Vol. 6(11), pp. 99-106, December, 2014
DOI: 10.5897/JHF2014.0365
Article Number: F9DF58549453
ISSN 2006-9782 ©2014
Copyright ©2014
Author(s) retain the copyright of this article
http://www.academicjournals.org/JHF
Journal of Horticulture and
Forestry
Full Length Research Paper
Prediction of Osyris lanceolata (Hochst. & Steud.) site
suitability using indicator plant species and edaphic
factors in humid highland and dry lowland forests
in Kenya
Mary Gathara1,2*, Paul Makenzi2, James Kimondo1 and Gabriel Muturi1
1
Kenya Forestry Research Institute, Nairobi 20412-00200, Kenya.
Environmental Science Department, Egerton University, Njoro 536-20115, Kenya.
2
Received 8 August, 2014, Accepted 20 November, 2014
Osyris lanceolata (African Sandalwood) belongs to the family Santalaceae that hosts some of the most
valuable species for perfumery oil extraction. In India and Australia, Santalum album and Santalum
spicatum are well developed for perfumery oil extraction through establishment of commercial
plantations. In Africa, O. lanceolata has attracted significant attention as potential perfumery oils
extraction species. However, African Sandalwood exploitation is through unsustainable smuggling
from natural forests and woodlands. Since sustainable production of O. lanceolata oils is only feasible
through establishment of commercial plantations, there is need to understand ecological requirements
of the species before the remaining natural stands disappear. The aim of this study was to determine
plant species and edaphic factors that can predict African Sandalwood site suitability for domestication
programs. Sample plots with and without O. lanceolata were selected from natural stands in a humid
highland forest and a dry lowland forest, vegetation sampled using nested-intensity plots and soils
sampled in the plots simultaneously. Vegetation data was recorded according to species abundance.
Soil samples were analyzed for nutrients, texture and moisture retention. Canonical Correspondence
Analysis using CANOCO software was used to determine species association and relationship between
species to soil variables. In the highland forest, O. lanceolata clustered with Rhus natalensis and six
other species, and was correlated to soil nitrogen, moisture and clay. In lowland forest, O. lanceolata
clustered with R. natalensis and Hypoestes forskahlii but did not correlate with any of the soil variables.
The clustering of African Sandalwood with R. natalensis in both forest types suggests strong predictive
capacity of R. natalensis for O. lanceolata site suitability in humid and dry areas. Inconsistence of O.
lanceolata relationship with soil variables in the two study sites provides opportunity for further studies
in different soil types.
Key words: CANOCO, domestication, edaphic, hemi-parasites, species association, African Sandalwood.
INTRODUCTION
Osyris lanceolata (African Sandalwood) is an evergreen
hemi-parasite that belongs to the family Santalaceae
(Maundu and Tengnas, 2005, Irving and Cameron,
2009). The family hosts culturally and commercially
100
J. Hortic. For.
important species that have long been used for herbal
medicine, religion and perfumery oil industry
(Tshisikhawe et al., 2012, Subasinghe, 2013). Species
such as Santalum album and Santalum spicatum have
long been exploited for perfumery oil and are now more
developed commercially with plantations of S. album
showing an increasing trend in Australia, China, India, Fiji
and Sri Lanka (Subasinghe, 2013). In recent past, trade
in African Sandalwood oil has also increased because of
ready markets in Asia and Europe (CITES Cop 16).
However, trade in African Sandalwood is unsustainable
because materials are smuggled from natural stands and
without clear domestication programs (Mukonyi et al.,
2011). Moreover, exploitation of the species for herbal
medicine has also increased (Tshisikhawe, 2012) leading
to its decline in natural stands (Githae et al., 2011).
Arising from this concern, African Sandalwood is now
listed as threatened species under USF and WS (2013).
Since sustainable production of O. lanceolata oils is only
feasible through establishment of commercial plantations,
there is need to identify predictive abiotic and biotic
variables for its occurrence before the remaining natural
stands disappear.
