Estimation of Population of Ten Selected Forest Tree Species Used by Communities around Kalinzu Forest Reserve, South-Western Uganda

doi:10.4236/ojf.2012.24025

Paper Menu >>

Journal Menu >>

Open Journal of Forestry

2012. Vol.2, No.4, 207-212

Published Online October 2012 in SciRes (http://www.SciRP.org/journal/ojf) http://dx.doi.org/10.4236/ojf.2012.24025

Estimation of Population of Ten Selected Forest Tree Species

Used by Communities around Kalinzu Forest Reserve,

South-Western Uganda

Adalbert Aine-Omucunguzi1, Grace Kagoro Rugunda2, Dominic Byarugaba2

1Department of Biology, Kabale University, Kabale, Uganda

2Department of Biology, Mbarara University of Science and Technology, Mbarara, Uganda

Email: [email protected]

Received January 23rd, 2012; revised April 12th, 2012; accepted April 29th, 2012

Local communities depend on Kalinzu Forest Reserve (KFR) for plant resources. This resource utilization

affects the population of tree species in the forest. This study set out to estimate the population of ten tree

species in the forest. Results of this study are hoped to provide a basis for studying future changes in

population dynamics of the species. The ten species selected were: Newtonia buchananii, Cynometra al-

exandrei, Teclea nobilis, Prunus africana, Entandrophragma exelsum, Sapium ellipticum, Diospyros

abyssinica, Zanthoxylum gilletii, Rytiginia kigeziensis and Spathodea nilotica. Their selection was based

on the results of a study (Aine-Omucunguzi et al., 2010) about utilization of plant resources by the local

people around KFR. Species with high demand from the local people were selected. Alternate nested

quadrants along five line transects, were used to estimate the population. The plants were categorized into

three classes as trees (>5 cm diameter at breast height), saplings (2 cm root collar diameter - 5 cm diame-

ter at breast height) and wildings (<2 cm root collar diameter). Population, population density, relative

density, frequency, and relative frequency of each species were determined. Generally, the number of

wildings, saplings and trees of each species increased away from the forest edge inwards. For all the spe-

cies, wildings had the highest population density followed by saplings and then trees. Teclea nobilis had

the highest number of individual per hectare, followed by Newtonia buchananii, while Rytiginia kigezien-

sis and Spathodea nilotica had the lowest number of individuals per hectare.

Keywords: Communities; Kalinzu Forest Reserve; Line Transect; Nested Quadrants; Population; Tree

Species

Introduction

Kalinzu Forest reserve occupies parts of the counties of

Bunyaruguru, Ruhinda and Igara in Bushenyi district, south-

western Uganda, with a population density 328 persons km−2. It

was gazetted for a reserve in 1932 with an area of 461 km2, but

some of the land was given off for tea estates development in

1954. The reserve now covers an area of 137 km2 (Howard,

1991). It lies on the eastern side of the Great Western Rift Val-

ley at the edge of the escarpment overlooking Lake Edward.

Kalinzu Forest Reserve is located between latitude 0˚17'N and

0˚30'N and between 30˚00'E and 30˚07'E and at an altitude

ranging from 1200 to 1500 m above sea level. Kalinzu forest is

one of the Afromontane forests of high biodiversity value in the

Albertine rift. It is a home to twelve endemic species. These

include nine butterfly species, one mammalian species and two

tree species.

Local communities depend on Kalinzu Forest Reserve for

plant resources. This resource utilization has an impact on the

population of tree species in the forest. This study set out to

assess the variation of wildings, saplings and trees of ten se-

lected species with distance from the forest edge, and to esti-

mate the population of each species, in order to provide data,

that will be used as a basis to study future changes in popula-

tion dynamics of the species. The ten species selected for the

study were: Newtonia buchananii, Cynometra alexandrei, Te-

clea nobilis, Prunus africana, Entandrophragma exelsum,

Sapium ellipticum, Diospyros abyssinica, Zanthoxylum gilletii,

Rytiginia kigeziensis and Spathodea nilotica. Their selection

was based on the results of a study (Aine-Omucunguzi et al.,

2010) about utilization of plant resources by the local people

around Kalinzu Forest Reserve. Results of the study indicate

that the above species are destructively and frequently har-

vested by fringe communities. The current study estimated the

population of these species and analyzed it in terms of their

harvesting demands, with a view of providing information that

can be used as a basis to study future population dynamics of

the species.

