ArticlePDF Available

Land use impact on Vitellaria paradoxa C.F. Gaerten. stand structure and distribution patterns: A comparison of Biosphere Reserve of Pendjari in Atacora district in Benin

  • Université Nationale d'Agriculture de Porto-Novo (Bénin)

Abstract and Figures

The shea tree, Vitellaria paradoxa, is a socio-economically important tree for the rural population in parts of West Africa. Our study assessed the current status of this native tree species with regard to increasing human pressure in northern Benin. We compared distribution of adult shea trees, seedlings and saplings in farmed lands with protected areas in the Biosphere Reserve of Pendjari (BRP). At our study site near BRP, agricultural activities foster recruitment of shea trees by regularly cropping of vegetation cover. Furthermore, traditional farming practices preserve adult individuals thus permitting regular fruit harvests. Consequently, most of the tallest and largest individuals of shea trees are found in framed lands. In contrast, the highest density of juvenile trees including seedlings (dbh <5cm) and saplings (dbh 5–10cm) occurred within BRP. Saplings were negatively affected by farming activities. Furthermore, spatial point pattern analysis revealed differences in the spatial structure of juveniles. Juveniles showed significant aggregations at small scale (<20m) in BRP as well as significant and positive small-scale associations with adult trees. This contrasts with farmed lands where we did not find such spatial patterns at similar small scale but only a weak aggregation between juveniles and absence of association (attraction) of adults to juveniles. Although our analyses indicate that shea trees are rather well preserved, we conclude that the observed severe reduction of saplings in farmed lands is likely to negatively impact the long-term viability of the tree population. Therefore agroforestry practices must consider the preservation of sapling populations in farming areas for long-term conservation.
Content may be subject to copyright.
Land use impact on Vitellaria paradoxa C.F. Gaerten. stand
structure and distribution patterns: a comparison
of Biosphere Reserve of Pendjari in Atacora district
in Benin
Bruno Agossou Djossa ÆJakob Fahr ÆT. Wiegand Æ
B. E. Ayihoue
´nou ÆE. K. Kalko ÆB. A. Sinsin
Received: 7 November 2006 / Accepted: 2 October 2007 / Published online: 27 October 2007
ÓSpringer Science+Business Media B.V. 2007
Abstract The shea tree, Vitellaria paradoxa, is a
socio-economically important tree for the rural pop-
ulation in parts of West Africa. Our study assessed
the current status of this native tree species with
regard to increasing human pressure in northern
Benin. We compared distribution of adult shea trees,
seedlings and saplings in farmed lands with protected
areas in the Biosphere Reserve of Pendjari (BRP). At
our study site near BRP, agricultural activities foster
recruitment of shea trees by regularly cropping of
vegetation cover. Furthermore, traditional farming
practices preserve adult individuals thus permitting
regular fruit harvests. Consequently, most of the
tallest and largest individuals of shea trees are found
in framed lands. In contrast, the highest density of
juvenile trees including seedlings (dbh \5 cm) and
saplings (dbh 5–10 cm) occurred within BRP. Sap-
lings were negatively affected by farming activities.
Furthermore, spatial point pattern analysis revealed
differences in the spatial structure of juveniles.
Juveniles showed significant aggregations at small
scale (\20 m) in BRP as well as significant and
positive small-scale associations with adult trees.
This contrasts with farmed lands where we did not
find such spatial patterns at similar small scale but
only a weak aggregation between juveniles and
absence of association (attraction) of adults to
juveniles. Although our analyses indicate that shea
trees are rather well preserved, we conclude that the
observed severe reduction of saplings in farmed lands
is likely to negatively impact the long-term viability
of the tree population. Therefore agroforestry practices
must consider the preservation of sapling populations
in farming areas for long-term conservation.
Keywords Benin Conservation Shea tree
Spatial analysis
Traditional agroforestry parkland systems, where
farmers grow annual crops in fields with scattered
trees, are among the most widespread systems in sub-
Saharan Africa (Teklehaimanot 2003). The agrofor-
estry system in northern Benin as well as in Mali and
Burkina Faso is built around the selection of desired
B. A. Djossa (&)B. E. Ayihoue
´nou B. A. Sinsin
´des Sciences Agronomiques /UAC, Laboratoire
d’Ecologie Applique
´e, BP 910, Abomey-Calavi, Cotonou,
Littoral, Be
J. Fahr E. K. Kalko
Institute of Experimental Ecology, University of Ulm,
Ulm, Germany
T. Wiegand
Department of Ecological Modelling, UFZ Leipzig, Halle,
Leipzig, Germany
E. K. Kalko
Smithsonian Tropical Research Institute, Balboa, Panama
Agroforest Syst (2008) 72:205–220
DOI 10.1007/s10457-007-9097-y
tree species and individuals (Maranz and Wiesma
2003). The indigenous parkland farming system
consists of alternating cycles of cultivation and
fallows where natural regeneration of woody plants
occurs around preserved trees (Boffa 1999; Lovett
and Haq 2000) like shea trees (Vitellaria paradoxa).
Therefore, phytodiversity in parklands and old fal-
lows is mostly dominated by tree species that are
useful for the local population. For example, density
of the economically important shea tree averages 15
mature trees per hectare in Mali (Ruyssen 1957). It
constitutes up to 70% of the woody vegetation in
some areas of Benin (Agbahungba and Depommier
1989) and may reach over 80% in parts of northern
Ghana (Lovett and Haq 2000) and Burkina Faso
(Boffa 1999). Shea trees also dominate our study area
in the Pendjari region situated in northwestern Benin.
Shea trees are a valuable source of a variety of
non-timber forest products namely fruit and the
nutrient rich seeds which serve as the basis for
cooking oil as well as the production of shea butter.
The latter is of substantial economic value as it is
used for exportation because is ingredient of many
cosmetics (Tran 1984). Women mainly exploit fruits
and seeds. Shea tree products procure about 80% of
their incomes in Mali (
and in Benin. Agbahungba and Dopommier (1989)
reported that women gained more income from shea
trees than from work in cotton fields.
Although there is an apparent good preservation
status of adult shea trees in farmed lands, the tree is
on a constant decline on the population level. The
Sudanian savanna zone, for instance, had the highest
density of shea trees in the 1940s, with a population
of 230 trees ha
(Chevalier 1946). This density has
been seriously reduced nowadays to a minimum of
11 trees ha
(Nikiema et al.2001) most likely
because of profound changes in farmers’ practices.
Recently, Kelly et al.(2004) studied the impact of
farmers’ practices on size class distribution and
spatial pattern of shea trees in Mali and reported that
the spatial pattern of V. paradoxa is becoming
progressively aggregated with a gradient ranging
from cultivated fields to fallow and finally to forest.
Similar trends can be observed on the local level.
With regard to the socio–economic importance of this
plant in northern Benin in general and particularly in
areas surrounding the Biosphere Reserve of Pendjari
(BPR), shea trees are better preserved compared to
other woody plants. However, increasing demographic
pressure shortens the fallow cycle and leads to changes
in farming patterns, in particular small- to large-scale
conversion of land into cotton fields. Both factors
seriously affect the regeneration of shea trees. As this
important tree species is not cultivated by the local
population, it depends heavily on natural regeneration,
where seed dispersal by flying foxes (Chiroptera) and
birds (Ruyssen 1957; Jackson 1968; Ayensu 1974)
plays a crucial role (Killick 1959). Flying foxes are
often the only remaining group of medium-sized
animal dispersers in these areas as similar to many
areas in Africa, large birds and mammals are often
hunted out (Fa et al.2000,2002,2005; Thibault and
Blaney 2003).
In our study we assessed and compared demographic
and morphological (dendrometic) characteristics of
shea trees including density, tree height, crown area
and general distribution patterns in farmed lands versus
protected areas. We hypothesized that intense land use
negatively affects the density of shea trees, leading to
modifications in their distribution pattern because
agricultural activities might require less vegetation
coverage. Diminution in tree density and soil fertiliza-
tion for agricultural needs favor adult trees’ height and/
or crown areas as well as natural regeneration because
seedlings demand light. We also assumed (hypothe-
sized) that adult shea trees should show a clumped or
random distribution pattern in protected areas and more
regular distributions where they are heavily influenced
by humans’ activities because of farming practices that
go with a reduction in vegetation density. Criterions like
need of space, growth speed, health, morphology, and
productivity lead to a selective cutting of individuals of
shea trees in farmed lands (Lovett and Haq 2000).
Juvenile individuals might show similar distribution
pattern like adult trees because they grow around adult
trees and are treated similarly. As the cropping of
vegetation leads to the preservationof only few selected
juvenile plants this should also influence the spatial
distribution pattern of juvenile plants around adults.
Overall, we expect that adult and juvenile shea trees
should show a positive association (attraction) in
protected areas without cropping and a negative
association (repulsion) where cropping of the vegeta-
tion occurs through agricultural activities.