African Sandalwood has wide ecological distribution in
Africa (Beentje, 1994, Mwang’ingo et al., 2003,
Tshisikhawe et al., 2012, International Plant Names Index
website (www.ipni.org/). The species can parasitize over
300 species of plants from herbaceous weeds, grass,
multi-stem shrubs and trees. Usually, it is found in
association with various hosts such as Dodonea viscosa,
Tecomaria capensis, Catha edulis, Apodytes dimidiata,
Brachytegia spiciforms, Rhus natalensis and Casuarina
equisetifolia (Mwang’ingo et al., 2010). In Kenya, the
species grows naturally in both humid highland and dry
lowland forests (Maundu and Tengnas, 2005) that differ
in altitude, vegetation types, soils and climatic variables
(Sombroek et al., 1980). However, the effect of abiotic
and biotic variables diversity on African Sandalwood
distribution is not well studied, thus limiting site suitability
prediction capacity for O. lanceolata domestication. The
objectives of this study were therefore to determine plant
species that associate strongly with O. lanceolata in
humid highland and dry lowland forests and to determine
soil variables that may influence the occurrence of the
species in natural stands.
MATERIALS AND METHODS
Study sites
The study sites were Gachuthi humid highland forest and Kibwezi
dry lowland forest (Figure 1). Gachuthi forest occurs in agro-climatic
zone III (Sombroek et al., 1980), at an altitude range of 2040 to
2200 m above sea level with temperatures ranging from 12 to 25°C
and mean annual rainfall range of 990 to 1500 mm. The soils in this
forest are nitosols that are derived from volcanic rocks (Okalebo et
al., 2002). The characteristics of these soils include high clay
content (more than 35%), good moisture-storage capacity, good
aeration, and high organic matter content. Cation exchange
capacity and the percentage base saturation range from low to
high. The soils are acidic (pH < 5.5) due to the leaching of soluble
bases (Okalebo et al., 2002). The natural vegetation of Gachuthi
forest is dominated by Calodendrum capense, Ehretia cymosa,
Maytenus
undata,
Teclea
simplicifolia,Vangueria
madagascariences, Warburgia ugandensis and Zanthoxylum
usambarense.
Kibwezi forest lies in a semi-arid region (agro climatic zone V) in
south eastern Kenya (Figure 1) within an attitude range of 900 to
1015 m above sea level with a temperature range of 19 to 30°C and
mean annual rainfall ranges between 250 and 350 mm (Sombroek
et al., 1980). The soils in this forest are classified as sandy loams,
gravely volcanic and clayey (Okalebo et al., 2002). Acacia
commiphora woodland is the dominant vegetation type. Dominant
trees include Acacia xanthophloea, Acacia tortilis, Adansonia
digitata, Balanites aegyptiaca and Commiphora species.
Vegetation data collection and soil sampling
A reconnaissance visit in both forests was undertaken where O.
lanceolata was found to be more abundant at the edges than deep
in the forest and a sampling framework was designed. The forest
edges were found to be fairly heterogeneous over short distances.
Subsequently, transects measuring 600 m were laid using a linear
tape measure. Modified nested-intensity plots (Barnett and
Stohlgren, 2003) were then laid along each transects. To avoid
spatial autocorrelation (Tiegs et al., 2005; de Knegt et al., 2010), a
distance of ≥ 50 m was adopted between any two plots. In total, 24
plots were sampled in each site. In Gachuthi forest, 7 plots
randomly fell in plots with O. lanceolata and 17 in plots without O.
lanceolata. In Kibwezi forest, 18 plots were with O. lanceolata and 6
plots without O. lanceolata.
A modified nested intensity plot consisted of a main plot “A”
measuring 5 by 20 m, a middle sub-plot “B” measuring 2 by 5 m
and four sub-plots “C” of 1 by 1 m (Figure 2). Normally, the 1 by 1 m
sub-plots are located near the corners of the main plot but their
location was modified in this study to be close to O. lanceolata trees
located at the middle of the main plot (Figure 2). Vegetation data
was captured in terms of trees from the main Plot A, shrubs from
Sub-plot B and herbaceous species and grass in Sub-plot C. The
species were identified in the field, using published keys (Beentje,
1994). If the species could not be identified, its vernacular name
was used and a specimen collected for identification at the national
herbarium. Species found in each of the three sub-plots were
tabulated in appropriate tables and their frequencies recorded for
further analysis.
