t crop production, there is a need to test alternative 

cost-effective soil fertility improvement 

techniques. An option identified by the farmers 

through participatory rural appraisal (PRA) was the 

'lse of green manures notably sunnhemp (Crotalaria 

juncea), velvet bean (Mucuna pruriens) and some 

agroforestry tree prunings such as Sesbania sesban. 

These practices were part of the farming systems 

before mineral fertilizers and most elderly farmers 

still recall and appreciate the usefulness of the two 

green manure species. Crotalaria and Mucuna have 

shown to be excellent N2-fixers in a wide range of 

enVironments (Bowen, et al. 1988; Kolar et al. 1993; 

MacColl, 1990; Yost et al. 1985). Green manures 

have the potential to accumulate up to 250 kg N ha- 1 

yr-l (Giller and Wilson, 1991) resulting in cereal 

grain yield increases of 600 - 4100 kg ha- 1 (Peoples 

and Herridge, 1990). The use of a green manure 

may also boost the levels of soil organic matter resulting 

in improved soil structure, better root proliferation 

and soil water holding capacity. This would 

in tum increase crop vigour and yields. There is a 

need therefore to evaluate the beneficial effects of 

these green manures on farmers' fields in Muswishi 

Agricultural Camp in Chibombo District and document 

the results. 

Project Objectives 

The overall objective was to address the soil fertility 

problem using green manures. 


Specific objectives were to: 

• Test the viability of velvet bean and sunnhemp as 

soil fertility improvement legumes within the 

farming system. ., 

• Assess the beneficial effects of incorporating green 

manuring into the cropping system. 


The field experiment was conducted at Muswishi 

Agricultural Camp in Chibombo District of Central 

Zambia. The Mushemi soil series at Chibombo is 

described as a Fine Kaolinitic Isohyperthermic Oxic 

Paleustalf (Soil Survey, 1992). The surface soil was 

sampled to a depth of 20 cm, dried and ground to 

pass through a 2 mm sieve. This was then analyzed 

for soil pH (in 0.01 M CaCh), total nitrogen, organic 

carbon (Wakley and Black, 1934), exchangeable 

cations (1.0 M ammonium acetate, pH 7.0) and 

available phosphorus (Bray and Kurtz, 1945). The 

soil sample was also analyzed for particle size using 

the pipette method. Selected chemical properties of 

the soil are given in Table 1. 

Field Work in the 1998/1999 Season 

The following treatments were applied: Maize (zero 

nitrogen), sunnhemp sole crop, velvet bean sole 

crop, maize/sunnhemp intercrop, maize/velvet 

bean intercrop, maize (fertilized) and a grass 

(natural) fallow. Triple super phosphate was applied 

to unfertilized maize (zero nitrogen) at 50 kg 

P20S ha- 1 • Compound 0 (N P K S : 1020108) was 

applied to the other maize/GM intercrop treatments 

at 100 kg ha- 1 . No fertilizer was added to the 

sole green manure treatments. The natural fallow 

was left intact. Maize hybrid variety MM604, velvet 

bean (Mucuna cv. W. NIRS 16)· and sunnhemp 

(erotala ria Juncea cv. NIRS 4) were planted. 

Maize and velvet bean were planted to give a plant· 

population density of about 44,000 plants per hectare 

while sunnhemp was planted at a seeding rate 

of about 20 kg ha- 1 • Plot sizes were 12 m x 8 m for 

all treatments. All maize plots (except unfertilized 

maize) were top dressed with urea at a rate of 23 kg 

N ha- 1 . 

Dry matter yields of sunnhemp and velvet bean 

were determined just before flowering and all the 

above ground biomass was ploughed under. The 

maize was harvested at physiological maturity from 

the fertilized and unfertilized plots and the grain 

weight determined. 

Field Work in the 1999/2000 Season 

To see the benefits of the green manure, the experiment 

was continued in the 1999/2000 season. 

Maize was planted in all plots including the natural 

grass fallow plot which had been left intact the previous 

season. All cultural practices were the same 

as in 1998/99. All plots (except unfertilized maize) 

were top dressed with urea at a rate of 23 kg N ha- 1 • 

Mid way through the season, soil samples were collected 

from all the plots and were analyse

Table 2. Grain yields of maize and stover yields of velvet beans and sunhemp grass fallow (Table 3). The mean grain 

Name of Farmer Maize Velvet bean Sunnhemp yield of the unfertilized maize was less 

or Farming Group Grain Yield Stover Yield than 1 t ha·l . 'There were no significant differences(p=O.05) 

between the maize grown 

No Fert Fertilized Inter crop Sole crop Intercrop Sole Crop 

after the -green manures and the fertilized 

--------------kg ha-'-------------- 

maize. During the drought sp~ll, it was 

Kanakishiwa Club 933 1533 1931 2133 2883 3183 

noted that the maize grown after the green 

2 Mr.B.B.Muteto 866 1200 1865 2067 2550 2850 

manure species.was less affected than that 

3 Mr Chenje 750 950 1798 2000 2150 2450 

after the grass fallow or even the fertilized 

4 Muswishi Women Group 978 1026 2142 2563 3211 3561 maize. Incorporation of sunnhemp bio 

5 Mr Kamilo 880 1200 1878 2080 2550 2850 mass increased not only the yield of maize 

6 Mr Maputa 867 2000 1865 2067 3350 3650 but also increased the organic matter con 

7 Kalangwa Women Club 920 1200 1918 2120 2550 2850 tent of the soil at the experimental sites 

8 Rural R. Centre No Data' 1866 2556 3527 3926 (Tables 3 and 4). No other effects of green 

9 Shana'ngombe No Data 1864 2864 3262 3269 manures on soil properties were observed, 

perhaps because a single season's inputs 

10 Chipaba Women Club 1027 2733 2025 2227 4083 4383 

were not sufficient to greatly change the 

11 Mr. Katiti John 1333 2467 2331 2533 3817 4117 

measured properties. 

12 Mukuyu Women Club 800 961 1798.0 2015.6 2150.8 2864.6 

13 Mr Manyelekete L. 867 1133 1864.9 2693.5 2458.6 

• ·No Data, the maize was grazed by goats 

growth was reduced. As a result, maize was 

ploughed into the soil together with the green manures. 

Originally, it was expected that the maize 

should have been left standing. To obtain a crop of 

maize from the intercrop, the planting dates of the 

maize and the green manures should be staggered 

with the maize being planted first (Gilbert, 1998). 

Sunnhemp had generally higher stover yields than 

velvet bean, as can be seen from Table 2. 

Results of the 1999/2000 season 

To see the benefits of the green manures, maize was 

planted in all plots at all demonstration sites. Unfortunately, 

there was a drought immediately after 

planting which adversely affected germination at 

most sites, though gap filling was done in most 

fields. Nevertheless, the maize at three sites ( Kanakashiwa 

club, Kamilo's and Shana'ngombe's farms) 

established well. 

The unfertilized maize had the lowest dry matter 

and grain yield followed by the maize after the 

Table 3. Influence of green manures on the dry matter and grain 

yield of maize 

Treatment Dry matter yield Grain yield 

Maize after Grass Fallow 

Unfertilized Maize 

Fertilized Maize 

Maize after Velvet Bean 

-Maize after Sunnhemp 

Maize after MaizeIVelvet Bean Intercrop 

Maize after MaizelSunnhemp Intercrop 

LSD (0.05) 

%CV 

2720bc 

1884c 

4137ab 

4891a 

4523a 

4409a 

3922ab 

kg ha-' 

1271 bc 

877c 

2104ab 

2367a 

2608a 

1969ab 

1902ab 

1545 842 

43.1 47.5 

2786.9 Results of the Cost Benefit Analysis 

Analysis of costs and benefits shows that 

the treatment with sole sunnhemp had the highest 

gross margin and returns to capital (Table 5 and 

Figure 1). In terms of gross margin, sole velvet bean 

was second while fertilizer was third in profitability. 

Intercropped sunnhemp was second to sole sunnhemp 

in profitability using the return to capital criteria. 

This demonstrated the superiority 'of using 

sunnhemp as a fertility enhancing technology to 

substitute for mineral fertilizer. The results also 

show that even though using chemical fertilizer 

raises the gross benefits and margin, the return to 

Table 4. Influence of green manures on soil fertility improvement 

Treatment %C pH P K Ca Mg 

CaClz ppm ----me%--- 

Unfertilized Maize 0.54b 4.6 18.3 0.15 0.82 0.22 

Maize after Grass Fallow 0.59ab 4.6 16.3 0.16 1.02 0.24 

Maize after Velvet Bean 0.59ab 4.7 14.6 0.15 0.86 0.23 

Maize after Sunnhemp 0.66a 4.6 16.2 0.17 0.92 0.22 

LSD (0.05) 0.08 ns ns ns ns ns 

%CV 16.4 4.7 36.8 17.0 35.1 35.2 

Table 5. Summary table of cost·benefit analysis. Zambia K/ha) 

Treatment 

Fallow 571950 

Intercropped velvet bean 886050 

Sole velvet bean 1065150 

Fertilizer 963000 

No Nitrogen Fertilizer 394650 

Intercropped sunnhemp 855900 

Sole sunnhemp 1173600 

I US $ - Zambia K 2500 (199912000) 

Economic indicator 

Gross Benefit Total Costs Gross Return to 

Margin Capital 

224000 347950 1.6 

266000 620050 2.3 

266000 799150 3.0 

224000 739000 3.3 

91000 303650 3.3 

182000 673900 3.7 

182000 991600 5.4 


'"~ 

v 

'" ~ 

~ 

1400000 

1200000 

1000000 

800000 

600000 

400000 

200000 

0 

~ 

.£ 

iii 

IL 

"0 

CI> 

0 

0 

o ill 

~> 

~ 

... 

ill 

> 

CI> 

"0 

en 

Illl 8"055 Benefit 

• Total Costs 

o 8"05s PJargi1 

Figure 1. Cost·benefit analysis of soil fertility improvement technologies Muswishi, 

1998·2000 

the capital invested is comparable to not applying 

fertilizer. 

Problems encountered, lessons, limitation of the 

work completed and farmer assessment of the 

technology 

Whilst farmers appreciated the increase in yield due 

to the green manures, they pointed out it was very 

difficult for them to justify weeding green manure 

crops. This is because other farm operations compete 

for the same labour. It is possible that this 

problem could be overcome by planting th~ green 

manures in the same plots where maize has already 

been planted, at first weeding of the maize crop. 

Some farmers also complained about the labour involved 

in ploughing the green manures under, especially 

where there is no animal draught power. 

Conclusions 

The soils of Muswishi Agricultural Camp were very . 

low in soil organic matter content rendering them 

infertile for most crops without external nutrient 

sources. 

These results show that incorporating the green manures 

into the local farming systems has beneficial 

effects by increasing maize yields. However, the 

green manures alone may not be sufficient to provide 

all the nutrients needed for the maize crop to 

full maturity. The maize grown aIter the green manures 

need supplementing with some inorganic fertilizer 

at the top dressing stage. The advantage is 

that the rate is less than the recommended one. Further, 

green manures are mainly used as a source of 

nitrogen while other elements like phosphorus and 

potassium may have to be added to avoid depleting 

the soil further of these essential nutrients. 

The number of green manure species 

used in the demonstration was 

limited. More species should be 

tried in the area to see if they can 

perform better and provide farmers 

with a wider selection. 

Analysis of costs and benefits 

showed that the treatment with 

sale sunnhemp had the highest 

gross margin and return to capital. 

~ 

"0 I 

CI> It was sup~rior to all others. Sale 

~ 

CI> en 

~ 

0 

IL 

t 

'" 

g-I "0'" 

velvet bean had the second highest 

CI> Z 

IL 

~en en 

gross margin, while fertilizer was 

0 

Z ~ 

.E third in profitability . 

Treatment 

The methods employed in this 

project allowed farmers to participate 

in the project from inception 

to conclusion. By farmers identifying 

the problem of soil fertility themselves and suggesting 

that green manures be used and then seeing 

how they were used to alleviate the problem meant 

that farmers felt they owned the project and results. 

As such, they not only contributed land to the project 

but also labour, which was essential. Researchers 

also made sure that whatever maize was obtained 

from the project was returned to the farmers 

after yield determination. 


We wish to thank the Farm Level Research Applied 

Methods in Eastern and Southern Africa 

(FARJ\l1ESA) and the Tropical Soil Biology and Fertility 

(TSBF) for partially funding this research project. 

Many thanks are also extended to the entire 

Soil Fertility Team at Mt. Makulu and the Field Site 

Working Group at Chibombo District for diligently 

executing the field trials. We would also like to 

thank Mr. Lewis Bangwe for cond ucting the economic 

analysis. 

References 

Bowen, W.T., J.O. Quintana, J. Pereira, D.R. Bouldin, 

W.s. Reid and D.T. Lathwell, 1988. Screening 

legume green manures as nitrogen sources to 

succeeding non-legume crops. Plant and Soil 

111:75-80. 

Bray, R.H, and L.T. Kurtz, 1945. Determination of 

total, organic and available forms of phosphorous 

in soils. Soil Science 59:39-45. 

Gilbert, R.A., 1998. Undersowing green manures 

for soil fertility enhancement in the maize-based 

cropping systems of Malawi. In: Waddington, S. 

R., H.K. Murwira, J.D.T. Kumwenda, D. Hikwa 

176 

Grain le!jumes and Green Manures for ~oil Fertility in Southern Africa

and F. Tagwira (eds.) 1998. Soil Fertility Research 

for Maize-Based Farming Systems in Malawi and 

Zimbabwe. Proceedings of the Soil Fert Net Results 

and Planning Workshop held from 7 to 11 

July 1997 at Africa University, Mutare, Zimbabwe. 

Soil Fert Net and CIMMYf-Zimbabwe, 

Harare, Zimbabwe, pp. 73-80. 

Giller, K.E., and K.J. Wilson, 1995. Nitrugen fixation 



Kolar, J.S., H.5. Grewal and B. Singh, 1993. Nitrogen 

substitution and higher productivity of a ricewheat 

cropping system through green manuring. 

Tropical Agriculture 70:301-304. 

MacColl, D., 1990. Studies on maize at Bunda, Malawi. 

III. Yield in rotations with pasture legumes. 


Peoples, M.B., and D.F. Herridge, 1990. Nitrogen 

fixation by legumes in tropical and subtropical 


SAS Institute, 1985. SAS user's guide: Statistics. Version 

5 edition. Cary, NC, USA: SAS Institute Inc., 

956 pp. 

Soil Survey, 1992. Keys to Soil Taxonomy, 5th edition. 

SMSS technical monograph No. 19. Blacksburg, 

Virginia, USA: Pocahontas Press, Inc., 556 

pages. 

Steel, R.G.D., and J.H. Torrie, 1980. Principles and 

procedures of statistics. McGraw-Hill Book Co. 

Inc., Ne:v York, USA. 

Wakley, A., and I.A. Black, 1934. An examination of 

the Degqareff method for determining soil organic 

matter and a proposed modification of the 

chromic acid titration method. Soil Science 37:29 

38. 

Yost, R.S., D.O. Evans, and N.A. Saidy, 1985. Tropical 

legumes for N production: Growth and N 

content in relation to soil pH. Tropical Agriculture 

62:20-24. 


EFFECT OF SURFACE APPLICATION AND INCORPORATION OF 

SUNNHEMP AND VELVET BEAN GREEN MANURES ON THE 

PRODUCTION OF FIELD CROPS IN ZAMBIA 

J. MULAMBU, K. MUNYINDA, S. NGANDU and 0.1. LUNGU 

Department of Crop Sciences, University of Zambia, 

P. O. Box 32379, Lusaka, Zambia 

Abstract 

The use ofgreen manure crops for soil fertility improvement in Zambian cropping systems is becoming increasingly important 

as a cheap source of biologically fixed N. Green manure crops can be used alone or supplemented with mineral 

fertilizer to reduce the cost of crop production. A field study on green manure placement was carried out at the University 

of Zambia field station during the 2001/2002 cropping season, following a request from small-scale growers 

through the Land Management and Conservation Farming (SCAFE) project to evaluate green manure placement. The 

experiment was designed to evaluate the most cost effective way to apply green manure biomass in small scale cropping 

systems and to determine the most appropriate green manure crop for use in these cropping systems. The experiment 

was conducted in two cropping seasons. The first phase (2001/2002 cropping season) consisted of production and placement 

of the biomass. The second phase (2002/2003 cropping season) ~onsisted of growing a test crop of maize over the 

biomass treatments applied. In the first cropping season, five varieties ofgreen manure crops were compared to a grass 

fallow. 

The results of the first cropping season are described. Biomass production of 7.5,5.8 and 4 t ha- 1 was highest (P < 0.05) 

for the sunnhemp spp. Crotolaria zanzibarica and Crotolaria juncea, and velvet bean variety W. Somerset respectively. 

The percent N content of the biomass was highest (P < 0.05) for velvet bean variety W. Green (3.2%) followed 

by Crotolaria juncea (2.98%). Sunnhemp spp. produced significantly (P

(SCAFE) requested an evaluation of green manure 

biomass placement. The experiment was designed 

to evaluate the most cost effective way to place biomass 

and to determine the most appropriate green 

manure crop. 

and total nitrogen and phosphorus yield were 

evaluated in the selection of the green manure 

crops. This paper reports the results of phase one of 

the trial to evaluate the most appropriate green manure 

crop among the five that were tested. 


The experiment was planned to be conducted over 

two cropping seasons, with the first season 

(2001/2002 cropping season) consisting of pro

it 

4 

a···· · ... 

- 

:.!! 

e..... 3.5 

c· 3 

.c.z 

GI- 2.5 .C .J 

c 

0 .W.G 

0 

2 

c 1.5 

GI 

01 .Gr 

.. 

0 

!:: 0.5 

Z 

0 

C.Z C.J W.G W.S W.St Gr 

SOU rce of Nitrog en 

Figure 2. Comparison of Ncontent among various types of 

biomass. 

Means followed by the same letter are not significantly different. 

 

~ 

a 

~ 

E 

"Q 150 b 

ell 

u 

::> b 

.C.J I 

.W.G 

"C 100 

E bc .W.S I 

Q. 

.W.St' 

Z 50 - c 

to 

0 0 

I 

'" 

~ Cl ~ en 

0 0 

~ ~ ~ 

Source of Nitrogen 

.... 

"- 

Cl 

III G r 

Figure 3. Total Nproduced by various types of biomass. 



The major contribution of the green manure crops 

evaluated was found to be their supply of nitrogen 

to the subsequent crop. For small-scale producers, 

sunnhemp species were likely to supply adequate 

nitrogen for cereal crops. hI the case of velvet bean 

varieties, half of the total N required by cereal crops 

would have to be supplemented from organic and 

mineral sources. The amount of P contained in the 

biomass was insignificant for normal plant growth. 

External sources of P have to be added to crops at 

the recommended rates. 

Future Research Needs 

There is need to quantify the nu.trient contribution 

to the soil by the green manure crops, i.e. evaluate 

the below ground effect of the green manure fallows. 


The authors would like to acknowledge the Land 

Management and Conservation Farming (SCAFE) 

B 

'", 

a 

.J::. 7 .C.Z 

D. 

e 

ab .C.J 

C! 

~ 

.W.G 

5 

"C .W.S 

41 

U 4 

::> 

.W.St 

"C 

3 

.Gr 

E 

c. 

D. 

2 

!! 

0 

I 0 

f"! , C) f/) 

C) 

U U 

0 

~ 

~ ~ 

Source of Phosphorous 

Figure 4. Total Pproduced by various types of biomass. 


SIDA/MAFF Project for the funding that made this 

study possible. Great thanks are also expressed to 

the Crop Science Department of the University of 

Zambia, for the support rendered during the study. 

References 

Allison, F.E., 1973. Soil organic matter and its role in 

crop production. Elsevier, New York, USA. pp. 

450-456. 

Bowen W. T., J. O. Quintain, J. Pereira, D. R. 

Bouldin, W. S. Reid and D. J. Lathwell. 1998. 

Screening legume manures as nitrogen sources 

to succeeding non legume crops. Plant and Soil 

75-80. 

Elliot, L.F. and Papendick, RJ. 1986. Crop residue 

management for improved soil productivity. 

BioI. Agric. Hortic . 3:131-142. 

Faris, M.A., 1986. Plow down effects of different forage 

legume species, cultivars, cutting strategies 

and seeding rates on the yields of subsequent 

crops. Plant and Soil 95: 419-430. 

Giller, K.E., and K.J. Wilson, 1995. Nitrogen fixation 






Rattray, A.G.H. and B.s. Ellis, 1952. Maize green 

manuring in Southern Rhodesia. Rhodesia Agricultural 

Journal 49:188-197. 

Yost, R.s., D.O. Evans and N.A. Saidy, 1985. Tropical 

legumes for N production: growth and N 

content in relation to soil pH. Tropical Agriculture 

62:20-24. 


181


Legume Benefits on Maize Productivity and Soil Properties 

To Walter Mupangwa, et al. 

Q: Can the smallholder farmers accept the 

application of high rates of basal fertilizers, e.g. the 

250 kg/ha Compound D used in your study? Is it 

not advisable to look at a lower range, e.g. 150 kg/ 

ha of compound fertilizer? 

A: Our soils are low in P yet legumes have a 

relatively high P demand. If the dairy farmers 

invest in improving soil fertility, they can recover 

such costs from milk sales or livestock sales. For 

meaningful biomass from legumes to be produced, 

the forage legumes have to be fed, i.e. adequate 

nutrients should be available. 

C: High rates of manure and basal fertilizer, e.g. 250 

kg ha- 1 of basal fertilizer, seem to make it difficult to 

extend otherwise good technologies to some of OUr 

resource poor farmers. Isn't it advisable to look at a 

range of say 150 - 250 kg ha- 1 to allow extension 

personnel to target their different clientele? 

To Bonaventure Kayin:1mura, et al. 

Q: What is the method of incorporation used? Can 

this method be used on a large area? 

A: Farmers in ShuI-ugwi acknowledged that it can 

be done in their fields. 

Q: Mucuna without inputs failed dismally in Soil 

Fert Net trial plots in Murehwa/Wedza in the 

1996/1997 season and was almost written off then. 

What were the soil characteristics of the study sites? 

Did you add any basal fertilizers? 

A: Single superphosphate was added at 200 kg/ha. 

These soils are sandy and were used for maize 

cropping by the farmer. 

C: Follow-up clarification about the Soil Fert Net 

mucuna trials. The experiments tha t failed were 

specifically sited on exhausted and fallowed fields. 

They produced little or no biomass. The aim with 

that work was to test rehabilitation strategies. 

Spatial deployment issues on farm are very 

important for performance. 

Q: Is the moisture difference between late and early 

incorporation really significant? Since 2001/2002 

was very dry, maize might respond to moisture 

benefits of late incorporation, but maize showed 

main benefits with early incorporation. 

A: The moisture difference between early and late 

incorporation is not statistically different at 5% but 

numerically early incorporation has an advantage of 

releasing nutrients early, which might outdo 

moisture content effects on yields. 

C: For your figures, there is need to keep 

consistency in the axes, species 1 ==sunnhemp; 

species 2 == Crotalaria. Sunnhemp is a species of 

Crotalaria. Please use the scientific name to reduce 

confusion. 

Grain leglmtS and Green Manures for Soil Fertility in Southern Africa 183

PERFORMANCE OF GREEN MANURES AND GRA1N LEGUMES ON 

SEVERELY ACIDIC SOILS IN NORTHERN ZAMBIA t AND THEIR 

EFFECT ON SOIL FERTILITY IMPROVEMENT 

COST AH MALAMA and KENNETH KONDOWE 

Misamfu Regional Research Centre, P. O. Box 410055, Kasama, Zambia 

E-mail: misamfu@zamnet.zm 

Abstract 

Green manures have been used in various parts of Zambia, especially where soils are not acidic. TherethR green manures 

have been reported to produce large amounts of biomass that leads to improved soil fertility once incorporated into 

the soil. We assessed the production of above-ground biomass by two green manures and the grain production oftwo 

grain legumes to see how they affect the fertility of an acidic Ultisol and an acidic Alfisol. Green manures were incorparated 

at flowering while grain legume residues were incorporated after harvesting the grain. On the Ultisol, sunnhemp 

produced the most above-ground biomass (2800 kg ha- l ) and velvet bean produced 2000 kg ha- l . Soyabean grain production 

was 792 kg ha- l and cowpea grain yield was just 9.2 kg ha- 1 . However, on the Alfisol velvet bean produced the highest 

above-ground biomass (2100 kg ha- 1 ) and sunnhemp produced 2000 kg ha- l • Grain yield was highest for soyabean 

(1313 kg ha- l ) and lowest in velvet bean (83 kg ha- l ). Velvet bean constantly produced high above-ground biomass on 

both soil types. Thus it can be used as a green manure on both soils. The results show that cowpea might be unsuitable 

for grain production on the Ultisol while soyabean can be used for grain production. Cowpea seems inferior on both soil 

types, while sunnhemp and velvet bean appear to be ideal for the production of biomass on both acid soils. Thus, these 

two green manures can be promoted for soil fertility improvement on these acid soils in northern Zambia. 

Key words: Green manures, soil acidity, Al saturation, P-fixation 

Introduction 

Soils in Northern Zambia are general acidic, infertile 

and of low productivity. As in other parts of Southern 

Africa, nitrogen is the nutrient most limiting 

crop production on these soils. Use of the mineral 

fertilizers needed to increase crop yields has become 

an almost impossible option for smallholder farmers 

due to the escalating prices resulting from the 

removal of subsidies on fertilizers and other agricultural 

inputs. This has seen many farmers resorting 

to biological methods of soil fertility management. 

Green manure use is one way to increase the basket 

of options for small-scale farmers. Several green manure 

legumes have been identified for use in Southern 

African cropping systems. However, the boundary 

conditions under which they perform best have 

not been ascertained. 

There is need to establish the soil and climatic conditions 

for legume adaptation so they can be used to 

improve soil fertility in specific environments. Increasing 

human population densities and the resultant 

pressure on land limits the growing of legumes 

for green manure in some areas as farmers have to 

grow crops that ensure they are food secure. The 

increase in human population has seen intensification 

of agriculture without replenishment of depleted 

nutrients. Population pressure has led to expansion 

of agricultural activities into marginal lands 

resulting in crop production declines. 

Research on practical options that are affordable to 

farmers such as intercropping of green manures 

with other crops to maximize area under cultivation 

is necessary, as well as exploring the use of grain 

legumes for home consumption and for soil fertility 

improvement. Greert manuring was an integral part 

of some local farming systems before inorganic fertilizers 

became widely used. Most elderly smallscale 

farmers recall arid appreciate the usefulness of 

two green manures, Crotalaria spp. and MucurUJ. 

spp., which have shown a high potential to fix atmospheric 

nitrogen symbiotically in a wide range of 

environments (Bowen et al. 1988; Kolar, et al. 1993; 

MacColl, 1990; Yost et al. 1985). 

Green manures have been reported to possess the 

potential to accumulate up to 250 kg N ha- 1 yr'!. 

(Giller and Wilson, 1991; Peoples and Herridge, 

1990). This amount of N leads to an increase in yield 

of cereals, reported to be between 600 and 4100 kg 

ha- I (Peoples and Herridge, 1990). The use of organic 

manures has been shown to improve soil organic 

matter (Mwale et al. 2000) in non-acid soils of 

southern Zambia. The improved organic matter 

status in tum leads to improved soil structure and 

better root aeration leading to improved water 


o 

') 

holding capacity of the soil. This directly causes an 

increase in crop vigour and grain yields. 

The objective of this project was to e\'aluate the biomass 

production of sunnhemp and velvet bean and 

the grain production of cowpea and soyabean on 

two acid soils of Northern Zambia. 

The work was designed to specifically: 

• Determine the influence of soil characteristics on 

legume establishment, growth and biomass production, 

• Assess the contribution of grain legumes and 

green manures to soil fertility. 

Materials and methods 

This experiment was conducted for two agricultural 

seasons: 2001/2002 and 2002/2003. The worked reported 

in this paper is for the 2001/2002 season. 

The experiment was conducted on the Misamfu soil 

series at Misamfu Regional Research Centre (10 0 10' 

S, 31 0 12' E) on an Ultisol and at Mungwi District 

(10 0 10' S 31 0 15' E) on an Aliisol. A composite soil 

sample was collected at 0-20 cm soil depth from 

each site of the trial. The soil sample was dried and 

ground to pass through a 2 mm sieve. The following 

properties were analyzed: pH (in 0.01 M CaCh), Al 

saturation, exchangeable acidity, P (Bray 1) total nitrogen 

(Kjeldahl), organic carbon (Walkley and 

Black, 1934), exchangeable cations (1.0 M ammonium 

acetate, pH 7.0). Particle size was also determined 

using the Pipette method. Table 1 shows the 

soil chemical data. 

Sunnhemp, velvet bean, cowpea and soyabean sole 

crop treatments were planted. Velvet bean was 

planted to give a plant population density of 44000 

plants ha·1, sunnhemp was drilled at a seeding rate 

of about 20 kg ha·1, while cowpea and soyabean 

were also drilled at about 80 kg seed ha·l . The plot 

sizes were 5 x 5 m. The design was an RCBD replicated 

three times. 

The above ground biomass of sunnhemp was determined 

at the flowering stage and then ploughed un- 

Table 1. Initial chemical soil properties of a 

composite sample (0·20 cm depth) of the 

experimental sites in northern Zambia 

Soil characteristics Misamfu Mungwi 

pH (in 0.01 M CaCh) 4.5 4.7 

Organic carbon ('¥o) 0.60 1.31 

Bray 1 P (mg kg') 1.07 2.65 

Exch. K(cmoltkg') 0.47 0.97 

Exch. Ca (cmoltkg') 0.28 0.84 

Exch. Mg (cmoltkg') 0.03 0.08 

AI saturation ('¥o) 20 10 

der while cowpea and soya bean grain were harvested 

at maturity along with the above ground biomass 

production for green manures and grain production 

for the grain legumes. 