In this study we assess how farming activities
affect structure and spatial distribution of shea tree
populations. We focus on stand density, distribution
206 Agroforest Syst (2008) 72:205–220
of age classes, tree height and crown area and
compare tree populations in areas of farmed lands
surrounding the BRP with those of plants from
protected areas inside the BRP. Ultimately, the main
goal of this study was to deliver data useful for
predicting the future of this important tree species in
the region and to make recommendations for its long-
term preservation.
Materials and methods
Study environment
Our study took place in the vicinity and within the BRP
(Fig. 1a, b). The region is located in the extreme
northwest of Benin in the district of Atacora (10°400
11°280N and 0°570–2°100E). It covers an area of
4,661.4 km
and is composed of the National Park of
Pendjari (2,660.4 km
), the hunting zone of Pendjari
(1,750 km
) and the hunting zone of Konkombri
(251 km
). The Pendjari is the only important river that
carries water all year round. It runs through the National
Park of Pendjari and the Pendjari hunting zone (Delving
et al.1989). Other small streams dry out in the dry
season including Magou, Bori, and Yapiti in the hunting
zone of Pendjari and Podiega in the National Park.
The park is located in the Sudanian zone with one
rainy season (April–May to October). The total
rainfall averages 1,000 mm with 60% falling between
July and September (Sinsin et al.2002) and one dry
season (November to March). During the rainy
season large parts of the park are flooded. The
dominant vegetation type is savanna intermingled
with some patches of dry forests composed of
deciduous trees (Sokpon et al. 2001). Temperature
varies from around 21°C during the night up to
around 40°C during the day. The annual mean varies
from 25 to 28°C during the cooler period of the dry
season and 30–33°C during the hot period of the dry
season. Relative humidity varies between 17 and 99%
(PAG2 2005). The border of Biosphere Reserve of
Pendjari is lined with many small villages. Population
density in this area is low compared with the whole
country. It totals about 213,000 persons leading to
13 person per km
(INSAE/RGPH 2002). In the area
surrounding the BRP Berba, Waama, and Gour-
´are the dominant ethnic groups. The main
activity of the local population is agriculture and
herding with the Fulani as the ethnic group with the
largest herds of cattle. Women mainly exploit non-
timber forest resources like fruits of shea trees.
Studied species
Vitellaria paradoxa C.F. Gaertn. (Sapotaceae), the
shea tree, is the source of vegetable oil obtained from
the kernels which is second in importance only to palm
oil in West Africa. Shea trees occur in a savanna belt
stretching across Africa from Senegal to the border of
Sudan and Ethiopia. A wide variety of economically
useful products are made from this tree. The oil
derived from the kernels of shea fruits (‘‘shea butter’’)
is by far the most important product. Women through a
laborious and time-consuming process traditionally
extract it. From early on, this oil has been an important
product in international trading activities in West
Africa. After the First World War, shea butter
contributed substantially to the flourishing trade
between West Africa and Europe, where it was used
in the production of vegetable margarine and in candle
making (Dudgeon 1922). In 1998 Benin exported
1,000 tonnes of shea butter and gained 400,000 US$.
From central to northern part of Benin this tree
species remains the most important source of non-
timber resources for the rural population. Fruits from
shea trees and the butter form a large portion of the
diet and income of local households.
Despite its socio–economic importance, the culti-
vation of this tree species is not yet mastered by local
populations who rely natural regeneration.
Inventory approach and data collection
Fourteen 1 ha-plots of *100 9100 m were installed
in homogeneous and comparable habitat types, com-
prising seven plots in farmed lands, i.e., fallows
(1–3 years old) in village territories where farming
activitieshad been previously conducted for a numberof
years and seven plots in protected areas inside the BRP
(Fig. 1c). Plots were established randomly. In protected
areas they were located in the Pendjari National Park
and in the buffer zone (hunting zone) of Pendjari.
An individual plot consisted of a rectangle of
100 m for each side (1 ha). If some individual shea
trees occurred next to the delineated plots, i.e., within
Agroforest Syst (2008) 72:205–220 207
20–30 m, they were also included in the census to
obtain comprehensive mapping of clusters. The
positions of all individuals were recorded with GPS
(GARMIN 76, Resolution 5 m). Plots in farmed lands
were named T01 to T07 but C01 to C03 and P01 to
P04 inside the BRP.
Data collection was conducted from September to
October that coincided with the end of the rainy
season allowing re-growth of all burned shea saplings
and seedlings.
All V. paradoxa individuals were mapped and
measured within each plot except seedlings with
Fig. 1 Study area, (a) Benin in West Africa, (b) map of Benin with BRP, and (c) with the positions of study plots in the BRP. T
indicates plots in the village area, Cplots in the hunting zone and Pplots of the Pendjari National Park
208 Agroforest Syst (2008) 72:205–220
dbh B1 cm as many of them are typically burned in
the following dry season. The diameter at breast
height (dbh) of the trees was measured at 1.30 m
above ground with ribbon ‘‘pi’ (specific ribbon used
by foresters to measure tree dbh) of 1-cm precision.
Seedlings were measured near ground level. Trees
with dbh [10 cm were grouped in the adult category
according to foresters’ classification (Rondeux 1999)
and all trees with a dbh \10 cm were grouped in the
juvenile category, which comprises seedlings (dbh
1–5 cm) and saplings (5 cm Cdbh \10 cm). Addi-
tionally, crown area radius was measured with ribbon
of 5 m length scaled in centimeter (precision 1 cm).
One operator put one end of the ribbon on the tree
and second operator push the second end toward the
vertical end of the crown area. Tree height was
measured with ‘‘dendrome
`tre’’ SUUNTO with for-
mula h=L[tg (ß
)-tg (ß
)], were Lmeans
distance between operator and the tree, ß
angle of observation toward the top of the tree, and ß
means angle of observation of the foot of the tree.
Coordinates were determined with a handheld GPS
receiver (GARMIN, GPS 76, Resolution 5 m).
Coordinates were used to analyze the spatial distri-
bution pattern of the shea tree species.
Spatial analysis
We used the pair-correlation function (Stoyan and
Stoyan 1994) and G(y), the distribution of nearest
neighbor distances y(Diggle 2003), to investigate (a)
the type of distribution pattern shown by adult and
juvenile plants of V. paradoxa (univariate point
pattern) and (b) the type of association between adult
and juvenile individuals (bivariate point pattern). The
pair-correlation function g(r) is an intensity-normal-
ized neighborhood density function indicating the
density of neighboring trees as a function of distance
rfrom an average tree (Wiegand and Moloney 2004)
divided by the intensity kof the pattern. The pair-
correlation function is a non-accumulative version
of Ripley’s K-function (Ripley 1981) satisfying
g(r)=(1/2 pr)dK(r)/dr. Hence, it does not integrate
the ‘memory’ (= trend at the beginning) of small-
scale’s second-order effects to larger scales, as does
Ripley’s K(Wiegand and Moloney 2004). Addition-
ally it is more intuitive than an accumulative measure
(Stoyan and Penttinen 2000) because it has a direct
interpretation as normalized neighborhood density
The pair-correlation function and the distribution
of nearest neighbor distances Gcan also be formu-
lated for bivariate point patterns. The bivariate g
is the normalized density of neighboring type 2 trees
(= pattern 2 represented here by juvenile trees) as a
function of distance rfrom an average type 1 trees (=
pattern 1 represented here by adult trees). The
bivariate Gdescribes the distribution of the distances
from type 1 trees to its nearest type 2 neighbors. Note
that g(r) and G(y) measure different features of the
point pattern (Diggle 2003). The estimator of gand
and the edge correction used are provided in the
appendix. G(y) and G
(y) were calculated without
edge correction (Diggle 2003).
To determine statistical significance of the observed
g(r), g
(r), G(y)orG
(y), 1% simulation envelopes
of an appropriate null model were generated by 999
replicate Monte Carlo simulations of the null model.
To this end, we used the 995th largest and the 995th
smallest value, for example, g(r), taken from 999
simulations of a null model as simulation envelopes
(Wiegand and Moloney 2004). In case that g(r) was for
a given scale routside the simulation envelopes, the
null hypothesis was rejected at this scale; i.e., for an
univariate point patterns with intensity k,g(r)[k
indicates clumping, while values of g(r)\kindicates
regularity. For a bivariate point pattern with intensity
of pattern 2, g
indicates a positive
association (attraction) whereas g
a negative association (repulsion).
