The influence of light and temperature on the germination of two Ugandan medicinal trees Torunn Stangeland1*, John R. S. Tabuti2 and Kåre A. Lye1 1 2 Department of Ecology and Natural Resource Management, Norwegian University of Life Sciences, PO Box 5003, NO-1432 Ås, Norway and Department of Botany, Makerere University, PO Box 7062, Kampala, Uganda Abstract For reasons of the problems of establishment of some Ugandan trees in certain environments, we investigated the influence of temperature and light on germination and seedling growth of two locally threatened medicinal trees, Hallea rubrostipulata and Sarcocephalus latifolius, to facilitate their establishment. Field and controlled laboratory experiments were carried out to investigate the species germination requirements and seedling growth. Both species needed light to germinate. Hallea rubrostipulata had a temperature optimum of 25C with 79% germination, while for S. latifolius, the total germination after 28 days was close to 60% at temperatures from 20 to 35C. Seedlings of S. latifolius survived well at 35C, while those of H. rubrostipulata died at this high temperature. Conversely, seedling of S. latifolius died at the low temperature of 15C. However, in field experiment S. latifolius failed to germinate in the available degraded environments, probably because of predation and because the soil is not able to retain water long enough to support seedling growth. We, therefore, conclude that in this part of Uganda, nursery assistance is needed to establish healthy populations of Sarcocephalus and many other endangered trees. Key words: conservation, germination, Hallea rubrostipulata, medicinal plants, Sarcocephalus latifolius, seedling growth Résumé En raison des problèmes que connaissent plusieurs arbres ougandais pour s’établir dans certains environnements, nous avons recherché l’influence de la température et de la lumière sur la germination et la croissance des plantules de deux arbres médicinaux localement menacés, Hallea rubrostipulata et Sarcocephalus latifolius, afin de faciliter leur *Correspondence: E-mail: torunn.stangeland@umb.no établissement. Des expériences de terrain et d’autres contrôlées en laboratoire ont été réalisées afin de découvrir les exigences des espèces pour leur germination et la croissance des plantules. Les deux espèces ont besoin de lumière pour germer. La température optimale pour H. rubrostipulata était de 25C avec 79% de germination, tandis que pour S. latifolius, la germination totale était près de 60% après 28 jours à des températures qui allaient de 20 à 35C. Les jeunes plants de S. latifolius survivaient bien à 35C alors que ceux de H. rubrostipulata mouraient à cette haute température. Par contre, les plants de S. latifolius mouraient à la basse température de 15C. Pourtant, sur le terrain, S. latifolius n’a pas réussi à germer dans les environnements dégradés qui étaient disponibles, probablement à cause de la prédation et parce que le sol n’était pas à même de retenir l’eau assez longtemps pour permettre la croissance des jeunes plants. Nous en concluons donc que, dans cette partie de l’Ouganda, il faut une aide en pépinière pour établir des populations saines de Sarcocephalus et de nombreux autres arbres en danger. Introduction Most rural communites in developing countries rely on medicinal plants for their health care (Tabuti, Lye & Dhillion, 2003; Hamilton, 2004). Unfortunately, important medicinal trees are threatened by overexploitation and land use changes (Hamilton, 2004; Kala, Farooqueen & Dhar, 2004). In Uganda, for example, the forest cover has decreased from 13.7% to 3.6% of total land area during the last century (Arinaitwe, Pomeroy & Tushabe, 2000). Important trees on which local livelihoods depend need to be conserved, which requires a good understanding of their seed and germination ecology (OryemOriga, 1999; Peters, 1999; Jäger & van Staden, 2000;  2007 The Authors. Journal compilation  2007 Blackwell Publishing Ltd, Afr. J. Ecol., 46, 565–571 565 566 Torunn Stangeland et al. Oryem-Origa, Kasenene & Magambo, 2004). However, such knowledge on indigenous trees in tropical Africa is very limited. Many factors, including temperature and light, influence seed germination and seedling establishment. Optimum