Both soils are acidic but the Mungwi soil is slightly 

more fertile than the Misamfu soil, as seen from the 

available P and pH (Table 1). Green manures established 

well at both sites. Soyabean established well 

on both soil types but cowpea establishment was 

bad on the two sites because it was attacked by 

pests. 

Above-ground biomass produced by the two green 

manures was similar on both soil types (Table 2). 

Mungwi site produced a higher grain yieid of soyabean 

than the Misamfu site while cowpea yield was 

poor throughout (Table 3). The green manures 

tested on these acidic soils in northern Zambia seem 

to have the potential to produce adequate biomass 

to allow a cereal crop to produce sufficient grain 

yield. Breman and Reuler (2002) reported a cowpea 

above-ground biomass of 2800 kg ha·l . On the acid 

soils of Northern Zambia, from 2000 to 2767 kg ha- 1 

sunnhemp above-ground biomass was produced 

(Table 2). The total content of N accumulated by 

legume green manures during N2-fixation has been 

measured by various authors. Up to 250 kg N ha- 1 

yr- 1 was reported to accumulate in green legumes 

(Giller and Wilson, 1991; Peoples and Herridge, 

1990). Assuming an average 3% N concentration, 

then from our experiments,sunnhemp was able to 

produce about 59 kg N ha- 1 on the Mungwi soil and 

83 kg ha- 1 N on the Misamfu soil, while velvet bean 

produced 63 and 60 kg N ha- 1 respectively on these 

two soil types. Since commercial fertilizers are expensive, 

a smallholder farmer would be able to produce 

enough maize grain yield to meet food security 

by planting the green manure, because they 

would be able to supplement part of the N fertilizer 

requirement of 120 kg N ha- 1 recommended for 

maize in Zambia (McPhillips, 1987). 

The grain yield of cowpea was very low on the 

Misamfu soil due to low fertility. However, soyabean 

performed much better on that same soil. 

Cowpea was also diseased and this was largely responsible 

for the low grain yield obtained. On the 

Table 2. Means of green 

Table 3. Means of grain legume 

manure above ground biomass grain yields (kg ha in Zambia 

(kg hal) in Zambia Treatments Misamfu Mungwi 

Treatment Misamfu Mungwi Cowpea 9.2 83.3 

Sunnhemp 2767 1967 Soyabean 792 1313 

Velvet bean 2000 2100 CV ('¥o) 115.8 36 

CV (%) 46.09 13.17 Probability 0.18 0.03 

186 


Mungwi soil, cowpea and soyabean produced a 

higher grain yield compared to the Misamfu site. 

This again follows the fertility trend of the two soil 

types as shown in Table 1. 

Cowpea production was well short of potential. The 

potential rainfed production of cowpea has been 

reported to be 1200 kg ha'\ of cowpea grain, in addition 

to the 2800 kg ha- 1 of folder or green manure, in 

the Sudanian savannah (Breman and Reuler, 2002). 

However, on the acid soils of Northern Zambia, less 

than 100 kg ha- 1 was produced (Table 3) . This could 

be due to high Al saturation common in these soils 

(Table I), which might affect root-rhizobium symbiosis 

involved in N2 fixation, as well as to low 

available P leading to the low grain yield. According 

to Breman and Reuler (2002), legumes will 

flourish under conditions of poor N but available P. 

The acid soils of Northern Zambia are low in both N 

and in available P (this is due to P fixation by these 

acid soils). 

Soyabean grain yield was relatively higher on the 

more fertile Mungwi soil than the more acid soil 

(Table 3). In the less acidic soils of southern Zambia, 

average grain yield of 2000 kg ha'\ with rhizobium 

applications have been recorded (McPhillips, 1987). 

Thus even under acid soils, reasonable yield of soyabean 

grain can be achieved as long as seed is inoculated 

prior to planting. 

Conclusion 

Despite the soils being acidic, establishment of 

green manures and soyabean was good. Mungwi 

soil, being slightly fertile than Misamfu soil, produced 

a higher soyabean grain yield. Cowpea grain 

yield on both sites was low, not because of the acid 

soil, but due to pest infestation which is a major 

problem in the cultivation of cowpea in Northern 

Zambia. The benefit due to the green manures will 

be assessed in the next season. 

References 

Bowen, T. 1997. The 1995/96. Fertilizer verification 

trial. Econ9mic analysis of results for policy discussion. 

Malawi Ministry of Agriculture and 

Livestock Development, Lilongwe, Malawi. 22 

pp. 

Bowen, W.T., Quintana, J.O., Pert'ira, J., Bouldin, D. 

R., Reid, W.5. and Latwell. D.T. 1988. Screening 

legume green manures as nitrogen sources to 

succeeding non-legume crops. Plant and Soil 

111:75-80. 

Breman, H and Reuler van, H. 2002. Legumes: 

When and Where an Option? In: Vanlauwe, B., 


Diels, J., Sanginga, Nand Merckx, (eds). -Integrated 

Plant Nutr.ient Management in sub-Saharan 

Africa. CABI International, Wallingford, UK. 

Gilbert, R.A. i 988. Understanding green manures 

for soil fertility enhancement in the maize-based 

cropping systems of Malawi. In: Waddington, S. 

R., Murwira, H.K., I

Abstract 

AGRONOMIC EFFECTIVENESS OF PHOSPHATE ROCK PRODUCTS, 

MONO-AMMONIUM PHOSPHATE AND LIME ON 

GRAIN LEGUMES IN SOME ZAMBIAN SOILS 

OBED I. LUNGU and KALALUKA MUNYII'JDA 

University of Zambia, School of Agricultural Sciences, Lusaka, Zambia 

Phosphorus deficiency severely limits crop yields in some Zambian soils. Where P fertilizer is not applied, yields ofgrain 

legume crops are reduced by as much as 50% from the optimal yields obtained with adequate fertilization. Some of the 

soils are inherently low in P while many others are depleted of P from lack of, or low application of P fertilizer. Recapitalization 

of soil with P fertilizer requires a heavy investment, which makes greater utilization of local phosphate 

rock resources an economically attractive strategy. Additional savings on the cost of Pfertilizers could come from utilizing 

accumulated P from previous applications. However, there is also limited information on the contribution of residual 

P to the nutrition of the subsequent crop. Field trials were conducted at nine sites in two Agro-ecological Regions to 

test the agronomic effectiveness of acid treated phosphate rock and to evaluate the crop response to residual P from the 

previous season. The treatments comprised six rates (0, 40, 80, 120, 160 and 200 kg P20S ha· l ) in On-Station experiments 

and three rates (0, 60 and 120 kg P20S ha- l ) in On-Farm experiments, replicated four and two times respectively 

and arranged in a randomized complete block design. Phosphorus was applied in the first year of the trials (2000/2001) 

as Mono-ammonium phosphate (MAP) and Partially Acidulated Phosphate Rock (PAPR). All experimental plots received 

70 kg K20 ha- l and 26.4 kg S ha- l . The disparity in N content between the two P products was balanced using 

urea on PAPR-treated plots, and subsequently all treatments On-Station received a topdressing of 200 kg N ha- l while 

On-Farm the control and 60 kg ha- l rates received 120 kg N ha- l and the 120 kg P20S ha- l got 200 kg N ha- l . Maize, sorghum, 

cowpea, groundnut, soybean, cotton and sunflower were planted at the different sites according to the importance 

of the crop in the region. In the second cropping season (2001/2002) the planting rows in which the fertilizer was 

banded were maintained, and the residual P was evaluated from the higher rates of P application (120, 160 and 200 kg 

P20S ha- l ) where the P applic'ltion was not repeated. The plots were split so that one half was limed and the other half 

left un-limed. The crops were rotated, and groundnut and soybean followed maize, cowpea followed sorghum. In the 

first year, there was q significant (p < 0.05) yield increase .to P application, regardless of P source, and 80 kg P20S ha- l 

appears to be the optimum rate for all crops. Application of P increased legume grain yields by more than two times the 

yields obtained from fields where all the major nutrients were applied except P. PAPR is as good as MAP in providing P 

to plants and improving yields of crops. On sandy soils, an application of more than 120 kg P20S ha- l as MAP depressed 

yields of legume crops. In the second year, recurrent P fertilizer application at rates of 40, 60 and 80 kg P20S ha- l was 

as effective as residual P fertilizer from the application of 120, 160 and 200 kg P20S ha- l the previous season. The PAPR 

was significantly a superior source of P for the legumes than MAP when lime was not applied. This study suggests that 

applying slowly-available and simply-processed PAPR in amounts sufficient fo·r several seasons in combination with 

readily available N fertilizer may provide a strategy to re-capitalize soil with phosphorus and improve crop yields. 

KetJ words: Residual fertilizer phosphorus, groundnut, soybean, cowpea, cereal-legume rotation 

Introduction 

verse environmental impacts associated with nitrogen 

use, the nutrient balance should be improved 

Crops in general respond quickly and quite dra by promoting P and K fertilizer use. 

matically to N, having a visible effect on crop production. 

For this reason the use of N fertilizer is The increase in both the costs of fossil energy and 

popular with farmers and very often even to the the world-wide demand for N fertilizer in food prodisadvantage 

of other equally essential nutrients duction are major reasons for the rekindled interest 

such as P and K. World fertilizer consumption by in biological nitrogen fixation (BNF) as an alternanutrients 

during the past 36 years up to 1995 (FAO, tive, or at least a supplement to the use of inorganic 

1994,1996) shows a consistent N:P20S ratio of nearly N fertilizers. Nitrogen fixation, whether biological 

2.5:1 , illustrating the dominance of N in total fertil or industrial, is a highly energy-consuming process, 

izer use. In Zambia, this ratio is 3:1 (FAO, 1996). To and inadequate sources of energy is one of the maimprove 

fertilizer use efficiency and minimize ad- jor limiting factors to achieving optimum BNF. 


Among BNF systems, symbiotic systems involving 

legume/bacteria associations have the highest N2 

fixing capability because N2-fixing microorganisms 

are supplied directly from the hos} plant with carbohydrates 

as a ready source of energy for N2 fixation. 

Therefore, root nodulation and N2 fixation are 

more complete and efficient when all the essential 

plant nutrient elements are available in sufficient 

quantities to the macrosymbiont. This fact is not 

always appreciated, and legumes are generally 

thought to be so well endowed that that they will 

fix N2 regardless of their non-N nutrition status. 

Phosphorus plays a critical regulatory function in 

photosynthesis and carbohydrate metabolism of 

leaves and P deficiency can limit growth, particularly 

during the reproductive stage of the crop. In 

the N2 fixation reaction involving the catalyzing enzyme 

complex nitrogenase, energy in the form of a 

r~d uctant Adenosine Tri-phosphate (A TP) is essential. 

Ciaquinta and Quebedeaux (1980) reported 

that the level of P supply during this period regulates 

the starch/sucrose ratio in the source leaves 

and the partitioning of photosynthates between the 

source leaves and the reproductive organs. This effect 

of P on partitioning of photosynthate is presumably 

responsible for the insufficient photosynthate 

supply to nodulated roots of phosphorusdeficient 

legumes and the occurrence of nitrogen 

deficiency symptoms in N2-fixing legumes receiving 

deficient levels of phosphorus (Marschner, 1986). 

Root infection with Versicula-Arbsucular (V A) mycorrhizae 

(Aguilar et al. 1979) not only increased P 

uptake from soil, but also VA aided the establishment 

of bacteria that fix N2 in soils that are low in 

available phosphorus. 

PR would be one way to provide the PR at low cost, 

but this mode of application was not effective with 

Zambian PR. In current field trials, simply processed 

partially acidulated PR (PAPR) was utilized. 

The main objective of this study was to evaluate the 

agronomic effectiveness of P APR produced from 

simply-processed phosphate rock products in soils 

of varying soil chemical properties, for grain legumes. 


The field trials were conducted in two Agroecological 

Regions of varying rainfall, length of 

growing season and soil properties, as shown in 

Figure 1. In the first year (2000/1 cropping season), 

seven trials were conducted consisting of four On 

Station and thr,ee On-Farm experiments. Three On 

Station trials were planted in Agro-ecological Region 

II at Kafuku Farm Institute in Mukonchi, University 

Farm (UNZA) and Magoye Cotton Development 

Trust (COT) on Mushemi, Chelstone and Nakambala 

soil series respectively, One On-Station 

trial was planted in Region I at Lusitu. All the onfarm 

trials were planted in Region II at Chibwe on 

Mushemi soil series, Colden Valley Agricultural Research 

Trust (CART) on Makeni soil series and at 

Magoye Mwanachingwala village on Nakambala 

soil series. The sites were selected for their low 

available phosphorus fertility status. The initial soil 

test values for P and pH are shown in Table 1. All 

the soils were slightly acid, to acid, and deficient in 

plant available phosphorus. Therefore, crop response 

to applied phosphorus fertilizer was expected 

at all these sites. 

Phosphorus deficiency is a major factor limiting 

crop production in the tropics, presumably 

because of the fixation of 

phosphate by iron and aluminum oxides. 

Much more P fertilizer, therefore, 

is required to meet crop requirements 

over and above the quantities that are 

fixed. The cost of fertilizers is often 

the reason for inadequate fertilization. 

In the second cropping season (2001/2002), the tri 

Tentative Distribution olSoil Series in the PAPR Project Implementatfon Area of Zambia 

Many countries in Sub-Saharan Africa 

(--_._......_..... 

"\ 

are rich in phosphate rock (PR)-the 

! LEGEND 

primary raw material for the produc 

-Sou Series 

tion of phosphate fertilizers. Because 

I 

-Maur,a 

of low local demand and the global """""'" I 

i 'A'~J 

surplus of P fertilizers, these deposits 

"':J~ I 

Mu~lill 

have not been developed. Technical, 

()IhoQr 8on.~ 

economic and conducive policy re 

-La...... 

l~ ~ J 

gimes are needed to initiate tapping of 

Se.t, I: 5.010,0.0 ._.... 100&1 .....,__.... ,..., 

these resources and providing them at 

low cost. Direct application of ground 

Figure 1. Agro-ec%gica/ Regions of Zambia 

190 


Table 1. Reduction of soil acidi.ty after liming at experimental sites. 

Acidity (cmol kgl) lime added 

Site A1 3 • + H" A1 3 • kg hal Initial After liming 

1. Chibwe 0.20 0.10 360 5.1 5.3 

2. GART 1.00 0.66 1500 4.1 5.2 

3. Magoye COT 0.30 0.24 510 4.1 4.5 

4. Mwanachingwala 0.34 0.28 450 4.4 5.1 

5. lusitu 0.12 0.10 450 4.3 5.4 

als were extended to three other sites - one site in 

Eastern Province (Region II) and two in Northern 

Province (Region III). 

Phosphorus was applied as mono-ammonium phosphate 

(MAP) as the reference fertilizer and as partially 

acidulated phosphate rock (PAPR) produced 

at the Pilot Plant of the School of Mines at the University 

of Zambia from Chilembwe phosphaterbck. 

The trials were designed as On-Farm, or On-Station 

in which the plots were 100 m2 and 22.5 m2 respectively. 

The P application rates were 0, 60 and 120 kg 

P20S ha· J in On-Farm trials and 0, 40, 80, 120, 160 

and 200 kg P20 S ha- 1 in On-Station trials. All treatments 

were replicated fouT times in a randomized 

complete block design. The fertilizer was banded in 

the planting furrow below the seed at planting. All 

treatments received adequate amounts of K, S as 

recommended for the particular sites (70 kg and 

26.4 kg ha- 1 respectively). Nitrogen was applied as a 

basal application at 24 kg N ha- 1 at On-Farm sites 

and 44 kg N ha- 1 at On-Stati0n sites. The test crops 

were maize; sorghum, sunflower and legumes 

(soybean, groundnut and cowpea). A peat-based 

inoculum was applied to soybean at planting using 

the recommended rate. 

The test crops were grown for . two seasons, and in 

all cases improved varieties of test grain legumes 

were planted according to the suitability and importance 

of the legume in the locality of the trial. In Onstation 

trials in Region II, soybean variety Kaleya 

and groundnut variety MGV4 were planted. These 

varieties are well adapted to Region II and are very 

responsive to inoculation with Rhizobium. The 

MGV 4 groundnut variety is tolerant to soil acidity 

and has low pod failure (Pops) in these soils. Soybean 

was grQwn at Kafuku Farm Institute and 

UNZA Farm, groundnut at Magoye CDT. Cowpea 

variety Bubebe was grown at Lusitu in Region 1. 

The variety was grown because of its earliness and 

high yields, the former attribute being particularly 

important in this drought prone Region. 

Soybean was drilled at an inter row spacing of 75 

cm. Groundnut was grown at an inter- and intrarow 

spacing of 75 cm and 10 cm respectively. The 

spacing for cowpea was 75 cm between rows and 10 

em between plants. 

pH 

During the second cropping season (2001/2002), ~he 

planting furrows from the first season were maintained. 

However, the crops were rotated around the 

plots at each sit~ . No further applications of P were 

made to the higher rates of P application (120, 160 

and 200 kg P20S ha- I ), and the residual effects were 

evaluated from these treatments. Other nutrients, N, 

P, K and S including the application of inoculum 

were repeated as in the first season according to the 

current fertilization practice. An absolute control 

treatment in which no fertilizer was applied was 

included for sites where space permitteQ. adjacent to 

the current trial. 

The treatments were split, and a lime treatment was 

included to evaluate its effect on crop growth, especially 

on the acid soils at Chibwe, GART, Magoye 

(both on-station and on-farm) and Lusitu sites. Each 

original treatment plot was split into two equal subplots, 

and one half was limed while the other was 

not limed. The amount of lime applied was calculated 

based on the exchangeable aluminium values 

(Table 1). 

The lime was broadcast on the surface and then 

worked into the soil by light cultivation using hand 

hoes before planting. Crop growth was monitored 

during the season, and some plant growth parameters 

were recorded. Crop management both in the 

first and second cropping seasons was carried out 

according to the conventional agronomic practices 

for these crops. 


Although various test crops were evaluated, only 

the results for the grctin legumes are presented and 

discussed in this paper. These results are discussed 

according to crop across trial sites. 

Soybean 

At Kafuku Farm Institute (sited on Mushemi soil 

series), a response to soybean biomass and grain 

yield was obtained in the second cropping season 

only with the application of PAPR at 40 kg P20S 

ha- 1 . There was a tendency for the residual effect of 

both MAP and PAPR to decrease with increasing P 

level, reaching a minimum when P was applied at 

160 kg P20S ha- 1 and subsequently increasing at the 

highest rate of P application. This is illustrated in 

Figure 2, showing the effect of source and level of P 

on soybean grain yield. 

The soils at Chibwe On-Farm sit-e · were similar to 

those at Kafuku Farm Institute except that the soils 

were higher in initial soil P. Consequently in the 

first cropping season (2001/02), there was no yield 


191

3500 

.f"'3000 

ca 

,c2500 

CI 

"" - 2000 

.PO 

~ 

'ti 

>. 1500 • MAP 

.PAPR 

c 1000 

~ 

(!) 500 

0 

o 40 80 120 160 200 

P level (kg P 2 0 S ha- 1 ) 

Figure 2_ Effect of source and level of P on soybean grain yield at 

Kafuku Farm Institute. Means followed by the same letter are not 

significantly different. 

response to applied P on soybean regardless of P 

source. The soil available P was adequate to meet 

the nutrient demand of the crop. 

In the second cropping season, there was no response 

of biomass and grain yield on the control 

treatment to lime application. This is because the 

initial soil pH for this site of 5.1 was high and consequently 

the exchangeable aluminium of 0.1 cmol 

kg-I was low to be detrimental to plant growth. Liming, 

therefore, did not reduce exchangeable aluminium 

any lower than was already in the soil to negatively 

influence plant growth (Table I, Figures 3 and 

4). 

High biomass yields of 1.7 and 1.3 times over the 

control were obtained for the recurrent and residual 

appiication of MAP respectively. Similarly, high 

grain yields of 1.6 times over the control treatments 

were obtained for both the fresh and residual application 

of MAP. Liming increased biomass and grain 

yields largely in the fresh than in the residual application 

of MAP. 

;::-- 5000 ,...,~=-.........~~----_~~ 

~ . 4500 .1--------.......",~----4;,..-~ 

~ 4000 

;- 3500 

Qj 3000 

': 2500 

~ 2000 

E 1500 

.S! 

..0 1000 

500 

-tV 

0 0 

~ 

o 13.5 27 60 120 

P level (kg P 2 

0 S ha- 1 ) 

.PO L 

.1'0 UL 

.MAP L 

.MAP UL 

• PAPR L 

ill PAPR UL 

Figure 3. Effect of source of P, level of Pand lime on soybean 

biomass yield at Chibwe On·farm site. Means followed by the same 

letter are not significantly different. 

With PAPR·there was a response to P for soybean 

biomass and grain yields only to residual application 

of the fertilizer. Highest biomass and grain 

yields were obtained at the lower rate of 13.5 and 27 

kg P20S ha- 1 for biomass and 27 kg PiOs ha- 1 for soybean 

grain yield. Both the biomass and grain yields 

decreased with increased rate of P so that there was 

no response to P at the ·highest rate (120 kg P20S 

ha- 1 ) with and without liming for grain yield and 

without liming for biomass. The effect of lime for 

MAP was similar, except at the lowest rate of P for 

grain and biomass yield. The reduction of soybean 

grain yield with increasing rate of residual P application 

suggested adequacy of soil P for crop production. 

The residual effect of P for PAPR was thus 

more effective than that of the fresh and residual 

application of MAP because adequacy in soil P for 

plant growth -:vas reached at a lower rate of P application. 

On Makeni soil series at GART, higher grain yields 

of soybean were obtained in the first cropping season 

with the recommended level of P application of 

60 kg P20S ha- 1 for MAP than at the improved technology 

level of 120 kg P20S ha- 1 . This is shown in 

Figure 5. The yields were 2.4 times more than the 

control treatment. In the case of P APR, the soybean 

yields were similar for both the recommended and 

improved technology with a two-fold increase in 

soybean yield compared to the unfertilized control. 

There was a significant reduction (p

2000 ~""""''7:ll!=''''''''__1[::''J!II ;>g!'~~L1!!2:~~ 

1800 t"-:":"-:-~""='....:r:~ 

-'Iv 1600 T-'~---'''';::''::~~ 

.s:: 1400 f-~""'""':-:=:~~ 

Cl 

==- 1200 +:::~-.~~~ .PO 

~ 1000 +.'1'"""--",.:-.,..,...-=:;;, 

Q.I 

.MAP 

';;;:' 800 t-"'~"'-:-'7-:'-- .PAPR 

600 

400 

200 

o 

o 60 120 

P Level (kg P~s ha- 1 ) 

Figure 5. Response of soybean grain yield to level and wurce of P 

at GA RT. Means follows by the same letter are not significantly 

significant. 

1600 

1400 

above-ground biomass was not translated into 

higher grain yields, suggesting poor photosynthate 

partitioning to pods. This was despite Soil P increasing 

with P application, reaching ad~quate levels for 

plant growth at greater than 60 kg P20S ha- l for both 

sources of P. The low shelling percentage indicated 

that there was need for liming. PAPR was 2.5 times 

more effective in increasing the level of soil P at the 

highest rate of P applied (120 kg P20s ha- 1 ). For 

MAP, soil P decreased at this high application rate. 

The higher rates of P applied as PAPR tended to increase 

soil P values and in tum tended to produce 

higher grain yields and increased shelling percentage. 

In the second cropping season, adding nutrients 

other than P (N, K, S) increased biomass yield over 

the absolute control, indicating that these nutrients 

were limiting. The effect of P on biomass yield depended 

on lime. Lime depressed biomass yield of 

the absolute zero control treatment, while increasing 

the yield of the PO control where P was not applied 

(Figure 9). The significant depression of biomass 

yield of 1.4 times with liming was probably 

due to a Ca/Mg imbalance. Magnesium was low in 

these soils, and therefore adding an excess of Ca 

through P APR and lime probably offsets the balance. 

The comparative higher yields of the unlimed 

compared to the limed absolute control treatment 

was because the exchangeable aluminium was low 

and not detrimental to plant growth even though 

the soil pH was strongly acidic. Overall, liming increased 

the pH from 4.4 without liming, to 5.1 wi'th 

liming. The change of pH with lime was confined 

mainly to the topsoil. The pH was higher for PAPR 

than MAP when P was applied at 60 kg P20S ha- 1 

and in the sub soil of the limed plots at both 60 and 

120 kg P20 S ha- 1 (Table 2). This suggests a liming 

effect of PAPR that did not occur at the higher rate 

of PAPR. 

Response of biomass yield to fresh applications of P 

over the absolute control with liming and the PO 

control without liming that were observed for MAP 

without liming and with liming for PAPR was corroborated 

by the low levels of soil Pfor the Absolute 

zero and PO controls (Figure 10). Although residual 

application of MAP and PAPR with and 

without liming did not increase soil P beyond that 

of the control treatments, response to residual P 

was, however, observed for the two fertilizers. The 

response was obtained for MAp without liming. For 

PAPR, the residual effect was greater and more effective 

than that of MAP without liming. 

Cowpea 

At Lusitu On-Station site, there was a significant (P 

> 0.1) response of cowpea grain yield to application 

of P above 80 ~g P20S ha- 1 with P APR and above 120 

kg P20S ha- 1 with MAP (Figure 11) in the first cropping 

season. The yield response to P was consistent 

with the inherent P deficiency in the soil at this. site 

and therefore the need for P application to increase 

yields. This is corroborated by the available soil P 

values which increased to levels adequate for plant 

growth with application of P above 80 kg P20S ha- 1 

for MAP and above 40 kg P20S ha- 1 for P APR 

(Figure 12).PAPR was more effective in increasing 

2500 ,....,.,.,.....,..,..., 

'0 1500 

'ii 

'>, 

UI 1000 

UI 

IV 

E 500 

.2 

In 

0 

....I ....I ....I ....I ....I 

o 

;:) 

o o 

;:) 

o CD N 

.... o 

N 

.... 

P and Lime (Level kg PzOs ha- 1 ) 

The soil P was lowest for the limed non-P fertilized 

control compared to the absolute control whether 

limed or unlimed. Crop production of maize and 

groundnut during the 2000/01 and 

Figure 9. Effect of source of P, level of Pand lime on groundnut 

biomass yield at Mwanachingwala On-Farm site. Means followed by 

the same letter are not significantly different. 

2001/02 cropping seasons depleted soil P Table 2. Response of soil pH to application of lime 

compared to the absolute control. Recur P level Depth (cm) 

rent applications of P as MAP or PAPR kg PzO~ hal (0-15) 

increased soil P by 3.7 and 4.9 for MAP MAP MAP PAPR PAPR MAP 

and PAPR with liming and by 2.8 and 4.2 l Ul l Ul l 

for MAP and PAPR without liming re 60 4.9 bcd 4.1 fgh 5.5 a 4.6 cde 3.9 gh 

spectively. The increase in soil available 120 

P occurred primarily in the topsoil, espe- CV _ 4.35 % 

5.3 a 4.5 cde! 5.7 a 4.1bcd 4.6 cde 

cially with PAPR with liming and to a LSD _ 0.4412 

lesser extent with MAP without liming. Means followed by the same letter are not significantly different. 

The increase in soil.P was highest for Key 

PAPR with or without liming. 

MAP L . MAP Limed MAP UL· MAP Unlimed 

PAPR L . PAPR Limed 

PAPR UL . PAPR Unlimed 

(15·30) 

MAP PAPR PAPR 

Ul l Ul 

3.9 h 4.7 cde 4.5 cdef 

4.1 fgh 5.2 ab 4.1 fgh 

194 


~'tn 

30 

25 

..:..: 20 

~ 

.PO Limed 

.PAPR Limed 

15 

EmAbsO Un limed 

11. 

·0 10 • PO Unlimed 

en 

.MAP Unllmed 

5 l1li PAPR Unlimed 

o 

o 60 120 

P Level .( kg P 2 0 S ha·') 

Figure 10. Effect of source of P, level of Pand lime on soil available 

Pat Mwanachingwala On-Farm site. 

80 ,~~~----~~~__----~ 

70 ~~---,~--~~~~~~ 

- 60 -+----4:~--~----::c;-;-;;::-:-;::;-:;:--l 

3r ,--------- 

0, 50 -~---F-------'\_--;;-7.7---~-; - PAPR (0-15) 

.§.. 40 +.-----t'--~--'---~---~ - PA P R ( 15-30) 

11. 

- MAP (0-15) 

30+--+--------~~~~~~ 

·0 - MAP (15-30) 

en 20+-~--~~~~~~~~~ 

10 +-~~~~~------~~- 

O 

+---~--._--._--.---._--, 

o 40 80 120 160 200 

P level (kg P20S ha- 1 ) 

Figure 12. Effect of source and level of Pon soil available Pat 

Lusitu. 

the level of soil P at all application rates. Yields of 

more than 2 t ha- l were obtained with the application 

of at least 80 kg P20S ha- l compare,s.:l to less than 

1 t ha- I without P application. PAPR was more eff~ctive 

than MAP in increasing cowpea yields. This is 

indicated by the difference between the response 

area graphs of PAPR and MAP (Figure 11). 

In the second cropping season, response to application 

of MAP was obtained only at 80 kg P20S ha- l , 

while response to P APR was observed at a lower 

rate of 40 kg P20S ha- l . There was a response to residual 

application of P only with PAPR applied at 

more than 120 kg P20S ha- I . This is attributed to the 

grea ter effectiveness of P APR because . of the slow 

release of P from P APR. 

Liming increased soil pH from 4.3 to 5.5 in the 

limed control treatments. Application of lime decreased 

cowpea grain yield by as much as 1.8 compared 

to the unlimed control. This was probably 

due to an imbalance of nutrients for normal plant 

growth. The yields were comparatively higher in 

the unlimed control treatment because the exchangeable 

aluminium of 0.1 cmol kg-I was too low 

to reduce crop yields. The effect of lime on cowpea 

~ 

"I 

..c '" 

2500 

2000 

~ 

..::0: 1500 

~ 

"0 

.. 