However, because of simultaneous inference (i.e.,
we tested rat several spatial scales simultaneously),
type I error may occur if the value of the chosen
statistic is close to a simulation envelope (i.e., the
null model may be rejected even if it is true;
Loosmore and Ford 2006). We therefore combined
the common simulation envelop method with a
goodness-of-fit test (Diggle 2003) recently advocated
by Loosmore and Ford (2006). In short, the goodness-
of-fit test collapses the scale dependent information
contained for instance in g(r) into a single test
statistic u
which represents the total squared devia-
tion between the observed pattern and the theoretical
result across the distances of interest (i=1 for the
data, and i=2,...1,000 for the simulations). If the
rank of u
is larger than 990, the summary statistic
value, u
, calculated for the observed pattern for the
Agroforest Syst (2008) 72:205–220 209
chosen statistic (Gor g) over a specified range of
distances is not different than those calculated for
random instances of the hypothesized spatial process
model on a 1% level.
For each treatment the data from the replicate plots
were combined into one overall, mean weighted pair-
correlation function (Diggle 2003; Riginos et al.
2005). Due to the low density of shea trees in the
study area, we were not able to perform point pattern
analysis for most of the individual plots. However,
combining the information about the point-to-point
distances from all replicate plots increased our
sample size considerably. For more details see
Null models
As we choose our plots in uniform and homogeneous
areas without apparent environmental heterogeneity,
we applied complete spatial randomness (CSR;
Diggle 2003; Wiegand and Moloney 2004) as null
model to assess the spatial structure of the patterns of
adult and juvenile trees.
In the bivariate pattern, when focusing on the
relation between juvenile and adult trees, we assessed
independence between both patterns (Goreaud and
Pelissier 2003). To this end we applied a random
toroidal shift of pattern 2 thus conserving the
individual spatial structures of the two patterns, but
breaking their possible dependence.
In all analyses we used a grid size of 1 m
, which
is a fine enough resolution for our questions
(Wiegand and Moloney 2004), and a ring width of
2 m. Because the point density was low, a smaller
ring width would produce rugged plots of g(r)
and wider rings would not allow separation of
different distance classes. We calculated the statistics
up to a scale of 50 m taking into account plots size
(100 9100 m).
Data analysis
To analyze the density of shea trees we averaged the
total number of individuals per age class plus
standard deviation (X
¯±SD) in the seven 1 ha plots
in the farmed lands and the seven others in protected
areas of the BRP. We distinguished three age classes
(seedlings, saplings, and adults) to compare densities
between zones (farmed lands and protected area in
BRP). We used Chi-square test to compare the
proportion of individuals in each class between the
two zones. In addition, we choose a One-Way
ANOVA to depict possible differences in age, height,
and crown area class distribution between farmed
lands and protected areas.
For all point pattern analyses we used the software
(Programita, Wiegand and Moloney 2004 improved
version of 2007). Standard statistical analyses were
conducted with SigmaStat 3.1 and Statistica 6.0.
Spatial analysis was performed for first order inten-
sity (Fig. 3) with ArcView GIS 3.2.
Impact of land use on tree density
We mapped and measured (dbh, height and crown
area) 178 adult and 346 juvenile trees in an area of
14 ha. Shea trees’ population structure is presented
per hectare in the protected areas (BRP) and in the
farmed lands (Table 1).
The proportions of juvenile trees (seedlings and
saplings) differed significantly between farmed lands
and protected areas (Table 1). There were more
seedlings and saplings in BRP than in farmed lands
but more adult shea trees in these areas than in BRP.
Impact of land use on stand structure of shea trees
Recruitment and number of young shea trees were
high in farmed lands and in protected areas (Fig. 2a).
However, the number of saplings was drastically
reduced in the farmed lands. The distribution of tree
diameters showed an opposite trend in farmed lands
compared to the protected areas. The protected area
was characterized by large numbers of medium
height trees (2–4 m). The tallest and hence oldest
trees (4–6 m and [6 m), however, were mainly found
in the farmed lands (Fig. 2b). Those trees also
attained larger crown areas (2–4 and 4–6 m) than
those occurring in the protected areas (Fig. 2c).
The distribution of diameter classes (dbh) revealed
a very low number of saplings (dbh 5–10 cm) in
farmed lands leading to a statistically significant
210 Agroforest Syst (2008) 72:205–220
difference between both zones. A similar result was
found for the crown area of the class of the largest
individuals. There were no differences in tree height
when data analysis was performed with a One-Way
ANOVA (Table 2).
Impact of land use on spatial distribution patterns
of shea trees
Typical examples of the spatial distribution of
juvenile and adult trees in the two zones show
differences in spatial distribution between zones with
more juveniles near adult trees in plot from BRP
(Fig. 3). Spatial analysis of distribution pattern of
shea trees revealed differences between farmed lands
(Fig. 4a, c, e) and protected areas (Fig. 4b, d, f).
The pair-correlation function of adult trees in the
farmed lands showed only a weakly significant
aggregation (the GoF for scales r=0–50 m yielded
a rank of 997) with peaks around 6 and 15 m
(Table 3, Fig. 4a). However, for the protected areas
the CSR null model was overwhelmingly rejected
(Fig. 4b), here the adult trees showed a strong
aggregation up to 3 m and notable aggregation up
to 10 m. Consequently, the GoF test yielded a rank of
1,000. The distribution of nearest neighbor distances
were in both zones significantly different from the
CSR null model (yielding rank 1,000), but comparing
the distribution of the farmed lands (small insert in
Fig. 4a) with the distribution of the protected areas
(small insert in Fig. 4b) shows that adult trees in the
farmed lands have much less nearest neighbors at
small distances up to 4 m. Thus, adult shea trees in
the farmed lands have at small scales a clearly more
regular distribution than shea trees in the protected
areas. This finding supports the hypothesis that in the
farmed lands only selected large shea trees are
The spatial pattern of juveniles was similar in the
farmed lands and protected zones (Fig. 4c, d); in both
cases there was clear aggregation with the GoF test
yielding ranks of 1,000 for both statistics gand G.
The aggregation had a range of some 12 m.
The spatial relationship between juveniles and
adults differed strongly between farmed lands and
protected areas. Juveniles at the farmed lands were
independent from adults (Fig. 4e; the GoF test for
gyielded a rank of 988), whereas the null model
Table 1 Distribution of shea trees in each 1 ha plot within the protected area (BRP; n= 7) and in Farmed lands around villages (n= 7) in Atocora district in Benin
Age classes dbh (cm) Farmed lands (X ±SD
) Number of shea trees BRP (X ±SD
) Number of shea trees Statistical analyses (Yates correction)
Seedlings (1–5) 19.7 ±28.4 118.0 21.3 ±13.8 128.0 v
=11.99, df =1, P=0.005
Saplings (5–10) 4.5 ±3.8 27.0 12.2 ±11.1 73.0
Adults ([10) 20.8 ±18.6 125.0 8.8 ±6.9 53.0 v
=18.13, df =1, P\0.01
Total 270.0 254.0
Standard deviation
Agroforest Syst (2008) 72:205–220 211
of independence was overwhelmingly rejected for
the protected areas (Fig. 4f; the GoF test for g
yielded a rank of 1,000). Juveniles were signifi-
cantly clustered around adults up to some 6 m
(Fig. 4f). Similarly, the GoF test for the distribution
of nearest neighbor distances yielded a rank of 999
for the farmed lands and 1,000 for the protected
Table 2 One-Way ANOVA on dbh, tree height and tree crown radius classes of farmed lands and BRP areas in Atocora district in
Analysis of variance
Marked significant effect with P\0.05
Effect Effect Effect Error Error Error FP
SC df MC SC df MC
dbh 1–5 cm 168.7 1 168.7 5,490.2 10 549.0 0.3 0.6
dbh 5–10 cm 645.3 1 645.3 802.7 10 80.3 8.0 0.01
dbh 10–20 cm 85.3 1 85.3 971.7 10 97.2 0.9 0.4
dbh 20–30 cm 21.3 1 21.3 121.3 10 12.1 1.7 0.2
dbh 30–40 cm+ 10.1 1 10.1 36.8 10 3.7 2.7 0.1
Height \2 m 6.7 1 6.7 3,911.5 10 391.2 0.02 0.9
Height 2–4 m 4.1 1 4.1 1,302.8 10 130.3 0.03 0.8
Height 4–6 8.3 1 8.3 1,121.7 10 112.2 0.1 0.8
Height 6 m+ 1.3 1 1.3 66.3 10 6.6 0.2 0.7
Crown radius \2 m 60.7 1 60.7 9,269.5 10 926.9 0.1 0.8
Crown radius 2–4 m 108 1 108 632.7 10 63.3 1.7 0.2
Crown radius 4–6 m 4.1 1 4.1 6.2 10 0.6 6.6 0.03
Significant differences (P\0.05) are indicated in bold
Farmed lands BRP
Classe de diamètre[cm]
Stem density [tree / ha]Stem density [tree / ha]
Stem density [tree / ha]
Farmed lands BRP
Farmed lands BRP
height < 2 height 2-4 height 4-6 height 6+
Class of height [m]
crown < 2 crown 2-4 crown 4-6
Class of crown [m]
Fig. 2 Frequency of
diameter (a), height (b), and
crown radius (c) of shea
trees in farmed lands and in
BRP in Atacora district in
Benin. Statistical
significance difference is
marked with an asterisk
212 Agroforest Syst (2008) 72:205–220
Fig. 4 Results of the point
pattern analyses. The main
figures show the results of
the univariate analyses of
adults (a,b), juveniles
(c,d), and the bivariate
analyses of juveniles around
adults (e,f). The small
insert figures show the
corresponding results for
the distribution function of
nearest neighbor distances.