Soil samples were collected under O. lanceolata trees and the
main plot measuring at depths of 0 to 25 cm and 25 to 50 cm using
a soil auger, bulked, homogenized according to their different
depths and stored in polythene bags. Soil samples were taken to
Kenya Forest Research Institute (KEFRI) soil laboratory for
analysis. The analysis included soil moisture, texture, pH and
electro conductivity, nitrogen, phosphorous and potassium. Soil
moisture and texture was determined using improved hydrometer
method for soil particle size analyses, pH and Electro
Conductivity(E.C.) values were determined with glass electrode, pH
meter Model 691 and E.C. meter Model TOA Cm-20s (Lawal
*Corresponding author. E-mail: mwgathara@gmail.com, Tel: +254 712 809 563.
Author(s) agree that this article remain permanently open access under the terms of the Creative Commons Attribution
License 4.0 International License
Gathara et al.
101
Figure 1. Geographical location of Gachuthi humid highland and Kibwezi dry lowland forests in Kenya.
and Girei, 2013). Total nitrogen was determined using Kjeldahl
method with Skalar Block Digester System, Model SA 5640 as
described by Okalebo et al. (2002). Available phosphorus was
analyzed using UV spectrophotometer method (Olsen et al., 1982)
with UV Spectronic Model 21-Milton Roy Co. Potassium was
determined specto-photometically (Okalebo et al., 2002) using
flame photometer, Model Corning M 410.
Data analysis
Species frequency data was combined into a single MS Excel©
spreadsheet and used as species data. Soil nutrients, texture and
moisture content data was then saved as a single MS Excel ©
spreadsheet and used as environmental data. The two data sets
were used in Canonical Correspondence Analysis using CANOCO
version 4.15 (Ter Braak, 1997) that relates species to measured
environmental variables (Palmer, 1993). This relationship is shown
graphically in biplots where lengths of the arrows reveal the relative
influence of a measured variable to a species. In our case, plant
species associations (clustering) was established from the species
data and relationship between species and measured soil variables
determined by using soil data as the environmental variable data in
the analysis.
102
J. Hortic. For.
20 m
A (20 x 5) m
C
C
5m
B (2 x 5) m
C
C
Figure 2. A modified nested-intensity sample plot used for field vegetation data collection. The star indicates
approximate location of Osyris trees in the sample plots.
RESULTS AND DISCUSSION
those reported in related studies (Mwang’ingo et al.,
2010; Githae et al., 2011).
Comparison of species occurrence in plots with and
without O. lanceolata in Gachuthi and Kibwezi forests
In Gachuthi forest, 16 herbaceous species were found in
plots with O. lanceolata and 24 herbaceous species
found in plots with no O. lanceolata (Table 1). In Kibwezi
forest, 11 herbaceous species were found in plots with O.
lanceolata and 6 herbaceous species found in plots
without O. lanceolata (Table 1). In Gachuthi, there were 3
grass species in plots with O. lanceolata and 5 grass
species in plots without O. lanceolata (Table 1 In Kibwezi,
there were 2 species of grass found co-occurring with O.
lanceolata and only 1 grass species was found in plots
without O. lanceolata only (Table 1). Twenty-two shrubs
were found in plots with O. lanceolata and Twenty-one
shrubs found in plots without O. lanceolata in Gachuthi
(Table 2). This was in contrast with 14 and 11 shrubs
found in plots with and without O. lanceolata in Kibwezi
respectively (Table 2). Fourteen and seventeen tree
species were found in plots with and without O.
lanceolata in Gachuthi forest, respectively as compared
to 21 tree species in plots with O. lanceolata and 7 tree
species in plots without O. lanceolata in Kibwezi forest. In
total, there were 55 species in plots with O. lanceolata as
compared with 67 species without O. lanceolata in
Gachuthi forest. This was in contrast to 48 species in
plots with O. lanceolata and 25 species in plots without
O. lanceolata in Kibwezi. Results of the study reveal
inconsistence of trends in species co-occurrence with O.
lanceolata between the two sites. The higher number of
species found in highland humid forest is consistent with
high species diversity of such forests when compared to
lowland dry forests as influenced by variation in altitude,
rainfall, temperature and soils (Sombroek et al., 1980).