Methods

Nested quadrants along line transects were used to estimate

the population of the species. Five line transects of 3.5 km each,

with an interval of 1000 m were constructed using a tape meas-

ure. Each transect begun 30 m from the forest edge to eliminate

edge effects. The plants were categorized into three classes as

trees (>5 cm diameter at breast height), saplings (2 cm root

collar diameter - 5 cm diameter at breast height) and wildings

(<2 cm root collar diameter) (Alder & Synnot, 1992; Nakibuka,

1994; Eilu, 1995). Eight alternate nested plots of 20 m × 20 m

with an interval of 500 m were constructed on each of the tran-

sects and the number of trees of each of the species was

A. AINE-OMUCUNGUZI ET AL.

counted. Inside each of the 20 m × 20 m plot, a 10 m × 10 m

plot was made and saplings of each of the selected species were

counted. Inside each of the 10 m × 10 m plot, a 5 m × 5 m plot

was made and wildings of each species were counted. In total,

an area of 1.6 hectares was covered for trees, 0.4 hectares for

saplings, and 0.1 hectares for wildings. The root collar diameter

was measured using vernier calipers while diameter at breast

height (DBH) was obtained by measuring the circumference of

trees and saplings at breast height, and dividing the values ob-

tained by pie (3.14).

The number and frequency of individuals in each category

was estimated for each species by relating the number in the

area sampled to the area of the whole forest (137 km2). The

population, population density, relative density, frequency, and

relative frequency of each species were calculated using the

following formulae:

Population of a species = population of wildings + popula-

tion of saplings + population of trees in the forest;

Population density = population of a speciestotal area sam-

pled;

Relative density = (density of a speciesdensity of all species

sampled) × 100;

Frequency = number of plots with a species/total number of

plots sampled;

Relative frequency = (Frequency of a species/total frequency

of all species) × 100.

Results

Species Distribution

120 plots were sampled. Of these, Teclea nobilis was found

in 112 plots, Newtonia buchananii in 103 plots, Diospyros

abyssinica in 97, Prunus africana in 94, Entandrophragma

exelsum in 90, Cynometra alexandri in 76, Sapium ellipticum in

73, Spathodea nilotica in 69, Rytiginia kigeziensis in 56, and

Zanthoxylum gilletii in 10 plots.

Throughout the distance of 3500 m covered during this study,

wildings of Newtonia buchananii were more than saplings and

trees. The number of wildings increased from 26 at 0 m to 43

wildings at 500 m, and then decreased to 37 wildings at 1000 m.

The number then increased steadily to 99 wildings at 3500 m.

The number of sapling was higher in plots deeper in the forest

than those close to the forest edge. For each of the plots, sap-

lings were more than the trees, except at 1000 m, where the two

were equal (18 individuals). The number of Newtonia bu-

chananii trees decreased slightly from 8 trees at 0 m to 5 trees

at 500 m. The number increased to 18 trees at 1000 m, and then

decreased to 14 trees at 1500 m. The number then increased

steadily to 43 trees at 3500 m into the forest (Figure 1). Wild-

ings of Newtonia buchananii had the highest frequency, fol-

lowed by saplings, and trees had the lowest frequency (Table

1).

The number of Cynometra alexandrei wildings increased

steadily from 5 individuals at 0 m to 66 individuals at 3500 m.

The number of saplings decreased from 11 saplings at 0 m to 0

at 1000 m, and then increased steadily to 25 saplings at 2500 m.

It then decreased to 23 saplings at 3000 m, and then increased

to 34 saplings at 3500 m into the forest. The number of Cy-

nometra alexandrei trees decreased from 6 individuals at 0 m to

3 individuals at 1000 m into the forest. The number then in-

creased steadily to 35 trees at 3500 m into the forest (Figure 2).

Figure 1.

Variation of Newtonia buchananii with distance from forest edge.

Figure 2.

Variation of Cynometra alexandrei with distance from the forest edge.

Saplings of Cynometra alexandrei had the highest frequency in

the forest, followed by wildings and then trees (Table 1).