Filled circles: statistic
calculated from the data,
gray solid lines: simulation
envelopes being the 995
lowest and highest values of
the 999 simulations of the
null model, and the solid
horizontal black lines show
the average statistic taken
from the 999 simulations of
the null model (i.e., the
theoretical expectation)
Fig. 3 Example of two
plots showing the
distribution of adults (open
circle) and juveniles (closed
circles) of shea trees in the
farmed lands (a) and in the
BRP (b) in Atacora district
in Benin
Agroforest Syst (2008) 72:205–220 213
Shea tree population under anthropic disturbances
Our study confirmed that land use impacts woody
vegetation, particularly the distribution of shea trees
in the Pendjari region. Small shea trees (dbh \10
cm) were still frequent in the farmed lands, most
likely because the local population intentionally
preserves adult, reproductive shea trees. This allows
natural regeneration around the adult trees (Boffa
1999; Lovett and Haq 2000). In agricultural areas still
farmed in more traditional ways, overall vegetation
coverage is low because of regular cropping. Laris
and Wardell (2006) confirmed this finding with their
study on fires in Mali when they stated that the
expansion of the area under agriculture coupled with
a reduction in the fallow cycle is a more probable
cause of tree cover decline. This facilitates establish-
ment of shea tree seedlings. However, we found that
farming activities have a negative effect on saplings
(Fig. 3a). The overall density of juveniles was lower
in farmed lands than in protected areas. This result is
in accordance with results from a study carried out in
Mali, an other West African country, where individ-
uals of small age classes occurred at lower numbers
in village parklands and young fallows when com-
pared to old fallows and forests where individuals of
small age classes were found in larger numbers
(Kelly et al.2004). Seedlings need light for growth so
that disturbance should foster regeneration; in
addition, low density of woody plants in farmed
lands reduces intra and interspecific competition and
soil fertilization for agriculture also provide nutrients
for seedlings. Despite these conditions we found an
important reduction of the density of saplings (dbh
5–10 cm) in these areas.
At least two explanations can be given for those
findings. First, fallow cycles are more and more
shortened because of increasing demographic pres-
sure where farmers constitute the largest part of the
growing human population. Second, when farmers
return to the fallows, they often cut the majority of
shea trees that are\2 m in height preserving only the
tallest ones (personal observations) because only
these individuals could start fruit production in a near
Although the juvenile trees are easily identified by
the local population and although the socio-economic
importance of this plant is known, the farmers cut
some of the juveniles of shea trees together with other
juveniles of undesired trees when cropping the
vegetation cover. They preserve mainly the mature
shea trees as they are of higher immediate value as
they regularly produce large amounts of fruit.
Furthermore, planning of the local population is
mostly limited to their daily needs but very often not
directed into the long-term future, knowing for
example that it takes a long time until shea trees
become productive. Indeed shea trees take 15–
25 years to grow to maturity. They can live up to
200–300 years (Hall et al.1996). Therefore, the
Table 3 Summary of point pattern analyses of shea trees and null models for juveniles, adults, and juveniles/adults in farmed lands
and BRP in Atacora district in Benin
Pattern Results of spatial point pattern
analyses at specific scales
Rank GoF test for 0–50 m Figures
12 gG
Farmed lands
Adult Weak aggregation, peaks at 6 and 14 m 997 1,000 4a
Juvenile Clear aggregation up to 11 m 1,000 1,000 4c
Adult Juvenile Distribution of juvenile trees is
independent from adult trees,
but weak aggregation in G
988 999 4e
Adult Clear aggregation up to 13 m 1,000 1,000 4b
Juvenile Clear aggregation up to 12 m 1,000 1,000 4d
Adult Juvenile Positive association (attraction) at scale 0–5 m 1,000 1,000 4f
gpair correlation function, Gdistribution of nearest neighbor distances
214 Agroforest Syst (2008) 72:205–220
current reduction of the density of sapling by
agricultural practices does not immediately translate
into a decline of the adult population but will
certainly be noticeable within a few decades when
the adult trees have reached old age and will
subsequently die without adequate replacements.
Our data also revealed that most of the tall trees
with large and extensive crowns occur in farmed
lands. These findings are supported by Lovett and
Haq (2000) who found in the Gonja districut of
northern Ghana that anthropic selection of shea trees
through elimination of other undesired trees as well
as selected shea trees taking into account criteria such
as spacing, growth, size, age, health, and productivity
lead to an overrepresentation of large trees on farmed
lands in comparison to smaller trees. This situation
underlines the lack of appropriate management of this
valuable natural resource where only exploitation is
put into the foreground without consideration of long-
term conservation of this resource for future gener-
ations. Explanation oriented toward the necessity of
preserving enough saplings around adult trees in
order to maintain similar shea tree population density
have to be done with local human population in these
areas to guaranty long-term conservation of this plant
Spatial distribution of shea tree with regard to
land use
Regarding the spatial distribution pattern of shea
trees under human impact, we documented a weakly
aggregated distribution pattern of adult trees and no
association between adults and juveniles in the
farmed lands. This result shows that human influence
on this plant species does not produce, at least now,
regular distribution patterns or repulsion between
adult and juvenile trees. Because as hypothesized,
aggregation/attraction goes with absence or minimum
disturbance whereas regular/repulsion means intense
disturbance. Adult trees were much less aggregated
than those in protected areas and juveniles were
significantly clustered around adults in the protected
areas. Thus, human impact made the patterns clearly
more regular.
Since juveniles are the offspring of adult trees,
we expected a clumped distribution pattern of
juveniles and a positive association between
juveniles and adult trees under protected conditions
because juveniles are supposed to establish around
adult trees. Fruit bats in these areas actively disperse
shea seeds, as shea fruit constitute a key resource
for them (B. A. Djossa, submitted data). Inside the
protected areas of the BRP where plots were
established, humans do not harvest fruit. Fruits are
removed only by animals (primates, elephants,
rodents, birds, and bats) that consume part of the
fruit crop which is produced during a season. As the
animals do not fully deplete the tree, enough fruits
and hence seeds remain available as source for
natural regeneration. Frugivores that then take the
fruit and hence the seeds away and eat it in close
proximity of the tree leading to the positive
association of parent tree and juvenile. Here the
parent tree acts so to speak as a source. The
clumped distribution of the juveniles might then be
the results of feeding roosts where animals drop
then several seeds on the forest floor. These seeded
fruit cannot be carried too far from parent trees and
cannot pass alimentary canal so that seeds are
simply dropped in the vicinity. Some seeds germi-
nate even next to the parent tree, thus creating a
clumped distribution. Taking humans into account
with their use of shea fruits, they usually harvest
almost all fruit and thus reduce natural regeneration
because there are few fruits left. In addition, when
natural regeneration occurs with the remained seed
bank, agricultural activities also reduce saplings.
Kelly et al.(2004) carried out a similar study in
Mali and concluded that the spatial pattern of V.
paradoxa became progressively more aggregated
from cultivated fields to fallows and then to forest.
Kelly et al.(2004) concluded that differences in
aggregation between sites were mostly explained by
the intensity of fruit production and subsequent
recruitment whereas increasing regularity of tree
distribution found in farmed lands and fallows was
due to human intervention. In our study we only
considered fallows and protected areas (equivalent
to forest in their study) because we were concerned
with natural regeneration of the target tree species
whereas Kelly et al.(2004) did not analyze seed-
lings (diameter girth at breast height from 20 cm
equivalent to dbh [5 cm). Our study had better
resolution on population structure.
We found the largest individuals mainly in farmed
lands whereas the smallest individuals occurred more
Agroforest Syst (2008) 72:205–220 215
frequently in protected areas. Our spatial pattern
analysis provides evidence that disturbance may
cause more regular distributions. In accordance with
our findings, a regular distribution of trees was found
in Mali at the most cultivated site whereas trees were
more aggregated at the site with fewer disturbances.