Also, the species found in O. lanceolata plots are among
Abiotic and biotic factors associated with occurrence
of O. lanceolata in Gachuthi and Kibwezi forests
Although, O. lanceolata was found co-existing with many
species in both sites (Tables 1 and 2), CCA biplots
(Figure 3a and b) revealed that the species could only
cluster with a few species in each of the two sites. This
suggests some of the species that coexisted may have
little or no functional associational roles. Our findings are
not surprising since studies on host preference of O.
lanceolata have demonstrated that the species has a
wide range of hosts but a few are more effective in its
establishment and early growth (Mwang’ingo et al., 2005;
Kamondo et al., 2007). The clustering of O. lanceolata
with R. natalensis in both sites is consistent with the
coexistence of both species in natural environments
(Githae et al., 2011; Teklehaimanot et al., 2012) and
effectiveness of R. natalensis as host species for O.
lanceolata (Mwang’ingo et al., 2005; Kamondo et al.,
2007). Therefore, we opine that R. natalensis is a good
tree indicator for O. lanceolata site suitability. Since O.
lanceolata also coexists with numerous herbaceous and
grass species (Githae et al., 2011; Teklehaimanot et al.,
2012), the functional association of the species with
Glycine wightii, Gutenbergia condifolia and Microglossa
pyrifolia at Gachuthi forest and Hypoestes forskahlii at
Kibwezi is subject of further studies to provide a more
effective stratification of O. lanceolata hosts among trees,
shrubs and herbaceous species.
Relationship between species with soil nutrients (N, P,
K), clay sand, silt, moisture, pH and EC revealed a
contrasting trend between sites. In Gachuthi forest, O.
lanceolata occurrence was correlated to nitrogen, clay
Gathara et al.
103
Table 1. Herbaceous (H) and grass (G) species found in plots with O. lanceolata (With Osyris) and those without Osyris (No
Osyris) in Gachuthi and Kibwezi Forests. Species occurrence is denoted by √ whereas species absence is denoted by ×.
Plant species
Abutilon mauritianum
Achyranthes aspera
Ageratum conyzoides
Asparagus racemosus
Barlelia acanthoides
Bidens pilosa
Chenopodium pumilio
Chloris sp
Cissus quadrangularis
Clematis brachiata
Commelina benghalensis
Conyza sumatrensis
Cyathula sp
Cynodon dactylon
Cyperus sp
Cyphostemma maranguense
Digitaria abyssinica
Duosperma kilimandscharicum
Fuarstia Africana
Galinsoga parviflora
Glycine wightii
Gutenbergia condifolia
Hyparrhenia rufa
Hypoestes forskahlii
Ipomea wightii
Justicia diclipteroides
Ocimum gratissimum
Oplismenus hirtellus
Oxalis obliquifolia
Pennisetum clandestinum
Periploca linearifolia
Seddera hirsute
Setaria verticillata
Sida tenuicarpa
Solanum incanum
Zehneria scabra
Total
Plant form
H
H
H
H
H
H
H
G
H
H
H
H
H
G
G
H
G
H
H
H
H
H
G
H
H
H
H
G
H
G
H
H
H
H
H
H
and moisture in contrast to lack of such relationship in
Kibwezi forest. The natural distribution of O. lanceolata in
Kenya (Maundu and Tengnas, 2005; Githae et al., 2011;
Mukonyi et al., 2011) and the soil maps of the range
(Sombroek et al., 1980) revealed a great soil diversity in
the range. Since our study was only restricted to two sites
with two soil types, further studies with more
representative soil types may be required to elucidate on
edaphic factors that may influence O. lanceolata
distribution.