The number of wildings of Teclea nobilis decreased slightly

from 26 wildings at 0 m to 24 wildings at 500 m. The number

then increased steadily to 108 wildings at a distance of 3500 m

into the forest. The number of Teclea nobilis saplings increased

steadily from 8 saplings at 0 m to 37 saplings at 1000 m. The

number decreased to 20 saplings at 1500 m, and then increased

steadily to 53 saplings at 3500 m into the forest. The number of

tree increased from 28 at 0 m to 48 trees at 1000 m. The num-

ber then decreased to 43 trees at 1500 m, and then increased

steadily to 78 trees at 3500 m into the forest (Figure 3). Wild-

ings of Teclea nobilis had the highest frequency followed by

saplings, and then trees (Table 1).

The number of Prunus africana wildings increased from 9 at

0 m to 20 wildings at 500 m. The number decreased sharply to

7 wildings at 1500 m. It then increased steadily to 53 wildings

at 3500 m. The number of saplings decreased from 22 at 0 m to

8 saplings at 1500 m. This number increased steadily to 32

saplings at 3500 m. The number of Prunus africana trees in-

208

A. AINE-OMUCUNGUZI ET AL.

Figure 4.

Variation of Prunus africana with distance from the forest edge.

Figure 3.

Variation of Teclea nobilis with distance from the forest edge.

Table 1.

Total population of each species in the forest.

Scientific name Number in forest Population density Relative density Frequency Relative frequency

Teclea nobilis 82,816,500 5430.6 22.3 166,113 12.77

Newtonia buchananii 73,425,438 5359.8 19.8 164,143 12.62

Cynometra alexandrei 42,427,188 3096.9 11.4 122,872 9.44

Diospyros abyssinica 37,161,250 2712.5 10.0 140,254 10.78

Prunus africana 30,140,000 2200 8.1 149,416 11.48

Entandrophragma exelsum 29,266,625 2136.6 7.9 134,432 10.33

Zanthoxylum gilletii 27,408,563 2000.6 7.4 137,086 10.54

Sapium ellipticum 26,115,625 1906.3 7.0 119,275 9.17

Rytiginia kigeziensis 11,371,000 830 3.1 82,628 6.35

Spathodea nil otic a 11,353,875 828.8 3.1 84,940 6.53

Population of a species shown in Table 1 is the sum of the population of wildings, saplings and trees for each species.

creased steadily from 3 at 0 m to 32 trees at 3500 m (Figure 4).

Wildings of this species had the highest frequency, followed by

saplings, and trees had the lowest (Table 1).

The number of Entandropgragma exelsum wildings was 10

individuals at 0 m. This decreased slightly to 10 wildings at 500

m, and then increased steadily to 53 wildings at 3500 m. The

number of saplings was 5 at 0 m. This increased to 11 saplings

at 500 m, and then decreased to 5 saplings at 1000 m. The

numbers increased to 10 saplings at 2000 m, and then decreased

slightly to 9 saplings at 2500 m. The number then increased to

24 saplings at 3500 m. The number of tree at 0 m was 3. This

decreased to 2 trees at 500 m, and then increased steadily to 29

trees at 3500 m (Figure 5). Wildings of Entandropgragma

exelsum had the highest frequency, followed by saplings, and

then trees (Table 1).

The number of wildings of Sapium ellipticum was 14 at 0 m.

This remained constant up to 500 m, and then decreased

sharply to 2 wildings at 1000 m. The number then increased

steadily to 43 wildings at 3500 m. The number of saplings of

Sapium ellipticum was 4 at 0 m; this decreased to 1 sapling at

500 m, and then increased 3 saplings at 1000 m. The number

decreased to 2 saplings at 1500 m and then increased to 11

saplings at 2000 m. It then decreased slightly to 10 saplings at

2500 m, from where it increased steadily to 27 saplings at 3500

m. The number of Sapium ellipticum trees was 1 at 0 m. This

increased to 2 trees at 500 m, and then decreased again to 1 tree

at 1000 m, from where the number increased steadily to 23

trees at 3500 m (Figure 6). The wildings of Sapium ellipticum

had the highest frequency, followed by saplings, and trees had

the lowest frequency (Table 1).

The number of Diospyros abyssinica wildings decreased

from 11 at 500 m and remained constant up to 1000 m. The

number then increased steadily to 55 wildings at 3500 m. The

number of saplings decreased from 16 at 0 m to 0 at 1000 m.

The number then increased steadily to 34 saplings at 3500 m.