The lack of a spatial association between juvenile
and adult trees in farmed lands may also be caused by
the fact that trees in fields are generally much larger
than in protected areas with extensive crown areas
that affect seedlings negatively through out-shading
since V. paradoxa is considered a light demanding
species. Consequently, seedlings could be negatively
affected when they are too close to adult trees
because of competition for light and nutrients (Hall
et al.1996). The Janzen–Connell model that states
that seeds fall under parent trees and subsequent
seedlings are under the pressure of predators and
pathogens of parents that reduce there density (Janzen
1970 and Connell 1971; Condit et al.1992, 1996;
Hyatt et al.2003) is not applicable here because most
seeds are collected by women so that seed predation
by domestic animals is unlikely and pathogens like
vegetal parasites (Tapinanthus sp.) were found
exclusively on adult trees but not on seedlings. Inside
the protected areas (BRP) the absence of a clear
aggregation in the distribution pattern of individuals
and the association of juvenile and adult trees at large
scales can be explained by the Janzen–Connell model
because seeds are removed by seed predators such as
rodents and primates (i.e., baboons) which are often
seen eating fruits of shea trees. This supports
colonization hypothesis of Janzen–Connell that states
the negative correlation between the amount of seed
transported from parent trees and distance leading to
a reduction of density of offspring.
Conservation status
In terms of the conservation status of shea trees, the
negative effect of intense local agricultural practices
by humans might push the distribution pattern of the
trees from aggregation to regularity (univariate) or
from attraction to repulsion (bivariate), finally lead-
ing toward the distribution of only few isolated
individuals. As we still found random (univarite) and
independent (bivariate) distribution patterns of shea
trees in farmed lands, we conclude that shea trees are
still rather well preserved. However, additional
studies are crucial to appreciate the population
dynamics of shea trees linking for example age class
with seed production, seedling establishment and
finally viability of juveniles in time (in French
‘matrice de transition’). This information is neces-
sary to appreciate in more detail the conservation
status of this plant species in this region.
Many authors reported that trees in tropical
ecosystems show in general random or clumped
distribution pattern (Condit et al.1992,2000; Wills
et al.1997; Hubbell et al.1999). Our findings are in
accordance with these findings. Aggregation was
mostly limited to neighborhood scales of about 12 m.
This may be related to seed dispersal where seeds are
dispersed for short distances away from the parent
tree and thus rapidly decrease in density with
increasing distance to the source (Harper 1977; Howe
and Smallwood 1982; Willson 1993 in Nathan and
Casagrandi 2004). In the BRP, the annual burning of
the vegetation can also play an important role with
regard to the distribution pattern and survival rates of
seedlings, in particular in years with a lot of dry
biomass. Bruner et al.(2001) pointed out the effec-
tiveness of park in biodiversity conservation and
negative effects of fire with regard to biodiversity
conservation. Fire is annually used inside the BRP as
a management tool mainly for renewal of resources
(i.e., pastures) for native herbivores such as antelopes
(see Sinsin and Saı
¨dou 1998). Prescribed burning is
used at the end of the rainy season when the biomass
still not very dried in order to avoid intense fires that
could be detrimental to plants, soil microorganisms as
well as a large number of invertebrates and verte-
brates. Nikiema (2005) reported from Burkina Faso
that early fire (end of rainy season) is less violent.
However, fire is a problem for the fragile seedlings of
shea trees (Grime 2001;Ho
¨lzel 2005) because if the
seedling is not completely burned it has to renew its
stem growth and leaves each season, generally
starting from near ground level. This slows down
firm seedling establishment and growth. Even for
mature trees, the annual vegetation burning causes
the smaller tree crowns as noticed in BRP compared
with farmed lands where vegetation fire also exist but
with less biomass to burn and thus fewer damage.
Moreover, similar to what Laris and Wardell (2006)
reported from Mali, early fire is used by local
population living around BRP to reduce risks of
216 Agroforest Syst (2008) 72:205–220
damaging trees preserved or planted by them. Our
findings are supported by Hjerpe et al.(2001) who
reported that 5 years after a cyclone and fire had
impacted a rain forest reserve on the Tafua peninsula
Savai’I in Samoa, canopy cover increased less in
burned area than in cyclone destroyed area. The fire
favored fast growing pioneer species more than the
cyclone leading to higher mean numbers of species
per plot in the unburned area compared to the burned
area. We derive from those results that the small size
of the shea trees (dbh and height of productive trees)
found in BRP zones compared to farmed lands could
also be attributed in part to the effects of the annual
fire that stresses trees and slows down their growth
even if fire stress has also been recognized to produce
positive and fast phenological activities on shea trees
(Hall et al.1996) so that leaves renewing and
flowering start short after vegetation burning.
This study assessed the distribution of shea trees
(V. paradoxa) in protected areas (BRP) and in farmed
lands. Overall, shea trees are well preserved in
parklands. Furthermore, traditional farming activities
foster the growth of young shea trees (seedlings) by
reducing potential competition with other tree spe-
cies. This leads to parklands that are dominated by
shea trees. As a consequence of the traditional
agricultural practices, most of the large, mature
individuals are in farmed lands. Point pattern analysis
of the mapped tree locations in plots in farmed lands
revealed random distribution patterns for adult and
juvenile trees and no association of juveniles to
adults. In plots located in the BRP, we found
significant small-scale aggregation of juveniles and
significant and positive small-scale associations of
juveniles to adults. The absence of small-scale
aggregations in plots of farmed lands demonstrates
to our opinion the influence of this type of land-use.
The low number of saplings (dbh 5–10 cm) in farmed
lands compared to protected lands is likely to have
long-term consequences on the viability of shea tree
populations and may in the long run even lead to a
collapse of the V. paradoxa population. As a
suggestion to managers we recommend the estab-
lishment of meaningful management plans for shea
trees in the region, particularly the village territories
surrounding the BRP for maintenance of the current
shea tree density and a better preservation for
saplings. This is the only way to guarantee the
presence of this economical important plant species
also for future generations.
Acknowledgments We are very grateful to BIOTA West
Africa for financial support during this work and to members of
stochastic department of University of Ulm for contributing on
statistical aspect. We also want to thank the manager team of
the BRP for creating a good working environment. Finally we
want to express our gratitude to all farmers in villages
surrounding the BRP that collaborated and accepted that we
used their fields for data collection.
Appendix: combining the data from individual
mapped replicate plots into mean, weighted O(r),
and G(y) functions
For statistical analysis it is common to map several
replicate plots of a larger point pattern under identical
conditions. In this case the resulting second-order
statistics of the individual replicate plots can be
combined into average second-order statistics (Diggle
2003). This is of particular interest if the number of
points in each replicate plot is relatively low. In this
case the simulation envelopes of individual analyses
would become wide, but combining the data of
several replicate plots into average second-order
statistics increases the sample size and thus narrows
the confidence limits. Average second-order statistics
are also an effective way of summarizing the results
of several replicate plots.
When the patterns are strict replicates of an
underlying process, the corresponding estimates
KiðrÞof the K-functions from plots iare identically
distributed and a reasonable overall estimate can be
obtained by simply averaging the individual K-
functions (Diggle 2003: Eq. 4.20 at page 52). Using
the grid-based estimator of Programita, the resulting
estimator of the O-ring statistic is given in the
appendix of Riginos et al.(2005).
However, because Ripley’s K-function K(r)is
defined as a ratio of expected number of points in
circles [= kK(r)] divided by the intensity k, a better
strategy may be to pool separately estimates of kand
kK(r). The resulting average K-function is a weighted
average of the individual estimates ^
KiðrÞ, where the
weight is the number of points in plot idivided by the
total number of points in all replicate plots (Diggle
Agroforest Syst (2008) 72:205–220 217
2003: Eq. 8.11, page 123). Note that the resulting
average K-function is also an appropriate estimator if
the replicates would be differentially thinned versions
of a common underlying process (Diggle 2003: p 123).
Using the grid-based estimators of Programita and
following the notation in Wiegand and Moloney (2004)
(their Eq. 11), the numerical estimator of the bivariate
pair-correlation function g
where n
is the number of points of pattern 1, R
is the ring with radius rand width wcentered in the
ith point of pattern 1, Points
[X] counts the points of
pattern 2 in a region X, and the operator Area[X]
determines the area of the region X.
To integrate the data of Ndifferent replicates into a
single weighted pair-correlation function, the formula
for one replicate (Eq. A1) is extended by calculating,
for each spatial scale r, the average weighted number of
points of pattern 2 taken over all Nreplicates and the
average weighted area taken over all Nreplicates:
where i
is the ith point of pattern 1 and replicate j,n
the number of points of pattern 1 and replicate j, and
is the total number of points of pattern 1 in
all replicates. Equation A2 simplifies to:
Following the strategy of Diggle (2003) to pool
separately estimates of k
and k
(r) [and analo-
gously estimates of k
and k
(r) for estimating the
pair-correlation function] the overall intensity k
estimated as k
is the total number of points of pattern 2 in all N
replicates j, and Athe total area
of all replicates jwith area A
The univariate estimator of g(r) is calculated in a
manner analogous to the bivariate functions by
setting pattern 1 equal to pattern 2.