temperature for germination varies among species. Some have a wide range (Sparg, Kulcarni & van Staden, 2005), while others have a very narrow range (Yirdaw & Leinonen, 2002). The growth rate is the fastest in small seeded species (Fenner & Thompson, 2005; Warren & Adams, 2005; Cardillo & Bernal, 2006). Grime, Mason & Curtis (1981) found that most seeds weighing <0.1 mg were largely photoblastic. Most of positively photoblastic seeds react negatively to the low R : FR ratio of leaf filtered light (Fenner & Thompson, 2005). In this study, we investigated the effect of temperature and light on germination and seedling growth of two locally important medicinal trees Hallea rubrostipulata (K. Schum.) J. F. Leroy (syn. Mitragyna rubrostipulacea) and Sarcocephalus latifolius (Sm.) E.A. Bruce, (syn. Nauclea latifolia) both of the family Rubiaceae. The bark of H. rubrostipulata was ranged as the most important treatment against malaria in a recent ethnomedicinal survey in Southern Uganda (Ssegawa & Kasenene, 2007). They found that there is a need to propagate H. rubrostipulata in nursery to conserve the natural population in the forest; but repeated germination experiments in two local nurseries at Malabigambo Forest have failed (Eilu, 2007). Sarcocephalus latifolius is a shrub or small tree in seasonally moist soils of woodland savannas extending from Senegal to Uganda. In the beginning of this study, a focus group discussion with seven traditional healers from Gadumire ranked S. latifolius as one of the five most important medicinal plants, but getting difficult to find. The study of Tabuti (2007) confirms that it is overharvested and locally threatened. evergreen oligotrophic rainforest, previously dominated by Podocarpus. The Reserve has a mean annual maximum temperature of 25–27.5C, a mean annual rainfall of 1300–1500 mm, and there are 90–100 days of rain per year. Sarcocephalus latifolius fruits were collected in March 2005 from two individual trees growing in disturbed savanna woodland in Kaliro District (102¢23¢¢N, 3328¢53¢¢E; and altitude 1065 m). The fruit is an irregularly globose berry, 3–8 cm in diameter, containing thousands of minute seeds immersed in a pinkish flesh. Kaliro is situated in East Uganda south of Lake Kyoga. It has a mean annual maximum temperature of 30–32.5C, mean annual rainfall of 1250–1300 mm, and there are 90–100 days of rain per year. Seeds were collected from few trees because mature trees with ripe seeds are very rare. The Hallea seeds are also difficult to collect as they are 20 m up in the canopy, they easily fall to the ground as they ripen, and are impossible to find on ground because of small size. Materials and methods Germination experiment Material Seeds of H. rubrostipulata were collected towards the end of May 2005 from one tree (056¢60¢¢S, 3135¢19¢¢E; and altitude 1140 m). Each fruit is a capsule about 1-cm long containing numerous minute winged seeds. The species grows in a dense swamp forest in Sango Bay Forest Reserve, Rakai District close to Lake Victoria and near the border to Tanzania. The climax vegetation of the Reserve is Field growth experiment In March 2005, a field germination experiment with S. latifolius was conducted in Gadumire in Kaliro District. Fruits of S. latifolius were soaked in water to separate the minute seeds from the dry fruits as recommended by Katende, Birnie & Tengnäs (1995) and then air dried in the shade. Seeds were sown in an agricultural fallow (>2 years) in an area dominated by disturbed wooded savannah. We tested for grazing, shade and different degrees of soil disturbance. The species failed to germinate and the experiment was repeated in November 2005 using the same plots but with different treatments. The experiments were performed in the long (March) and short (November) rainy seasons. The minute seeds were stored dry at room temperature for 3 (H. rubrostipulata) or 6 months (S. latifolius) before sowing. The germination study began in September 2005 and lasted 4 weeks. The average seed weights were 0.00712 mg for H. rubrostipulata and 0.0233 mg for S. latifolius (n = 50). The seeds were incubated in five controlled environment cabinets at constant temperatures of 15, 20, 25, 30 and 35C under 12 : 12 h light ⁄ dark