The number of tree dcreased from 18 at 0 m to 17 trees at 500

m. The number then increased steadily to 51 trees at 3500 m

(Figure 7). Wildings of Diospyros abyssinica had highest fre-

quency, followed by saplings, while trees had the lowest (Table

1).

A. AINE-OMUCUNGUZI ET AL.

Figure 5.

Variation of Entandrophragma exelsum with distance from the forest

edge.

Figure 6.

Variation of Sapium ellipticum with distance from the forest edge.

Figure 7.

Variation of Diospyros abyssinica with distance from the forest edge.

The number of wildings of Zanthoxyllum gilletii was 4 at 0 m;

and increased to 7 wildings at 500 m. This decreased to 6 wildings

at 1000 m and remained constant up to 1500 m. The number

then increased steadily to 49 wildings at 3500 m. The number

saplings was 9 at 0 m. This remained constant up to 500 m; and

increased to 15 saplings at 1000 m. This decreased to 6 saplings

at 1500 m, and then increased steadily to 27 saplings at 3500 m.

The number of trees of Zanthoxyllum gilletii was 14 at 0 m; this

decreased to 6 trees at 500 m; and then increased steadily to 57

trees at 3500 m (Figure 8). Wildings of Zanthoxyllum gilletii

showed the highest frequency, followed by saplings. Trees of

this species showed the lowest frequency (Table 1).

The number of wildings of Rytiginia kigeziensis increased

from 0 at 0 m to 4 wildings at 500 m; this decreased to 2 wild-

ings at 1000 m. The number then increased steadily to 23 wild-

ings at 3500 m. The number of saplings increased from 1 to 5 at

1500 m; and remained constant up to 2500 m, and then in-

creased to 13 saplings at 3500 m. The number of Rytiginia

kigeziensis trees increased from 2 at 0 m to 3 trees at 500 m.

The number decreased to 1 tree at 1000 m, and then increased

to 6 trees at 2000 m, from where it dcreased to 5 trees at 2500

m, and then increased to 7 trees at 3500 m (Figure 9). The

frequency of wildings of Rytiginia kigeziensis was the highest,

followed by saplings. Trees had the lowest frequency in this

category (Table 1).

The number of wildings of Spathodea nilotica was 0 from 0

m to 1500 m into the forest. The number then increased steady

until it was 29 at 3500 m into the forest. The number of sap-

lings decreased from 1 at 0 m to 0 at 500 m; this remained con-

stant up to 1000 m. The number increasedd to 10 saplings at

2000 m and remained constant up to 2500 m, from where it

increased to 13 saplings at 3500 m. The number of Spathodea

nilotica trees was 8 at 0 m and remained constant up to 500 m.

It then decreased to 3 trees at 1000 m; the number then in-

creased steadily to 21 trees at 2500 m. It then decreased slightly

to 20 trees at 3000 m, and again increased to 28 trees at 3500 m

Figure 8.

Variation of Zanthoxyllum gilletii with distance from the forest edge.

Figure 9.

Variation of Rytigini a kigezi ens is with distance from the forest edge.

210

A. AINE-OMUCUNGUZI ET AL.

(Figure 10). Wildings of Spathodea nilotica had the highest

frequency, followed by saplings, while trees had the lowest

(Table 1).

Estimation of Population of Each Species

For all the species, wildings had the highest population den-

sity, followed by saplings, while trees had the lowest popula-

tion density. Teclea nobilis had the highest population density

of 5430.6 individuals per hectare. Newtonia buchananii fol-

lowed with a population density of 5359.8 individuals per hec-

tare. Cynometra alexandrei had a population density of 3096.9

individuals per hectare. Diospyros abyssinica had a density of

2712.5 individuals per hectare, Prunus africana followed with

a density of 2200 individuals per hectare. Entandrophragma

exelsum had 2136.6 individuals per hectare, Zanthoxylum gil-

letii had a density of 2000.6 individuals per hectare, Sapium

ellipticum followed with a density of 1906.3 individuals per

hectare, Rytiginia kigeziensis had a density of 830 individuals

per hectare and Spathodea nilotica had the lowest population

density of 828.8 individuals per hectare (Figure 11). The over-

all frequency of the species followed a trend similar to that of

population density.

Figure 10.

Variation of Spathodea nilotica with distance from the forest

edge.

Population density for each category was obtained by dividing the

number of individuals in the category by the total area sampled.