Because we did not use edge correction for the
distribution function Gof nearest neighbor distances
we simply summed up the frequency distributions of
single replicated and normalized the frequency dis-
tribution after combining all replicated to yield the
normalized G
(y) and G(y).
Literature cited in appendix
Diggle PJ (2003) Statistical analysis of point patterns,
IInd edn. Arnold, London
þ::: þnN
! !
þ::: þnN
12ðrÞ¼ P
1;i1ðrÞ þ ::: þP
1;i1ðrÞ þ ::: þP
218 Agroforest Syst (2008) 72:205–220
Riginos C, Milton SJ, Wiegand T (2005) Context-
dependent negative and positive interactions between
adult shrubs and seedlings in a semi-arid shrubland.
J Vegetation Sci 16:331–340
Wiegand T, Moloney KA (2004) Rings, circles
and null-models for point pattern analysis in ecology.
Oikos 104:209–229
Agbahungba G, Depommier D (1989) Aspects du parc a ka-
rites1-neres (Vitellaria paradoxa Gaertn. f. Parkia
biglobosa Jacq. Benth) dans le sud du Borgou (Benin).
Bois et Forets des Tropiques 222:41–54
Ayensu EA (1974) Plant and bat interaction in West Africa.
Ann Mo Bot Gard 61:702–727
Boffa JM (1999) Agroforestry parklands in sub-Saharan
Africa. FAO conservation guide 34. FAO, Rome
Bruner AG, Gullison RE, Rice RE, da Fonseca GAB (2001)
Effectiveness of parks in protecting tropical biodiversity.
Science 291:125–128
Chevalier A (1946) L’arbre a
`beurre d’Afrique et l’avenir de sa
culture. Ole
´agineux 1:7–11
Condit R, Ashton PS, Baker P, Bunyavejchewin S, Gunatilleke
S, Gunatilleke N, Hubbell SP, Foster R, Itoh A, Lafrankie
JV, Lee HS, Losos E, Manokaran N, Sukumar R, Ya-
makura T (2000) Spatial patterns in the distribution of
tropical tree species. Science 288:1414–1418
Condit R, Hubbell S, Foster R (1992) Recruitment near con-
specific adults and the maintenance of tree and shrub
diversity in a neotropical forest. Am Nat 140:261–286
Connell JH (1971) On the role of natural enemies in preventing
competitive exclusion in some marine animals and rain
forest trees. In: den Boer PJ, Gradwell PR (eds) Dynamics
of populations. Centre for agricultural publishing and
documentation, Wageningen, pp 289–312
Delvingt W, Heymans JC, Sinsin B (1989) Guide du parc
national de la Pendjari. Communaute
´Economique Eu-
´enne (C.E.E.), Bruxelles
Dudgeon WR (1922) The agricultural and forest products of
West Africa. Murray, London
Fa JE, Garcia Yuste JE, Castelo R (2000) Bushmeat markets on
Bioko Island as a measure of hunter presence. Conserv
Biol 14:1602–1613
Fa JE, Peres CA, Meeuwig J (2002) Bushmeat exploitation in
tropical forests: an intercontinental comparison. Conserv
Biol 16:232–241
Fa JE, Ryan SF, Bell DJ (2005) Hunting vulnerability, eco-
logical characteristics and harvest rates of bushmeat
species in afrotropical forests. Biol Conserv 121:167–176
Goreaud F, Pelissier R (2003) Avoiding misinterpretation of
biotic interactions with the intertype K-12-function: pop-
ulation independence vs. random labeling hypotheses.
J Vegetation Sci 14:681–692
Grime JP (2001) Plant strategies, vegetation processes and
ecosystem properties, 2nd edn. Wiley, Chichester, UK
Hall JB, Aebischer DP, Tomlinson HF, Amaning EO, Hindle
JR (1996) Vitellaria paradoxa: a monograph, school of
agricultural and forestry sciences. Publication Number: 8,
University of Wales, Bangor, UK, p 105
Hjerpe J, Hedena H, Elmqvist T (2001) Tropical rain forest
recovery from cyclone damage and fire in Samoa. BIO-
TROPICA 33(2):249–259
¨lzel N (2005) Seedling recruitment in flood-meadow spe-
cies: the effects of gaps, litter and vegetation matrix. Appl
Vegetation Sci 8:115–224
Hubbell S, Foster R, O’Brien S, Wechsler B, Condit R, Harms
K, Wright S, De Loo Lau S (1999) Light gaps, recruitment
limitation and tree diversity in a neotropical forest. Sci-
ence 283:554–557
Hyatt LA, Rosenberg MS, Howard TG, Bole G, Fang W,
Anastasia J, Brown K, Grella R, Hinman K, Kurdziel JP,
Gurevitch J (2003) The distance dependence prediction of
the Janzen-Connell hypothesis: a meta-analysis. Oikos
INSAE/MPPD (2002) Rapport du troisie
`me recensement
´ral de la population et de l’habitat. RGPH-3, Be
Jackson G (1968) Notes on West African Vegetation. 3. The
seedling morphology of Butyrospermum paradoxum
(Gaertn. f.) Hepper. J West Afr Sci Assoc 13:215–222
Janzen DH (1970) Herbivores and the number of tree species in
tropical forests. Am Nat 104:501–528
Kelly BA, Hardy JO, Bouvet J-M (2004) Temporal and spatial
genetic structure in Vitellaria paradoxa (shea tree) in an
agroforestry system in southern Mali. Mol Ecol 13:
Killick HJ (1959) The ecological relationship of certain plants
in the forest and savanna of central Nigeria. J Ecol
Laris P, Wardell AD (2006) Good, bad or ‘necessary evil’?
Reinterpreting the colonial burning experiments in the
savanna landscapes of WestAfrica. Geogr J 172(4):271–290
Loosmore NB, Ford ED (2006) Statistical inference using the
Gor Kpoint pattern spatial statistics. Ecology 87:1925–
Lovett PN, Haq N (2000) Diversity of the Sheanut tree
(Vitellaria paradoxa C.F. Gaertn.) in Ghana. Genet
Resour Crop Evol 47:293–304
Maranz S, Wiesman Z (2003) Evidence for indigenous selec-
tion and distribution of the shea tree, Vitellaria paradoxa,
and its potential significance to prevailing parkland
savanna tree patterns in sub-Saharan Africa north of the
equator. J Biogeogr 30:1505–1516
Nathan R, Casagrandi R (2004) A simple mechanistic model of
seed dispersal, predation and plant establishment: Janzen-
Connell and beyond. J Ecol 92:733–746
Nikiema A, van der Maesen LJG, Hall JB (2001) The impact of
parkland management practices on plant resources diversity.
In: Teklehaimanot Z (ed) Improved management of agro-
forestry parkland systems in Sub-Saharan Africa. EU/INCO
Project Contract IC18-–CT98-0261, 3rd Annual Progress
Report, University of Wales Bangor, UK, pp 57–66
Nikiema A (2005) Agroforestry parkland species diversity:
uses and management in semi-arid West Africa (Burkina
Faso). Dissertation, Wageningen Universty, The Nether-
lands, p 102
PAG2 (2005) Plan d’Ame
´nagement Participatif et de Gestion
2004–2013. Parc National de la Pendjari, Be
pp 83 +Annexes
Agroforest Syst (2008) 72:205–220 219
Ripley BD (1981) Spatial statistics. Wiley, New York
Rondeux J (1999) Mesure des arbres et peuplements forestiers.
Presses Agronomiques de Gembloux, Belgium, p 521
Ruyssen B (1957) Le karite
´au Soudan, Premiere Partie, Aire
´ographique du karite
´en Afrique et au Soudan.
L’Agronomie Trop 11:143–171
Sinsin B, Tehou AC, Daouda I, Saidou A (2002) Abundance
and species richness of larger mammals in Pendjari
National Park in Benin. Mammalia 66(3):369–380
Sinsin B, Saı
¨dou A (1998) Impact des feux contro
´s sur la
´des pa
ˆturages naturels des savanes soudano-
´ennes du ranch de l’Okpara au Be
´nin. Annales des
Sci Agronomiques du Be
´nin 1:11–30
Sokpon N, Biaou H, Hounhyet O, Ouinsavi C, Barbier N
(2001) Inventaire et caracte
´risation des formations ve
tales du complexe national de la Pendjari, zone
´tique de la Pendjari et de l’Atacora: re
´gion de
Konkombri. Rapport, UNB/Be
´nin III, pp 56 +Annexes
Stoyan D, Stoyan H (1994) Fractals, Random shapes and point
fields. Methods of geometrical statistics. Wiley, Chichester
Stoyan D, Penttinen A (2000) Recent application of point
process methods in forest statistics. Stat Sci 15:61–78
Teklehaimanot Z (2003) Improved management of agrofor-
estry parkland systems in Sub-Saharan Africa. EU/INCO
Project Contract IC181-CT981-0261, 3rd Annual Progress
Report, University of Wales Bangor, UK, pp 57–66
Thibault M, Blaney S (2003) The oil industry as an underlying
factor in the bushmeat crisis in central Africa. Conserv
Biol 17(6):1807–1813
Tran T (1984) Shea butter—more than just a cosmetic ingre-
dient. Drug Cosmet Ind 54:54–61 and 114–116
Wiegand T, Moloney K (2004) Rings, circles and null1-models
for point pattern analysis in ecology. Oikos 104:209–229
Wills C, Condit R, Foster R, Hubbell S (1997) Strong den-
sity—and diversity—related effects help to maintain tree
species diversity in a neotropical forest. Proc Natl Acad
Sci 94:1252–1257
220 Agroforest Syst (2008) 72:205–220
... V. paradoxa is native to Gambia, Senegal, Guinea, Côte d'Ivoire, Mali, Burkina Faso, Niger, Ghana, Benin, Togo, Nigeria, Cameroon, Chad, Central African Republic, South Sudan, Uganda Democratic Republic of the Congo, and Ethiopia [39,[51][52][53][54][55][56]. The species is typical of the West African savanna, but also in the southern Sahel (Figure 1d). ...