.:;., 

c: 

'" ... 

r.:> 

1000 

500 

0 

0 40 80 120 160 200 

P Level (kg P 205 ha.- 1 ) 

Figure 11 . Response of cowpea grain yield to source and level of P 

at Lusitu 

-.;- 400 .,----- - ------------- 

III 

.c 350 +----------------------- 

Cl 

~ 300 +-------~--~~~------- ~______~ 

" '___~L......._~::::=::::~~~___- - MA P L 

ai 250 + 

.>' 200 +---/-~L----=........:~;;;:_----~- - MA P UL 

c:: 

-PAPRI,. 

'~ 150 -PAPR UL 

Cl 

III 100 +----------------------- 

41 

0.. 50+---- --------------- 

:: 

o 0 +---.---,----,--.--,--, 

U 

o 40 80 120 160 200 

P level (kg P 2 0 S ha- 1 ) 

Figure 13. Effect of source of P, level of Pand lime on cowpea 

grain yield at Lusitu. 

grain yield at higher rates of P was different for 

MAP and PAPR. Cowpea grain yields increased 

with lime application for MAP, while in the case of 

PAPR the highest yie.lds were obtained without liming 

(Figure 13). Lime increased cowpea grain yield 

by over 50% compared to the unlimed control. 

Cowpea grain yields were about nine times lower in 

2001/02 compared to the 2000/01 cropping season 

because of the severe drought experienced in Agroecological 

Region I at Lusitu. 

The first season groundnut was planted in the 

2001/02 cropping season with an On-Farm and On 

Station trial in Petauke and Msekera Research Station 

respectively. At Petauke, there was a response 

to P only at the higher rate of P application of 120 

kg P20S l)a- l for PAPR (Figure 14). The yield at this 

level of P increased two-fold compared to the absolute 

control. Other nutrients apart from P were limiting 

in the absolute control. The level of soil P at 

this site was high so that the groundnut yield of the 

non-P fertilized control and P applied at 60 kg P20s 

ha- l was similar. 


195

3000 

"0 

]i 

>. 

,5 

~ 

C> 

800 

600 

400 

200 

0 

o 60 120 

.PAPR 

Figure 14. Effect of source and level of Pan groundnut yield at 

Petauke On·Farm site. Means followed by the same letter are not 

significantly different 

At Msekera Research Station, there was a response 

to P for MAP at 40 and 80 kg P20S ha·J and the highest 

rate of P application of 200 kg P20s ha· J , Response 

to PAPR was only obtained at 80 kg P20S 

ha· 1 (Figure 15). MAP was superior to PAPR, especially 

at the lower rate of P application, possibly because 

the PAPR maintained a higher rate of soil P 

compared to P APR. 

Conclusions 

The biomass and grain yields of the test legumes 

more than doubled with the application of P. 

Simply processed PAPR (50% acidulated with concentrated 

H2S04) was agronomically as effective as 

MAP and had an even better effect than MAP on 

acid soils. Soil recapitalization can be achieved with 

PAPR rather than with MAP because it does not depress 

plant growth at higher rates. There is greater 

soil residual P with PAPR than with MAP. PAPR 

maintains a higher level of soil available P than 

MAP, especially at the higher level of P. The optimal 

P application rate was 80 kg P20S ha· 1• The results 

of the PAPR study have indicated that nutrients 

other than P are limiting grain legume production. 

There is therefore a need to identify those that 

are limiting. 

«;- 2500 

.r:. "' 

.. 

t:n 2000 

::. 

'0 1500 

Qj 

':;" 

1000 

,5 

~ 

(!) 500 

0 

o 40 80 12() 160 200 

P Level (kg P 2 0 S ha·1) 

.AbsO 

.PO 

.MAP 

.PAPR 

Figure 15. Effect of source and level of Pon groundnut grain yield 

.at Msekera Research Station. Means followed by the same letter 

are not significantly different. 

References 

Aguilar de, C.A-G, R. Azcon, and J.M. Barea. 1979. 

Endomycorrhizal fungi and Rhizobium as biological 

fertilizers for Medicago sativa in normal 

cultivation. Nature 279:325-327. 

Giaquinta, R.T. and B. Quebedeaux. 1980. Phosphate-induced 

changes in assimilate partitioning 

in soybean leaves during podfilling. Plant Physiology 

65: Suppl., 119. 

FAO (Food and Agriculture Organization of the 

United Nations). 1994. Fertilizer data diskettes. 

F AO (Food and Agriculture Organization of the 

United Nations). 1996. Fertilizer data diskettes. 

Marschner, H. 1986. Mineral Nutrition of Higher 

Plants. Academic Press, New York, USA. 

196 


THE EFFECT OF PHOSPHORUS AND SULPHUR ON GREEN MANURE 

LEGUME BIOMASS AND THE YIELD OF SUBSEQUENT 

MAIZE IN' NORTHERN MALAWI 

ATUSAYE B. MWALWANDA', SPIDER K. MUGHOGHO', 

WEBSTER D. SAKALA? and ALEX R. SAKA 3 

1Bunda College of Agriculture, P. O. Box 219, Lilongwe 

2Maize Commodity Team, Chitedze Agricultural Research Station 

P. O. Box 158~ Lilongwe 

3Department of Agricultural Research and Technical Services, Ministry of Agriculture 

and Irrigation, P. O. Box 30134, Lilongwe 3, Malawi 

Abstract 

A study on the effect of phosphorus and sulphur on biomass production of green manure legume crops and that of the 

green manures on subsequent maize yield was conducted during the 1999/2000 and 2000/2001 growing seasons. The 

study was undertaken at two sites in Northern Malawi. The main objective of the investigation was to evaluate the response 

ofgreen manure legume crops to Phosphorus and 5ulphur application in terms of dry matter production and the 

manure's effect on subsequent maize yield. 

Three legume green manure crops: Mucuna pruriens, Cajanus cajan, and Tephrosia vogelii and one cereal crop, 

Zea mays, were planted as sub-plots and each crop received three rates of P and 5 fertilizer; 0, 20 kg P20S and 4 kg 5, 

and 40 kg PzOs and 8 kg 5 per hectare. At each site, the experiment was replicated five times using five farmers' plots in 

a 2*4"'3'" split-split plot arrangement in a Randomized Complete Block Design. 

There were significant differences between the four crops (P= 0.001) in biomass production. Mucuna prurie~ outperformed 

the other crops (5380 kg ha-1),followed by Tephrosia vogelii (5258 kg ha-1), Zea mays (4972 kg ha-1) and Cajanus 

cajan (2669 kg ha- 1 ) respectively. Fertilizer application significantly increased (P=O.OOV biomass production. 

The lowest amount of biomass was recorded from the treatment without fertilizer input, and the highest was recorded 

from the treatment with the highest fertilizer rate. The two sites were significantly different (P=O.OOl) for biomass production. 

Mean biomass yield at Nchenachena (5650 kg ha-1) was higher than from Champhira (3490 kg ha- 1 ). 

In the subsequent growing season, the maize grain and total dry matter produced were significantly different at (P < 

0.01) and (P = 0.001) respectively and were attributed to the type of crop preceding them. Maize grain and biomass after 

Mucuna pruriens was the highest (1104 kg ha-1for grain and 5170 kg ha-1for biomass). This was followed by those 

after Cajanus cajan (880 and 4430 kg ha- 1 ), Tephrosia vogelii (785 and 3915 kg ha- 1 ) and then Zea mays (627 kg ha- 1 

for grain and 3059 kg ha-1for total dry matter produced) _ 

Key words: Green manure legume, short-term fallow, phosphorous, sulphur, maize, northern Malawi 

Introduction 

The problem -of declining soil fertility in smallholder 

farms is recognized as the fundamental cause of declining 

per capita food production in African agriculture 

(Gilbert, 1998; Smaling, 1993; Mokwunye et 

ai., 1996). Major causes of declining soil fertility include 

continuous monocropping, low use and inappropriate 

application of organic or inorganic fertilizers, 

lack of fallows, and lack of proper soil and water 

conservation practices. These factors have contributed 

to lower average crop yields in most smallholder 

farmers' fields. 

One intervention identified to address the problem 

of soil fertility decline, particularly in low-input and 

limited land resource base agricultural systems, is 

the use of a short-term fallow system with fast 

growing herbaceous legumes planted in rotation 

with major food crops, such as maize (Zea mays L.). 

Some of the promising legume green manure crops 

in such a system include pigeonpea (Cajanus cajan), 

fish bean (Tephrosia vogelii), 5esbania sesban anJ velvet 

bean (Mucuna pruriens). Sole cropped green manure 

legumes have the potential to accumulate up 

to 250 kg N ha- 1 yrl (Giller and Wilson, 1991) resulting 

in subsequent cereal yield increases of 600-4100 

kg ha- 1 (Peoples and Herridge, 1990). 

Grain tegumes and Green Manures for Soil Fertility in Southern Africa 197

Most interventions involving green manure crops 

have emphasized the role of the legumes in the 

maize-based cropping systems without looking at 

improving the legume itself. It is iffiportant to note 

that for the leguminous crops to fix nitrogen, they 

need good phosphorus and sulphur nutrition, in 

addition to other nutrients. The aim of supplying 

. green manure legume crops with phosphorus, sulphur 

and zinc is to boost early root development 

that would take up soil nutrients for plant development, 

and subsequent biomass production and biological 

nitrogen fixation (BNF). With low soil nutrient 

contents, most legume manure crops do not produce 

sufficient quantities of biomass to supply the 

required levels of nutrients upon mineralization 

(Palm et al. 1997). Many organic materials when applied 

in modest amounts, i.e. 3-5 t ha·1 dry matter 

contain sufficient N to meet the requirements of a 2 

t maize crop. However, they cannot. supply the P 

requirements of maize, hence legumes must be supplemented 

by P in areas where P is deficient (Palm, 

1995). Application of inorganic fertilizer to legumes 

would thus improve biomass production and nutrient 

recycling, thereby releasing higher amounts of 

plant nutrients upon decomposition. 

The objectives of the experiment were (a) to evaluate 

the effect of phosphorus and sulphur application 

on biomass production by three legume green manure 

crops; MUCllna pruriens, Cajanus cajan and 

Tephrosia vogelii and (b) to screen a green manure 

legume crop that can result in higher yields for the 

subsequent maize crop. 


Experimental sites 

The on-farm, farmer-managed, researcher-designed 

experiment was conducted in two Extension Plan 

ning Areas (EPAs) of Mzuzu Agricultural Develop 

ment Division in Northern Malawi. The sites were 

Champhira EPA in Mbawa Rural Development Pro 

ject and Nchenachena EPA in Rumphi Rural Devel 

opment Project. 

Soils of Champhira (Loudon series) are classified as 

weakly Ferallitic Latosols and those of Nchenachena 

(N chenachena series) are Ferrisols (Young and 

Brown, 1962). Champhira lies at an elevation rang 

ing from 1216 to 1338 ill above sea level and located 

12° 24' Sand 33° 40' E while Nchenachena is 1216 to 

1307 m above the sea level and located at 10° 30' S 

and 33° 50T 

Experimental design 

The experiment was laid out in a split-split plot ar 

rangement in a randomized block design. The two 

sites of the experiment were the main plots. In Year 

1 (1999 - 2000), three green manure legume crops; (i) 

Pigeon pea, hybrid variety ICP 9145 (Cajanus cajan 

(L) Mellsp.), (ii) Velvet bean (Mucuna pruriens) and 

(iii) Fish bean (Tephrosia vogelii) and (iv) Maize hybrid 

MH 18 (Zea mays (1.)), were the sub-plots. The 

sub-plots measured 15 m long with five ridges 

spaced at 0.90 m apart (67.5 m2). There were three 

sub-sub plots for each crop with five ridges each 5m 

long and spaced at 0.90m (22.5 m2). Treatments for 

sub-sub plots were (i) without phosphorous and 

sulphur, (ii) 20 kg phosphoru~ ha·1 plus 4 kg sulphur 

ha- 1 and (iii) 40 kg phosphorus ha··1 plus 8 kg 

sulphur ha- 1 • Plant density was as shown in Table 1. 

Immediately after harvest (3 rd week of May and 2 nd 

week of June, 2000 for Champhira and Nchenachena 

respectively) in year 1 (1999-2000), the green 

manure and m,aize stover were ploughed into the 

soil in the individual treatment plots. In Year 2 

(2000 - 2001), a maize crop (MH 18 hybrid) was 

grown in all the plots to assess the residual effect of 

the green manure legume crops. 

Soil sampling 

Soil sampling was done at each site before the start 

of the experiment. Soil samples were randomly 

taken from 0-15 cm and 15-30 cm soil depths from 

each of the smallholder-farmers' plots using an auger. 

From each farmer's plot, five samples were 

taken at each of the soil depths. The soils from the 

same depths with the same farmer were mixed and 

after several splits, about 500 g of the soil was obtained 

and stored in plastic bottles. The initial soil 

samples were for characterizing the two sites. These 

samples were analyzed for general soil physical and 

chemical properties (Table 2). All soil samples were 

air-dried, sieved through a 2 mm sieve and stored 

in plastic bottles before laboratory analyses. 

Plant sampling 

Three plants from the middle ridge of each treat 

ment plot were sampled eight weeks from planting 

and at mature harvest for both seasons. These sam 

ples were oven-dried at 65°C for 48 hours, then 

ground to powder (passed through a 0.1 mm sieve) 

using an electric grinder and stored in plastic bot 

tles. The samples were analyzed to determine nitro 

gen and phosphorus in the plant tissue. 

Biomass was estimated at harvest for both the leg 

ume and maize stover after the end of the first sea- 

Table 1. Spacing of the crops between and within ridges (em) 

Crop Within ridges Between ridges Plants per station 

Maize 50 90 2 

Mucuna 15 90 1 



198 


Table 2. Initial properties of soils at the two sites before the 

start of the experiments 

Parameter Champhira Nchenachena 

Soil pH (1 :2.5 H2O) 5.3 5.26 

Organic carbon % 0.56 1.032 

Total nitrogen % 0.2 0.634 

C:N Ratio 3.21 1.63 

Mehlich·3 P(ppm) 0.054 0.078 

Clay % 30.5 26.7 

Silt % 8.3 13.3 

Sand % 61.2 60.0 

son and for maize stover orily at harvest for the second 

season. Grain yield was also determined at 12.5 

% moisture content at harvest of both seasons. During 

the first growing season, only maize grain yield 

was recorded. This was because Mucuna pruriens 

and Tephrosia vogelii were incorporated at flowering, 

when the crops attained their potential maximum 

dry matter production. Pigeonpea grain yield was 

very poor, mainly due to poor crop establishment, 

and as such the available data were inadequate for 

statistical analyses. The net plot from which yield 

data was obtained was a four-metre section of two 

of the innermost ridges from the five ridges of the 

sub-sub plots i.e. 2 * 0.90 m * 4 m (7.2 m2). Rainfall 

data was recorded from Champhira and Nchenachena 

meteorological centres (Figure 1). 

Treatment management 

Agronomic operations 3uch as weeding, harvesting, 

and sampling, was done at almost the same time for 

each ·site. Planting was earlier in Champhira EPA 

(within the 3 rd week of December) than in Nchenachena 

EPA (2 nd week of January) owing to differences 

in the onset of the rainy season between the 

two sites (Fig. 1). In the first season, fertilizer was 

applied two weeks after planting using 23:21 :0:4S 

compound fertilizer. The rates were derived using 

the following calculations: 

From a 50 kg bag of fertilizer, there is 21 % P20S and 

4 % S that translates to: 

(0.21 * 50 kg) = 10.5 kg P20S. Similarly, for S = (0.04 * 

50 kg) = 2 kg S. 

Quantity to apply per unit area using the fertilizer 

formulation at hand was obtained from the simple 

proportion calculation below: 

Example 20 kg P20S ha·1 treahnent 

= (20 kg P20sha·1 * 50 kg bag-I) / 10.5 kg P20S bag- I 

= 95 kg ha·1 of the fertilizer. In this there is approximately 

4 kg S. 

The fertilizer was banded along the ridge. Weeding 

was done twice: before applying fertilizer and eight 

weeks from planting. 

In the second year of the trial, 50 kg N ha·1 was top 

dressed using a high analysis straight fertilizer, 

Urea, in all the maize.plots. To get the 50 kg N ha·I, 

the following calculations were done: Each bag of 

Urea has 46 % N,that translates to 23 kg N. The required 

quantity =(50 kg N ha- 1 * 50 kg bag- 1 )/23 kg 

bag- I = 108.7 kg ha·1 Urea. A banding method was 

used. 

The purpose of the second season was to evaluate 

maize yield in response to the incorporated green 

manure legume crops. The nutrients released from 

decomposition of incorporated green manures were 

expected to have a residual nutrient replenishment 

effect. 

Data from the experiment was statistically analyzed 

using the Genstat 5 Release -3.2, (1995) computer 

package. 


First Season Results 

Characterization of the soils at the two experimental 

sites showed that soils at Nchenachena had higher 

percent total nitrogen and organic carbon than soils 

at ChamphiJ;"a (Table 2). The soils at Nchenachena 

have been cultivated for less time than those at 

Champhira. 

Rainfall recorded during the study period. During 

both growing seasons, Champhira received earlier 

rainfall, but it stopped about one month earlier than 

at Nchenachena. This meant different times of 

planting at the two sites. Total annual rainfall was 

higher at Nchenachena (932 mm) than Champhira 

(558 mm) during the first season but in the second 

season the difference was not substantial, i.e. 1061 

mm for Champhira and 1120 mm for Nchenachena. 

However, Champhira EPA received more rainfall 

than normal during the second season (Figure 1). 

Nitrogen and phosphorus in plant species at harvest. 

The four crop species showed significant differences 

(P= 0.001) in N content of their tissues at 

harvest (Table 3). The highest mean was for TephrDsia 

vogelii followed by Mucuna pruriens, Cajanus cajan 

and Zea mays. 

Leguminous crops fix ahnospheric nitrogen in their 

tissues thereby ensuring the supply of this important 

nutrient for their metabolism. Maize relies on 

inherent soil nitrogen and the external supply of 

this nutrient element. There were also Significant 

differences (P < 0.05) in the content of phosphorus 

of the four crops at harvest. Maize had the highest P 

content followed by Tephrosia vogelii, Mucuna pruriens 

and then pigeonpea (Table 3). 


199

~ ~~----.-~r-~-------------r~~ 

Aoo 

350 

Enl 

~250 

:!axJ 

c: 

iii 150 

0:: 100 

50 

O+-~~~~~~~Lr~~~~~~.y~~~ 

O~o""~~ ~

of inorganic fertilizer and the crops, the green manure 

legumes appeared to respond better than 

maize to the increasing rate of inorganic fertilizer 

(Table 6). The differences in response to inorganic 

fertilizer application among the crop species are 

graphed in Figure 3. 

Biomass production in the first year (199912000 season). 

The four crops produced different amounts of 

biomass across the twe sites (P= 0.001). Mucuna pruriens 

had the highest biomass followed by Tephrosia 

vogelii, Zea mays and Cajanus cajan (Table 7). In general, 

all legumes except pigeonpea outperformed 

maize. The difference in dry matter production between 

maize, mucuna and tephrosia was not significant. 

Pigeonpea had lowest dry matter, probably 

due to poor crop establishment. 

With nitrogen being a limiting plant nutrient in 

most Malawian soils (Kumwenda and Gilbert, 

1998), green manure legume crops are likely to outperform 

cereals in their dry matter production. Ad- 

Table 6. The effect of rate of inorganic fertilizer on N 

accumulated (kg ha I dry matter) in different crop species at 

harvest 

Crop spedes 

Rates of inorganic fertilizer (kg hal) 

Nil S 20 kg P20S 40 kg 

+ 4 S P20S + 8 S 

Zeamays 40.8 53.4 54.3 

Mucuna pruriens 124.8 162.0 169.9 

Cajanus cajan 50.3 70.8 96.0 

Tephrosia vogelii 105.2 172.9 190.2 

~.. 

250 T---r=======c=~==============~~~ 

200 

~ 150 

~ 

c 

~ 100 

o 

~ 

Z 

50 

o 

maize mucuna pigeon pea tephrosia 

Crops species 

Figure 3. Nitrogen (k hat) accumulated by crop species as affected 

by rate of inorganic fertilizer 

Table 7. Dry matter production (kg hat) of crops at harvest in 

the first growing Season (1999·2000) 

Crop Species Champhira Nchenachena Mean 

Maize 4204 ' 5741 b 4972 ' 

Mucuna pruriens 4352 ' 6407 b 5380 ' 

Pigeonpea 1906 b 3432 ' 2669 b 

Tephrosia vogelii 3497 • 7020,b 5258. ' 

SED ± 576.6 

CV % 20.0 

Key: SED ±. Standard error of difference; CV. Coefficient of variation; 

NB: Means followed by same letters are not statistically different 

ditionally, some legumes explore a deeper volume 

of soil owing to the~r tap root system and therefore 

can extract nutrients that may have been leached to 

lower soil depths. Mucuna pruriens produces a 

dense vegetative cover, mainly leafy biomass, that 

include a mass of creeping stems, during its vegetative 

growth stages. Kumwenda and Gilbert, (1998) 

found similar results.' Mucuna pruriens had the 

greatest mean biomass and a greater response to the 

added phosphorus than Cajanus cajan and Tephrosia 

vogelii. The other green manure legumes crops, pigeonpea 

and Tephrosia vogelii, have a slower early 

growth rate and tend to lose much of their leafy biomass 

by the time they attain physiological maturity 

(Giller and Cadish, 1995; Sakala, 1994). 

Application of inorganic fertilizer increased biomass 

production. There were significant differences (P = 

0.001) in ciomass production among the crops due 

to inorganic fertilizer. The higher-rate of inorganic 

fertilizer applied had the highest mean dry matter 

produced followed by the second rate and the treatment 

without any inorganic fertilizer gave the lowest 

dry matter yield (Table 8). This positive response 

to inorganic fertilizer application is indicative of the 

need to supplement the green manure crops with. 

external nutrients and that the soils need nutrients. 

Maize had the best response to the inorganic fertilizer 

applied, giving a difference of 2764 kg ha·J between 

the treatment without inorganic fertilizer and 

the treatment with 20 kg PzOs + 4 kg S of inorganic 

fertilizer. This was followed by Tephrosia vogelii, 

(2050 kg ha·1), Mucuna pruriens (1069 kg ha·1) and 

pigeonpea (1005 kg ha·l ). The difference in biomass 

produced between treatments that received 20 kg 

PzOs + 4 kg S and those that received 40 kg PzOs + 8 

kg S was generally lower. The apparently smaller 

response to inorganic fertilizer application by green 

manure legume crops compared with maize is because 

legumes are relatively independent of external 

nutrient supply, particularly nitrogen, and 

hence require relatively smaller doses. 

Second Season Results 

The type of preceding crop significantly (P < 0.01) 

influenced maize stover nitrogen content at harvest. 

Maize grown after Mucuna pruriens had the highest 

Table 8. The effect of rate of fertilizer on biomass production (kg 

ha·t) 

across the sites 

Rate of fertilizer (kg hat) Champhira Nchenachena Mean (kg hal) 

No fertilizer 2357 4133 3245' 

20 kg pzDs + 4 kg S 3908 6013 4961 b 

40 kg P20S + 8 kg S 4204 6803 5503 b 

SED ± 306.2 

CV % 21.2 

Key: SEO:t. Standard error of difference; CV. Coefficient of variation; NB: Means 

followed by same letters are not statistically different. 


201

stover N, followed by after pigeon pea, Tephrosia 

vcgelii and maize respectively (Table 9). 

This could be because Mucuna pruriim$ contributed 

the highest N through its easily mineralizable leafy 

residues. Generally, maize following green manure 

legumes had higher stover N than the continuously 

grown maize. This was expected since the green 

manure legume plots provided more N that the subsequent 

maize crop benefited from after mineralization. 

Maize grain yield. The type of preceding crop significantly 

influenced (P < 0.01) maize grain yield 

across the two experimental sites. The highest maize 

grain yield followed Mucuna pruriens (1104 kg ha- 1 ), 

then after Cajanus cajan (880 kg ha- 1 ), Tephrosia vogelii 

(785 kg ha- 1 ) and continuous Zea mays (627 kg 

ha- 1 ) (Table 10). 

These results are consistent with those by Kumwenda 

and Gilbert (1998). Their studies on green 

manure legumes in rotation with maize on exhausted 

soils of Malawi found that maize grain 

yield was the highest after a Mucuna pruriens fallow, 

followed by Crotalaria spp. and Tephrosia voge/ii fallows. 

Maize grain yield average for the second season 

could have been higher but lack of timely weeding 

at Nchenachena EPA and high rainfall received 

during the 2000/2001 season, at Champhira EPA 

adversely affected crop growth. 

The rate of fertilizer in the first season significantly 

influenced (P < 0.015) maize grain yield. The highest 

Table 9. The effect of a preceding crop on N content in maize 

stover at harvest 

Fallow sequence 

Maize stover nitrogen (kg ha"j 

Maize after maize 

37.7' 

Maize after Mucuna pruriens 

69.5 b 

Maize after pigeonpea 

55.911> 

.Maize after Tephrosia vogel;; 

44.6'b 

SED ± 

8.79 

CV % 

26.8 

Key: SED ±. Standard error of difference; CV, Coefficient of variation; 

NB: means followed by same letters are not statistically different 

Table 10. The effect of preceding crop species on subsequent 

maize grain yield (kg ha·1) 

at Champhira and Nchenachena 

Fallow sequence Champhira Nchenachena Mean 

Maize after maize 986 268 627' 

Maize after Mucuna pruriens 1783 424 110411> 

Maize after Pigeon pea 1402 359 880' 

Maize after Tephrosia vogelii 1273 297 785' 

SED ± 126.3 

CV % 23.5 

Key: SED±, Standard error of difference; CV, Coefficient of 

variation; NB: Means followed by same letters are not statistically 

different; Sites 1and 2 are Champhira and Nchenachena 

maize grain was from the treatment that received 20 

kg P20S + 4 5 ha- 1 followed by 40 kg P20S + 8 kg 5 

ha- 1 , The treatment without inorganic fertilizer had 

the least mean maize grain (Table 11). This suggests 

an optimum fertilizer rate at 20 kg PiOs + 4 5 kg 

ha- 1 . The higher rate of inorganic fertilizer, 40 kg 

P20S + 8 kg 5, slightly reduced grain production. 

High N fertilizer promotes succulence and more 

vegetative plant material at the expense of reproductive 

organs. 

The response of the second season maize crop to the 

application of inorganic fertilizer showed that yield 

of maize grain after Mucuna pruriens was the highest 

followed by that of pigeonpea, Tephrosia vogelii and 

maize (Figure 4). Maize grain yield after maize and 

pigeonpea was depressed at the higher rate of fertilizer 

application (40 kg P20S + 8 kg 5 ha- 1 ). This may 

be because the compound inorganic fertilizer used 

has nitrogen, ~hich when applied at high rates 

tends to enhance vegetative growth and succulence 

at the expense of grain production. The other possible 

reason, particularly for maize, is that maize residues 

incorporated at harvest after the first season 

had a low N content, unlike Mucuna pruriens, pigeonpea 

and Tephrosia z:ogelii (Table 3). The low N 

content could have caused net immobilization of 

nutrients, particularly nitrogen, from its residues. 

Crop residues of low quality, i.e. less than 2 % nitrogen 

can result in poor growth of the succeeding cereal 

crop since the N requirement of the crop is not 

in synchrony with N mineralization (Nandwa et al., 

1995). 

Table 11. Effect of inorganic fertilizer rate on maize 

grain yield 

Rate of fertilizer 

Maize grain yield 

-----(kg ha'j---- 

No fertilizer 708' 

20 kg P205 + 4 kg S 921 b 

40 kg P205 + 8 kg S 918 b 

SED ± 78.8 

CV % 29.4 

Key: SED ±, Standard error of difference; CV, Coefficient of variation; 

NB: means followed by same letters are not statistically different 

1400 

1200 

~ 1000 

Maize grain yield after the green manure legumes 

was higher compared with that after maize, but 

among the three green manure legume crops it was 

. m.aize after Mucuna pruriens that gave the highest 

grain yield. The likely reason is that Mucuna pruriens 

produced the most biomass, besides having 

residues that had a relatively high N content. 

Conclusions 

The green manure legume crops tested in this study 

responded to the application of inorganic fertilizer 

and an increased rate of inorganic fertilizer produced 

more dry matter. The response to fertilizer 

was greatest between the unfertilized control and 

the lower rate of fertilizer application (20 kg P20S + 

4 kg S ha- I ). 

Among the three candidate green manure legume 

crops tested, Mucuna pruriens is the superior, giving 

the highest dry matter production. It is therefore the 

best candidate green manure legume crop for improving 

soil fertility. Another good candidate was 

Tephrosia vogelii, which at the time of incorporaiion 

gave the highest nitrogen content in its tissue. 

Maize gave the greatest response to inorganic fertilizer 

among the four crops tested whereas Tephrosia 

vogelii responded most favourably among the green 

manure legume crops. Cajanus cajan was the least 

responsive. Maize residues however, had the least 

content of nitrogen, a factor disc;.ualifying it as a potentiat 

green manure crop in low input systems and 

soils with low fertility. Maize grain yield after the 

short-term fallow was higher after the green manure 

legume crops than after maize. Maize grain 

and biomass produced was highest after Mucuna 

pruriens. 

Inorganic fertilizer, particularly at the lower rate, 

had a positive residual effect on maize grain yield. 

The higher rate of inorganic fertilizer gave the most 

remarkable positive effect on total maize biomass 

production. 


Application of modest amounts of phosphorus and 

sulphur fertilizer (20 kg P20S + 4 kg S ha- I ) to green 

manure legume crops should be adopted to improve 

the litter quality of these organic fertilizers as 

well as enhance their growth and subsequent dry 

matter production. 

Straight inorganic fertilizer sources of phosphorus, 

sulphur and nitrogen should be used in further 

studies to isolate the individual effects of these nu- 

trient elements as well as their interactive effects. 

Other critical nutrient elements such as zinc and 

molybdenum have to be tested in experiments 

where green manure legumes are screened for response 

to inorg'anic fertilizers. 

Mucuna pruriens should be promoted as a potential 

soil fertility-improving crop where soil fertility is 

low. The effect can be seen. in as short a fallow as 

one season. 