Figure 11.

Comparison of population density of wildings, saplings and

trees per species.

Discussion

The number of plants in each plant category generally in-

creased away from the forest edge inwards. There are two pos-

sible explanations to this trend. First, it could be due to edge

effects where micro-climatic conditions affect the growth of the

tree species under study. Secondly, it could be due to the har-

vesting behaviour of people in the forest fringe communities. It

is likely, that people tend to avoid walking long distances in the

forest and thus first harvest tree resources they come across as

soon as they enter the forest. This is probably the reason why

plots that were deep in the forest generally had more individu-

als than those that were close to the forest edge. The low num-

ber of wilding near the forest edge can be attributed to two

factors; first, as people move and harvest tree resources, the

wildings are either cut to pave way or stepped on and eventu-

ally die. Second, some trees are harvested before they bear

seeds and this reduces the number of wildings. Some plots deep

in the forest however had fewer individuals than those close to

the edge. This is because, at times harvesters are interested in

plants of a particular size or level of maturity, and in such cases,

they tend to ignore those that are close to the forest edge and go

to harvest deep in the forest. This was particularly common

with species of medicinal value. This mode of harvest for in-

stance explains why Diospyros abyssinica had 16 saplings at a

distance of 0 m and no saplings at 1000 m into the forest (Fig-

ure 7). The variation of saplings of Prunus africana with dis-

tance (Figure 4) can also be attributed to the mode of harvest

used by the local people, where the removal of bark affects the

reproductive performance of trees.

For species like Rytiginia kigeziensis, Spathodea nilotica,

Prunus africana, and Zanthoxylum gilletii whose bark is of

medicinal value, the trend may be attributed to the fact that bark

used for medicinal purposes is mainly harvested from mature

trees and in this case, harvesters have to particularly look for

such trees, in which case they may be forced to move deep into

the forest, if the trees are not near the forest edge. This, for

example, explains why Zanthoxylum gilletii had 14 trees at 0 m

and 6 trees 500 m into the forest (Figure 8). Some of the plants,

especially wildings are destroyed during the harvesting process.

This may for instance explain why Sapium ellipticum had 14

wildings at a distance of 0 m and 02 wildings at 1000 m (Fig-

ure 6). Resource harvesters normally cut the understorey to

create paths. It’s mainly the wildings and saplings that are af-

fected. Other plants are destroyed when trees are cut down.

It is important to note however, that this plant distribution

cannot entirely be attributed to human disturbance. There are

many other factors that influence plant distribution. They in-

clude; the nature of the soil, forest canopy cover, altitude, and

mode of dispersal among others. These factors vary with in the

forest, and therefore shape the distribution of plants in different

parts of forest. If a species is shade tolerant, majority of its

individuals will be found under thick canopy cover, whereas

most individuals of a light tolerant species will be found where

there are gaps. This agrees with findings of (Fayle et al., 2009).

Likewise, a species that is dispersed by self explosion will have

most of its individuals confined to one area of the forest, while

a species that is dispersed by animals may have its individuals

widely spread in the forest. (Bleher et al., 2002) emphasize that

the mode of dispersal of a species influences its distribution in a

forest. Soil conditions influence the rate of germination. More

individuals germinate where the soil conditions are favourable

A. AINE-OMUCUNGUZI ET AL.

212

(Nazre et al., 2009). It is important to note that tree populations

with low densities and clumped distributions are at risk in the

face of increased human dependence on forest products.

Frequency is a good measure of species distribution. Species

with high frequencies are frequently encountered in the forest,

and are widely distributed, while those with very low frequen-

cies are either confined to particular parts of the forest, or are

scarce. This kind of distribution is usually due to high rate of

harvest, nature of soil, mode of dispersal, or altitude. The high

frequency of Teclea nobilis (Table 1) for instance, suggests that

it is widely distributed in the forest, while the low frequency of

Spathodea nilotica is an indication that it is not common in the

forest. Comparing the frequency of wildings, saplings and trees

of a given species gives an insight into the stage of the plant

that is affected most by harvesting. The low frequency of trees

in this study suggests that they are the most affected by har-

vesting, and the high frequency of wildings suggests that they

are the least affected by harvesting, while the saplings are in-

termediate. The high frequency of wildings for most of the

species suggests that they are widely distributed in the forest

and that their numbers are still high due to low levels of harvest.