... In 1946, 230 shea butter trees/ha were reported in the Sudanian savanna zone [62]. An average of 7 trees/ha was determined in Uganda, while in Benin, five trees/ha were counted [54,63]. The decline (c) (d) The shea fruit can be described as a globose to ellipsoid berry, measuring 4-5 cm × 2.5-5 cm, with a weight of 20-30 g. ...
... They bears stipulated rudimentary leaves (Figure 1b,c) [37,49,50]. [39,[51][52][53][54][55][56]. The species is typical of the West African savanna, but also in the southern Sahel (Figure 1d). ...
Full-text available
Vitellaria paradoxa (C.F.Gaertn.) is a multipurpose tree species distributed in a narrow band across sub-Saharan Africa. The species is integrated into cropping and agroforestry systems as a nutritional and economic resource, which provides a range of environmental services. Integration of the species into land-use systems provides an essential source of livelihoods and income for local populations. The economic potential of the shea butter tree derives from its edible products, which also serve cosmetic and pharmaceutical applications. To understand the current state of knowledge about V. paradoxa, this paper summarizes information about the ecology, population structure, and genetic diversity of the species, also considering compositional variation in the pulp and kernels, management practices, and efforts towards its domestication. Despite the great potential of the shea butter tree, there are some gaps in the understanding of the genetics of the species. This review presents up-to-date information related to the species for further domestication and breeding purposes.
... V. paradoxa is native to Gambia, Senegal, Guinea, Côte d'Ivoire, Mali, Burkina Faso, Niger, Ghana, Benin, Togo, Nigeria, Cameroon, Chad, Central African Republic, South Sudan, Uganda Democratic Republic of the Congo, and Ethiopia [39,[51][52][53][54][55][56]. The species is typical of the West African savanna, but also in the southern Sahel ( Figure 1d). ...
... In 1946, 230 shea butter trees/ha were reported in the Sudanian savanna zone [62]. An average of 7 trees/ha was determined in Uganda, while in Benin, five trees/ha were counted [54,63]. The decline (c) (d) in shea tree population densities may be the result of firewood and charcoal production, agricultural development, and biofuel crop production, contributing to removing shea trees from the farms [64,65], but it is important to note that classification criteria (e.g. ...
Full-text available
Vitellaria paradoxa (C.F.Gaertn.) is a multi-purpose tree species distributed in a narrow band across sub-Saharan Africa. The species is integrated into cropping and agroforestry systems as a nutritional and economic resource, which provides a range of environmental services. Integration of the species into land-use systems provides an essential source of livelihoods and income for local populations. The economic potential of the shea butter tree derives from its edible products, which also serve cosmetic and pharmaceutical applications. To understand the current state of knowledge about V. paradoxa, this paper summarizes information about the ecology, population structure, and genetic diversity of the species, also considering compositional variation in the pulp and kernels, management practices, and efforts towards its domestication. Despite the great potential of the shea butter tree, there are some gaps in the understanding of the genetics of the species. This review presents up-to-date information related to the species for further domestication and breeding purposes.
... Les niveaux des pressions anthropiques exercées sur une ressource varient en fonction des mesures de protection prises, lesquelles sont fonction à leur tour du mode d'affectation des terres. En effet, les modes d'affectation des terres sont reconnus comme des facteurs déterminant la structure des peuplements forestiers et leur état de conservation (Djossa et al., 2008). Ainsi, l'évaluation de l'influence des modes d'affectation des terres sur un peuplement forestier permettra d'apprécier efficacement les types de pressions que subit une espèce forestière mais aussi de connaître leur impact sur la structure du peuplement. ...
... La présente étude vise à évaluer l'influence des modes d'affectation des terres sur les caractéristiques dendrométriques, les types morphologiques et la phénologie de B. costatum dans la Réserve de biosphère de la Pendjari (RBP). Considérant que le statut écologique des peuplements d'une espèce forestière varie suivant un gradient de mesures de protection (Assogbadjo et al., 2006 ;Djossa et al., 2008 ;Fandohan et al., 2010), nous avons émis l'hypothèse que les caractéristiques dendrométriques, les structures en diamètre, les types morphologiques et la phénologie de l'espèce varient suivant les modes d'affectation des terres. ...
Full-text available
Bombax costatum est une espèce agroforestière menacée de disparition du fait des fortes pressions anthropiques qu’elle subit par le prélèvement de son calice. La présente étude a pour objectif d’évaluer l’influence des modes d’affectation des terres sur les caractéristiques dendrométriques, les types morphologiques et la phénologie de B. costatum dans la Réserve de biosphère de la Pendjari (RBP). Quatorze placeaux de 200 m × 200 m répartis dans quatre modes d’affectation des terres (zone d’occupation contrôlée, chaîne de l’Atacora, zone cynégétique de la Pendjari et Parc national de la Pendjari) de la RBP ont été inventoriés. Les données telles que le diamètre à hauteur de poitrine, la hauteur totale, le nombre d’individus, l’aspect de l’écorce et la couleur des fleurs ont été collectées. Les caractéristiques structurales de B. costatum ont été évaluées à trois niveaux : les paramètres dendrométriques dont la densité, le diamètre moyen, la surface terrière et la hauteur moyenne ; les structures en diamètre ; la distribution de fréquence des différents types morphologiques obtenus (aspect de l’écorce, couleur de la fleur). Pour tester les différences entre les modes d’affectation des terres, une ANOVA, une analyse log-linéaire et une analyse de covariance ont été effectuées respectivement sur les paramètres dendrométriques et les types morphologiques suivant les modes d’affectation des terres. Le mode d’affectation des terres a un effet significatif (P < 0,05) sur la densité des individus adultes, le diamètre moyen, la hauteur moyenne, les types morphologiques et la phénologie de B. costatum. La densité des individus adultes est plus élevée dans les zones d’occupation contrôlée et la chaîne de l’Atacora. Il ressort des résultats que l’espèce a encore un potentiel semencier qui assure sa pérennisation mais qu’une utilisation contrôlée est nécessaire pour sa conservation durable.
... ( Table 4) In contrast, shade tolerant species such as pepper and yam were predominantly cultivated on farms located in Site 1 where V. paradoxa and total woody plants exhibited aggregate spatial patterns. Although spatial regularity was not observed in this study, the ndings of spatial randomness and aggregation have been similarly reported in West African parklands (Djossa et al., 2007;Kelly et al., 2004). This goes to suggest that regular patterns are more of a product of strong human manipulations in silviculture systems and may rarely occur in semi-natural (e.g., traditional parklands) or natural systems (Condit et al., 2000;Hubbell, 1979;Mason et al., 2007). ...
Full-text available
Agroforestry parklands are an age-old traditional land use practice that integrates crop cultivation under scattered woody plants. This practice is widespread in West African savannas providing many essential ecological and socio-economic benefits to people such as food, fuelwood and medicine. Currently, parklands are decreasing due to changes in agriculture and land use practices, often associated with human population growth. Understanding spatial patterns as well as identifying reliable methods of sampling to estimate density of woody plants is necessary for sustainable management of parklands. In this study, five relatively easy-to-use plotless sampling methods were applied to estimate density of woody plants using field and simulated datasets with known spatial patterns from field assessments. Results of spatial indices tests indicated that woody plants in parklands exhibited two spatial patterns: i.e., aggregate and random, the latter being the dominant pattern observed in field datasets. Based on relative measure statistics (i.e., RRMSE and RBIAS), the ordered distance (OD), point-centered quarter (PCQ) and closest individual (CI) methods performed well when woody plants were located in a random pattern while the variable area transect (VAT) method was better at estimating density under patterns of spatial aggregation. Overall, OD and VAT methods are recommended for density estimation in parklands because they are relatively more accurate, less biased, practical and computations are easy to undertake.