Gachuthi Forest
Osyris
No Osyris
√
√
√
√
√
√
√
√
×
×
√
√
×
√
√
×
×
×
×
√
√
√
√
√
×
√
×
√
×
×
×
√
×
√
×
×
√
√
√
√
√
√
√
√
√
√
√
√
×
√
×
×
√
√
√
√
×
√
×
√
×
√
×
×
√
√
×
×
√
√
√
√
19
29
Kibwezi Forest
Osyris
No Osyris
√
×
√
×
×
×
√
√
√
√
×
×
×
×
×
×
√
√
×
×
×
×
×
×
×
×
√
√
√
×
×
×
×
×
√
×
×
×
×
×
×
×
×
×
×
×
√
√
√
×
√
×
×
×
×
×
×
×
×
×
×
×
√
×
×
×
×
√
√
√
×
×
13
7
Conclusion
In Gachuthi, Osyris clustered with R. natalensis and six
other species whereas in Kibwezi, it clustered with R.
natalensis and H. forskahlii. Therefore, O. lanceolata site
suitability for domestication can be predicted using R.
natalensis. CCA biplots showed clearly that O. lanceolata
in Gachuthi forest positively correlated to soil nitrogen,
moisture and clay whereas in Kibwezi forest; the species
did not have a relationship with any of the soil variables.
104
J. Hortic. For.
Table 2. Shrub (S) and tree (T) species found in plots with O. lanceolata (With Osyris) and those without Osyris (No Osyris)
in Gachuthi and Kibwezi Forests. Species occurrence is denoted by √ whereas species absence is denoted by ×.
Plant species
Acacia brevispica
Acacia mearnsii
Acacia robusta
Adenium spp
Antidesma venosum
Aspilia mossambicensis
Balanites maughamii
Calodendrum capense
Cassipourea malosana
Celtis Africana
Clausena anisata
Clutia abyssinica
Combretum sp
Combretum sp
Commiphora baluensis
Commiphora eminii
Commiphora spp
Crotalaria mauensis
Croton dichogamus
Croton megalocarpus
Cussonia hostii
Diospyros consolatae
Dodonaea viscose
Dombeya burgessiae
Dombeya kirkii
Ehretia cymosa
Elaeodendron buchananii
Erythrococca bongensis
Euclea divinorum
Euphorbia candelabrum
Euphorbia scheffleri
Fagaropsis angolensis
Ficus vasta
Grewia similis
Grewia spp
Haplocoelum foliolosum
Helichrysum sp.
Heteromorpha trifoliate
Hibiscus diversifolius
Hibiscus fuscus
Hymenodictyon parvifolium
Indigofera swaziensis
Juniperus procera
Lantana trifolia
Leucas grandis
Leucas spp
Lippia javanica
Maerua oblongifolia
Maytenus senegalensis
Plant form
S
T
T
S
T
S
T
T
T
T
T
S
T
T
T
S
T
S
S
T
T
T
S
S
S
T
T
S
T
T
S
T
T
S
S
T
S
T
S
S
T
S
T
S
S
S
S
S
S
Gachuthi Forest
Osyris
No Osyris
×
×
×
√
×
×
×
×
×
×
√
√
×
×
√
√
√
√
×
√
√
√
√
×
×
×
×
×
×
×
×
×
×
×
√
√
×
×
×
×
×
×
×
×
×
×
√
×
×
×
×
√
√
√
√
√
√
√
×
×
×
×
√
×
×
×
√
×
×
×
×
×
√
√
×
×
√
√
×
×
×
×
×
√
×
√
√
√
√
√
×
×
√
√
×
×
×
×
Kibwezi Forest
Osyris
No Osyris
√
×
×
×
√
×
×
√
√
×
√
√
√
×
×
×
×
×
×
×
×
×
×
×
×
√
×
√
√
×
√
×
√
×
×
×
√
×
√
×
√
×
√
√
×
√
×
×
√
√
×
×
×
×
×
×
√
×
√
×
√
×
×
×
√
√
×
×
√
×
√
×
×
×
√
×
×
×
√
√
√
×
√
√
×
×
×
×
×
×
√
√
×
×
√
√
√
×
Gathara et al.