using Philip master (TDL 36W ⁄ 830) with photosynthetic  2007 The Authors. Journal compilation  2007 Blackwell Publishing Ltd, Afr. J. Ecol., 46, 565–571 The influence of light and temperature Seedling growth experiment The study of seedling growth started in mid-January 2006 and lasted 12 weeks. Seeds were sown on saturated filter paper in 9-cm Petri dishes and placed in a controlled environment cabinet at 25C. After germination, when the seedlings had unfolded the two cotyledons, the seedlings were transferred to pots (8 cm) using a standard potting compost mixed with perlite and placed in five growth chambers at constant temperatures of 15, 20, 25, 30 and 35C under 12 : 12 h light ⁄ dark. Light sources were Osram Powerstar HQI-BT 400W, Osram, Munich, Germany supplemented by incandescent lamps to lower the R : FR ratio. PPFD values were 200 lmol m)2 s)1 (light to medium shade), R : FR 1.6 and relative humidity 90%. Five seedlings of each species were harvested, dried and weighed at the time of planting (time 0). After week 4, four seedlings of S. latifolius and five of H. rubrostipulata plants from each temperature were harvested, separated into shoots and roots, dried at 70C for 48 h and then weighed. Harvesting continued every second week until the twelfth week after planting. One-way ANOVA test for percentage germination versus temperature was run for both species. Two-way ANOVA was used to test if germination was significantly different between the two species. The Minitab statistical program was used for the tests. Results Germination In the field experiments, no seeds of S. latifolius were observed to germinate. However, during the laboratory experiment, close to 60% of the same batch of seeds germinated at 20, 25, 30 and 35C (Fig. 1). Germination was completed within 15 days at the temperatures 25 and 30 C. Total germination took more days (25) at 20 and 35 C (Fig. 2). At 15 C, no seeds germinated. Hallea rubrostipulata had optimum germination (78.8%) at 25C (Fig. 1). At 20 and 30 C, germination were 59.6 and 67.2C respectively. Most of the seeds had germinated within 6 days and germination was completed by the ninth day at 25 and 30C (Fig. 2). No seeds of H. rubrostipulata germinated at 35C. At the low temperature of 15C, the capability of H. rubrostipulata to germinate was low (25%) and delayed. For both species, very few seeds germinated in the dark (0.8% of Hallea at 20 and 25C; 5.6% and 4% of Sarcocephalus at 20 and 25C respectively). 100 80 70 60 50 40 30 Relative growth rate and statistical analysis 20 Relative growth rates (RGR) at the different temperatures and at different phases of seedling growth were determined by taking the natural logarithm of change in dry mass (W ) of shoots or roots to change in time (t) as follows (Hunt et al., 2002): 10   W2  W1 RGR ¼ ln : t2  t1 Hallea rubrostipulata Sarcocephalus latifolius 90 % Germination photon flux density (PPFD) at 400–700 nm values approximated 130 lmol m)2 s)1. Five pseudo replicates of 50 seeds of each species were germinated at each temperature in 9-cm Petri dishes on water saturated double filter paper. To prevent evaporation, the Petri dishes were placed in polyethylene bags. Every second day for 28 days, newly germinated seeds were counted and removed to another Petri dish. Germination was defined as radicle emergence. To determine the effect of continuous darkness, five pseudo replicates of each species were wrapped in aluminium foil and incubated in the same cabinets. The seeds under dark treatment were counted under safe green light. Distilled water was added regularly to the dishes as needed. 567 0 15 20 25 30 35 Temperature (°C) Fig 1 Temperature effect on total per cent germination after 28 days of Hallea rubrostipulata and Sarcocephalus latifolius sown at five temperatures (15, 20, 25, 30 and 35C) and 12 : 12 h light ⁄ dark. Bars show standard errors of the mean (n = 5)  2007 The Authors. Journal compilation  2007 Blackwell Publishing Ltd, Afr. J. Ecol., 46, 565–571 568 Torunn Stangeland et al. 100 15°C 20°C 25°C 30°C 35°C 80 Dryweight shoots (mg) % Germination 80 60 40 20 60 40 20 0 0 100 60 Dryweight shoots (mg) 80 % Germination 80 15°C 20°C 25°C 30°C 35°C 40 20 60 40 20 0 0 2 4 67 9 11 13 15 17 19 21 23 25 28 0 Days after