Long-term studies of the residual effects of green 

manure legume crops on soils, as well as on cereal 

crop yields, should be conducted to ascertain sufficient 

information about the benefits of the shortterm 

fallow system. 


We are grateful to the Rockefeller Foundation for 

providing funds for the research. 

References 

Eaglesham, A.RJ. and A. Ayanaba. 1984. Tropical 

stress ecology of Rhizobia, root nodulation and 

legume fixation. In: Subba Rao, N.5. (ed.) Current 

Oevelopments in Biological Nitrogen Fixation. 42 p. 

Gilbert, RA. 1998. Growth Characteristics and N 

balance of Maize - Green manure intercropping 

systems in Malawi. In: J.D.T. Kumwenda and M. 

K.M. Komwa (ed.) Annual Report for 1997/98 

season for the Cereals Commodity Group, volume 

1. Chitedze Agricultural Research Station, 

Lilongwe, Malawi. pp. 217-223. 

Giller, K.E. and K.J. Wilson, 1991. Nitrogen Fixation 

in Tropical Systems. CAB International, Wallingford, 

England. p. 43. 

Giller, K.E. and G. Cadish. 1995. Future benefits 

from biological nitrogen fixation: An ecological 

approach to agriculture. Plant and Soil 174:225~ 

277. 

Kumwenda, J.D.T. and R Gilbert. 1998. Legume 

biomass production and maize yield response in 

legume manure rotations in Malawi. In: JD.T. 

Kumwenda and M.K.M. Komwa (ed.) Annual 

Research Project Report for the 1997/98 season 

for the Cereals Commodity Group. pp. 148-159. 

Kumwenda, J.D.T. and R, Gilbert. 1998. Biomass 

production by legume green manures on exhausted 

soils in Malawi: A Soil Fertility Network 

Trial. In: Waddington, S.R; Murwira, H.K.; 

Kumwenda, J.D.T.; Hikwa, 0; Tagwira, F. (eds). 

~oil Fertility Research for Maize-Based Farming Sys 


203

tems in Malawi and Zimbabwe. Proceedings of the 

-Soil Fert Net Results and Planning Workshop 

held from 7 to 11 July 1997 at Africa University, 

Mutare, Zimbabwe. Soil Fert Net and CIMMYT 


Mokwunye, A.U., A. de Jager, and E.M.A. Sma ling 

(ed.). 1996. Restoring and maintaining the productivity 

of West African soils: Key to sustainable 

development. Miscellaneous Fertilizer Studies. 

14. International Fertilizer Development 

Centre, Lome, Togo. pp. 12-31. 

Nandwa, S.M., J.M. Anderson and P.O. Seward. 

1995. The effect of placement of maize stover and 

N fertilisation on maize productivity and N use 

efficiency. Mimeo, 19 pp. 

Palm, CA. 1995. Contribution of Agroforestry trees 

to nutrient requirements of intercropped plants. 

Agroforestry Systems 30:105-12~. 

Palm, CA., R.J.K. Myers, and S.M. Nandwa. 1997. 

Combined use of organic and inorganic nutrient 

source for soil fertility maintenance and replenishment. 

In: R.J. Buresh et al (ed:) Replenishing 

Soil Fertility in Africa. SSSA special publication 

No. 51. SSSA, Madison, Wisconsin, USA. pp. 

193-217. 



Agriculture. Advances in Agronomy 44:155-223. 

Sakala, W.O. 1994. Crop management interventions 

in traditional maize pige,?npea intercropping 

systems in Malawi, MSc. Thesis, Bunda College 

of Agriculture, University of Malawi. 

Smaling, E.M.A. 1993. An agroecological framework 

for integrating nutrient management, with special 

reference to Kenya. PhD Thesis. Wageningen 

Agricultural University, Wageningen, Netherlands. 

. 

Young, A. and P. Brown. 1962. The PhYSical Environment 

of Northern Nyasaland. Ministry of 

Land and Physical Planning, Lilongwe, Malawi. 

38pp. 

204 


MANAGEMENT OF AN ACID SOIL USING MINE TAILINGS 

AS LIME FOR SOYBEAN PRODUCTION 

LACKSON K. PHIRI, MOSES MWALE and MLOTHA I. DAMASEKE 

Mt. Makulu Research Station, Soil Fertility Programme, 

Private Bag 7, Chilanga, Zambia; E-Mail: genetics@zamnet.zm 

Abstract 

Soil acidity is the most limiting constraint in acid soils of the high rainfall area (Region III) of Zambia. These soils are 

not suitable for growing most arable crops including soybean (Glycine max L.) without adjustment of soil pH. Nampundwe 

mine tailings (NMT) are an abundant and readily available dolomitic limestone source in Nampundwe area, 60 

km west of Lusaka. The objective of this work was to evaluate the effectiveness of NMT in managing soil acidity for soybean 

production in Region III of Zambia in comparison with commercial agricultural lime (AgLi). Results for the 

1995/96 season show that AgLi significantly (P

1965). The tailings and AgLi were analyzed for their 

chemical characteristics using established methods 

(Page et al. 1982) as shown in Table 1. 

Plot size used was 0.5 m x 5 m and lhe treatments 

were arranged as a randomized complete block design 

with four replicates. The treatments were control, 

NMT and AgLi (reference). The quantity of 

lime applied was based on exchangeable aluminium 

and calculated using the formula by Kamprath 

(1967); where one t of lime = 2.0 x Exchangeable AI. 

The quantity of lime applied was 2.0 t ha-1of soil. 

Lime was applied to the soil surface by hand and 

incorporated into the soil with a hand hoe and rake 

during November 1995, before the onset of the 

rains. Soybean seed (Cv, SCI) was inoculated and 

planted with an initial 30 kg N ha·1as O-compound 

to boost initial crop growth. Four weeks after germination, 

the soybeans were thinned. At physiological 

maturity, the grain yield was assessed from a net 

plot area of 2.0 m 2 • The grain yield was adjusted to 

12.5% moisture content. After harvesting, soil samples 

were collected from all plots, air-dried and 

ground to pass through a 2 mm sieve and analyzed 

as above (Page et al 1982). The plots were maintained 

to study the residual effects of the liming material 

applied in the previous season. 

The trial was repeated in the 1996/97 season, with 

some modifications. A new location within the same 

experimental field, having the same soil type, was 

planted to determine the optimal liming rate for the 

study soil. Liming rates used were 0, 1.0,2.0, and 3.0 

t ha-1. All other management practices were similar 

to those used in the 1995/96 season except that the 

variety of soybean used was 'Santa Rosa' instead of 

'SCI'. Statistical analysis of the data was done using 

Proc GLM in SAS (SAS Institute, 1995). The·treatment 

means were compared -using the least significant 

difference method (Steel and Torrie, 1980). 

Results and Qiscu'ssion 

NMT did not significantly (P = 0.05) increase soil 

pH nor decrease exchangeable aluminium of the 

study soil in the first year of the study (1995/96 season) 

(Table 2). Comparatively, AgLi significantly 

increased soil pH (P > 0.05) and reduced exchangeable 

aluminium more than NMT and the control 

treatment. However, the increase in soil pH was still 

below the optimal soil pH range of 5.0 and 6.5 

(Mapiki 1997, unpublished) for most crops, including 

soybean. AgLi significantly (P < 0.05) contributed 

more exchangeable calcium to the study soil 

than NMT and the control treatment. The NMT did 

not significantly (P > 0.05) contribute exchangeable 

magnesium to the study soil during the first season, 

as expected. Application of calcitic limestone has 

been known to increase soil pH and exchangeable 

calcium more quickly than dolomitic limestone. Ananthanarayana 

and Hanumantharaju (1993) in their 

study on efficacy of different liming materials in 

neutralizing soil acidity reported that calcium oxide, 

calcium hydroxide and calcium carbonate increased 

soil pH and reduced soil acidity more quickly than 

dolomitic limestone. Particle size and neutralizing 

value are some of the factors that contribute to the 

reaction of lime in the soil (Coleman and Thomas, 

1967). AgLi has a 63 ).lm mesh size and higher neutralizing 

value than NMT. Therefore, AgLi was 

more soluble and reactive, increased soil pH and 

reduced exchangeable aluminium of the study soil. 

During the 1996/97 season (second year of the 

study), both NMT and AgLi significantly (P = 0.05) 

increased soil pH of the study soil over the control 

treatment. NMT significantly (P < 0.05) increased 

exchangeable magnesium over AgLi and the control 

treatment (Table 3). However, NMT gave significantly 

(P < 0.05) higher exchangeable magnesium 

than AgLi and the control treatment. This was expected 

because the magnesium content in NMT -is 

higher than in AgLi, whilst AgLi maintained its superiority 

in significantly (P < 0.05) increasing exchangeable 

calcium. NMT also significantly increased 

exchangeable calcium and soil pH. Soil pH 

rose to 4.8 after NMT, which is the establish~d critical 

pH value for soybean production for most soils 

in the Zambian high rainfall area (Munyinda 1984). 

Table 1. Chemical characteristics of NMT and Agli 

used in the study 

ELEMENT UNIT AgLi NMT 

Ca (%) 32 14.2 

Mg (%) 1.8 10.1 

Free cyanide (%) < DL

Table 3. Residual effects of liming sources on soil pH, 

exchangeable calcium and magnesium for Mufulira soil 

series during the 1996/97 season 

Treatment Soil pH (CaCh) Ca Mg 

(cmolc kg 1 ) 

Control 4.3b 0.45b O.lOb 

Nampundwe tailings 4.8a 0.82b 0.30a 

Agricultural lime 5.0a 1.3a 0.18b 

Means in columns followed by the same letter are not 

significantly different at P - 0.05. 

Both NMT and AgLi produced a linear increase in 

soil pH with increasing rates of liming (Figure 1). 

However for AgLi, no further increase in soil pH 

occurred beyond 2.0 t ha ot • The soil pH at the new 

site was raised more quickly than that at the old site 

established in 1995/96. It appears that management 

practices of the research field in which the trial site 

was located could be the reason for differences in 

the overall soil chemistry of the two locations belonging 

to the same study soil. The pH result for 

1996/97 strongly indicates that 2.0 t ha ot is the optimal 

liming rate for this soil. This confirms the finding 

by Munyinda (1984) that 2.0 t ha oJ was the optimal 

liming rate for the Mufulira soil series. 

The liming materials did not significantly increase 

soybean grain yield in the 1995/96 season (Figure 

2). However, the residual effect of both NMT and 

AgLi applied in the 1995/96 season did not significantly 

(P < 0.05) increase soybean grain yield. It appears 

that despite the increase in soil pH of Mufulira 

soil series from 4.3 to 5.0, and the increase in exchangeable 

calcium by . AgLi and exchangeable 

magnesium by NMT, the soil conditions for healthy 

growth of soybean were not attained. Mufulira soil 

series is dominated by the kaolinite type of clay 

mineral with a substantial amount of amorphous 

iron and aluminium oxides. Since tmder most conditions 

complete neutralization is not achieved 

when acid soils are limed, hydroxyl compounds of 

aluminium and iron could remain (Coleman and 

Thomas, 1967). There is a time lag between a change 

in soil pH and subsequent changes in the concentration 

of aluminium (Mtmyinda, 1984) before signifi- 

cant yields are realized. It follows that though .the 

soil pH of Mufulira soil series was increased from 

4.3 to 4.8, the alumirlium polycomplexes were not 

instantaneously. reduced to their chemical inert 

forms, which reduce slowly over several years. 

Thus, at a soil pH 4.8 that was attained by NMT it 

appears that monomeric exchangeable aluminium 

that remained tm-neutralized could have affected 

the soybean yield. Coleman and Thomas (1967) indicated 

that the products of complete neutralization 

that are attained at a pH higher than 8.3 aTe exchangeable 

calcium and magnesium and inert hydroxides 

of aluminium and iron. Therefore, to attain 

near or complete neutralization of soil acidity for 

soils of the Mufulira series requires a longer period 

of residual effect than the two-year period of this 

study. 

NMT increased soybean grain yield linearly with 

increasing liming rates, with the maximum yield of 

1.1 t ha ot obtained at3 t ha o 1Jime (Figure 3). The soybean 

grain yield obtained . is similar to yields obtained 

by Coma et al. (1990) at the liming rate of 2.0 

t ha ot on a related soil series. NMT significantly increased 

soybean grain yield over AgLi and the control 

treatment. The trend in soybean yield after the 

AgLi treatment was not consistent. At 2 t ha oJ AgLi 

produced a lower grain yield than NMT. The depressed 

soybean yield could have been due to the 

loss of AgLi through water erosion in two AgLi 

treated plots that had a slope. 

.E'" 

'"~ .., 

Qi 

'>' 

c 

'n; 

(j, 

c 

Q) 

'" 

.g, 

0 

en 

3000 

1500 

0 

01995116 D19H111 I 

Control Agric. lime Mine Tailings 

Treatment (tlha) 

Figure 2. Effects of liming sources on soybean grain yield for 

1995/96 and 1996/97 seasons 

6 

5.5 

5 

4.5 

4 

a 3.5 

'5 3 

en 2.5 

2 

1.5 

1 

0 .5 

o 

IB Agrie. LIme 111 Mine Tailings I 

control 2 3 

lime rates (tlha) 

Figure 1. Effects of liming rates on soil pH, planted in 1996/97 

season 

_ 1200 

i .. 1000 

.., 

800 a; 

'>' 

c 600 

i! 

0> 

iii 

400 

~ 200 

0 

r/l 

0 

control 2 3 

Lime rale (tlha) 

Figure 3. Effect of liming rates on soybean grain yield (Cv. Santa 

Rosa) for the new site planted in 1996/97 season 


207

The significant soybean yield could have been due 

to differences in management of the research fields 

where the trials were located. It appears the location 

used for the 1996/97 trial to test the rate effect of the 

two liming materials could have been managed so 

that soil fertility was improved and that influenced 

soybean yield. 

Conclusion and Recommendation 

The residual effect of NMT increased the soil pH 

and exchangeable magnesium of the study soil. 

However, the increase in soil pH and exchangeable 

magnesium did not increase soybean grain yield. It 

is recommended that a study of the residual effect 

be conducted for a longer time for NMT to completely 

react to conclusively evaluate its effectiveness 

for soybean production in acid soils of the high 

rainfall area of Zambia. 

Acknowledgement 

We acknowledge with many thanks the financial 

support from Zambia Consolidated Copper Mine 

(ZCCM), without that this work would not haVe 

been carried out. We thank the Government of the 

Republic of Zambia for logistical and administrative 

support given during the execution of this work. 

References 

Adams F., 1984. Crop response to liming in the 

southern United States. In: F. Adams (ed.) Soil 

Acidity and Liming. 2nd Ed. Agronomy 12:211-265. 

Day, P.R 1965. Particle fractionation and particle 

size analysis. In: c.A. Black et al., (Eds.) Methods 

of Soil Analysis. Part 1. Agronomy. American Society 

of Agronomy, Madison, WI, USA. 

De Oliveira E.L., and M.A. Pravan 1996. Control in 

soil acidity in no-tillage system for soybean production. 

Soil and Tillage Research 38:47-57. 

Goma H., Phiri S., and A. Mapiki 1990. Liming 

needs of some benchmark soil series of the high 

rainfall zone of Zambia. Soil Productivity Research 

Programme Annual Report. Misamfu Research 

Center, p.o. Box 410055, Kasama, Zambia. 

Gnmdon N.J. 1982. Acid soil amendments, soil 

chemistry and plant growth. Proceedings of the 

International Conference on Fertilizer Usage in the 

Tropics (Fertrop). Kuala Lumpur, Malaysia. 

Kamprath, E.J. 1967. Soil Acidity and response to 

Liming. International soil testing. North Carulina 

State Univ. Agric Exp. Station Tech. Bull. 4. Raleigh, 

NC, USA. 

SAS Institute, 1995. SAS User's Guide: Statistics. 

Version ed.> Statistical Analysis System Institute, 

Cary, New York, USA. 

Soil Taxonomy and Agrotechnological Transfer Report 

1985. Proceedings of the X1 U ' International 

Forum on Soil Taxonomy and Agrotechnology 

Transfer. Zambia. 

Soil Survey Staff 1992. Keys to Soil Taxonomy. 5 th 

Edition. SMSS Technical Monograph No. 19. Pocahontas 

Press, Inc, Blacksburg, Virginia, USA. 

556 pp. 

Steel RG.D., and J.H. Torrie 1980. Principles and 

Procedures of Statistics. McGraw-Hill Book Co. 

Inc., New York, USA. 

Mengel E.A., and K. Kirkby 1987. Liming and its 

calcium nutrition. 4 th Ed. International Potash 

Institute, Bern, Switzerland. 

MW1yinda K 1984. Relationship between pH 

(CaCh) and aluminium saturation. Proceedings 

of The Seminar on Soil Productivity in the High 

Rainfall Areas of Zambia, Lusaka, 8t11-10 th February 

1983. 

208 

Grain legumes and Green Manures for Soil Fertility in Southern Afries


Improving the Productivity of Grain Legumes and Green Manures 

To Obed I. Lungu and Kalaluka Munyinda 

Q: Was the reduction in yields after adding MAP 

not due to N2 fixation reduction by the presence of 

N in the MAP? 

A: No, all treatments received optimal N, and N2 

fixation was not being evaluated in this study. 

PAPR has no N, but these treatments received N 

from either urea or ammonium nitrate. 

Q: 

1) Were these trials conducted on station or on 

farm? 

2) If on farm, do farmers still get the benefits of 

PAPR if they plant late or weed less effectively? 

3) What is the cost of getting P20S per unit of 

nutrient by applying PAPR compared to DAP, 

considering the bulk of PAPR is inert material? 

A: 

1) The trials were researcher-designed and farmer 

managed. They were both on farm and on station. 

2) PAPR is just a substitute or alternative to MAP, 

TSP, etc. Farmers would therefore manage their 

crops with PAPR in the conventional way. 

3) At only 10% P20S in PAPR, the material is bulky. 

The product that will be promoted will be 

beneficiated by a physical process to at least 15% 

P20S as already demonstrated. 

Additionally, the utilization of PAPR is being 

promoted for areas close to the deposits of PR. 

Q: Should RP work still be considered a priority 

given that so much work has been done, but these 

materials are most often not available nor 

sufficiency reactive? 

A: Yes, PR work is still needed, but there should be 

a shift from basic research on-station to promotion 

and demonstration on-farm. Soils are acutely 

deficient in P, and adequate P application is in 

excess of 80 kg P20S ha·1, which is beyond what 

small farmers can afford especially since imported 

fertilizers in the region cost at least four times their 

cost outside Africa. The local product would be 

developed and made available to farmers if there 

was the political will and the policy and 

institutional support. 

To Atusaye Mwalwanda, et al. 

Q: I do not think the effect of S increase was 

studied, when looking at your treatments? 

A: The fertilizer source used was unfortunately a 


compound fertilizer 23:21:0:4S, hence it is difficult to 

isolate the individual effects of nutrients. However, 

the response to the inorganic fertilizer is an 

indication of the deficiency of the elements in the 

soils at the study sites. 

Q: In your first conclusion your attribute an increase 

in yield to P and S, but what is the economic return 

for the extra yield versus the cost of nutrients? 

A: Economic analysis was not done in this study. 

There is need to do that analysis to get the benefit 

from using inorganic fertilizer compared with not 

using. 

To Lackson K Phiri, et al. 

Q: Ca and Mg are usually leached into the sub-soil. 

In your assessment of the residual effect of the 

liming materials tested on soil Ca and Mg levels, 

what was the justification for tracing the two bases 

down to 20 cm soil depth only? 

A: I agree that there is the possibility of Ca and Mg 

leaching in a high rainfall envirorunent such as 

where the trial was located. But the trial was only in 

the second year so our interestwas to see what was 

happening in the rooting depth (0 - 20 cm) and later 

assess the leaching of these cations down the soil 

profile. Unfortunately funding ended prematurely. 

Q: Your results do not mention the effect of 

increasing lime material on available phosphorous, 

but only mention pH and exchangeable AP+ effects. 

Why did you not mention P availability since it is 

very much influenced by AP+? 

A: Certainly P is critical. Indeed in the middle of 

the trial in the 1995/96 season, P deficiency 

symptoms were observed such that an additional 20 

kg P20s/ha had to be added to correct P deficiency. 


C: How much yield benefit after a fallow is 

necessary for the system to yield more than with 

two years of cropping? The naive answer is "twice 

as much". But maize next year is not worth as much 

as maize in your hand - an effect economists call 

"discounting". This means the yield improvement 

must be more than two times for the fallow to be 

viable. On the other hand, the fallow may take less 

labour and may benefit long term-fertility or reduce 

weed populations, which implies that less than 

209

twice the yield is necessary. We have to consider 

these effects clearly and cost them. 

C: Both leaving land fallow with non-food legumes 

or growing maize after maize may'face a problem of 

adoption. There are agronomic management 

systems that minimize competition and improve the 

compatibility of components in intercropping 

systems. Growing legumes under early maturing 

maize may give better opportunities to minimize 

risk and increase profitability per unit land area 

over seasons. 

C: Almost all the papers presented on legumes 

(grain or green manure) assume that the crop 

residue is incorporated into the soil by farmers. But 

often livestock is a major component of the systems, 

and farmers give priority to their livestock rather 

than soil fertility when it comes to decision-making. 

We have to reconsider our work to integrate and 

evaluate it as it affects the whole production system. 

C: Various nutrient deficiencies occur in tropical 

soils and this may confound responses to the 

application of only a limited number of nutrients. 

In practice it is important to apply sufficient levels 

of all nutrients except the one being investigated to 

obtain predictable and explainable responses. 

There are many options with providing N, e.g. 

manures, N2 fixation, N fertilizer, if all other 

nutrients are adequate. 

C: Often in the presentations we hear the statement 

"the treatments were not significantly different, but 

they were different". If we are going to ignore the 

results of our statistical tests, why bother doing 

them? Also we should not accept a significant 

difference as meaning a treatment is viable - the 

question is, does it show a significant economic 

benefit? 

210 

Grain Legumes and Green Manures for Soil Fertility in Southern Afriea

Abstract 

EVALUATION AND PROMOTION OF VARIOUS CLASSES OF ANNUAL 

LEGUMES WITH FARMERS IN 'CHIOTA, ZIMBABWE 

DORAH MWENYE 

AREX, Marondera District, PO Box 150, ('V1arondera, Zimbabwe 

Nitrogen is one of the most limiting crop nutrients in crop production. It is important, therefore, to' identify and utilize 

all available sources of nitrogen, particularly those that are readily available and generally affordable by resource poor 

farmers. Green manure and cereal - legume rotations can be practiced to supply nitrogen. 

Green manures must be managed well to produce a significant fertilizing effect on the following crop. This is particularly 

important for the smallholder farmer who is sacrificing a food crop for one year through this practice. Smallholder 

farmers have practiced cereal - legume rotations for many years using legumes such as groundnut, bambara nut and 

field bean . Oespite this practice of crop rotation, the problem 'of low soil fertility has persisted resulting in low maize 

yield. Generally, the use of mineral fertilizers has declined over the years due to high input costs. Hence, the objective of 

this work was to introduce better N-fixing legumes to improve soil fertility. Velvet bean, sunnhemp and soyabean were 

introduced. 

Farmer participatory research methods were used for three years. Research and extension worked with identified farmer 

groups in a multidisciplinary approach. Th~ evaluation and promotion of technologies was carried out during farmer 

feedback sessions and field days. 

The major problems encountered in the promotion of these legumes included the unavailability of seed material locally, 

and lack of knowledge on the management aspects of the legumes to attain the optimum biomass. Legume cereal rotations 

were widely accepted, whilst green manures were accepted to a lesser extent. The results from the pilot project indicated 

that 29% of the farmers used the green manure and 57% the soyabean + cereal rotation. Maize yields from different 

sites increased by 15-70% for green manures and by 40-100% for the rotations. 

Farmers take up technologies within given domains so there is a need to come up with green manure or legume fertility 

packages for different farmers in their agro-ecological zones. An impact assessment would best indicate the results of this 

multidisciplinary approach. 

Key words: Annual legumes, farmer participatory research and extension, technology promotion, adoption potential, 

impact 

Introduction 

Nitrogen reserves in the soil are difficult to build 

due to its liability to leaching losses. It is important, 

therefore, to identify and utilize aU sources of nitrogen, 

particularly those that are readily available and 

generally affordable by resource poor farmers. 

Green manure and cereal legume rotations can be 

practiced as important ways to supply nitrogen. 

Green manures must be managed well to produce a 

significant fertilizing effect on the following 

(usually cereal) crop. This is particularly important 

for the smallholder farmer who is sacrificing a food 

crop for one year through this practice. Smallholder 

farmers have practiced cereal-legume rotations for 

many years, particularly focusing on grain legumes 

such as groundnut, bambara nut and edible bean. 

The area allocated to bambara nut is so insignificant 

that it can give little impact (AGRITEX, 1998-2002). 

Despite crop rotation, the problem of low soil fertility 

persists and maize yields continue to decline. 

Generally, the use of inorganic fertilizers has declined 

since the early 1990s, mainly due to high input 

costs. Hence there was need to introduce better 

performing legumes that have good nitrogen fixation 

abilities. For green manures, velvet bean 

(Mucuna spp.) and sunnhemp (Crotolaria juncea) 

were introduced. With cereal-legume rotations, soyabean 

was introduced as a new legume. 

A pilot project was set up with Soil Fert Net to 

evaluate Best Bet soil fertility technologies with 

farmers from Chiota Communal Area of Mashonaland 

East Province in Zimbabwe. The major goal 

of the project was to expose approximately 4000 


Table 1. Maize and Grain Legumes Production Statistics 

·-Results from the Chiota Pilot Pro·ect 

CROP 1998·1999 1999·2000 2000·2001 2001·2002 

Area Yield Area Yield Area' Vield Area Yield 

(ha) (t/ha) (ha) (t/ha) (ha) (t/ha) (ha) (t/ha) 

MAIZE 8784 0.5 7125 1.8 6973 2 7733 0.2 

GROUND NUT 546 0.5 439 0.6 399 0.8 420 0.1 

SOYABEAN 5 0 12.8 0.6 9 0.8 28 0.1 

EDIBLE BEAN 420 0.6 163 0.8 280 0.8 680 0.4 I 

Source ·[AREX· Fortnightly Crop and livestock Reports] 

farmers to the Best Bet soil fertility technologies in 

two years. Table 1 shows production statistics from 

the pilot area during 1998-2002. 

Data from Fortnightly Crop and Livestock Reports 

AREX (Division of Agriculture Research and Extension) 

indicate production levels of major cereals and 

legumes before the inception of the project in 1998 

1999, during the project and after the project in 

2001-2002. 

Methods Used 

Farmer participatory research 

Farmer participatory research methods were used. 

These methods call for a systematic dialogue between 

farmers, research and extension. In participatory 

research, scientists work with informants 

(farmers providing information) and experimenters 

(farmers who perform experiments and evaluations) 

(Bellon, 2001). The community usually identifies 

these farmers with the assistance of extension staff. 

The group extension method is one of the extension 

methods used. Farmer group members were selected 

based on the following criteria: 

- farmers' ability to grow a variety of crops 

- farmers' reputation and workmanship 

- Sex, age 

- Land holding. 

In 1998-1999, a Participatory Rural Appraisal was 

undertaken to find out about farmers' understanding 

of the soil fertility status in their areas. Research 

and extension facilitated the identification of suitable 

interventions from a list of existing technologies. 

During the project cycle, monitoring and 

evaluations were carried out through demonstrations, 

field days and farmer feedback sessions. Research, 

extension and farmers participated at all 

stages. 

Demonstrations and field days were also used as 

evaluation and promotion sessions. Demonstrations 

were research designed but farmer managed. Each 

host farmer served as a replicate of the experimental 

Figure 1. Layout of demonstration plots 

MAIZE/ MAIZE 

O.lha. 

MAIZE/LEGUME 

O.lha. 

unit. Each host farmer had a single plot measuring 

0.2 ha (Figure 1). 

Demo-plots of maize after a grain legume and 

maize after a green manure were compared with the 

farmer practice of planting maize after maize. In 

some cases more than one grain legume was established. 

Twenty-three sites of cereal - grain legume rotations 

and 10 sites Qf green manures were established. 

Soya bean, groundnut, bambara nut and cowpea 

were established in rotation as sole crops. Velvet 

bean and sunnhemp were established as either 

intercrops or as sole crops. Maize yield from the 

demo plots was compared. 

Field days were held at all established sites. The 

field days served as sessions for the sharing and exchange 

of ideas between farmers, research and extension. 

In some cases, farmer feedback sessions 

were arranged. 

Results 

Farmer participation 

The farmers who used at least one of the technolo 

gies after a year were considered to be adopters. 

These farmers were both from within the groups 

and outside the groups. Field days played an im 

portant role in the promotion of the technologies. In 

some cases, farmers used more than one of the tech 

nologies on offer. Adoption of the technologies also 

depended on the socio-economic status of the 

farmer. 

More farmers adopted the cereal-legume rotations 

(57%) compared with 29% of farmers that adopted 

the green manure technology (29%). 

Demonstration plots 

Several shortcomings occurred during implementa 

tion at some sites. The results from these plots were 

discarded. For example, plots with different pre 

establishment treatments were compared (unlimed 

plots were compared with limed plots). This in 

creased the number of factors, thus complicating the 

demos from the farmers' point of view. Variability 

in the results occurred due to different management 

abilities of the farmers and the competency of the 

extension agent, even though pre-planting demon 

strations had been held. 

212 


Table 2. Maize yields from the cereal· grain legume rotations, 

2000-2001. 

Maize/ Maize/soya Maize Maize Maize/ 

maize (t/ha) /groundnut /bambara Cowpea 

(t/ha) (t/ha) (t/ha) (t/ha) 

Site 1 3.3 5_8 5.4 

Site 3 2.6 4.2 4.1 

Site 3 4.0 4.5 4.3 

Site 4 4.3 4.6 4.5 

Site 5 2.0 4.0 3.8 

Site 6 1.5 3.0 2.7 

Site 7 2.0 3.5 2.5 

Site 8 1.6 3.0 

Average 3.2 4.7 4.2 

L-. 