Table 2. Contd.
S
S
S
T
T
T
T
T
T
S
S
S
S
T
T
S
T
S
T
S
S
T
S
S
S
T
T
Maytenus undata
Microglossa pyrifolia
Mystroxylon aethiopicum
Mystrxylon aethiopicum
Nuxia congesta
Ochna ovate
Olea europaea ssp. Africana
Pappea capense
Pittosporum viridiflorum
Plectrunthus barbatus
Pterolobium stellatum
Pterolobium stellatum
Rhus natalensis
Ritchiea albersii
Schrebera alata
Scutia myrtina
Steganoteenia oraliacea
Syphorstermma viminale
Teclea simplicifolia
Trimeria grandifolia
Triumfetta tomentosa
Turraea abyssinica
Vangueria madagascariensis
Vernonia brachycalyx
Vernonia lasiopus
Warburgia ugandensis
Zanthoxylum usambarense
Total
a
×
√
√
×
√
×
√
×
√
×
√
√
√
×
√
√
×
×
√
√
√
√
√
√
√
√
√
36
√
√
√
×
√
×
√
×
√
×
√
√
√
√
√
√
×
×
√
√
√
√
√
√
√
×
√
38
×
×
×
√
×
√
√
√
×
×
×
×
√
×
×
×
√
√
√
×
×
√
×
×
×
×
×
35
×
×
×
×
×
×
√
√
√
√
×
×
√
×
×
×
×
√
×
×
×
×
×
×
×
×
×
18
b
Figure 3a. CCA biplot of first and second axes showing species association and relationship between species with
soil variable at Gachuthi humid highland forest. The first two axes explain 53.3% of species-soil variables relations.
The circle highlights species that clustered with Osyris lanceolata.Species are abbreviated by the first 8 letters of
their genus name shown in Tables 1 and 2. CCA biplot of first and second axes showing species association and
relationship between soil variable at Kibwezi dry lowland forest. The first two axes explain 51.6% of species -soil
variables relations. The circle highlights species that clustered with Osyris lanceolata. Species are abbreviated by
the first 8 letters of their genus name shown in Tables 1 and 2.
105
106
J. Hortic. For.
Due to the limited number of sites used in the current
study, we recommend further studies on relationship
between soil variables and O. lanceolata occurrence in
natural ecosystems.
Conflict of Interest
The authors have not declared any conflict of interest.
ACKNOWLEDGEMENTS
This study was funded by KEFRI through Drylands
Forestry Research program in support of Mary Gathara’s
MSc study and we are grateful for the financial support.
We appreciate the assistance extended by Mr. Eliud
Macharia and Ms Margaret Nduta in field data collection
and Mr. Shadrack Odhiambo in soil analysis. We are also
grateful to Mr. Bernand Kamondo for assisting us in
editing the manuscript before submission.
REFERENCES
Barnett DT, Stohlgren JT (2003). A nested-intensity design for
surveying plant diversity. Biodivers. Conserv.12:255-278.
Beentje HJ (1994). Kenya trees, shrubs and lianas. Nairobi: National
Museums of Kenya. Cites, Cop16 proposal and listings in appendix II,
Bangkok (Thailand).
De Knegt HJ, van Langevelde F, Coughenour MB, Skidmore AK, de
Boer WF, Heitkonig IMA, Knox NM, Slotow R, van der Waal C, Prins
HHT (2010). Spatial autocorrelation and the scaling of speciesenvironment relationships. Ecol. 91:2455-2465.
Githae EW, Gachene CKK, Odee DW (2011). Implications of in situ
conservation of indigenous species with special reference to Coffea
arabica L. population in Mount Marsabit Forest, Kenya. Trop.
Subtrop. Agroecosyst. 14:715-722.