sowing Fig 2 Cumulative per cent germination of Hallea rubrostipulata (above) and Sarcocephalus latifolius (below) sown in Petri dishes and incubated at five constant temperatures and 12 : 12 h light ⁄ dark for 28 days. Bars show standard errors of the mean (n = 5) Seedling growth Both species had the best growth at 30C, and this was increasing linearly with time (Figs 3 and 4). For S. latifolius, the growth in terms of dry weight of the shoots was almost twice as high as the dry weight of roots, while for H. rubrostipulata dry weight of roots and shoots was almost similar. After 12 weeks, dry weight of Sarcocephalus roots and shoots were almost ten times higher than that of Hallea. Sarcocephalus latifolius had its second best growth at 35C, while Hallea died at this high temperature. In both species, RGR was the highest and showed the clearest trends in the first 4 weeks (Fig. 5, Table 1). Sarcocephalus latifolius showed a linear increase with temperature between the temperatures of 15 and 30C in both 28 42 56 70 84 Days after planting Fig 3 Dry weight of shoots and roots of Hallea rubrostipulata grown in pots at five constant temperatures and 12 : 12 h light ⁄ dark and harvested every second week from week 4 to week12 after planting (n = 5). Bars show standard errors of the mean root and shoot. Hallea had a somewhat lower and broader optimum of RGR between 25 and 30C (Fig. 5). For the periods after 42 days, the RGR of S. latifolius declined, while for H. rubrostipulata it still increased for 14 days and then declined (Table 1). After 9 months in greenhouse at the same environment, the two species had about the same height of 50 cm. Two-way ANOVA for per cent germination versus species and temperature showed that there was effect of species on some of the temperatures (F = 39.14, P < 0.001). Discussion As S. latifolius seeds failed to germinate under disturbed conditions in wooded savanna, constraints like water stress and herbivory could have affected germination and  2007 The Authors. Journal compilation  2007 Blackwell Publishing Ltd, Afr. J. Ecol., 46, 565–571 The influence of light and temperature 569 25 15°C 20°C 25°C 30°C 35°C 600 Shoots Roots 20 RGR (% per day) Dryweight roots (mg) 800 400 200 15 10 5 0 25 0 800 RGR (% per day) Dryweight roots (mg) 20 600 400 15 10 5 200 0 0 10 0 28 42 56 70 84 Days after planting Fig 4 Dry weight of shoots and roots of Sarcocephalus latifolius grown in pots at five constant temperatures and 12 : 12 h light ⁄ dark and harvested every second week from week 4 to week 12 after planting (n = 4). Bars show standard errors of the mean survival. As termites were abundant in and around the experimental field, they may have eaten or collected the small seeds of S. latifolius. The shaded plots were more protected against drought and sunburn, but lack of germination here could be because of low R : FR (Grime et al., 1981; Yirdaw & Leinonen, 2002). This implies that the species cannot be established easily without nursery assistance. As it is well known that species with seeds weighing <0.1 mg usually require light to germinate (Grime et al., 1981; Fenner & Thompson, 2005), it was not unexpected that few seeds of S. latifolius and H. rubrostipulata germinated in darkness (0.8% of Hallea at 20 and 25C; 5.6% and 4% of Sarcocephalus at 20 and 25C respectively). Although both species overlapped in terms of their preferred temperature for germination and early seedling establishment (20–30C), S. latifolius had a higher 15 20 25 30 35 Temperature (°C) Fig 5 Relative growth rate (RGR) mg g)1 day)1 of Hallea rubrostipulata (above) and Sarcocephalus latifolius (below) during the first 28 days of the experiment temperature range than H. rubrostipulata and this reflected their natural ecological adaptations. The natural habitat where S. latifolius grows in Uganda has higher temperature ranges than those of Sango Bay Forest Reserve according to Atlas of Uganda (Anonymous 1967). Hallea rubrostipulata managed to establish at the lower temperature of 15C, while S. latifolius failed. Simon et al. (1976) found that germination of tropical species declines dramatically at about 14C and ceases at 10C. This is consistent with our findings for H. rubrostipulata, but not with our findings for