Table 3. Maize yields from the green manure demonstration plots, 

2000-2001. 

Maize/maize Maize/velvet Maize/ 

(t/ha) bean sunnhemp 

(t/ha) 

(t/ha) 

Site 1 2.1 3.7 

Site 2 1.6 3.5 

Site 3 1.5 3.5 

Site 4 2.0 2.0 

Site 5 2.5 3.0 

Site 6 1.1 1.4 

Site 7 0.9 1.0 

Site 8 1.1 1.3 

Site 9 2.0 2.0 

Site 10 2.5 3.0 

Average 1.5 1.6 1.4 

Maize after a legume outperformed maize after 

maize at all sites (Table 2). With the green manure 

technology, farmers preferred sole cropping to in 

ter-cropping, mainly because of the constraints en 

countered during harvesting. The demonstrations 

were held for two years only and no arrangements 

were made for the third year because project fund 

ing had terminated. The data collected was from the 

sole cropping. 

The average grain yield of maize after maize was 

out-yielded by maize after a green manure (Table 

3). Results from Table 3 indicate that the yield in 

crease over the two years did not compensate for 

the yield lost during the first year. 

Farmer feedback sessions 

Attendance at field days was overwhelming. Atten 

dance ranged from 50 to 160 people at some sites. 

At least 14% and up to 60% of the targeted farmers 

used annual legumes (such as velvet bean and soya 

bean) as soil fertility interventions. However, farm 

ers cited the following setbacks (AGRlTEX, 2000; 

2001): 

Table 4. Participation of Chiota farmers in legume production, 2000 

2001. 

LEGUME TOTAL NO. OF AOOPTERS WITHIN AOOPTERS 

PARTICIPATING THE GROUP OUTSIDE 

FARMERS 

THE GROUP 

No. % No. 

ROTATION 1433 818 57 381 

GREEN MANURES 631 185 29.3 94 

- Lack of planting material. 

- Lack of knowledge on the utilization of legumes 

for human consumption and stock-feed. 

- Lack of knowledge on either uses of velvet bean 

and sunnhemp. 

- Lack of knowledge on the residual nutrient levels 

because of the rotation and green manure. 

- The concept of input reduction costs was not properly 

demonstrated. 

. 

- Generally, management of the green manure was 

poor. 

"Adopters" were considered to be those farmers 

who used the technology. The table above shows 

the total number of adopters over two years. The 

number is expected to rise through farmer-to 

farmer contacts. 

There was need to repeat the demonstration in the 

second year, but this could not be carried out due to 

unfavourable weather conditions and other socio 

economic circumstances. However, the following 

issues in management of green manures are to be 

considered for future demonstrations: 

• Sowing and site selection - Plant populations 

were low due to plant destruction by wild a:nimals. 

This resulted in very low biomass. 

• Fertilizer use - without fertilizing it is not possible 

to achieve a closed green stand and biomass 

quantities obtained will be low. Farmers considered 

that it was not practicable to fertilize fallows. 

• Incorporation - rarely was the green manure incorporated 

at the best stage. The method of incorporation 

was also not ideal because the green 

manure was not fully covered and the environmental 

conditions were not always good. 

• Seed procurement - Seed was not readily 

available locally. 

The pilot project built a sense of awareness amongst 

the farmers. An impact assessment will reveal the 

best steps forward. 


213

Recommendations and Conclusion 

In view of the setbacks given, other options can be 

tried out. These include: 

Biomass transfer - plant material can be transferred 

from its place of growth in other fields and be incorporated 

into the soil as green manure. 

Green manuring with roots - the green material can 

be used as fodder whilst roots remain in the soil. 

The benefit from this practice depends primarily on 

the quantity of root material remaining in the soil 

after harvest. In work done at Makoholi on sandy 

soils, maize yields obtained after sunnhemp tops 

were removed were similar to those obtained when 

everything was incorporated (Nyakanda, 1996). 

The farmer participatory research and extension 

methods used strengthened farmer research and 

extension linkages. This linkage is important in the 

promotion of technologies. However strong back-up 

from policy makers is required to ensure continued 

implementation of the technologies. There is also 

need to carry out an economic analysis of the results 

and give farmers some feedback. In most cases the 

researcher is interested in just the results and no 

feedback is normally given back to other stakeholders. 

On the other hand, extension normally uses 

narrative and qualitative analysis of data, which is 

not always ideal. 

Farmers are concerned about the dollar they have·in 

the pocket today rather than the three dollars they 

may have tomorrow, whereas benefits derived from 

rotation and green manures are derived 2-3 years 

later. Our recommendations should also consider 

farmers' circumstances. Farmers differ in their 

socia-economic status and they mainly take up technologies 

that suit their circumstances. Farmers require 

continued support as back up services from 

both research and extension if the promotion of. 

various annual legumes is to make an impact. Additionally 

a more sustainable approach should be considered. 

At present, all our efforts are left hanging 

or have been suspended. Now that farmers are 

aware of the technologies, what is the way forward? 

Despite the shortcomings, the pilot project in Chiota 

Communal Area played a pivotal role in implementing 

the farmer participat?ry research methods. 

The experiences can be adopted by extensionists 

and researchers working with resource-poor farmers. 

Promotion of annual legumes with farmers requires 

a participatory approach at all stages of the 

project cycle. 

Acknowledge~ents 

This work was supported by a grant from the 

Rockefeller Foundation to AGRlTEX in conjunction 

with the Soil Fert Net. Staff in the Soil Fert Net coordination 

unit provided technical backup. 

References 

AGRlTEX 2000. Chiota soil fertility project mid season 

and end of season report. 1999-2000. 

AGRITEX 2001. Chiota soil fertility project mid season 

and end of season report 2000-2001. 

AGRITEX 2002. Fortnightly crop and livestock reports 

1998-2002. 

Bellon, M.R. 2001. Participatory research methods 

for technology evaluation. A manual for scientists 

working with farmers. CIMMYT, Mexico, 

OF, Mexico. 93 pp. 

Nyakanda, C. 1996. Green manuring as a soil ameliorant 

in sandy soils of Zimbabwe-1996. 

214 


FINANCIAL AND RISK ANALYSIS TO ASSESS THE POTENTIAL 

ADOPTION OF GREEN MANURE TECHNOLOGY 

IN ZIMBABWE AND MALAWI 

MULUGETTA MEKURIA and SHEPHARD SIZIBA 

CIMMYT-Southern Africa Regional Office, PO Box MP 163, 

Mt Pleasant, Harare, Zimbabwe 

M. Mekuria@cgiar. org S. Siziba@cgiar. org 

Abstract 

Smallholder farmers in southern Africa face acute food insecurity as the productive capacity of their soils have declined. 

These resource poor farmers cannot afford to use inorganic fertilizers. Consequently, Soil Fertility Management and 

Policy Network researchers in Southern Africa engaged in developing best-bet soil fertility technologies have 

recommended green manure technologies as possible options to improve soil fertility to increase maize yield. 

In the sandy soil sites in Malawi and Zimbabwe, agronomic results showed that the incorporated mucuna biomass 

substantially improves maize yields. Mucuna was grown and the biomass incorporated in the first growing season and 

maize was then grown in the two subsequent seasons. 

The objectives of this paper are: (1) to verify the financial incentives to farmers for investing in mucuna as a green 

manure technology (as defined above) by quantifying its Net Present Value (NPV), (2) to assess the possible effects of 

likely changes in the cost and benefit elements of the technology on the NPVs and (3) to compute the risk that farmers 

might face in investing in mucuna (i.e. chances of getting a negative NPV). Analysis was done for cases where land is 

fallowed (land fallowing farmers) and where maize production is forgone in the first year for non-fallowing farmers, who 

would be prospective users of mucuna. 

Financial analysis reveals that under the current inputs and output prices there are positive payoffs to investing in 

mucuna for both Malawian'and Zimbabwean farmers. Considering uncertainty, land constrained farmers face more risk 

than' land adequate farmers in adopting mucuna. In both countries, the NPVs are sensitive to changes in the 

discounting rates, maize grain price and yield. This implies that policy interventions to decrease the discounting rate 

and/ or an increase in output price would create incentives to access credit to invest in soil fertility technologies and 

increase farmers' income, respectively. Research efforts to improve maize yield response to mucuna resitiue 

incorporation would greatly contribute to increased profitability and make the technology more attractive. 

Key words: Financial analysis, Net Present Value, green manure, technology adoption, policy 

Introduction 

The Problem Setting 

About 70% of the population of Southern Africa 

resides in rural areas where they subsist on small 

farms (Mabeza-Chimedza, 2000) . For many of these 

resource poor farmers, maize is the staple crop. 

Problems of food insecurity persist in the region, 

with the 2001-02 year being one of the worst in 

history. Over 13 million people are currently 

threatened with food shortages in Lesotho, Malawi, 

Mozambique, Swaziland, Zambia and Zimbabwe 

(SADC Food Security Bulletin July, 2002). 

Crop yields in Sub Saharan Africa have been 

declining steadily over the years. It is becoming 

increasingly acknowledged that low and declining 

soil fertility reduces maize yield even when normal 

rainfall is received (Kumwenda et al. 1996). The 

soils have become depleted of nutrients in many 

smallholder systems where nutrient outputs exceed 

nutrient inputs, leading to mining of soil nutrients 

(Buresh et al. 1998). Reduced access to inorganic 

fertilizer sources explains this depletion of soil 

nutrients from smallholder farming systems. 

Increased human population pressure has resulted 

in land scarcity and shortened fallow periods 

limiting the use and benefits from traditional soil 

fertility management practices like cattle manure, 

fallowing and slash and burn in replenishing soil 

nutrients. Reduced access to inorganic fertilizers 

derives from unfavourable grain / fertilizer price 

ratios, driven by poor infrastructure, and unsuitable 

input and product pricing policies, and uneven 

Grain legumes and Green Manures for Soil Fertility in Southern Afriea 215

performance of private sector companies (Mwangi, 

1997). 

Proposed Interventions 

Researchers have attempted to come up with 

alternative soil fertility management strategies that 

can arrest the decline in soil fertility in southern 

Africa. An inventory of soil fertility practices known 

as 'best bets" being suggested to farmers include 

among others, inorganic fertilizer regimes, lime use 

in acidic soils, improved animal manures, grain 

legumes for rotations, green manure legumes, 

agroforestry / trees in cropland, and biomass 

transfer systems. Each of these technologies has its 

advantages and limitations depending on the 

biophysical conditions in which it is applied and 

socio-economic attributes of the farmers. It is 

generally accepted that the extent and intensity of 

technology adoption is governed by how well 

technology attributes fit into farmers' farming 

systems, its profitability and contribution to reduce 

risk and income variability (Mekuria and Waddington, 

2002). 

Green manures look promising for such poor farmers. 

Green manure biomass is rich in N and therefore 

has potential of supplying N; the most commonly 

limiting nutrient to cereal production in both 

Zimbabwean and Malawian smallholder soils. Results 

from trials conducted in both Malawi 

(Kumwenda 1998; Sakala, 1998) and Zimbabwe 

(Muza 1998) concur in confirming that Mucuna prurims 

compared to other green manure species produces 

the highest amount of biomass. 

Mucuna pruriens is a herbaceous legume that has the 

ability to fix nitrogen from the air into the soil. A 

healthy mucuna plant can grow vigorously to produce 

10 t/ha of biomass and can leave over 100 kg 

N /ha in the soil. If it is to be used solely as a green 

manure, mucuna can be incorporated early when it 

is still green or late after flowering if the seeds are to 

be harvested. To achieve fertility benefits, mucuna 

biomass has to be incorporated and in the following 

seasons maize is grown. It is generally agreed that 

maize yield increases are obtained for a maximum 

of three years (S Waddington, 2002 personal communication) 

thereafter the biomass is almost totally 

degraded. For farmers to benefit from mucuna, they 

have to use it on fallowed land or have to forgo one 

season of maize production. Thus, the mucuna technology 

requires additional labour and some modest 

cash amounts for seed purchase in the first year. To 

some land constrained farmers, it means forgoing 

one season's harvest of maize. The decision to adopt 

mucuna can be difficqlt for poor smallholder farmers 

who are pressed with the need to produce the 

staple maize crop every season and at the same tiJ?e 

are faced with declining soil fertility and maize yields 

every year because of continuous cropping. 

Objectives of the Study 

The paper seeks to assess the potential returns and 

incentives for investing in mucuna as a soil -fertility 

management technology by smallholder farmers in 

Zimbabwe and Malawi. Additionally, the paper 

undertakes risk analysis to assess how some changes 

in the cost and benefit elements of this technology 

will influence the attractiveness and therefore likely 

investment in mucuna as a soil'fertility technology. It 

draws lessons for policy intervention and research 

strategies. 

Data Sources 

Maize yield data from on-farm trials generated for 

three seasons from 1997 to 1999 in Malawi (provided 

by Webster 'Sakala of DARTS Malawi and 

summarized in, Sakala et al. 2001) and Zimbabwe 

(provided by Tendai Gatsi and Lucia Muza of 

Agronomy Institute, DR&SS and summarized in 

Muza, 2002) (see Appendix 1) was used in the 

analysis. In Zimbabwe, the on-farm trials were 

conducted in two communal sites, Chihota and 

Zvimba, which are typical smallholder farming areas 

in the north central sub-humid rainfall belt with poor 

sandy soils. In Malawi, the on-farm trials were 

conducted in two sites in central Malawi representing 

sandy soils of the country. Mucuna was grown in the 

initial year and incorporated early in both Zimbabwe 

and Malawi. The maize yields for the subsequent two 

seasons were compared to a control of continuously 

cropped maize. Secondary sources of information 

were used for prices, costs and labour data. 

Growing of mucuna for incorporation of its biomass 

is a farm investment activity whereby effort and 

money are spent with the anticipation of a future 

stream of benefits as increased maize yields due to 

improved soil fertility. Before encouraging farmers to 

adopt mucuna, it is necessary to assess whether the 

incremental income from mucuna investment will be 

large enough to compensate them for the additional 

effort and risk they will incur. 

Analytical Procedures 

The study used financial cost-benefit analysis to 

determine the financial effects of using mucuna as a 

soil fertility technology on the smallholder farms. In 

addition, sensitivity analysis was employed to 

evaluate the risk associated with the technology. 

Financial Analysis 

For a sound financial analysis, it is important to 

properly identify costs and benefits of an investment 

activity (Gittinger, 1982). A 'with' and 'without' 

technology approach was used to capture the 

216 


incremental costs and benefits associated with 

mucuna as a green manure technology. 

The 'without' technology scenario was defined as 

growing maize continuously for three years without 

use of any soil fertility intervention, a practice that 

has become common among many poor 

smallholder farmers in Malawi and Zimbabwe. The 

'with' technology deals with a mucuna-maize-maize 

3-year crop rotation. 

The incremental costs for using mucuna are 

perceived to be different for farmers who are able to 

fallow some portions of their field and those who 

are not able to do so due to land constraint. For 

those who fallow land, the incremental costs are 

additional labour and seed costs associated with 

growing mucuna. For farmers who have to forgo 

maize production on the portion of their land 

planted to mucuna in the investment year, the 

incremental cost is the opportunity cost in terms of 

value of forgone maize production. Benefits were 

measured as the value of the incremental maize 

grain yields in the subsequent two seasons which 

was derived as the difference between continuously 

cropped maize yields and the maize yield after 

mucuna incorporation. To value the costs and 

benefits of mucuna, 1999 market prices of inputs 

and outputs in both countries were used (see 

Appendix 2). 

NPV 

The. Net Present Value (NPV) was used to quantify 

the net financial gains to farmers for investing in 

mucuna. NPV is the djfference between the sum of 

discounted benefits (incremental maize yields in 

two seasons) and the sum of discounted costs (cost 

of mucuna production/forgone maize grain 

harvest). NPV is an absolute measure of the present 

worth of an income stream accruing to the 

individual farmers; which is what the farmers are 

more interested in (Gittinger, 1982). 

The NPV was calculated per one hectare unit of 

land. If NPV is greater than zero it is worthwhile 

investing in mucuna and if it is less than zero it is 

not worthwhile investing in mucuna. The larger the 

value of NPV· the higher the financial returns to 

investing in mucuna. 

Sensitivity and Risk Analysis 

Sensitivity analysis shows to what extent the returns 

(NPV) to investing in mucuna is influenced by 

variations in the major quantifiable elemerits. This 

is important for identifying those elements whose 

changes have the largest impact on viability of the 

technology. Monitoring and management of such 

elements is critical in ensuring viability of the technology. 

Grain legumes and Green Manures for Soil Fertilitv in Southern Africa 

Risk analysis shows th~ probability out-comes af the 

NPV derived from the changes in its elements. The 

probability that the NPV would be negative can be 

used as an indieator of the degree of risk in adopting 

the technology. 

Both sensi tivity and risk analysis were perfomied 

using @RISK (a risk analysis software). @RISK uses 

distribution functions instead of constant values 

and can simulate 1000s of possible outcomes of 

NPV. Because of a lack of historic data, triangular 

distributions were used to estimate the distribution 

functions of all the parameters used in computing 

NPV (see Appendix 3). 

In sensitivity analysis the regression and correlation 

coefficients for the relationship between the changes 

in NPVs and changes in the cost and benefit elements 

are computed. The larger the regression coefficient 

of an NPV element · the more important it is 

in accounting for changes in the expected NPV. The 

correlation coefficient simply shows the nature of 

the relationship (positive or negative) between 

changes in expected NPV and the element. In fisk 

analysis, a cumulative distribution curve for the 

simulated NVP outcomes is generated. 

Results 

Financial Incentives 

For both farmers who fallow and those who can not 

fallow, the NPVs in both Zirrlbabwe and Malawi are 

positive, implying positive pays-offs .for investing in 

mucuna for the two types of farmers (Table 1). The 

NPV s are larger in Zimbabwe than in Malawi. This 

means that there is less additional returns for 

investing in mucuna in Malawi than in Zimbabwe, 

largely explained by the much higher maize yield 

responses to mucuna in Zimbabwe compared with 

Malawi (see Appendix 1). 

In Zimbabwe, investing into mucuna yields higher 

returns (NPV) to farmers who forgo maize 

production than for those who fallow their lands 

(Table 1). The difference between NPVs for the two 

types of farmer is accounted for by the difference in 

Table 1. NPVs hal for investing,in Mucuna as agreen manure 

(US Dollars) 

Zimbabwe 

Malawi 

(Discounting at 50%) (Discounting .at 34%) 

(in US$) Non· Fallowing Non· Fallowing 

Fallowing farmers fallowing farmers 

farmers 

farmers 

Total costs 37.11 57.46 56.64 50.76 

Total benefits 189.88 189.88 7a.ll 70.11 

NPV 152.77 132.42 13.48 19.35 

Source: calculated by authors from on·farm trial data 

217

the investment costs incurred. The investment cost 

.for farmers who forgo maize production 

(opportunity cost of the land in terms of forgone 

maize grain) is less than for those who fallow their 

lands (mucuna production cost). This is due to very 

low maize yields achieved on the soils that are 

degraded. 

In Malawi, the opportunity cost of forgoing maize 

production is more than the production cost (on 

labour and seed) of growing mucuna in the 

investment season. ' This explains why, unlike in 

Zimbabwe, investing in mucuna is more attractive 

(better NPY) for farmers who fallow than for 

farmers who have to forgo maize production. 

The fact that NPY is positive is not the only criterion 

or factor that farmers may cOf\Sider in their decision 

to adopt mucuna. The magnitude of the NPY and 

the riskiness of the technology are additional factors 

farmers may consider in adoption of the mucuna 

technology. 

Significance of NPV 

To assess the magnitude of the NPY and its 

significance to farmers, the maize grain equivalent 

value can be a simple benchmark indicator. In 

Zimbabwe the NPYs ha-] for both types of farmers 

are worth about 1.1 t of additional maize grain to 

the farmer's household over the mucuna-maizemaize 

rotation period (3 years) while in Malawi the 

NPYs ha-] are worth about 0.25 t of additional maize 

grain. Whether these additional pay-offs 'are 

worthwhile or not will vary from farmer to farmer 

depending on their various socio-economic 

characteristics. However, an additional tonne of 

For farmers who fallow land: 

#1 Maize Grain price (K/kg) 

#2 Year 1 yield increase (kg) 

#3 Discounting factor 

#4 Year 2 yield increase (kg) 

#5 Wage rate (K/day) 

#6 Mucuna labour (days). 

#7 Seed rate (Kg/ha) 

1#8 

Mucuna seed costs (K/kg) 

maize can be very meaningful for many farmers 

who are faced with food insecurity and limited 

choices. Note that the additional cash demands for 

labour and seed for investing in mucuna are very 

modest compared to input costs for growing maize. 

Sensitivity and Risk Assessment 

Farmers operate in an environment of risk and 

uncertainty. In reality the expected maize yield 

increase a'ssociated with mucuna, maize grain. price 

and labour costs are not \ fixed but subject to 

changes. This makes it necessary to subject NPYs to 

sensitivity analysis to take into account the 

uncertainties inherent in the elements of the NPYs. 

Results of risk analysis show that in general the 

technology is more risky to farmers who forego 

maize produ~tion than farmers who fallow in both 

countries (see Appendix 4). There is a 30% and 38% 

chance in Zimbabwe and Malawi, respectively, that 

the NPY is negative for farmers who forgo maize in 

the investment year. For farmers who fallow, 

chances that the NPY is negative are about 10% in 

both countries. The risk level of 30 and 38% of 

getting negative returns can be high to many 

smallholder farmers who are generally risk averse. 

This level of risk can be prohibitive to mucuna 

adoption by land constrained smallholder' farmers 

in Malawi and Zimbabwe who are pressed with the 

need to produce maize every season. Farmers who 

alreaady fallow land are therefore more likely to 

adopt mucuna as their investments in mucuna are 

less risky than for farmers who have to forego 

maize production. 

The sensitivity analysis reveals that changes in the 

benefit elements are more 

Table 2. Sensitivity of NPVs to changes in costs and benefit elements 

Malawi 

Zimbabwe 

important in determining 

the NPY than changes in 

For non· fallowing farmers 

the cost elements for 

Rank 

Element 

Regr Corr Rank 

Element Regr Corr fallowing farmers (Table 

#1 Forgone maize harvest (kg) ·0.67 ·0.72 #1 Forgone Maize harvest (kg) .0,60 .0.66 , 2). For farmers who fallow 

#2 Discounting factor 

0.40 0.49 #2 Year 1maize yield increase (kg) 0.55 0.57 land in Malawi, changes in 

the maize grain price have 

#3 Year 1 maize yield increase (kg) 0.36 0.42 #3 Year 2 maize yield increase (kg) 0.41 0.15 the most influence on 

#4 Year 2 maize yield increase (kg) 0.31 0.27 #4 Discounting Factor 

0.40 0.33 

#5 Maize Grain price (K/kg) 0.09 0.06 #5 Maize grain price (S/kg) 0.21 0.07 

0.64 0.52 #1 Discounting factor 0.64 0.47 

0.57 0.67 #2 Maize Grain price (S/kg) 0.57 0.42 

0.40 0.35 #3 Year 1 yield increase (kg) 0.45 0.37 

0.35 0.25 #4 Year 2 yield increase (kg) . 0.41 0.42 

·0:07 ·0.12 #5 Mucuna labour (hrs) 0.00 0.12 

0.00 ·0.06 #6 Land prep cost (S) 0.00 0.03 

0.00 0.08 #7 Wage rate (S/hr) 0.00 0.05 

0.00 ·0.06 #8 Seed rate (Kg/ha) 0.00 ·0.02 

#9 Mucuna seed costs (S/kg) 0.00 0.02 

NPY. The changes in year 

1 maize yield increment, 

discounting factor and 

year 2 maize yield 

increment are ranked in 

th,at order as the 

additional factors 

positively related to NPY. 

For farmers who fallow 

land in Zimbabwe, the 

discounting factor is the 

most important factor 

positively related to NPY 

218 

Grain Legumes and Green Manures· for Soil Fertility in Southern Africa

for mucuna. Other important factors are maize grain· 

price, Year 1 maize yield increment and year 2 maize 

yield increment, in that order of importance. 

For furmers who forgo maize production in the 

investment year, the most critical element in 

determining NPV is the maize yield forgone in both 

countries. This is a cost element and as expected is 

negatively related to NPV. This means that mucuna 

adoption can easily be more attractive where forgone 

maize yields are low and the converse is true. 

Holding other factors constant, mucuna would be 

more attractive on · those pieces of land where 

continuously cropped maize yields are already very 

low. For farmers who forgo maize production, maize 

grain price is the least important determinant of 

NPV. This means that an increase in the maize grain 

price does not increase the attractiveness of investing 

in mucuna for non-fallowing farmers as much as it 

does for farmers who fallow land. This is because an 

increase in maize grain price increases both the costs 

and benefits of investing in mucuna technology for 

non-fallowing farmers, hence minimizing the net 

effect on the NPV. 

Conclusions 

The pay-offs to investing in mucuna as a green 

manure in both Zimbabwe and Malawi were positive 

though modest in magnitude for both categories of 

smallholder farmers. After investing their labour and 

some modest amount of cash to buy mucuna seed~, 

farmers stand to gain a net present income worth an 

additional 1.1 t ha-1of maize over th~ 3-year mucunamaize-maize 

(investment-benefit-benefit) period in 

Zimbabwe and 0.25 t ha- 1 in Malawi. 

Although adoption of mucuna could generate higher 

returns (positive NPVs), it is necessary to look into 

the uncertainties inherent in the· NPV elemeiUS of 

mu.c.una technology, such as maize yield responses, 

prices and discounting rates. The study has shown 

that the mucuna technology is not free from risk. The 

risk of farmers encountering losses after investing in 

mucuna was substantial for the category of farmers 

who have to forgo one season of maize to grow 

mucuna. The chances for farmers to realize negative 

returns to their investment in mucuna were 

calculated to be 30% in Zimbabwe and 38% in 

Malawi. Mucuna has few other uses (the seed is not 

edible) and so has a low monetary value. The risk of 

negative returns can expose land-constrained farmers 

to increased food insecurity. These two aspects of 

mucuna technology could strongly deter its wide 

adoption by smallholder farmers, many of whom are 

land constrained and need to produce maize every 

season. 


The regression results from sensitivity analysis have 

shown that the maize grain forgone had the greatest 

influence on the expected NPVs for non-fallowing 

farmers in both Zimbabwe and Malawi. 

For such farmers, mucuna would give relatively 

better pay-offs on lands where maize yields are very 

low than where they are relatively high. In other 

words, mucuna pays off better for the nonfallowing 

farmers on those pieces of land where 

they sacrifice little grain by choosing to plant 

mucuna instead of maize. 

This implies that minimizing the amount of maize 

grain sacrificed by farmers in the first season would 

increase PCly-offs of mucuna to land constrained 

farmers. This calls for increased research efforts into 

ways of minimizing the amount of maize forgone in 

the first season for example by exploring intercrop 

arrangements of mucuna with maize. Research 

should also focus on improving the maize 'yield 

response to mucuna incorporation and the 

alternative end use possibilities for human 

consumption. 

Maize grain prices and discounting factors were 

ranked the most important determinants of 

expected NPVs for farmers who fallow in both 

countries. Policy instruments can be used to make 

these two economic parameters favorable for 

mucuna adoption. For example, in Zimbabwe 

where the cost of borrowing was very high, 

reducing the discount rate has the most significant 

effect in increasing attractiveness of investing in 

mucuna by farmers who fallow ·land. In Malawi, a 

policy measure to increase maize grain price would 

easily increase incentives for investing in mucuna as 

a soil fertility improving technology by those 

farmers who fallow land. In both countries, a 

combined effect of policy instruments to reduce the 

discount rate and to increase the grain price of 

maize would create more incentives for investing in 

mucuna as a soil fertility technology. 


The authors would like to acknowledge Webster 

Sakala of DARTS, Malawi and Lucia Muza of the 

Agronomy Institute, DR&SS for providing the 

agronomic data used in this paper. 

References 

Buresh, RJ. and K.E. Giller, 1998. Strategies to reple~ish 

soil fertility in African smallholder agriculture. 

In: Waddington, S.R et al. (eds.) Soil Fertility 

for Maize-Based Farming Systems in Malawi 

and Zimbabwe. Proceedings of the Soil Fert Net 

219

Results and Planning Workshop. Soil Fert Net 

and CIMMYf-Zimbabwe, Harare, Zimbabwe. 

p.13-19. 

Gittinger, J.P, 1982. Economic Analysis ofAgricultural 

Projects. Second Edition. Johns Hopkins University 

Press, Maryland, USA. 

Kumwenda, JD.T., S.R. Waddington, S.s. Snapp, R. 

B. Jones, and M.J. Blackie, 1996. Soil fertility 

management research for the maize cropping 

systems of smallholders in southern Africa: A 

review. NRG Paper 96-02. Mexico, D.F.: CIM 

MYT.36p. 

Mabeza-Chimedza, R., 2000. Agricultural production 

and food security situation in southern Africa. 

IDEAA Working paper 1/2000, Harare, 

Zimbabwe. 

Matabwa, c.J., and J. Wendt, 1993. Soil fertility 

management: present knowledge and prospects. 

In: (Mintha Ii et al. eds) Proceedings Conference on 

Agricultural Researchfor Development. p. 107-123. 

Mekuria, M. and S.R. Waddington. 2002. Initiatives 

to encourage farmer adoption of soil fertility 

technologies for maize based cropping systems in 

Southern Africa. In: Barrett, C.B., F. Place and 

Aboud, A.A. (eds) Natural Resource Management in 

African Agriculture: Understanding and Improving 

Current Practices. CABI, in AssoCiation with the 

International Centre for Research in Agroforestry 

(ICRAF), Wallingford, UK. 

Muza, L. 2002. Green Manure Use in Zimbabwe. 

Unpublished Report, Department of Research and 

Specialist Services, Harare, Zimbabwe. . 

Mwangi, W., 1997. Low use of fertilizers and low 

productivity in sub-Saharan Africa. Nutrient Cycling 

in Agroecosystems 47:135-147. 

SADC- 2002. Early Warning System Food Security 

Bulletin, July 2002, Harare Zimbabwe. 