Irving LG, Cameron DD (2009). You are what you eat: interactions
between root parasitic plants and their hosts. Adv. Bot. Res. 50:87138.
Kamondo B, Juma P, Mwangi L, Meroka D (2007). Domestication of
Osyris lanceolata in Kenya: Propagation, management, conservation
and commercialization. In Muguga Regional Research Centre.
Annual Report. July 2006 – June 2007.
Lawal HM, Girei HA (2013). Infiltration and organic carbon pools under
the long term use of farm yard manure and mineral fertilizer. Int. J.
Adv. Agric. Res. 1:92-101.
Maundu P, Tengnas T (2005). Useful trees and shrubs for
Kenya.Technical handbook edition No. 35.World Agroforestry Centre
–Eastern and Central Africa Regional Programme (ICRAF-ECA)
Nairobi, Kenya.
Mukonyi KW, Kyalo S, Lubia IK, Leitoro E, Mbaka RM, Lusweti AM,
Mutwiri FM (2011). Status of Osyris lanceolatain Kenya.Kenya
Wildlife Service Report.
Mwang’ingo PL, Teklehaimanot Z, Hall JB, Lulandala LLL (2003).
African Sandalwood (Osyris lanceolata): resource assessment and
quality variation among populations in Tanzania. Southern Afr.For. J.
199:77-88.
Mwang’ingo PL, Teklehaimanot Z, Lulandala LL, Mwihomeke ST
(2005). Host plants of Osyris lanceleota (African Sandalwood) and
their influence on its early growth performance in Tanzania. South
Afr. For. J. 203:55-65.
Mwang’ingo PL, Kibodya G, Mng’ong’o AR (2010). Oil yield and quality
variation between sexes in Osyris lanceolata (African Sandalwood)
and its value as a fodder plant in Tanzania Forests. J. For. Sci.
72:69-74.
Okalebo JR, Gathua KW, Woomer PL (2002). Laboratory methods of
soil
and
plant
analysis:
a
working
manual.
Second
Edition.TSBF.CIAT and SACRED Africa, Nairobi, Kenya.
Olsen SR, Sommers LE (1982). Phosphorus.In: A.L. Pageetal. (eds.)
Methods of soil analysis, part2.Agron.Mongr.9. 2nd ed. ASA and
SSSA, Madison, WI.
Palmer MW, (1993). Putting things in even better order: the advantage
of Canonical Correspondence Analysis. Ecol. 74:2215-2230.
Sombroek WG, Braun HMH, Van Der P (1980). Exploratory soil map
and agro-climatic map of Kenya.Kenya Soil Survey Nairobi.
Subasinghe SMCUP (2013). Sandalwood Research: A Global
Perspective. Department of Forestry and Environmental Science,
University of Sri Jayewardenepura, Nugegoda, Sri Lanka. J. Trop.
For. Environ. 3(01):1-8.
Teklehaimanot Z, Mwang'ingo PL, Mugasha AG, Ruffo CK (2012).
Influence of the origin of stem cutting, season of collection and auxin
application on the vegetative Propagation of African Sandalwood
(Osyris lanceolata) in Tanzania. South Afr. For. J. 201:13-24.
Tiegs SD, O'Leary JF, Pohl MM, Munill CL (2005).Flood disturbance
and riparian species diversity on the Colorado River Delta. Biodivers.
Conserv. 14:1175-1194.
Ter Braak CJF (1997). CANOCO – a FORTRAN for canonical
community ordination
by [partial]
[detrended] [canonical]
correspondence analysis, principal component analysis and
redundancy analysis (Version 4.15). Institute of applied computer
science, 95, Wageningen, the Netherlands.
Tshisikhawe MP, van Rooyen MW, Bhat RB (2012). An evaluation of
the extent and threat of bark harvesting of medicinal plant species in
the Venda Region, Limpopo Province, South Africa. Int. J. Exper. Bot.
81:89-100.
USF and WS (2013). US Forest and Wildlife Service notice to the world
import/export community on changes in CITES species listings.
Inspector, US Department of Fthe Interior. Fish and Wildlife Service.