S. latifolius where the lower germination limit was between 15 and 20C. The two tree species thus show an amazing adaptation of seed germination and early seedling growth to their prevailing environmental conditions. As Hallea germinated after 4 days and many seeds were contaminated by fungi within 4 days, especially at 20C, we conclude that rapid germination may be a way of  2007 The Authors. Journal compilation  2007 Blackwell Publishing Ltd, Afr. J. Ecol., 46, 565–571 570 Torunn Stangeland et al. Table 1 Relative growth rate (RGR) of shoots and roots of Hallea rubrostipulata (H) and Sarcocephalus latifolius (S) grown at five temperatures and harvested five times from 28th to 84th day after planting RGRb (mg g)1 day)1) Treatment (days)a T2 ) T1 (28 ) 0) T3 ) T2 (42 ) 28) T4 ) T3 (56 ) 42) T5 ) T4 (70 ) 56) T6 ) T5 (84 ) 70) a b Temp. (C) Shoots H. Shoots S. Roots H. Roots S. 35 30 25 20 15 35 30 25 20 15 35 30 25 20 15 35 30 25 20 15 35 30 25 20 15 69 166 145 158 74 )73 88 35 )29 12 –b 145 114 35 36 –b 21 54 66 2 –b 23 11 7 29 188 227 180 130 39 76 129 136 72 )169 103 90 84 86 –b 46 7 21 89 –b 27 20 2 90 –b 25 148 141 156 8 –b 51 1 11 )22 –b 197 146 21 53 –b 68 89 45 24 –b 60 11 30 38 139 179 155 133 )8 118 189 116 21 )92 87 142 93 116 –b 115 26 54 103 –b 65 36 2 )66 –b T is number of days from T1: the day of potting. No data indicate that the plants have died. escaping mortality by fungi, as suggested by Muhanguzi, Obua & Oryem-Origa (2002) for some Ugandan forest trees. The Sarcocephalus seeds germinated slightly later and were more resistant to fungal infection (about 50% of the seeds had germinated within 10 days at the same temperatures). There were striking differences between the species in biomass allocation to shoots and roots. While at the end of the experiment, Hallea had almost as much biomass in roots as in shoots, Sarcocephalus had only about half as much biomass in the roots as in the shoots. The slow growth of the Hallea seedlings and the relatively large allocation to roots might be a strategy for adapting to an oligotrophic environment. The high germination percentage (60% and 60–80%) after 3 or 6 months storage indicates that seeds of these species can be stored for considerable time, but the actual length of seed longevity must be studied further. The light treatment given was suitable for seedling growth (200 lmol m)2 s)1, R : FR = 1.6), which corresponds to light to medium-shade (Lee et al., 1996, 1997; full sun: 1528 lmol m)2 s)1, R : FR = 1.34; light shade: 400–600 lmol m)2 s)1, understory shade: 13 lmol m)2 s)1, R : FR = 0.20). Several studies have given the best seedling growth at moderate shade and high R : FR (Lee et al., 1996; Yirdaw & Leinonen, 2002). Moderate shade results in better survival during the dry season (McLaren & McDonald, 2003; Lemenih, Gidyelew & Teketay, 2004). To domesticate Hallea and Sarcocephalus successfully, it will be important to find or create habitats with the preferred light and temperature requirements of the species. Our field experiment indicates that natural establishment of S. latifolius in degraded environment is very difficult without nursery assistance. For Hallea, repeated germination experiments in the local nurseries at Malabigambo Forest also failed. Acknowledgements We are most grateful to Professor Ola M. Heide, who gave advice regarding experimental work in the growth chambers and to the group of healers in Gadumire including their leader Patrick Daire, who helped us in fieldwork. Dr Knut A. Hovstad, Dr Svein Dahle and Dr Toril D. Eldhuset at UMB assisted in analysing and discussing data. We also thank Paul Ssegawa, Peregrine Sebulime, Charles Halaiza and Dennis Kamoga who helped with logistics and assistance during the field trips. The work was supported by a grant from NUFU project 13 ⁄ 2002. References Anonymous (1967) Atlas of Uganda. Department of Lands and Survey, Entebbe, Uganda. Arinaitwe, H., Pomeroy, D. & Tushabe, H., Eds (2000) The State of Uganda’s Biodiversity 2000. National Biodiversity Data Bank, Makerere University, Kampala. Cardillo, E. & Bernal, C.J. (2006) Morphological response and growth of oak (Quercus suber L.) seedlings at different shade levels. For. Ecol. Manage. 