Sakala, W.O., JD.T. Kumwenda, Alex R. Saka and 

Vernon H. Kabambe 2001. The potential of green 

manures to increase soil fertility and maize yields 

in Malawi. Soil Fert Net Network Research Results 

Working Paper Number 7. Harare, Zimbabwe 8 p. 

Waddington, S. 2002. Personal communication, 

CIMMYT Zimbabwe. 

Appendix 1. On·Farm Trial Maize Yields (kg/ha) after mucuna in 

Zimbabwe and Malawi 

Year 0 Year 1 Year 2 

Zimbabwe: (1996/7) (1997/8) (1998/9) 

Maize after maize 225 454.7 260 

Maize after mucuna 0 1400.3 2456 

Malawi: 

Maize after maize 951.5 828 1075 

Maize after mucuna 0 2178 1381 

Appendix 2. Rates, Prices and Costs used in the cost benefit 

analysis (US$) 

Zimbabwe: 

Maize Grain price ($/kg) 4.5, 0.12 

Mucuna labour (hrs) 107.2 

Land prep cost ($) 900 23.7 

Wage raie ($/hr) 6.75 0.18 

Seed rate (Kg/ha) 80 

Mucuna seed cos.ts ($/kg) 7 0.18 

Malawi: 

Maize grain price (K/kg) 5 0.06 

Mucuna seed cost (K/kg) 26 0.31 

Wage rate (K/day) 26 0.31 

Mucuna labour (labour days) 84 

Mucuna seed rate (kg/ha) 80 

Note in 1999 lUS$ -ZM$38 and lUS$ -MK84 

220 

Grain legumes and Green Manures for'Soil Fertility in Southern Africa

Appendix 3. Distribution Functions for uncertain Mucuna cost and benefit elements used in simulating NPV distributions for Zimbabwean 

and Malawian farmers 

Zimbabwe 

Malawi 

Parameter Distribution function Parameter Distribution funciion 

Benefits: 

Benefits: 

Year 1 maize yield increase (kg) Triang.(O 945 2500) Year 1 maize yield increase (kg) Triang.(O 13502500) 

Year 2 maize yield increase (kg) Triang.(O 2196 2500) Year 2 maize yield increase (kg) Triang.(O 305 2000) 

Discounting factor Triang.(O.5 0.666 1) Discounting.factor Triang.(0.5 0.746 1) 

Maize Grain price ($/kg) Triang.(3 4.5 10) Maize Grain price (K/kg) Triang.(3512) 

Costs: 

Mucuna labour (hrs) Triang.(105 107.18 110) Mucuna labour (days) Triang.(82 8486) 

land prep cost ($) Triang.(800 900 1500) Wage rate (K/day) Triang.(24 26 35) 

Wage rate ($/hr) Triang.(5 6.75 10) Seed rate (Kg/ha) Triang.(78 80 82) 

I Seed rate (Kg/ha) Triang.(78 80 82) Mucuna seed costs (K/kg) Triang.(20 27 30) 

I 

i 

Mucuna seed costs ($/kg) Triang.(5 79) Maize harvest forgone (kg) Triang.(800 951 .5 1500) 

Maize harvest forgone Triang.(300 3131000) 

Costs: 

Appendix 4. NPV Cumulative distribution curves in Zimbabwe and Malawi generated by @risk4.5 

Prob. 

0.6 

10 15 

0.6 

0.4 

0.2 

0 

·2 8 10 12 14 

Thousand ZM$ 

Non-fallowing farmers ZIMBABWE Fallowing farmers 

Prob. 0.6 0.6 

0.6 0.6 

0.4 0.4 

0.2 0.2 

o 

0 

·5 20 

·10 .0 0 15 20 25 

ThousandMK 

NPV 

NPV 

Fallowing farmers MALAWI Non-fallowing farmers 


A SOCIO-ECONOMIC ANALYSIS OF LEGUME· PRODUCTION MOTIVES 

AND PRODUCTIVITY VARIATIONS AMONG SM.ALLHOLDER FARMERS 

OF SHURUGWI COMMUNAL AREA, ZIMBABWE 

CHARLES NHEMACHENA1, HERBERT K MURWIRA2, KILIAN MUTIR0 2 

and PAULINE CHIVENGE 2 

1Department of Agricultural Economics and Ext£!nsion, 

University of Zimbabwe, Box MP167, Mt Pleasant, 

2 Tropical Soil Biology and Fertility Institute of CIA T, Department of Soil Science and 

Agricultural Engineering, University of Zimbabwe, Box MP228, 

Mt Pleasant, Harare, Zimbabwe 

Abstract' 

The impacts of poor soil fertility in Zimbabwe's communal farming systems have great implications on the food security 

and livelihoods of communal households. This study identifies opportunities for using legumes in replenishing soil 

fertility to improve agricultural production in the communal sedor through an assessment of social and economic 

factors that affect legume production. The study also identifies the economic potential of green manures on farm. 

Interviews with individual farmers and focus group discussions were conducted to establish perceived roles for legumes 

in soil fertility improvement. Data were also collected from the on farm trials. Analytical tools such as frequency 

analysis, regression analysis, descriptive analysis and cost benefit analysis were used to test proposed hypotheses. 

The motives for legume production were indicated to be food, cash and sometimes soil fertility improvement. It was also 

shown that the area under legume production, legume crop prices and labour availability are important factors affecting 

legume productivity. Legume production as indicated by the area cropped, yield, income and home consumption is very 

low. The constraints raised by farmers of limited cropping area, lack of markets, seed unavailability and lack of sufficient 

labour greatly contribute to the low status of legumes in the smallholder cropping system. The potential exists to 

intensify the use of legumes in the communal areas. The approach required to do this needs to be holistic and take into 

account their multiple use purposes, input and output markets, and promote new legumes. 

Key words: socio-economics, grain legume, motives for legume production, soil fertility 

Introduction 

Diminishing soil fertility remains the most limiting 

biophysical constraint to smallholder agricultural 

production in Zimbabwe. Increasing scarcity of 

locally derived nutrient sources and the changing 

socio-economic environment has rendered soil 

fertility improvement in smallholder farming 

systems in semi-arid and sub-humid Africa more 

difficult and complicated. External options for 

improving soil fertility have failed over the years 

because of inconsistency with the current 

circumstances of the farmers. 

The major sources of N available to farmers include 

animal manure, mineral fertilizers, woodland leaf 

litter and termitarium soil. Cattle manure, which is 

the commonly used source of organic fertilizer, is 

often limited in its supply by lack of cattle among 

farmers. Where available it is often of low quality 

due to the poor state of the rangelands and lack of 

adequate proteins in the animals' diet. Use of 

mineral fertilizers, especially ammonium nitrate 

(34.5% N), which is the major alternative source of 

N, is limited in the communal sector due to high 

costs, unavailability, risk and low returns to 

investment due to poor crop prices. Furthermore, 

the traditional sources of N, which include 

woodland leaf litter, have been depleted due to 

rapid population increases. There is an opportunity 

for nitrogen-fixing legumes to be used as cheap 

alternative sources of soil fertility improvement to 

help reverse the worsening poverty in these farming 

systems. 

Though traditional legume crops such as groundnut 

are widely grown in the smallholder farming 

system, areas planted and yields are very low. Thus, 

there is need for intensive promotion of these crops 

for them to be significant sources of N to enhance 

agricultural production. This study explores the 

motivation behind legume production among 

smallholder farmers, the important factors affecting 

legume productivity and the economic potential of 


green manures under the current farming system in 

Shurugwi, Zimbabwe. It was hypothesized that 

consumption requirements rather than reasons for 

income and soil fertility improVement motivate 

legume production. 

Research Methodology 

Data Collection 

The study drew on primary data obtained through a 

farm survey of 100 randomly selected households in 

Mfiri ward of Shurugwi. The survey was carried 

out in January and February 2002. All survey 

information was collected by interviews with 

individual farmers using a standard questionnaire. 

Prior to the field survey, a pre-test of the 

questionnaire was undertaken to improve the 

questionnaire design and enhance quality of 

responses obtained from the farmers; Discussions 

were held with a group of 15 farmers to establish 

constraints being faced by farmers in legume 

production, and identify opportunities for 

increasing the role of legumes in soil fertility 

management. 

Data Analysis 

The data were analyzed using the Statistical 

Package for Social Scientists (51'55). Cross 

tabulations were used to determine important 

factors affecting area under legume production. 

Frequency analysis and descriptive statistics were 

used to analyze motives for production, and 

regression analysis and descriptive statistics were 

used for analyzing important factors affecting 

legume productivity. Measures of project worth 

(Net Present Value and Internal Rate of Return) and 

gross margin analysis were used to identify the 

economic potential of green manure legumes. 


Comparative assessment of motives for legume 

production among smallholder farmers 

There are three motives for legume production; 

household consumption (100% of households), sales 

(78%) and soil fertility (12% of households). 

All households interviewed indicated that they 

grow legumes for household food requirements. 

This shows that farmers are aware of the potential 

role of legumes as a source of food for the family. 

Legumes can be promoted in the farming systems 

as alternative sources of protein in place of animal 

protein sources. Svubure et al (2000) also found that 

the primary reason for producing legumes in 

Wedza and Buhera was for household 

consumption, although yield levels are very low. 

Cash was the second reason for growing legumes, 

with 78% of the respondents citing it as the motive 

for growing legumes. The relatively high 

percentage of farmers conscious of the cashgenerating 

role of legumes indicates that promotion 

of legumes can be built on this role. This wiII need 

efficient marketing structures for the commonly 

grown legumes. Hildebrand (1996) also found that 

both input and output markets are very important if 

farmers are to increase legume production for the 

market. Thus if farmers are assured of good output 

markets for their legumes, they are likely to increase 

production of legumes for both consumption and 

for the market. Currently there is no formal market 

for selling legume prod ucts in the area and most of 

the produce is sold in the local market at very low 

prices. 

In Mfiri, farmers do not deliberately grow legumes 

for soil fertility improvement, though there is some 

appreciation of a residual benefit through retained 

residues. Only 12% of the responses indicated that 

they grow legumes for soil fertility improvement. 

This is probably due to a scarcity of land resources 

resulting in legumes being allocated small pieces of 

land in relation to maize, the major food and cash 

crop. Although farmers indicated that they do not 

grow legumes for soil fertility improvement, they 

are aware of some soil fertility benefits of growing 

legumes. Eighty-six percent of farmers were 

deliberately and consciously aware that legumes 

add nutrients in the soil. The farmers reported 

notable changes in crop growth on areas where 

legumes were previously grown, espeCially with 

maize. The fact that farmers are aware that legumes 

add nutrients in the soil suggests that they are likely 

to increase their use of legumes in. soil fertility 

improvement if appropriate extension messages are 

provided. 

Economic analysis of grain and green manure 

legume options for soil fertility improvement 

Tables 1 and 2 show the gross margin analysis and 

measures of project worth (NPV and IRR) for the 

green manures. 

5unnhemp and mucuna had negative overall 2-year 

gross margins per hectare using the official grain 

prices from the Grain Marketing Board (GMB), the 

sole buyer of maize in Zimbabwe. Cowpea had the 

highest positive overall 2-year benefit, Z$ 16 198, 

compared to other options, maize without fertility 

inputs (Z$3 003), crotalaria (Z$3 542), mucuna (Z$-2 

701), and sunnhemp (-Z$3 384). Gross margins at 

local prices of grain, Z$36.39/kg, were much higher 

and positive overall for all options. At a GMB price 

of Z$18.34/kg, the NPV values for mucuna, 

crotalaria, sunnhemp and maize without fertility 

224 


Table 1. Gross margins from the different treatments at the gazetted price for grain sold to GMB (Z$18.34/kg) and at local market prices 

(Z$36.39/kg) 

Treatment Gross margin per ha (Z$lha) Overa. 2 year benefits per ha (Z $) 

First Year (legume) 

Second Year (maize) At GMBprice At local price 

At GMB price At local price 

(Z$18.34) (Z$36.39) 

Maize after maize. no fertility 1501.46 (maize) 16123.42 

Cowpea followed by maize 4526.44 26414.44 

Mucuna followed by maize ·13320.56 ·13320.56 

C. grahamiana followed by maize ·13320.56 ·13320.56 

C. juncea followed by maize ·13320.56 ·13320.56 

At GMB price At local price 

(Z$18.34) (Z$36.39) 

1501.46 16123.42 3002.92 32246.84 ' 

11671.50 36303.68 

. 16197.94 62718.12 

10619.31 34215.84 ·2701.25 20895.28 

16862.29 46603.70 3541.73 33283.14 

9936.09 32860.14 -3384.47 19.539.58 

Table 2. Net Present Values for the different treatments at the 

opportunity cost of capital (20% interest) 

Treatment Net Present Value (NPV) Internal Rate of Return 

(lRR) % 

GMB price Local price GMB price Local price 

Maize after maize. ·1786.47 20552.63 7% 84% 

no fertility 

(maize) 

Cowpea followed 8344.93 43690.62 37% 204% 

by maize 

Mucuna followed ·7800.75 8585.72 22% 21% 

by maize 

C. grahamiana ·3465.35 17188.40 9% 40% 

followed by maize 

IC. juncea followed ·8275.21 7644.27 24% 19% 

by maize 

inputs were all negative, except for cowpea. At a 

discount rate of 20%, the IRR was significantly 

improved by selling grain on the local market 

where the price of grain was higher. Using the 

discount rate of 120%, which is the current inflation 

rate for Zimbabwe, only the cowpea option had a 

small positive NPV. 

Econometric analysis of factors affecting legume 

productivity across smallholder farmers 

A verage yields per hectare for the commonly grown 

legumes were compared to the staple maize. As 

shown in Table 3 for the commonly grown legumes, 

grain yield levels are very low (ranging from 18 kg/ 

ha to 164 kg/ha). Maize grain yields range from 464 

kg/ha to 550 kg/ha. Although average maize yields 

are higher than those of commonly grown legumes, 

the yield levels of all crops are generally low. This 

might be due to low soil fertility and consistent dry 

spells in the area, allied with lack of working capital 

Table 3. Average crop yields per hectare for the past three seasons 

Crop Average yield per hectare for Approximate area under 

past three seasons (kglha) croplhousehold (Mean 

household size - 3.2ha 

1999 2000 2001 

Groundnut 154 164 146 15% 

Bambaranut 28 34 31 2% 

Cowpea 18 19 20 Intercropped with maize 

Maize 464 550 510 75% 

Source: survey data 

to buy purchased inputs such as imptoved seed and 

chemicals to control pests and diseaseS. Working in 

Wedza and Buhera, Svubure et al. (2000) also found 

that yield levels for legumes were very low and 

cited low and erratic rainfall -and poor soil fertility 

as the major factors contributing to low yields. 

Analysis of important factors affecting legume 

productivity 

To determine the important factors that affect 

smallholder farmers' legume productivity, a simple 

regression equation was estimated from the survey 

data. Only factors affecting groundnut productivity 

were analyzed because it is the major legume grown 

by communal farmers, accounting for about 20% of 

the total arable land area. 

The following model was used: 

Yieldgt = po+plareat+p2gpricet+p3mpricet 

+p4amount of labotir+ E i . 

Where, Yieldgt= groundnut yield per hectare in 

a given year (kg/ha) 

areat=area under legume production in a given 

year (acres) 

gpricet =selling price groundnut in a given 

year ($/kg) 

mpricet=selling price of maize in a given year 

($/kg) 

labour = amount" of permanent labour to work 

in fields 

po =constant parameter 

PI,P2, P3, p4, = coefficients of the variables 

Ei =disturbance or etror term 

From the results, 58% of the total variation of the 

groundnut yield per hectare for the 2001 season was 

explained by the regressors included in the model 

as indicated by the adjusted R-square. This 

therefore implies that the remaining 42% of total 

variation was unaccounted for by the regressors, 

but by other factors not included in the model, 

perhaps by land shortage, seed unavailability and 

natural variability of production due to rainfall 

patterns. Land atea, groundnut selling prices and 

labour availability were important factors affecting 

groundnut productivity (Table 4). 


225

Table 4. Summary statistics for OlS Estimation for factors 

affecting groundnut productivity 

Variable Coefficient t·velue Significance 

- 

Constant ·28.458 ·1.014 0.313 

Area under groundnut production 0.395 5.434 0.000 

Selling price of groundnut 0.466 6.508 0.000 

Amount of permanent labour to 0.162 2.332 0.022 

work in fields 

Selling price of maize ·0.028 ·0.415 0.679 

R·Square adjusted - 0.58 F- 34.82 Significance- 0.000 

Source: Survey data 

Conclusion 

Consumption requirements of the household were 

shown to be the primary reason for producing 

legumes, followed by income and lastly soil fertility 

reasons. Farmers do not deliberately grow legumes 

for soil fertility improvement despite the fact that 

they are consciously aware of some soil fertility 

benefits from growing legumes. The potential for 

expanding legume production has not been realized 

due to a shortage of land resources and lack of 

knowledge, among other factors. Legume 

productivity was generally low in the area. Area 

under groundnut production, labour availability, 

and groundnut output prices are important factors 

affecting legume productivity in the smallholder 

farming systems. Cowpea, which is a multipurpose 

grain legume, was more attractive compared to 

green manures. This indicates that legumes with 

multiple uses have a higher adoption potential than' 

green manures. 

There is need for extension services, nongovernmental 

organizations and marketing 

agencies interested in the production of legumes to 

provide incentives and improve access to inputs (e. 

g. seed, fertilizers) to encourage farmers to increase 

land area under legume production. Benefits of 

legumes need to be explained to farmers, especially 

by health and community workers, as they can help 

contribute to fight against malnutrition reported in 

many communal areas of Zimbabwe. Increasing the 

availability of credit can enable farmers to purchase 

the needed inputs in legume production such as 

hiring labour and purchasing seeds. This could help 

relieve farmers from some major constraints such as 

labour shortages ' and so increase legume 

productivity that would benefit farmers through 

more consumption, farm incomes and soil fertility 

improvement through effective maize-legume 

rotations. 

References 

Hildebrand G.L. no date. The status of technologies 

used to achieve high groundnut yields in 

Zimbabwe, In: Achieving High Groundnut Yields: 

Proceedings of an International Workshop, 25-29 

August, Laixi City, Shandong, China (Renard c., 

Gowda c.L.L., Nigam S.N., and Johansen C. 

eds). Patancheru, Andhra Pradesh, India: 

International Crops Research Institute for Semi 

Arid Tropics. pp. 101-104. 

Mapfumo P., Campbell B.M., Mpepereki S. and 

Mafongoya P. 2001. Legumes in soil fertility 

management: The case of pigeon pea in 

smallholder farming systems of Zimbabwe: 

African Crop Science Journal 9 (4):629-644. 

Mazhangara E., Anandajayasekeram P., Mudhara 

M., Martella D. and Murata M. 1997. Impact 

assessment of groundnut research and the 

enabling environment in Zimbabwe: 1960-2000, 

SACCAR, Gaborone, Botswana. 

Morris RA. and Garrity D.P. 1993. Resource capture 

and utilization in intercropping: water. Field 

Crops Research 34:303-317. 

Mwenye D. and Kuwaza C. 2001. Chiota Soil 

Fertility Promotion Project, End of season report 

for the period 1999-2000. Agritex Marondera, 

Zimbabwe. 

Svubure 0., Mpepereki S., and Makonese F. 2000. 

Legume production and farmer awareness of 

Biological Nitrogen Fixation (BNF) technologies 

in Communal areas of Zimbabwe: Survey of 

Wedza and Buhera. Biotechnology: Vol. 4 No.5. 

Department of Soil Science and Agricultural 

Engineering, University of Zimbabwe, Harare, 

Zimbabwe. 

Truskott K. 1985. Socio-economic factors related to 

food production and consumption: a case study 

of twelve households in Wedza communal land. 

Agritex, Harare, Zimbabwe. 

226 


Abstract 

LINKING TECHNOLOGY DEVELOPMENT AND DISSEMINATION WITH 

MARKET COMPETITIVENESS: PIGEONPEA IN THE SEMI-ARID 

AREAS OF'MALAWI AND TANZANIA 

JOSEPH RUSIKE, GABRIELE LO MONACO and GEOFF M. HEINRICH 

/eR/SAT, Matopos Research Station, PO Box 776, Bu/awayo, Zimbabwe 

Legumes have long been grown in smallholder farming systems throughout Southern and Eastern Africa in intercrops 

and rotations with cereals. Legumes play an important role as food and cash crops, livestock feed, as a soil fertility 

amendment through biological Nrfixation (BNF) and for firewood, Because of recent increases in international and domestic 

prices of inorganic fertilizers, there has been more interest to expand legume plantings and management in 

smallholder areas especially in the semi-arid areas in order to provide a low-cost supply of nutrients, This paper uses 

the sub sector approach to explore two hypotheses, First that farmer uptake of pigeon pea-based technologies is driven by 

improvements in input and output markets. Second that linking technology development and uptake pathways with increasing 

competitiveness of pigeon pea products in international and domestic markets drives adoption of improved crop 

management practices, thereby enabling farmers to capture the potential soil fertility benefits of pigeonpea. The hypotheses 

are tested using farm survey and case study data from Malawi and Tan zania. 

The analysis shows that pigeonpea markets are now highly globalized and competitive. Pigeonpeas from Malawi and 

Tanzania are losing their competitiveness to pigeonpea from Myanmar and yellow pea substitutes from Canada and 

France. To increase the competitiveness of African pigeonpea and pull technologies through the system, crop variety 

improvement, choice of variety, seed distribution, production practices and more-efficient marketing arrangements need 

to be established targeting the needs and competitive patterns of specific identified markets. 

Key words: Pigeonpea-based technology, sub sector approach, competitiveness, uptake pathways, globalization 

Introduction 

Legumes have long been grown in smallholder 

farming systems throughout Southern and Eastern 

Africa in intercrops and rotations with cereals, Legumes 

play an important role as food and cash crops; 

they also provide livestock feed and firewood, and 

improve soil fertility through biological nitrogen 

fixation (BNF). Despite these multiple benefits, 

most households only allocate between 10 and 30 

percent of their total cropped area to legumes, 

mostly for subsistence food requirements 

(Rohrbach, 2001; Twomlow, 2001 ; Freeman, 2001; 

Semgal,2001), Farmers explain that legume cultivation 

is limited by seed and land shortages, lack of 

money to buy mputs, high labor requii-ements, lack 

of cash markets, pests and diseases, and low yields, 

Starting in the mid-1990s, prices of inorganic fertilizer 

escalated because national currencies depreciated 

and subsidies were removed under struchual 

adjustment programs. The escalation of inorganic 

fertilizer prices has forced farmers and scientists to 

look for cheaper substitutes, Researchers have hy" 

pothesized that because legumes provide a low-cost 

means of supplying N to the cropping system, increasing 

the area under legumes and improving legume 

residue management will enable smallholder 

farmers to reduce inorganic fertilizer use and still 

maintain soil fertility, 

Researchers have identified pigeonpea as the "best 

bet" legume for semi-arid areas because the crop 

has a deep root system that makes it drought tolerant. 

It mobilizes unavailable soil phosphorus; it has 

high nitrogen fixation; it adds organic residues 

through leaf fall; it recycles nutrients lost from the 

rooting zone; and .it is semi-perennial, which reduces 

yield and production risk (Nene, 1991; Singh, 

1991), Research investments by national programs, 

ICRlSAT, and other CGIAR centers have resulted in 

the development and release of superior varieties 

and better crop management options, including 

intercropping combinations, plant spacing, patterns 

and dates of planting, fertilizer management, and 

control of weeds, pests and diseases (Silim, 

Johansen, and Chauhan, 1991; Silim, 1992; Soko et al 

1995; Daudi andMakina, 1995; and Mbwaga, 1995), 

But adoption of these technologies remains limited, 

Surveys have shown that for farmers to adopt improved 

technologies and intensify legume production, 

collateral investments are needed to improve 

input and output markets, In addition, legume intensification 

needs to target poor households for 

food for home consumption and wealthier house- 


holds for cash income from the sale of surplus produce. 

This paper analyzes opportunitie#s and constraints 

for improving the adoption and impact of pigeonpea 

technologies by linking technology development 

and dissemination with increasing competitiveness 

of pigeonpea products in domestic and international 

markets. The study draws on pilot studies 

carried out by ICRISAT. with national programs, 

NGOs, and private·sector partners in Malawi and 

Tanzania. The studies on pigeonpea in Malawi and 

Tanzania also provide lessons that can be applied in 

other countries and on other legumes such as 

groundnut, cowpea, lablab, mucuna, common bean, 

and bambaranut. 

Objectives 

This paper aims to analyze the determinants of 

competitiveness of pigeonpea in international markets 

and identify strategic interventions that can improve 

competitive advantage (and thus technology 

adoption). The specific objectives are to: 

• Describe the pigeonpea subsectors in Malawi and 

Tanzania; 

• Identify opportunities and constraints for increasing 

competitiveness of pigeonpea in international 

markets, including targeting of investments; 

• Assess the profitability of adopting new pigeonpea 

technologies; and 

• Draw lessons for other legume sub-sectors and 

semi-arid areas in Southern Africa. 

Research Approach: Theory, Hypotheses 

and Methods 

The conceptual framework that guides this study is 

derived from business strategic management theory 

and agricultural commodity subsector analysis. 

Over the past two decades, these have emerged as 

useful tools for analyzing the performance of agricultural 

markets, diagnosing constraints and institutional 

innovations to resolve them and prioritizing 

potential interventions. 

Conceptual Framework 

The marketing of agricultural products is subject to 

the law of demand and supply. Falling trader barriers 

and globalization of agricultural markets have 

led to standardization of consumer preferences, and 

price formation forces now operate at international 

rather than national or regional levels. Competition 

has intensified, implying lower returns to investments 

by farmers, pro~essors and traders. 

Porter (1980; 1985; 1986; 1990) has developed a conceptual 

framework for identifying major determinants 

of competitiveness in globalized industries. 

Industries achieve and sustain competitive advantage 

through innovation, including. new technologies, 

new product design, new production processes, 

new marketing approaches, and new ways of 

trading. To gain competitive advantage, a firm 

must perform activities in its value added chain better 

than its competitors. Because activities at any 

stage depend on activities in the upstream and 

downstream stages, value-added chains of different 

firms interact at the different stages of technology 

development-input supply-production-distributionconsumption 

sequence. Therefore, firms need to 

coordinate and harmonize their activities at different 

stages of the vertical chain to match supply and 

demand throughout the vertical stages at prices 

consistent with the costs of production of least cost 

producers. To be globally competitive, an agricultural 

and food industry cannot be organized in an 

ad hoc way. Vertical coordination is imperative. 

To analyze vertical coordination of a whole sub sector, 

identify opportunities for improving economic 

performance, diagnose constraints and prescribe 

technological and institutional changes to resolve 

the constraints one can use the subsector approach 

(Shaffer, 1973; Marion et al 1986; Staatz, 1996). Sub 

sector analysis views effective demand as the pump 

that pulls goods and services, including new technologies 

such as cultivars, nutrients, and farm 

equipment innovations through the vertical system. 

Therefore, the approach emphasizes understanding 

the dynamics of how demand is changing at both 

the domestic and international levels (including the 

evolution of different niche markets) and the implications 

of that evolution for sub sector organization 

and performance. 

Hypotheses 

This study explores three hypotheses: 

1. Tanzania and Malawi can increase the domestic 

and international competitiveness of their pigeonpea 

sub-sectors by pursuing niche markets; 

linking quality characteristics required by buyers 

in premium markets with farmers' choice of varieties, 

crop production management practices, 

harvesting, and post-harvest handling during 

the physical movement to markets; improving 

marketing efficiency, and reducing costs of production 

and transportation. 

2. Investments to increase the competitiveness of 

Tanzania and Malawi's pigeonpea sub-sectors 

will generate high payoffs if targeted to promote 

the use of best varieties, extension of crop management 

advice, market information systems 

and reduction of transaction and transportation 

costs. 

228 


3. If households are linked to produce specifically 

for cash domestic and export markets then there 

is significant adoption of technologies, which 

permits farmers to capture soil fertility improving 

benefits. 

Methods 

In-depth interviews were conducted with selected 

participants - traders, processors, policy makers, 

and others - in order to obtain their subjective 

evaluations and perceptions of constraints and opportunities. 

Additional interviews were conducted 

with traders, processors and government officials in 

India, United Kingdom, Kenya, Malawi and Tanzania 

during the 2001/02 cropping seasons to generate 

data on quantity demanded, quality standards 

required by international buyers, and competition 

from alternative suppliers and alternative products 

(Lo Monaco, 2002.). Rapid reconnaissance surveys 

were cond ucted in Tanzania and Malawi to follow 

the flow of pigeon pea down the marketing chain 

from international buyers to farmers. During the 

reconnaissance surveys, informal interviews were 

conducted with farmers, extension agents, rural 

traders, NCO representatives, crop assemblers, 

transporters, and government officials. Trader, 

farmer and key informant interviews were combined 

with an analysis of quantity and price data, 

relative price relationships, and gross marketing 

margins. Quantity and price data ·were collected 

from secondary sources, including ministries of agriculture, 

national statistical offices, the Food and 

Agriculture Organization (FAO) database, and published 

and unpublished ·reports. 

Overview of the Pigeonpea Sub-sector in 

Malawi and Tanzania 

Pigeonpea is widely grown in the semi-arid areas of 

Malawi and Tanzania, mostly as an intercrop with 

maize, sorghum and pearl millet; but also in hedges 

around fields and on soil conservation barriers 

along contours. This makes it difficult to obtain accurate 

estimates of planted area, yield and production. 

National statistics indicate that in Malawi; pigeonpea 

is planted on 180,000 ha, yields are about 

600 kg per hectare and annual production is about 

100,000 t (Ministry of Agriculture and Irrigation, 

2001). In Tanzania about 815,000 ha are planted to 

pulses, including pigeonpea, yields average 800 kg 

per hectare, and production is about 635,500 t 

(Ministry of Agriculture and Food Security, 2002). 