The low frequency of Cynometra alexandrei wildings com-

pared to other species is an indication that the population of this

species is on the decline. This could probably be attributed to

the high demand of this species for fuel wood. Most of the trees

of this species are harvested before fruiting and this has greatly

reduced its seed bank.

Conclusion and Recommendations

For all the species studied, the low number of trees implies

that there are fewer individuals contributing to the forest seed

bank. If the rate of tree harvest is not regulated, the seed bank

may in the long run be depleted. However, the high number of

wildings and saplings suggests that the population of most of

these species is generally stable. It’s worth noting however, that

human disturbance may in some cases increase resilience of

some species. Effective conservation of these tree species re-

quires strategies that specify the age and size of plants from

which to harvest, and the frequency of harvest. Since wildings

are normally destroyed during the harvesting process, it’s im-

portant to also specify parts of the forest in which harvesting is

allowed. These parts will then need effective monitoring. All

this will only be possible, if the local people are involved in the

decision making process and management of the forest.

Changes in population can be used to show the intensity of

harvesting of a species; however it was quite hard in this par-

ticular study, because no previous data was available to com-

pare populations of particular species in order to determine the

population decline or increase over time. It is therefore hoped

that the results of this study will provide a basis for studying

changes in the population of these ten species. Never the less,

the low population of Zanthoxylum gilletii, Rytiginia kigezien-

sis and Spathodea nilotica, can be attributed to the destructive

methods used to harvest their bark, which in most cases result

into death of the affected plants.

Species with low numbers are mainly those with medicinal

value and they have been massively and destructively harvested

for a long time. This has tremendously affected their distribu-

tion in the forest. They are now left very deep in the forest.

There is a need to sensitize the resource harvesters about re-

sponsible harvesting of plant resources. This will minimize the

destruction of the understorey during the harvesting process.

There is need to incorporate the resource users in the manage-

ment process of Kalinzu Forest Reserve if their negative im-

pacts on the forest are to be minimized. Their incorporation

requires a clear understanding of the socio-economic attributes

that shape their forest resource demands.

Acknowledgements

This work was funded by the Toyota Foundation of Japan

through the Kalinzu Forest Project. We are grateful to Professor

Takeshi Furuichi and Associate Professor Chie Hashimoto of

the Kalinzu Forest Project for the financial and technical help

they gave to us. We are also indebted to the local people around

Kalinzu Forest Reserve for all the assistance extended to us

during the study.

REFERENCES

Aine-Omucunguzi, A., Rugunda, G. K., & Byarugaba, D. (2010). Utili-

zation of forest plant resources by the local people around Kalinzu

forest reserve, south-western Uganda. African Journal of Ecology, 48,

819-821.

Alder, D., & Synnot, T. J. (1992). Permanent sample plot techniques

for mixed tropical forest. Teopical Forestry Paper No. 25, Oxford:

Oxford Forestry Institute.

Bleher, B., Oberrath, R., & Bohning-Gaese, K. (2002). Seed dispersal,

breeding systems, tree density and the spatial pattern of trees—A

simulation approach. Basic and Applied E cology, 3, 115-123.

doi:10.1078/1439-1791-00088

Eilu, G. (1995). The natural regeneration of canopy tree species fol-

lowing human disturbance in Bwindi impenetrable forest, Uganda.

Master’s thesis, Kampala: Makerere University.

Fayle, T. M., Chung, A. Y. C., Dumbrell, A. J., Eggleton, P., & Foster

W. A. (2009). The effects of rain forest canopy architecture on the

distribution of epiphytic ferns (Asplenium spp.) in Sabah Malaysia.

Biotropica, 41, 676-681. doi:10.1111/j.1744-7429.2009.00523.x

Howard, P. (1991). Nature conservation in Uganda’s tropical forest

reserves. Gland and Cambridge: International Union for Conserva-

tion of Nature.

Nakibuka, M. (1994). Some aspects of tropical rain forest regeneration

following encroachment in the Mabira forest reserve. Dissertation,

Kampala: Makerere University.

Nazre, M., Latiff, A., & Mohamad-Roslan, M. K. (2009). Effects of

topography and soil on the distribution of under canopy trees of Gar-

cinia (Guttiferae) in lowland forest of Peninsular Malaysia. Interna-

tional Journal of Botany, 5, 287-294.