222, 296–301. Eilu, G., Ssegawa, P., Obua, J. & Galabuzi, C. (2007) Progress report for the period October 2006 to March 2007. Available at:  2007 The Authors. Journal compilation  2007 Blackwell Publishing Ltd, Afr. J. Ecol., 46, 565–571 The influence of light and temperature http://www.plantlife.org.uk/international/plantlife-med-plantsprojects-allachy-uganda.htm (accessed on 19.11.2007). Fenner, M. & Thompson, K. (2005) The Ecology of Seeds. University Press, Cambridge. Grime, J.P., Mason, G. & Curtis, A. (1981) A comparative study of germination characteristics in a local flora. J. Ecol. 69, 1017– 1059. Hamilton, A. (2004) Medicinal plants, conservation and livelihoods. Biodivers. Conserv. 13, 1477–1517. Hunt, R., Causton, D.R., Shipley, B. & Askew, A.P. (2002) A modern tool for classical plant growth analysis. Ann. Bot. 90, 486–488. Jäger, A.K. & van Staden, J. (2000) The need for cultivation of medicinal plants in Southern Africa. Outlook Agric. 29, 283– 284. Kala, C.P., Farooqueen, N.A. & Dhar, U. (2004) Prioritization of medicinal plants on the basis of available knowledge, existing practices and use value status in Uttaranchal, India. Biodivers. Conserv. 13, 453–469. Katende, A.B., Birnie, A. & Tengnäs, B. (1995) Useful Trees and Shrubs for Uganda. Technical Handbook No 10. Reg. Soil Conserv. Unit., Nairobi. Lee, D.W., Baskaran, K., Mansor, M., Mohamad, H. & Yap, S.K. (1996) Irradiance and spectral quality affect Asian tropical rain forest tree seedling development. Ecology 77, 568–580. Lee, D.W., Oberbauer, S.F., Baskaran, K., Mansor, M., Mohamad, H. & Yap, S.K. (1997) Effects of irradiance and spectral quality on seedling development of two South Asian Hopea species. Oecologia 110, 1–9. Lemenih, M., Gidyelew, T. & Teketay, D. (2004) Effects of canopy cover and understory environment of tree plantations on richness, density and size of colonizing woody species in southern Ethiopia. For. Ecol. Manage. 194, 1–10. McLaren, K.P. & McDonald, M.A. (2003) The effect of moisture and seedling germination and seedling survival in a tropical dry forest in Jamaica. For. Ecol. Manage. 183, 61–75. Muhanguzi, H.D.R., Obua, J. & Oryem-Origa, H. (2002) Influence of light quality on the germination characteristics of seeds of selected pioneer, under storey and canopy tree species in Kalinzu forest reserve, Uganda. Ug. J. Agric. Sci. 7, 25–30. 571 Oryem-Origa, H. (1999) Fruit and seed ecology of wild robusta coffee (Coffea Canephora Froener) in Kibale National park, Uganda. Afr. J. Ecol. 37, 439–448. Oryem-Origa, H., Kasenene, J.M. & Magambo, J.S. (2004) Some aspects of wild robusta coffee seedling growth in Kibale National Park, Uganda. Afr. J. Ecol. 42(Suppl. 1), 34–39. Peters, C.M. (1999) Ecological research for sustainable non-wood forest product exploitation: an overview. In: Non-wood Forest Products of Central Africa: Current Research Issues and Prospects for Conservation and Development (Eds T. C. H. Sunderland, L. E. Clark and P. Vantomme). Food and Agriculture Organization, Rome. Simon, E.W., Minchin, A., McMenamin, M.M. & Smith, J.M. (1976) The low temperature limit for seed germination. New Phytol. 77, 301–311. Sparg, S.G., Kulcarni, M.G. & van Staden, J. (2005) Germination and seedling establishment strategies for Merwilla natalensis, a South African plant in high demand. S. Afr. J. Sci. 101, 205– 208. Ssegawa, P. & Kasenene, J.M. (2007) Plants for malaria treatment in Southern Uganda: traditional use, preference and ecological viability. J. Ethnobiol. 27, 110–131. Tabuti, J.R.S. (2007) The uses, local perceptions and ecological status of 16 woody species of Gadumire Sub-county, Uganda. Biodivers. Conserv. 16, 1901–1916. Tabuti, J.R.S., Lye, K.A. & Dhillion, S.S. (2003) Traditional herbal drugs of Bulamogi, Uganda: plants, use and administration. J. Ethnopharmacol. 88, 19–44. Warren, C.R. & Adams, M.A. (2005) What determines interspecific variation in relative growth rate of Eucalytus seedlings? Ecophysiology 144, 373–381. Yirdaw, E. & Leinonen, K. (2002) Seed germination responses of four afromontane tree species to red ⁄ far-red ratio and temperature. For. Ecol. Manage. 168, 53–61. (Manuscript accepted 24 October 2007) doi: 10.1111/j.1365-2028.2007.00900.x  2007 The Authors. Journal compilation  2007 Blackwell Publishing Ltd, Afr. J. Ecol., 46, 565–571