But the FAO estimates are considerably lower 

(Table 1). Plantings in Malawi are concentrated in 

Blantyre, Machinga, and Shire Valley regions. In 

Tanzania, pigeonpea is mostly grown in Mtwara 

and Lindi in the southern coastal areas, Babati in the 


Table 1. Pigeonpea area and p.roduction in Kenya, Malawi and 

Tanzania, 1980 to 2001 

.Area ('ODD hal 

Production ('ODD tl 

1980·82 mean 1999·01 mean 1980·82 mean 1999·01 mean 

Kenya 66 147a 29 45a 

Malawi 127 ~ 23 85 79 

Tanzania 37 65 23 47 

Source: FAOSTAT, a. 1996·98 average 

north, and Kondoa in the central region. 

Management practices vary widely within and between 

regions. Wide differences exist in choice of 

variety, tillage, planting methods, intercropping, 

spacing, soil water and fertility management, weeding, 

pest and disease control, harvesting, and postharvest 

management. For example within the same 

agroecological zone in Kondoa, better resourceendowed 

farmers grow as much as 5 ha of pigeonpea 

intercropped with maize for export markets, 

using improved varieties and science-based management 

practices. Poor households grow a few 

plants in homestead gardens for home consumption 

using local varieties and traditional practices. Because 

farmers cultivatesmall plots, they often plant 

crop mixtures in the same field to maximize returns 

to land and labor. In the main pigeonpea producing 

areas, 58 percent of the maize area is a maizepigeonpea 

intercrop, particularly in areas where pigeonpea 

is a cash crop. 

In the major pigeon pea growing regions, 90 percent 

of farmers grow the crop and 70 percent of farmers 

are "commercial", selling over half their production 

(Orr, Jere, and Koloko, 1997). There is a long marketing 

chain, with many intermediaries. Households 

sell to vendors who buy from door to door, or 

transport the grain to village markets for sale to intermediaries. 

All transactions are in cash, and by 

volume (bucket), not weight. The intermediaries 

sell to other intermediaries who then sell to traders 

for transport to the major towns and sale to large 

exporters by weight. Traders do not pursue grades 

and quality standards. They believe the market is 

not mature enough for trading in graded form, and 

that farmers may not produce a marketable surplus 

if grades and standards are introduced. Exporters 

clean, grade, pack, and ship it to international buyers. 

In Tanzania there is no milling of pigeon pea; 

the grain is exported 'raw'. Traders estimated that 

annual exports currently average 30,000 to 35,000 t, 

almost double the official estimates (Table 2) . Some 

exports are shipped through Kenya. Likewise in 

Malawi, traders estimated that about 30,000 tare 

exported annually, although official estimates are 

lower (Table 3). Traders estimated that as much as 

35 percent of Malawi's exports is grown in Mozambique, 

although this share has been declining begin 

\ 

229

Table 2. Pigeonpea production and exports (t) in Tanzania, 

1993·97 

1993 1994 

Production 38,000 34,000 

1995 1996 

42,000 '55,000 

Exports 6934 17,633 3594 17,430 

Source: TCFB for exports; FAD for production data 

Table 3. Pigeonpea production and exports, Malawi 

1997 

41,000 

15,489 

1994 1995 1996 1997 1998 1999 

Production 43,311 52,601 87,880 72.67218,400b 80,000b 

Whole 1209 13852 1506 7877 

Processed 6394 7709 6552 9704 

Total Whole 10343 24865 10866 21740 18400 b 19600 b 

Exports equivalent' 

• Whole equivalent calculated assuming arecovery yield of 70% for dhill 

b Only aggregate data available 

Sources: FEWS and FAD 

1994·97: Bvumbwe Research Station; Patel, 1998 

1998·1999: Malawi National Statistical Office 

ning in 1999/2000 because of traders buying directly 

in that country. The bulk of Malawi's exports 

are milled into fur dhal, thus adding value, and 

shipped to India. 

Trader interviews revealed that both domestic and 

international markets are very volatile because there 

is negligible consumption of dry pigeonpea in domestic 

markets; pigeonpea is mostly exported. The 

international market is highly globalized and dominated 

by India, the major producer and consumer. 

Tables 4 and 5). Export demand depends largely on 

production in India -- demand and prices for African 

pigeonpea are high when there is a poor crop in 

India and Myanmar. Trader interviews also show 

that there is a marketing window for exports from 

Tanzania and Malawi, which opens around August 

to September and closes in October or November. 

Subsequently prices drop because the crops in India 

and Myanmar are harvested. This is an opportunistic 

market. Tanzanian and 'Malawian traders need. 

to discover prices, obtain confirmed orders with 

Table 4. World pigeonpea production ('000 t), 1980·98 

1980·82 1990·92 1996·98 1996·98 

1% share) 

India 1983 2432 2420 83.8 

Myanmar 29 47 159 5.5 

Africa 165 254 250 8.7 

Rest of the world 56 72 58 2.0 

Total 2805.4 2805 2887 100 

Source: FADSTAT. 2001 

Table 5. World dry pea imports ('000 t), 1995·99 

1995 1996 1997 1998 1999 

European Union 2522 3838 1530 1882 1890 

India 173 155 282 257 366 

Total 3603 4743 2538 2845 3016 

Source: FAD. 2001 

specified prices before they start buying from farmers, 

and then buy the crop, transport grain to export 

centers, clean, pack, and export it before prices in 

India start to fall. Before declaring prices to farmers, 

traders take into account bagging costs, transportation, 

handling, cleaning, port charges, freight, 

local levies, corporate tax, corruption and harassment 

charges, and financial costs. An increase in 

any of these cost elements is passed down to smallholder 

farmers because the farm level-derived supply 

is highly inelastic in the short run. Exporters are 

reluctant to hold inventory stocks because of the 

high price uncertainty of the Indian market. Because 

Indian traders have monopolistic market 

power and can drive prices down, forward contracting 

with farmers is difficult since exporters cannot 

assure farmers the contracted prices. 

Market Participants' Assessment of Opportunities 

and Constraints to Increasing 

Competitiveness 

Trader interviews, analysis of volumes traded, and 

international prices indicate mixed prospects for 

increasing the long-term competitiveness of African 

pigeonpea exports. Historically there has been a 

strong export market demand, but this market is 

shrinking due to increased competition from other 

exporters (notably Myanmar) and substitution of 

pigeonpea with yellow pea (exported by Canada). 

In the past five years, Myanmar has more than 

quadrupled exports to India, driving down wholesale 

prices (Table 6). There also has been a sharp 

increase of competing yellow pea exports from Canada 

(Table 7). Because yellow peas have been used 

in the past as animal feed, they are being exported 

to India, the Middle East, and North Africa at extremely 

low prices. Because of. increasing price 

competition, the prospects for producing pigeon pea 

as a cash crop for the export market are diminishing 

(Table 8). hldeed, Tanzania and Malawi have lost 

market share and farm gate prices have declined 

compared to three years ago, when exports were 

increasing, production in India was poor, and 

Myanmar was still not competitive (Table 9). Traders 

reported that opportunities exist for exporting to 

niche markets in Europe. But the volumes are small, 

about 1000 to 1500 t annually, and markets get 

Table 6. Myanmar Table 7. Annual exports of dry peas 

pigeonpea (I) exports to from Canada ('000 t), 1997·2001 

India, 1999·2001 

Canada to Asia 

1999 73,430 1997·98 395 

2000 185,964 1998·99 700 

2001 293,934 1999·00 638 

Source: Directorate of Economics 2000·01 850 

and Statistics. Ministry of 

Agriculture, India 

Source: Agriculture and Agri·Food Canada, FAO 

230 


Table 9. Average 

Table 8. Pigeonpea import prices, US 

pigeonpea prices (US$/t) 

$ per onne C.I•• . f M urn bai 19952001 

paid by marketing agents 

September October November at the first assembly 

1995 375 415 400 stage, Malawi and 

1996 315 320 295 Tanzania, 1998·2002 

1997 n.a. n.a. 445 Malawi Tanzania 

1998 450 410 395 1998 483 478 

1999 325 300 310 1999 431 288 

2000 300 n.a. n.a. 2000 336 248 

2001 295 275 250 2001 139 136 

Source: The Pulse Importers AssociatIOn 

2002 154 

quickly saturated. Another possibility is to supply 

pigeonpea as green vegetables to Europe. The companies 

surveyed did not have experience with these 

niche markets. To expand exports, there is a need to 

target particular niches and develop ways of reducing 

prices. 

Trader interviews revealed that the major determinants 

of competitiveness in international markets 

are consistent quality and quantity, price, and timeliness 

of delivery, especially for the August 

November window. Buyers look for grain color, 

size and milling characteristics, including ease of 

dehulling, shape, cleanness, and uniformity. White 

grains are preferred and fetch premium price~. B~bati 

White from northern Tanzania and whIte PIgeonpea 

varieties from Malawi have a unique taste 

that Asian and European customers like; and this 

explains why exporting firms are still surviving. In 

terms of grain size, market requirements vary. Indian 

millers prefer small to medium-grained varieties 

such as Babati White, while European millers 

require large-sized grains. Moreover, size requirements 

can change rapidly from large to small from 

one year to the next because of shifts in milling technology. 

Compared to Myanmar and India, Malawi 

and northern Tanzania produce better quality pigeonpea 

(Table 10). However, pigeonpea from central 

and southern Tanzania is mostly red color and 

poor quality because of insect damage. Infestation 

begins in the field, during the flowering stage. Insects 

are carried over into storage, and cause high 

losses. Quality standards are largely determined by 

traders who buy, grade, and sort grain for specific 

markets. For farmers to obtain a high-quality crop, 

Table 10. Grain quality traits relevant for the milling industry 

Africa Myanmar Yellow pea 

Grain size Medium to large Small to medium Large 

Grain shape Round Round Round 

Ease of dehulling Low Fair Very high 

Cleanness High Low High 

Weeviled grains Fair High Low 

Homogeneity High Low High 

Average yields % 65·70 65·75 90 

various issues need to be addressed, including correct 

choice of variety, seed delivery systems for getting 

pure seed to farmers, crop management, pest 

and disease management, harvesting methods, 

post-harvest management and handling during the 

various stages from farm gate through assembly, 

transportation, cleaning, 9rading and packing for 

export. 

Traders cited several major constraints affecting the 

pigeonpea sub sector in Malawi and Tanzania: 

• Low yield 

• Poor quality 

• Low farm gate prices 

• High transport costs 

• Lack of information 

• Attitudes towards traders 

• Lack of domestic markets 

• Inconsistent government policies 

Yield 

Because yields are low, grain cannotbe delivered at 

competitive prices. This is partly because farmers 

use recycled seed of traditional varieties (low 

yielding, and susceptible to Fusarium wilt) and use 

poor crop management practices. Also farm gate 

prices are not high enough to attract investment 

from other competing activities -- farmers often 

view pigeonpea as a "wild" crop and focus their in 

vestments on other cash crops. 

Quality 

Pigeonpea from central and southern Tanzania is of 

poor quality. The varieties are not white-seeded, 

crop management (especially pest and disease con 

trol) is poor, harvesting and post-harvest manage 

ment are poor. Weevil infestation is a major prob 

lem. 

Farm gate price 

Farmers receive a much lower price than the prices 

offered by exporters at the factory gate. This is be 

cause of the large number of intermediaries and in 

efficient trading mechanisms. For example, export 

ers believe that they offer prices as competitive as 

anywhere in the world. Farmers believe that the 

prices they receive are too low for pigeonpea to 

compete with alternative activities. During the 

2000/01 marketing season, exporters in Malawi 

were offeririg MK 10/kg at the factory gate, while 

farmers received not more than MK 5/kg at the 

farm gate. The first middleman was making MK 1/ 

kg and the other intermediaries were earning at 

least MK 2/kg. Traders interviewed for this study 

indicated that farmers are justified when they com 

plain that farm gate prices are low. 


231

Transport costs 

Competitiveness is eroded by high transport costs 

and the short time available to buy the crop, move it 

to export centers, clean, pack, and 'ship grain to the 

markets before the export window closes. Because 

of poor infrastructure and short timing there is a 

need to ship large quantities of pigeon pea to export 

centers at the same time that either commodities 

such as cashew nuts and tobacco are being transported. 

Transport costs are high because roads are 

bad (high vehicle .depreciation and operational 

costs) and because few transporters operate, and set 

monopolistic prices. For example, transporting pigeonpea 

from Babati to Dar es Salaam cost 42,000 

TSh/t, the same as shipping costs from Dar es Salaam 

to Mumbai. Transport from Tunduru to 

Mtwara takes 24 hours to travel 265 km and is more 

expensive than sending goods fro~ Dar es Salaam 

to Durban. It costs US$ 95/t to transport pigeonpea 

by road to South Africa from Malawi for transshipment 

to international markets. If· Nacala port in 

Mozambique worked, transport costs would be only 

US$23 / t. Traders reported that failure to deliver 

products in time results in renegotiation of contracts 

and heavy financial losses. 

Lack of information 

Industry, exporters and farmers often lack information 

on production and quanUty available for sale in 

different areas, prices offered and quality standards 

demanded by different buyers, and transport options. 

Because of the lack of a market information 

system, there is high price uncertainty, which 

makes it difficult for exporters to procure pigeonpea 

and discourages farmers from investing in pigeonpea 

production as they do not what prices they will 

get. Because of lack of information, farmers, middlemen 

and large traders engage in strategic bargaining, 

further increasing transaction costs. 

Attitudes towards traders 

There are negative attitudes towards intermediaries 

and political rhetoric against traders, many of 

whom are ethnic minorities. 

Lack of domestic markets 

Few local companies manufacture pigeonpea food 

products for the domestic market and there is little 

domestic consumption of processed pigeonpea food 

products. If exporters are unable to sell the crop in 

export markets, they incur heavy losses. 

Government policies 

Pigeonpea production and trade are hampered by 

inconsistent government policies, including licensing 

requirements for traders, road haulage, district 

local government levies and cess. The regulations 

create opportunities for corruption and harassment 

and increase transaction costs. For example, the 

Tanzanian government declared that levies and cess 

should not exceed 5 percent of the farm gate price 

but today district rural councils charge levies of 

more than 25 percent. This directly results in farmers 

being paid less. Farm gate prices are indirectly 

reduced because traders are required to have several 

licenses. For example, a trader requires 6 to 7 

licenses to deal in cashew nuts. Traders often need 

to visit district by district to obtain licenses because 

of excessive bureaucratic controls and regulations . . 

Despite these constraints, traders argued that there 

are high payoffs to investments in the pigeon pea 

sub-sector. For example, in Malawi, 15 years ago 

there were only two firms processing pigeonpea. 

Today over 10 firms process and export pigeonpea, 

and at least 15 ,firms export raw pigeonpea. 

Farm-level Opportunities and Constraints 

Farmer interviews revealed that opportunities exist 

for expanding the production of pigeonpea both as 

a food security crop and as a cash crop, targeting 

niche export markets. But increasing production for 

the market requires greater use of quality seed of 

the right varieties (i.e., varieties with traits in demand 

in specific markets), and better crop management 

in order to achieve grades and standards required 

by international buyers. ICRlSAT and 

NARS scientists have developed improved, shortand 

medium-duration varieties, with white bold 

grain. These varieties are suitable for cultivation by 

small-scale farmers aiming to service the August-to 

November export market to India. Both on-station 

and on-farm agronomic trials show that the yield 

gains from the improved pigeonpea varieties vary 

from 27 to 190 percent (Figure 1). The marginal rate 

of return from adoption of the varieties ranges from 

500 to 1000 percent, which far exceeds the 100 percent 

hurdle rate of return that is required for widespread 

adoption by smallholders. But the perform 

"0 

~ 

2000 

1500 

1000 

500 

o 

Treatrrent 

DOn-farm 2000/1 

.On-station:2001/2 

Figure 1. Performance of new pigeonpea varieties in on·station and 

on·farm trials, Oodoma, Tanzania, 2000/01·2001/02. 

232 


ance of sorghum-pigeonpea intercropping technology 

is highly variable depending on varieties, soil 

type, rainfall, and crop management practices including 

methods of land preparation, crop residues 

management, manure application, planting time 

and methods weed, pest control, harvesting and 

post-harvest handling. Productivity gains from 

short-season varieties under farmers' conditions 

have been limited, largely because of high insect 

damage as these cultivars flower during the rainy 

season, when pest populations are high. In addition, 

short-season varieties are unsuited to the traditional 

practice of intercropping. Medium-duration 

varieties have given much higher gains because 

they flower during the dry period when pest incidence 

is low and therefore escape insect damage. 

In collaboration with TechnoServe, a US-based 

NCO, grain samples of the new varieties have been 

sent for test marketing in Europe and India. Several 

varieties have been identified that are high yielding, 

have characteristics demanded in international markets, 

and offer productivity gains even when 

planted late and grown without intensive weeding. 

NARS scientists have also developed a range of 

crop management options, (intercropping, planting 

date, spacing and plant arrangement) designed to fit 

the different resource endowments, investment 

strategies, and risk management practices of different 

smallholders. Use of these management options 

along with the new varieties will enable farmers to 

produce grain that fetches a premium in international 

markets. 

Farmers identified opportunities to expand pigeonpea 

cultivation in semi-arid areas, including maizepigeon 

intercrops in areas with less rainfall risk, and 

sorghum-pigeonpea intercrops in areas of higher 

risk. Cross margin analysis reveals that pigeonpeamaize 

and pigeonpea sorghum intercrops are the 

most profitable among the major competing cropping 

activities in Tanzania and third most profitable 

in Malawi (Tables 11 and 12). This explains why in 

Kondoa district in Tanzania, pigeonpea is now the 

major cash crop, following an expansion of research 

and extension over the last five years. Farmers used 

to grow pigeonpea on a small scale; production expanded 

when they adopted the Kombowa variety, 

developed at Ilonga Research Station, and intercropping 

and spacing technologies developed by 

the Selian Agricultural Research Institute. Because 

of increased availability of white-seeded mediumsize 

Kombowa grain, traders came in from the 

neighboring Babati district, where pigeonpea was 

already highly commercialized. Farmers found 

they could earn high cash income from pigeonpea, 

dna eKfi'dl1u\~d fi'l\JdU"L....IUf., dl.\·l-dL......i-rg- ~eli ffl\Jl."e' 

traders. Farmers have become much more receptive 

to new technology, adopting crop management 

practices such as rotating the maize--pigeonpea 

intercrop with lab lab, using mucuna as a cover crop, 

and adopting the Magoye ripper to incorporate crop 

residues into the soil to increase fertility. 

Farmers interviewed for this study also believe that 

significant opportunities exist to increase household 

food security by expanding pigeonpea cultivation. 

Legumes are commonly eaten as relish, along' with 

cereals. Cowpea and beans are the traditional legume 

crops but in most semi-arid areas, farmers cannot 

produce beans successfully because of drought. 

Pigeonpea is a better alternative, but most households 

do not plant pigeonpea because they are unfamiliar 

with the crop and do not know how to utilize 

it. Some varieties are bitter when dry and difficult 

to cook. 

Farmers reported several constraints to expanded 

pigeonpea production, including poor farming implements 

which results in inadequate land preparation 

and late planting, poor access to seed of im~ 

proved varieties, non-availability of.chemicals for 

spraying, poor farming knowledge, lack of e~tension 

agents, pests and diseases, and lack of reliable 

organized markets. 

Technological, Institutional and Policy Innovations 

with Potential to Increase 

Competitiveness 

Traders suggested that to increase competitiveness 

on the international market, constraints must be resolved, 

and available opportunities exploited. This 

will require innovative approaches. 

Table 11. Profitability of principal crops in Kondoa District, Tanza· 

nia, 2001/02 (Tanzania Shilling) 

Maize + Finger Sesame Sun· Sorgo Pearl Maize 

Pig'npea millet flower hum millet 

Gross margin 123,731 53,672 48,626 16,519 7,289 6,948 6,379 

(Sh/ha) 

Breakeven 32 66 82 47 83 53 51 

price (Sh/kg) 

Breakeven 133 541 224 960 907 873 635 

yield (kg/ha) 

Table 12. Profitability of principal crops in Chisepo Extension 

Planning Area, Malawi, 2000/01 

Tobacco Ground· Maize + Soybean Bambara Maize 

nut Pig'npea 

Gross margin 34,906 10,771 6,495 4,769 4,337 (626) 

(Kwacha/ha) 

Breakeven price 21 15 5 11 18 11 

(Kwacha/kg) 

Breakeven yield 198 249 330 155 340 1,144 

I (kg/ha) 

I 


233

Traders suggest that the most important intervention 

would be to promote the use of improved, 

high-yielding varieties with traits demanded in target 

markets. This can be achieved 'by expanding 

investments in breeder, foundation and certified 

seed production. They also recommend targeting 

investments to improve productivity by providing 

advice on crop management, harvesting and postharvest 

sorting through village-level demonstrations 

and farmer training. Traders argued that if 

farmers increase productivity and production, unit 

costs of grain assembly and transportation will fall. 

Farmers recommended farm level training on pigeonpea 

production (how to improve yields and 

quality), sale of inputs in retail outlets within walking 

distance, loan of small seed packs, more government 

extension agents, farmer to farmer extension, 

and direct participation by smallholders in marketing. 

Other recommendations included: improve market 

efficiency by establishing a market information system. 

This will lead to price premiums for quality, 

reduce transportation and transaction costs, and improve 

technical and operational efficiency of buying 

in the villages and transporting to export centers. 

Because of the decline in cash-cropping opportunities 

(in turn due to declining export markets), respondents 

argued for promoting pigeonpea for food 

security by familiarizing and encouraging people to 

eat it. To increase domestic consumption they rec- ' 

ommended training of farmers in better cooking 

methods using, for example, the radio to reach more 

households. Respondents also recommended that 

local industries be encouraged to exparid collateral 

investments in new product and market development 

such as using pigeonpea for livestock feed. 

Finally, recommendations were made for more 

enlightened taxing and licensing policies, removal 

of legislative and administrative barriers to trading, 

and measures to correct market imperfections. 

Summary and Conclusions 

The prospects for pigeonpea in international markets 

are mixed. Historically pigeonpea cultivation 

expanded because of export-led growth. Pigeonpea 

markets are highly globalized and dominated by 

India. The export window for Tanzanian and Malawian 

pigeonpea is shrinking because of competing 

exports from Myanmar, and substitution of pigeonpea 

with yellow peas exported by Canada and 

France. Farm gate pri~es are falling, so farmers 

preferentially invest in other crops (and non-farm 

activities) rather than pigeonpea. To 'increase competitiveness, 

the pigeonpea sub-sectors in Tanzania 

and Malawi need to reduce prices and look at particular 

niches. Traders identified several opportunities 

for increasing competitiveness: including promoting 

the use of high-yielding varieties with traits 

demanded in niche markets; extending crop management 

advice that is linked to producing grain 

with quality characteristics required by buyers; setting 

up marketing arrangements to encourage a premium 

for quality production, and reducing transaction 

and transport costs. At the (arm level, opportunities 

lie in farmer training to improve yields and 

quality, sale of inputs at retail outlets within walking 

distance, loans for small packs of inputs, more 

government extension agents, farmer to farmer extension, 

and direct participation by smallholders in 

marketing. Because of the decline in opportunities 

for producing for export, respondents argued for 

promoting pigeonpea for food security by familiarizing 

and encouraging people to eat it. To increase 

domestic consumption respondents recommended 

training of farmers in better cooking me.thods using, 

for example, the radio to reach more households. 

These initiatives, together with improvements in 

policy, will improve adoption of pigeonpea technologies, 

and thus help smallholder farmers improve 

food security, income, and soil fertility. 

References 

Daudi, A.T., and D.W. Makina, 1995. Screening pigeonpea 

lines for resistance to root-knot nematodes. 

In: Improvement of Pigeonpea in Eastern and 

Southern Africa, Annual Research Planning 

Meeting 1994, 21-23 September 1994, . Nairobi, 

Kenya. Silim, S.N., King, S.B., and Tuwafe, S. 

(eds) Patancheru, Andhra Pradesh, India, International 

Crops Research Institute for the Semi 

Arid Tropics. pp 1-4. 

Freeman, A. 2001. Malawi baseline survey report. 

In: Improving Soil Management Options for Women 

Farmers in Malawi and Zimbabwe. Twomlow S.}. 

and Ncube, B. (eds) Bulawayo, Zimbabwe: International 



Mbwaga, A.M., 1995. Fusarium wilt screening in 

Tanzania. In: Improvement of Pigeonpea in Eastern 

and Southern Africa- Annual Research Planning 

Meeting 1994, 21-23 September 1994, Nairobi, 


(eds) Patancheru, Andhra Pradesh, India, International 



234 


Mligo, J.K., 1995. Pigeonpea breeding research in 

Tanzania. In: Improvement of Pigeonpea in Eastern 

and Southern Africa - Annual Research Planning 

Meeting 1994, 21-23 September 1994, Nairobi, 


(eds). Patancheru, Andhra Pradesh, India, International 



Nene, YL., 1991. Pigeonpea research: Future strategies 

in Africa. In: Proceedings of the First Eastern 

and Southern Africa Regional Legumes (Pigeonpea) 

Workshop, 25-27 June 1990, Nairobi, Kenya. 

Sigh, Laxman Silim, S.N., Ariyanayagam, RP. 

and Reddy, M.V. (eds) Nairobi, Kenya: Eastern 

Africa Regional Cereals and Legumes (EARCAL) 

Program, International Crops Research Institute 

for the Semi-Arid Tropics. pp 1-4. 

Rohrbach, D. 2001. Zimbabwe baseline: Crop management 

options and investment priorities in 

Tsholotsho. In: Improving Soil Management Options 

for Women Farmers in Malawi and Zimbabwe. 

Twomlow S.J. and Ncube, B. (eds) Bulawayo, 

Zimbabwe: International Crops Research Institute 


Semgal, 2001. Same Survey Report. ICRISAT: Bulawayo, 

Zimbabwe. 

Singh, Laxman, 1990. Overview of pigeonpea improvement 

research: Objectives, achievements 

and looking ahead in the African context. In: Proceedings 

of the First Eastern and Southern Africa Regional 

Legumes (Pigeonpeq) Workshop, 25-27 June 

1990, Nairobi, Kenya. Sigh, Laxman, Silim, S.N., 

Ariyanayagam, R.P. and Reddy, M.V. (eds) Nairobi,Kenya: 

Eastern Africa Regional Cereals and 

Legumes (EARCAL) Program, International 

Crops Research Institute for the Semi-Arid Tropics. 

pp 1-4. 

Silim, S.N., C. Johansen and y.s. Chauhan, 1991. 

Agronomy of traditiQnal and new cropping systems 

of pigeonpea and potential for Eastern Africa. 

In: Pr(Jceedings of the First Eastern and Southern 

Africa Regional Legumes (Pigeonpea) Workshop, 

25-27 June 1990, Nairobi, Kenya. Sigh, Laxman, 

Silim, S.N., Ariyanayagam, R.P. and Reddy, M. 

V. (eds), Nairobi, Kehya: Eastern Africa Regional 

Cereals and Legumes (EARCAL) Program, International 

Crops Research Institute for the Semi- . 


Silim, S.N. 1992. Traditional and alternative pigeonpea-based 

cropping systems. In: Pigeonpea in 

Eastern and Southern Africa: Summary Proceedings 

of the Launching Meetings for the African Development 

Bank/ICRISAT Collaborative Pigeonpea Project 

for Eastern and Southern Africa, 17-18 March 1992, 

Narobi, Kenya, and 30-31 March 1992, Lilongwe, 

Malawi. Eds Silim, S.N., Tuwafe, S., and McGaw, 

E.M. Patancheru, India: International Crops Research 

Institute for the Semi-Arid Tropics. 

Soko, H.N., A.A. Likoswe, S. Tuwafe, and T. 

Kapewa, 1995. Pigeonpea improvement in Malawi. 

In: Improvement of Pigeonpea in Eastern and 

Southern Africa- Annual Research Planning Meeting 

1994,21-23 September 1994, Nairobi, Kenya. 

Silim, S.N., King, S.B., and Tuwafe, S. (eds) pp 1 

4. Patancheru, Andhra Pradesh, Indi~, International 

Crops Research Institute for the Semi-Arid 

Tropics. 

Twomlow, S.J. 2001. Zimuto (Zimbabwe) baseline 

survey report. In: Improving Soil Management Options 

for Women Farmers in Malawi and Zimbabwe. 

Twomlow S.J. and Ncube, B. (eds) Bulawayo, 

Zimbabwe: International Crops Research Institute 




Promotion, Economics and Adoption of Annual Legumes 

To Dorah Mwenye 

Q: How much was your project trying to promote a 

presumably tested and trusted package, and how 

much wete you asking farmers to test technologies? 

A: The project tried and tested 1) rotations of 

cereals/legumes, 2) intercrops of grain legumes and 

green manures, and 3) evaluated two green manure 

legumes (sunnhemp + velvet bean). Farmers tested 

all the technologies and evaluated them, but only 

those technologies acceptable by them are being 

promoted. 

Q: 1) To what extent are long-term benefits 

associated with different legume technologies being 

explained in current promotion efforts? 

2) What is the minimum investment required to 

kick-start legumes on degraded (abandoned) fields? 

A: 1) The long benefits in the promotion of 

soyabean rotations are in the provision of food and 

reduction in input costs. Green manures will 

complement the meager rates of fertilizers applied 

by farmers to their maize crop. 

2) The minimum input investment required will be 

provided, as a result of the soil analysis to assess the 

initial fertility status of the field. Further research is 

needed in this area. 

Q: There is need to look more at management of 

green manures, including the time of planting on 

green manure performance. Late planting would 

also reduce legume performance rather than 

attributing poor performance solely to the sandy 

nature of the soils. 

A: Agreed, the general management of green 

manures is important. All the demonstrations were 

set up on limed plots. 

Q: What are the causes for the non-availability of 

information on green manures to extension and the 

farming community and what measures should be 

put in place to improve the situation? 

A: Information generated st1lllies in the hands of 

researchers. One of the limiting factors is the weak 

linkage between research and extension. The 

measures to be put in place to improve the situation 

are discussed at the end of the conference. 

Q : When you add the element of utilization in your 

project it will help to increase the adoption rate of 

the legume crops. 

A: Yes I agree. An impact assessment will best 

reveal the way forward. One of the issues to be 

considered in the next step is utilization. 

To Mulugetta Mekuria and Shephard Siziba . 

Q: Your comments on macroeconomic policy were 

all criticisms of governments in southern Africa. 

What do you think IS the effect of agricultural 

policies in the EU and the USA? 

A: Yes agricultural development has been adversely 

affected by poor policies in Southern Africa. For 

example, declining investments in extension, 

agricultural research. It is also true that protection 

and subsidy policies of the OEeD will affect ou·r 

competitiveness. Hence the debate on free world 

trade. 

Q: Discounting rate came out as having a large 

effect on NPV. How do you measure discounting 

rate, for different groups of farmers? 

A: The discounting rate used for both farmers is the 

same. It is the going interest (borrowing) rate used 

by a public bank. 

Q: In the medium term we can forget subsidies. The 

USA has launched the GOA initiative and regional 

trade hubs. 

A: I think you are right. 

Q: The presenters cited the issue of policy being 

wrong. It is now three or four years since the SFNet 

Economics and Policy Working Group (EPWG) was 

constituted. What is the impact of EPWG in terms of 

making the policy right? 

A: Its not easy for researchers to influence policy 

change - policies are not favorable still. 

Q: Looking at sensitivity analysis, what policy 

instruments would you advocate for concerning 

green manures, and are there major 

recommendations from this conference? 

A: Profitability of green manure technology is 

sensitive to prices and input costs. 

Recommendations are that relatively high output 

market prices of maize favour adoption of mucuna. 

A low interest rate will also facilitate adoption of 

mucuna by farmers. 


237

To Charles Nhemachena, et al. 

Q: 

1) What target yields (or limits) carrb~ 

recommended for grain and green manure legumes 

lillder the current economic tools of analysis in 

order to give a positive feedback to research? 

2) Most of the economic evaluation seems to be 

based on data sets that do not represent the optimal 

practices/experimental designs. Are we sure we 

are not dismissing orupgrading technologies 

prematurely? 

A: 

1) Due to great diversity of the biophysical and 

socio-economic environments, it is difficult to 

recommend specific yields for grain and green 

manure legumes. Depending on the available 

conditions in an area, yield levels should be high 

enough to cover the costs of production incurred 

and give positive net returns to farmers. Generally, 

from a given grain and green manure legume, yield 

levels should be high enough to offer positive net 

returns to factors of production used. 

2) Economic evaluations of any undertaking or 

enterprise make assumptions or goes in the form of 

an abstraction from reality to lillderstand 

relationships between certain variables, holding 

other conditions constant. Like any scientific 

experiment, it has controls. In addition, provisions 

are given for possible outcomes if other factors 

previously held constant come into play, for 

example in sensitivity analysis, so I don't think we 

are dismissing or upgrading technologies 

prematurely. 

C: There is need for longer-term rotational trials to 

have a quantitative idea about longer effects of 

legumes on soil fertility. 

To Joseph Rusike, et al. 

Q: .Can we learn from what happened in Myanmar 

where pigeonpea production has increased 

substantially in a few years time? 

A: Agricultural sector growth is being actively 

promoted by the Myanmar Government. 

Q: What is the potential size of the pigeonpea 

market in India and the world? 

A: Total world pigeon pea production from 1980-82 

to 1996-98 is as follows: (in 1000 t). 

1980-1982 2,805.4 

1990-1992 2,805 

1996-1998 2,887 

Q: Grain yields for pigeonpea from most 

presentations are around 200 kg/ha. Is the yield 

potential for pigeonpea any higher than this for 

Zimbabwe? How do our potential yields compare 

with yield levels achieved elsewhere in the world, 

in places such as India? 

Q: I too am interested in the pigeon pea yields in 

ideal conditions. Seeds are also a constraint. There 

are seeds coming through donations, coming 

through NGO programs, etc. Maybe the Soil Fert 

Net can help with pigeonpea mfuture. 

A: Here is the performance of new pigeonpea 

varieties in on-station trials at Hombolo and 

Makutupora Research Stations, Tanzania, 2001 

2002. 

Variety Days to 50% 100-seed Yield 

flowering weight (kg/hal 

Farmer 183 16.2 476 

ICEAPOO053 165 16.4 604 

ICEAPOO040 162 20.0 667 

ICEAPOO020 163 18.8 752 

ICEAPOO068 85 14.0 1530 

Q: The market for pigeonpea in India will be 

saturated if the whole of SA DC grows pigeonpea. 

We need to diversity markets to include domestic 

markets. 

A: Yes I agree. 

Q: How do we promote technologies to farmers? 

Are the technologies introduced to the farmers one 

after the other or simultaneously. 

C: We can ask, should the technologies be 

introduced at once or should we choose which 

technologies to expose to farmers? One of the 

approaches we are using in Zimbabwe is to cluster 

technologies in an area. We give each farmer one or 

two technologies and give other different 

technologies to other farmers in the area to try. 

Involve all farmers in the area on implementation, 

monitoring, evaluation of all the technologies and 

hold field days. Allow farmers later to choose the 

technology they want to adopt. 

238 

Grain legumes and Green Manures for Sail Fertility in Southern Africa

Synthesis and Working Group Reports 

Synthesis Reports 

Reporters: 

Webster Sakala, Steve Twomlow, Aggrey Agumya, 

Ishmael Pompi, Moses Mwale, Wezi Mhango, Paul 

Mapfumo, and Joseph Rusike 

Introductory Key Papers 

The papers in this session were very broad ranging. 

Important points included: 

Targeting and Niches: 

• Targeting is useful, through decision trees and 

whole farm models, into niches on farm. Niches 

were thought to be important because not all 

technologies perform best in all environments. 

For example, mucuna does not grow well in waterlogged 

conditions. Another example on 

proper utilization of niches is the use of long 

duration pigeonpea. These do well in a wellextended 

rainy season. The system is well 

adapted to southern parts of Malawi because of 

the extended Chiperoni rains after the main 

rains. 

• From this work we have learned and need to 

accept that some technologies may not be useful 

and adaptable, e.g. alley cropping compared 

with soyabean in some parts of Zimbabwe. 

• Process research is to be encouraged so that we 

can develop a clear understanding on some factors 

that limit technology adaptation to marginal 

niches. 

Multiple value of green manure: 

• The papers also indicated that fast and quick 

adoption of some green manures and grain legumes 

is facilitated by their multiple uses including 

as food, feed, cash and firewood. 

• Technologies with higher multip'l~ values are 

likely to be adopted. Examples were pigeonpea 

because of its. cash attractiveness, and soyabean 

(not mainly for soil fertility but) because of the 

cash incentive. 

Socio Economic Factors: 

• When new technologies are being tested in new 

areas there is need to further understand the 

existing farming household systems in the area. 

• Once the systems have been fully studied, some 

aspects of the cropping systems can be incorporated 

into the technology, to better fit the technology 

into the existing system. 

• Conduct cos,t-benefit analysis as a routine. 

Scaling Up: 

A range of questions emerged during the discus 

sion: 

• Who does it and who is best placed to do it? 

• When is the best time to measure technology 

and method adoption? 

• What influences adoption? 

Synchrony: 

• We need to look more at the practicality of 

matching nutrient release and nutrient demand 

from combined organics and inorganics. 

Seed Issues: 

• Who provides the starter seed with annual legumes? 

• If seed is of no cleJ.r economic value (as with 

some green manures), how then do we sustain 

its supply? 


Four papers were presented: 

Nitrogen fixation, grain yields and residual N 

benefits of promiscuous soyabean tomaize under 

smallholder field conditions. kasasa et al. 

Interaction of inoculum and liming on yield and N 

fixation by soyabean grown on sandy soil:- A case 

study of Murewa District. Nemasasi et al. 

Response of different cultivars of bean to inoculation 

and nitrogen fertilizer application. Sikombe et 

al. 

Role of phosphorus and mycorrhizal fungi on 

nodulation and shoot nitrogen content in groundnut, 

pigeonpea and lab lab bean. Besmer et al. 

Reasons for Quantifying N Fixation: 

• There is a need for an understanding of the relative 

contribution of N-fixing components to the 

N cycle within ecological conditions. 

• To understand the amount of N2 fixed by legumes 

for the development of an efficient agricultural 

production system. 

• Evaluation of the symbiotic effectiveness of 

rhizobial inoculants and success of inoculation, 

or the N fixing capabilities of legumege1Ol.Qtypes 

in plant breeding programs. 


239

• Assessment of the potential benefits from the 

_input of fixed N2, residual effects on subsequent 

crops following the growth of legumes or effects 

on crops associated withlegu!11es. 

Need for a Reference Crop: 

-The assumption is that both non-fixing (control) 

.and fixing crops take up N from the soil in the same 

ratio. 

-Wrong choice of a.reference crop can either underestimate 

or overestimate the N2 fixed. 

• Reference crop combinations include wheat 

(bean), maize (soyabean), and non-promiscuous 

soyabean variety (soya bean). 

Why do legumes fix nitrogen?: 

• "Conventional wisdom" has it that legumes fix 

N for themselves. 

• However data here showed thaf maize yields 

after soya bean increased even where stover was 

removed. This appeared due to root exudates 

and build-up in SOM. 

Inoculum Use: 

The following aspects are important: 

1. Isolate 

2. Selection and authentication (Testing) 

3. Production for specific locations 

The possibility exists that applying exotic strains of 

rhizobium indiscriminately can suppress indige~ 

nous strains. 

Effect of N source on Bean grain yield: 

• CIAT 899 and the local isolate were comparable 

in Mbala and Lundazi, Zambia. 

• In Pembela, the native rhizobia strains at the 

trial site were as effective as other strains in increasing 

grain yields. 

• Soil available P determines to a large degree the 

nodulation in groundnut in semi-arid parts of 

Zimbabwe. 

• Enhancing AMF activity of native fungi promotes 

nodulation and increases shoot N concentration 

in lab lab bean. 

• To enhance N fixation, can we explore combinations 

of P and Rhizobia? 

Screening of Annual Legumes for Adaptation 

and Use 

Presentations covered indigenous herbaceous legumes, 

timing of legum~ incorporation, performance 

of short duration pigeon pea, risk diversification 

through legumes and simulating maize response to 

mucuna. 

Three of the five papers were directly devoted to 

screening. The other two dealt with issues relevant 

to screening. Only the paper on indifallows dealt 

with bringing on stream a new family of legumesthe 

indigenous ones . 

Indifallows: 

• These appear to be self-regenerating and well 

adapted. 

• Among the benefits for N2fixation andfarmer 

adoption, some indifallow species demonstrated 

high levels of N2 fixation, and the 

"Gwezu smell approach" participatory method 

for identifying legumes works well. 

• There was some concern that these species may 

not withstand grazing. 

• Suggestions .for the indifallow work included 

extending the study to Matebeleland and measuring 

how the population of species varies with 

the duration of the fallow. It is likely to change 

markedly. 

Effect of time of legume incorporation on maize 

yield: 

• Yields from early or late incorporation were not 

statistically different. Late incorporation 

spreads labour demand. This is consistent with 

the earlier SoilFertNet study over three seasons. 

• Soil moisture content from late incorporation is 

higher. 

• Water use efficiency is an additional benefit to 

N2-fixation. Inclusion of water use in the study 

objectives was commended. 

• Mucuna and C. grahamiana are recommended as 

best bets for Zimbabwe communal lands. C. grahamitma 

adapts better to degraded soils due to 

its strong roots. 

• Examine the method of incorporation-it might 

be a constraint for expansion of plot sizes. 

Short duration pigeonpea in Matebeleland: 

• The short duration types performed better in 

clayey than in sandier soils. 

• Concern was raised about the low yields and 

competition by weeds. 

• Also there was concern about the need for 

spraying against pests. Explore the effect of not 

spraying on performance. 

• The recommended time of incorporation is after 

harvesting the grain. 

• We also need to see how to improve access to 

literature about past work on legumes in the 

region. 

240 

Grain legumes and Green Manures for Soil Fertility in Southern Afriea

Screening: 

• New options identified were "indifallows" and 

a broadening of the suitability range for pigeon 

pea in southern Africa. 

• All three screening studies reported were 

started recently in the 2000/01 and 2001/02 seasons, 

yet benefits usually accrue after 3-5 years. 

Computer simulation of benefits could augment 

the experiments. 

• There was a general view that weshould continue 

screening but that first we should establish 

what has already been done by a thorough 

literature review and employing tools such as 

the Legume Expert System. 

• Biotechnology offers prospects of improving 

our ability to screen thousands of species. 

Risk diversification and computer simulation: 

• Adoption of legumes depends on return on investment 

and risk characteristics. 

• Computer simulation is undertaken to explore 

long-term trends. 

• The results (recommendations) from the APSIM 

simulations are consistent with current farmer 

practices except for the use of kraal manure in 

drier areas. 

• The extremely low rates of fertilizer application 

in serni-arid areas (18 kg/ha) reflect farmer 

aversion to risk in these areas. 

• Non-market benefits (remaining in the soil) are 

captured in the yield through APSIM. Values 

not captured are considered not relevant to 

farmers. 

Simulating maize yield response: 

• Computer simulation is undertaken to overcome 

the short-term perspective of most expetimental 

trials. The simulations reported covered 

a 46-year period. 

• APSIM, unlike many modeling tools, considers 

carry-over effects of variables including soil N. 

• Long-term simulation of Mucuna indicates large 

potential benefits over the long term. Increases 

of 100-200% in maize yield were reported. 

These benefits are not captured in short-term 

experiments. 

• So far, the simulated maize yields are showing 

satisfactory agreement 'with field trials, but 

more needs to be done about legumes. Consultations 

are underway to validate legume grain 

yields. 

General comments: 

We should look at soil fertility more broadly and 

give greater attention to other benefits from leg 

umes besides N-fixation, e.g. soil physical proper 

ties, water use efficiency, weed suppression and till- 

age effects. For example, pigeon pea's noted contribution 

to SOM is probably traceable to deep rooting, 

while chickpea increases the availability of P. 

With respect to soil nutrients, whereas P and N are 

adequately discussed; more consideration should be 

given to other nutrients such as zinc. 

A general question was, where should we start the 

screening; from the plant side or the rhizobium side? 

It is important that improved legume varieties be 

developed. This means that we need to involve/ 

collaborate with breeders in the screening. We also 

need to be involved in their work so that the soil 

fertility improvement, water use, weed suppression 

and other traits that we would like to see enhanced 

in legumes are given some attention by breeders. 

Identification of Best Bet Legumes for 

On-Farm Performance as Grain Legumes, 

Intercrops, Rotations, and Green Manures 

Eight papers and one poster were presented. 

Key findings from the presentations and discussion: 

• From the evidence presented in the different 

papers, it is clear that yield responses were 

greater when crop residues and manures were 

ploughed in, i.e. incorporated, compared to · 

when they were left on the surface as a mulch. 

Consequently, we may begin to see some conflicts 

between soil fertility and conservation 

farming which advocates that crop residues be 

left on the soil surface as a mulch. 

• Intercropping vs. rotation issues. Data presented 

by the authors suggest that in the short 

term (two or three seasons) a cereal-green manure 

rotation is less productive than a cereal/ 

grain legume intercrop. This is supported by the 

fact that many farmers that have participated in 

trials are more willing to adopt an intercrop approach 

to soil fertility amendment than a green 

manure-cereal rotation, particularly when land 

is scarce. 

• We had little information presented on the 

intercropping characteristics of the different legume 

and cereal varieties currently available. 

More work needs to be done on that. 

• Although many of the studies reported in this 

session were conducted on farmers fields, very 

Grain legumes and Green Manures for SQil Fertility in Southern Africa 

241

few studies characterized the household assets 

of the host farmers. Consequently it is difficult 

to assess the reasons why some farmers may 

favour one technology and others another. 

• Few papers presented actually showed clear 

hypotheses for the experiments. 

• It is evident from the papers presented that host 

farmers are beginning to develop their own local 

taxonomies. These need to be catalogued to 

enable wider dissemination. 

Suggestions arising: 

• There is a need to collate information from different 

trials in the region into GIS databases to 

look at soil, climate and social interactions. 

• Ex ante market studies on legumes are required. 

This will meet a growing need to assess market 

demands for legumes before they are promoted 

in an area. 

• Need a synthesis study on results ~e have to 

date concerning the relative merits and benefits 

of intercropping and rotations. 

• Combinations of inorganics and organics need 

further attention. More and detailed studies are 

required on the synergistic effects of organiC 

and inorganic fertility amendments. At the 

same time, work is required to develop simple 

and transferable messages. 

• Detailed economic analyses of many of the interventions 

bring into question their appropriateness 

for smallholder farming systems. If research 

intends that the smallholder farmer is to 

benefit from their work, it is essential that research 

take on a greater participatory emphasis 

in problem identification, development and 

evaluation of interventions. 

legume residues and concluded t fertilizer N 

applications were necessary for sustained production 

(based on one abnormal year). 

2. N availability/dynamics in soil: 

• Mineral-N in soil does not correlate well with N 

recovery by maize from preceding legumes nor 

with maize yield response. 

• Mineral-N dynamics suggest that mineralized 

N is flushed through the soil profile before 

maize roots are present to extract it, leading to 

poor synchrony of N availability and N uptake 

by maize. 

3. N recovery from legumes (and fertilizer) by 

subsequent maize crops: 

• Measured using 15N techniques by Chikowo et 

al. 

• Net N inputs from legumes were < 10 kg/ha for 

soybean, pigeonpea and erotolaria but> 80 kg/ 

ha for mucuna. 

• N recovery was always < 36%; being least for 

mucuna (12%) and greater for legumes with 

small N inputs. Their high percent recovery 

possibly being due to their low total N input. 

• N recovery from fertilizer was 2x N recovery 

. from mucuna, which had similar inputs (95 and 

84 kg-N fha, respectively). 

Issues from the questions and discussion: 

• Economics of green manures: What marginal 

increment/ yield gain is necessary for farmers 

to take up the technology? 

• The multiple uses of green manures need 'to be 

considered in maize/green manure-grain legume 

systems; e.g., animals that graze residues. 

Legume Benefits on Maize Productivity 

and Soil Properties 

Main issues from the three presentations in this 

session were: 

1. Maize response to legumes in rotations and 

intercrops. 

2. N availability / dyna.mics in soil as affected by 

green manures and grain legumes in rotations 

and intercrops. 

3. N recovery from legumes (and fertilizer) by 

subsequent maize crops. 

1. Maize response to legumes in rotations and 

intercrops: 

• In two studies, legumes gave very large maize 

yield increases; by 2-3x the yields without fertilizer. 

• BUT one study found only weak responses to 

Improving the Productivity of Grain Legumes 

and Green Manures 

Highlight points from the papers: 

1. Agronomic effectiveness of phosphate rock 

products, mono-ammonium phosphate and 

lime on grain legume productivity in some 

Zambian soils (abed I. Lungu and Kalaluka 

Munyinda) 

• Partially acidulated phosphate rock (PAPR) 

maintained a high level of soil P than mono ammonium 

phosphate (MAP) 

• Lime increased P effectiveness and legume biomass 

productivity 

• Optimal P application rate for legumes was 80 

kg P20S per ha 

• Simply processed PAPR (acidulated with sulphuric 

acid) was agronomically as effective as 

242 


MAP and even more effective than MAP on 

acid soils 

• Was greater soil residual P with PAPR than 

with MAP. 

2. The effect of P and S on biomass productivity of 

grain legume crops and subsequent maize grain 

yields in Malawi (AB Mwalwanda, Spider Mughogho 

and Webster Sakala) 

• Dry matter yields ranged from 2 t per ha with 

no P and S to >6 t per ha with 20-40 kg P per ha 

and 4-8 kg per ha S 

• Yields were Pigeonpea

• Conduct more research to measure and value 

other benefits 

• Develop policy instruments (ptjce increases and 

decrease in interest rates) to support green manures. 

2. To raise the economic potential of Green Manure, 

we need to address: 

a) Market constraints 

• Low yields 

• Poor quality 

• Gap, factory vs. farm gate 

• Lack of information 

• Government policies 

• Local industrial use. 

b) Farm constraints 

• Poor fit into cropping system 

• Poor input market 

• Communication problems (poor marketing). 

3. Learn from socio economic analysis 

• Policy and development planning is vital 

• Assess the conditions under which cereal legume 

rotations and intercropping are most feasible 

in the smallholder sector 

Appreciate that cowpea appears the most attractive 

of all legumes to Zimbabwe farmers. 

Working Group Reports 

Biophysical Work 

244 

1. Benefits of the Technologies and the Work 

Completed 

• Legumes (grain legumes as well as green 

manures) increase soil fertility and productivity 

and therefore the grain yields of subsequent 

maize crops. 

• Green manures tended to give more consistent 

and substantive effects on subsequent maize 

yields than did grain legumes. Green manures 

can increase maize grain yields by as much as 

385%, or 2.5 - 3.0 t/ha on farmer's fields in 

Malawi. 

• The yield benefit of N from a legume can be due 

to an increase in below ground biomass, as 

appears to be the case with soyabean. 

• Improvements in maize productivity have 

important consequences for diversification and 

food security. 

• Various types of economic benefits accrue, 

including the reduction in amounts and costs of 

inorganic fertilizer ·inputs and extra grain and 

cash. 

• Weed suppression may result (and can be especially 

beneficial with Striga). There may be important 

labour savings for weeding, especially 

in high rainfall areas. 

• There is a potential benefit with improved soil 

physical properties (aggregates, infiltration, porosity, 

soil loss/surface runoff, water use efficiency). 

• Fodder resources may increase also. 

2. Gaps and Limits with Existing Work and 

Knowledge 

Reviews and Synthesis: 

• The current reviews presented for Malawi and 

Zimbabwe cover green manures but not the 

grain legumes. 

• A review of Zambian work highlights the need 

for widespread dissemination of green 

manures. 

• Not all the lessons from the comprehensive 

work presented from other regions of Africa, 

particularly West Africa, are directly 

transferable to Southern Africa. The 

environments (and socio-economics) are 

different. 

Soil Science: 

• There is inadequate measurement of soil. physical 

properties and related issues like reduced 

tillage. 

• Non-nitrogen benefits of legumes on below 

ground biomass, texture, Ca/Mg balance and P, 

cations, need more attention, as does SOM and 

C sequestration. 

• We need to pin down the fate of N in the system. 

What goes into the plant anq what elsewhere? 

This also involves nutrient balances and 

partitioning of N in pools. 

• Far more information is needed on mycorhiza/ 

P interactions for us to provide appropriate 

recommendations. 

• The evaluation of long-term benefits and sustainability 

aspects need attention. 

Legume Germplasm: 

• The genetic base of species/provenances we 

have worked with has been too narrow. We 

need to screen more of these. What approac.hes 

should be used and where should screening be 

done? 

• Link up with breeders more often. They need to 

work on issues that we are already working on 

and we need to help focus their work onto new 

useful traits. 

• Need to inoculate with rhizobium/mycorhizae 

(where, which legumes, and with what?). 

• Alternative uses of green manures, as seeds and 

firewood, need to be determined. 


Systems and Networks: 

There are gaps on: 

-4> . Sharing of information within the Soil Fert Net 

and other networks. 

-4> Soil-Crops-Livestock integration. 

-4> Management of legumes, including time of residue 

incorporation, method, time of planting in 

intercrops, and seed maintenance. 

-4> Establishing the agro-ecological niches for the 

species and varieties. 

Research Methods and Interpretations: 

• People should synthesize their data fully. 

• We must avoid generalization with no data or 

evidence. Do we know where we haven't done 

enough research? 

• There are big gaps on information flows between 

research-extension-farmers. 

3. Strategies and Work for the Future 

Research Synthesis: 

• Review the literature on existing uses of grain 

legumes and green manures in different countries, 

including indigenous and private sector 

knowledge. Link with other disciplines, e.g. 

food technology and animal nutritionists. 

• Review and synthesize information on the management 

of legumes. Identify gaps that can then 

be researched. Develop a database. 

Research: 

• We need further work to identify the 

biophysical, socio-economic, and cultural 

conditions where the legumes perform best. 

• There is need to explore ways that may 

maximize the benefit from green manures as 

intercrops or in rotations. 

• Establish long-term trials to monitor soil physical 

properties, and evaluate other long-term 

benefits of green manures and grain legumes. 

• Plant breeders need to breed for the farmers using 

their conditions, e.g. breed in soils with low 

soil fertility. We need to collect legume materials 

from the breeders and screen them in different 

conditions. 

• Screening should be done for new species and 

not endlessly repeated for the same ones 

already done. More screening of indigenous 

legumes is needed, as they seem to have a 

positive role to play in soil fertility 

improvement. 

• Further research is required on nutrients other 

than Nand P, and on C sequestration. 

• When screening for BNF, we have to be more 

systematic and take plant samples from 

different geographic areas and see if legumes 

are nodulating in different soils. 

• There is a great need for more interdisciplinary 

Soils-Crops-Livestock 'interaction research, e.g. 

on feed quality and manure quality. Howserious 

are the sustainability issues where the cycle 

is not balanced? 

• There is a clear need to re-visit the ideas (widely 

used during the first few years of Soil Fert Net) 

of developing common experimental protocols 

and establish multi-Iocational trials. 

• Management issues should be considered to 

improve the chances for green manures, e.g. 

avoid very late planting. 

• Negative results should also be reported so that 

similar experiments are not repeated in future, 

e.g. it is all right to say "Legumes did not 

suppress Striga". 

Technology Promotion: 

• Rules of thumb will be useful on the economic 

yield increase of maize and other cereals following 

various classes of legumes. Should it be a 

two fold, three fold or five fold yield increase 

that we need for it to be economic? 

• GIS and modelling should be used more often 

for scaling up purposes. 

• We need to do more on the processing and utilization 

of legumes, including linkages with food 

technologists to encourage local utilization. 

Networking and Capacity Building: 

• Improve network linkages. We need to keep 

trying to develop an effective network databank 

for members to know what is happening, what 

has been done, etc. 

• Soil Fert Net and others should look at conducting 

more training courses on new techniques. 

• Networks can help with the setting up of 

screening experiments. 

• Build capacity to undertake our research more 

effectively, including how to measure and 

assess the longer-term benefits of legumes. 

• Attach graduate students to work on different 

issues with Government research, not necessarily 

using network funds. 

• Members should link more effectively with 

other existing networks e.g. ANAFE. 


Socio-Economics, Policy and Technology Promotion 

Results 

Benefits 

Research gaps 

Research strategies 

Application 

of GIS for 

targeting 

and scaling 

up 

Efficiency and 

speed. It saves 

resources. 

Multiple 

biophysical and 

economic data 

can be collected. 

Capacity lacking - 

and capital. 

both human 

Training of research and extension staff on the use or 

interpretation of GIS data. 

Decision 

guides 

(Decision 

support 

systems, 

DSS) 

Technology 

promotion 

approaches 

Several 

promising 

grain 

legumes and 

green 

manures 

have been 

developed 

and 

identified,. 

but adoption 

is low 

Gives flexible 

options for 

choice of 

technologies and 

their 

management. 

Farmer 

empowennent. 

Enhanced 

linkages between 

research, 

extension, 

farmers and 

other role players 

in technology 

promotion. 

Improve soil 

fertili ty and 

increase food 

production and 

income. 

Available guides are incomplete 

and consider mainly biophysical 

traits. 

Inadequate farmer characterization. 

There is ample data in the region, 

waiting for synthesis. 

Priority should be given to gap 

filling and synthesis of regional 

knowledge. 

There is limited multi-disciplinary 

integration and initiative for true 

linkages among the various actors 

and stakeholders. 

Limited capacity of extension/ 

research to extend "best bet" 

options. 

Lack of solutions to emergency 

problem of land degradation. 

The synthesis and development of guides should 

consider and incorporate sodo-economic factors, policy 

issues and current production trends. 

DSS and qualitative indicators should be tested, 

validated, refined and identified to target socio 

economic and biophysical niches and over larger areas 

for the integration of legumes. 

DSS should be geared not only towards increasing 

productivity, but also suggest ways that could 

minimize or distribute risk, decrease opportunity costs, 

identify niches and respect social values. 

Identify appropriate technologies that could be scaled 

up immediately to be effective in the short and long 

term. 

Improve researcher-extension-fanner linkages. 

Increase the capacity of fanners to organize themselves 

to have bargaining power in markets. 

Seed unavailability is a major Seed bulking. 

constraint. 

Enhance extension efforts. 

Limited knowledge among Conduct comprehensive adoption and impact: studies. 

extension staff. 

Focus on adoption barriers, and give feedback to 

Poor / limited farmer participation researchers to modify and improve the technology. 

in research and scaling up process. Technology adoption could be enhanced if farmers 

Lack of communication media with participate in the research and scaling-up processes. 

fanners (bulletins, pamphlets, that To minimize poor adoption, encourage linkage 

are developed in local languages, between farmers, scientists, extension. Develop 

and in forms that are easy to ownership and follow up the problem. 

understand). 

Unfavorable 

policy 

environment 

Need for 

more 

economic 

analysis of 

grain 

legumes and 

green 

manure 

technologies 

None 

Incentive to 

produce. 

Concrete and convincing data 

seems to be absent. 

Policy makers are unaware or do 

not believe or understand what is 

coming out of the research centres. 

Lack of advocacy, and poor 

linkages between policy makers 

and researchers. 

Limited market information. 

Limited productivity and quality. 

No organized markets. 

Financial returns not attractive to 

farmers. 

Produce and present convincing infonnation to 

influence the .decision making of policy makers. 

Demonstrate benefits of technology to policy makers 

under different scenarios. 

Simplify our findings, create a forum and facilitate 

opportunities so as to create awareness among policy 

makers. Invite them to targeted meetings. 

Create a forum where fanners will have access to 

policy makers (e.g. invite farmers to meetings and 

encourage discussion with policy makers). 

Proper financial analysis. 

Undertake empirical studies to demonstrate the 

profitability and sustainability of technologies across 

regions, locations and times. 

Promotion of local utilization. 

Develop market studies and provide market 

information on both inputs and outputs. 

Crea te farmer groups for marketing purposes. 

246

Grain Legumes and Green Manures for Soil Fertility in ... - cimmyt

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?