... es terres agricoles en particulier. Le lien entre la pression démographique, l e (voir par exempleDjossa et al., 2008;Nkeki, 2016). Il ressort de ces travaux de recherche que les migrations et la poussée démographique participent à la construction de nouveaux territoires dont il est important de comprendre les mécanismes de constitution et les principes de leur fonctionnement. ...
Full-text available
Les questions foncières sont en évolution rapide et engendrent des transformations sociales, économiques, culturelles et politiques particulièrement en Afrique de l’Ouest. Une diversité de travaux de recherche a animé ce débat scientifique et les connaissances mises au point restent éparses. Le présent article a fait une synthèse des principales thématiques de recherche abordées sur le foncier rural et a mis en évidence de nouvelles questions de recherche pour une gouvernance responsable des régimes fonciers en Afrique de l’Ouest. La recherche est basée sur l’analyse de contenu de 181 travaux de recherche collectés sur le web. Ils sont constitués des articles scientifiques (78%), articles de conférence (4%) livres et chapitres de livre (13%) et thèses de doctorat (5%) portant sur Afrique de l’Ouest ces 20 dernières années. Il ressort que les principaux travaux de recherche sur le foncier rural se sont focalisés sur les questions de droit, législation et politique foncière, le rapport entre le foncier-population, le développement rural, la sécurisation foncière. Très peu d’accent est mis sur le marché foncier, le genre et foncier, la gestion de conflits et le développement économique. Une analyse approfondie des contenus de chaque domaine a permis de dégager de nouvelles orientations de recherche pour une gouvernance foncière responsable. Mots clés : Afrique de l’Ouest, foncier rural, gouvernance, sécurisation foncière
... In Nigeria, shea trees occur naturally in the wild and thrive almost exclusively in the North and grow especially better in opened field with low tree densities (Odebiy et al., 2004;Djossa et al., 2008). In general, trees do not usually yield fruit until they are 20 years old, and do not reach full maturity until they are 45 years old. ...
... However, with the increase of the human demographic pressures, fruits, roots, leaves, bark, and wood of useful species are overexploited, thus jeopardizing the chances of regeneration of many species (Lokonon et al., 2018). Several studies carried out in West Africa have revealed a strong influence of the use and exploitation of native species on their regeneration patterns (Djossa et al., 2008;Schumann et al., 2010;Jurisch et al., 2012). ...
Parkia biglobosa (Jacq.) G. Don, Pterocarpus erinaceus Poir, Milicia excelsa (Welw.) C. C. Berg, Prosopis africana (Guill., Perrot. and Rich.) Taub., Afzelia africana Sm. and Khaya senegalensis (Desv.) A. Juss. are the most highly valued indigenous tree species in the agroforestry systems of the Ouémé catchment area. However, information on the population structure of these species is lacking, thus limiting the development of their sustainable conservation, utilization and restoration strategies. This study addressed this gap. It assessed the population structures and regeneration status of the six species from Don, Tan-Houègbo, Atchabita, Bétékoukou, Glazoué, Tchaorou, Zagnanado, Tévèdji, Sinaou and Bétérou along the catchment. Data were collected from 78 permanent rectangular plots (50 × 30 m) randomly installed within 10 provenances. Dendrometric data including diameter at breast height (dbh) of adult trees (dbh ≥ 10 cm), collar diameter, total height of seedlings and saplings, number of individuals per species according to adult, sapling and seedling were recorded. The population structure was described using ecological and dendrometric parameters (relative frequency, importance value index (IVI), mean densities, basal area, mean height), and diameter size-class distributions. Seedling:sapling and sapling:adult ratios were also computed and analyzed for determining regeneration patterns. Based on IVI, Parkia biglobosa (95.85%) and Khaya senegalensis (65.92%) were the most represented species in the catchment area. The analysis of variances showed that dendrometric parameters of the six species varied significantly between provenances. Seedling:sapling and sapling:adult ratios were
... In the present study, high tree density in harvesting site combined with fruit production indicated that this site provided the highest fruit yields compared to non-harvesting site. This is in line with Djossa et al. (2008) who reported that Vitellaria paradoxa and Parkia biglobosa (Jacq.) R. Br. ex G. Don are conserved during field clearing because of their use value. ...
The exploitation of non-timber forest products contribute considerably to local livelihoods for subsistence in developing countries. The study aims to assess the structure and fruit production of C. procera natural populations in harvesting and non-harvesting sites and to compare the structure characteristics and fruits yields of the two sites. The study was carried out in Burkina Faso on both, harvesting and non-harvesting sites. Eighty fruits bearing trees (40 per site) were sampled, from each, all the fruits were counted. Dendrometric parameters were measured in 80 plots of 1000 m². In each plot, diameter at breast height (dbh) and total height of trees were measured. Density, basal area, mean diameter and height were computed for each site. Analyses of variance were performed to test differences between sites based on measured variables. Weibull theoretical model was used on size class distribution to analyze trees structure in both sites. The results showed that the production of fruits reach a mean of 6462.73 ± 4237.51 fruits/ha and 9103.44 ± 7631.88 fruits/ha, respectively in non-harvesting and harvesting site. Significant differences (p < 0.05) were found for tree densities, basal area and tree height between sites with highest values in the harvesting site. The size class distribution of trees showed in both sites the existing of potential regeneration with the dominance of young trees (1 < c < 2) with some survival difficulties during the growth stages. C. procera populations exhibit a high potential of fruit production and natural regeneration but are influenced by pressures impact.
... Amoako and J. Gambiza decline in the number of A. leiocarpa in the successive diameter classes (>10-16 cm) in crop fields and fallows is an indication of high mortality of the species and non-protection from fire during the regenerative and establishment stages compared to V. paradoxa in crop fields. This is attributed to a high preference (Blench 2001;Djossa et al., 2008;Schumann et al., 2011) for V. paradoxa in crop fields over A. leiocarpa in the study area, thus less protection of A. leiocarpa (from fire and cropping). The reduction in the population of A. leiocarpa in burnt land use types is as well be attributed to its sensitivity to fire (Hennenberg et al., 2005;Schumann et al., 2011). ...
Full-text available
The West African savanna experiences a prolonged dry season with Harmattan winds which facilitate large and persistent biomass burning from November to April. The fires are mostly caused by humans, mainly in pursuit of the day to day livelihood activities. We examined how fire influences the population structure and abundance of two economically important woody species Vitellaria paradoxa (Shea tree) and Anogeissus leiocarpa (African Birch) in six land use types in the Guinea savanna, Ghana. We calculated the stand basal area, mean densities of juveniles and adult trees, Lorey's mean height of adult trees and Simpson's index of dominance. Eight diameter size classes of each species were analysed by comparing their observed distributions to a three-parameter Weibull distribution across the land use types. A total of 3366 individuals of A. leiocarpa (n = 1,846) and V. paradoxa (n = 1,520) were enumerated. The basal area of A. leiocarpa and V. paradoxa in sacred groves (16.9 m 2 ha − 1) and unburnt woodlands (20.6 m 2 ha − 1) was higher than the estimates in the other land use types. High mean densities of A. leiocarpa and V. paradoxa were found in sacred groves (22.7 ± 29.7 stems ha − 1) and fallows (15.3 ± 2.2 stems ha − 1), respectively. Mean density of juveniles of both A. leiocarpa (248.0 ± 89.1 stems ha − 1) and V. paradoxa (68.0 ± 29.7 stems ha − 1) were higher in unburnt woodlands than in the other land use types. A. leiocarpa was absent in fallows and burnt crop fields. An inverse J-shaped distribution was found in sacred groves for both species. The absence of A. leiocarpa in burnt crop fields and the decrease of some size classes of V. paradoxa in both burnt and unburnt crop fields indicate the need for sustainable conservation of both species. Furthermore, the inverse J-shape distribution found in sacred groves for both A. leiocarpa and V. paradoxa implies that these species thrive best with minimal anthropogenic disturbances. Although species conservation is achieved through conventional protection, traditional or cultural conservation practices which avoid the indiscriminate use of fire should be highly promoted to ensure sustainable conservation of species.
Full-text available
Boswellia dalzielii Hutch., an African frankincense tree, is a socio-economically important aromatic and medicinal tree. It is currently threatened by uncontrolled exploitation, and therefore requires action to ensure its sustainable management. This study assessed the population structure and regeneration of its natural stands across three land use types in Burkina Faso: woodlands, fallows and farmlands. Sixty, fifty and fifty 50 m × 20 m plots were established respectively in woodlands, fallows and farmlands. All the plots were surveyed for adult tree (dbh ≥ 5 cm) density, dbh, total height and health conditions. Data on regeneration density (dbh < 5 cm), source (generative, stem shoots, suckers), total height and collar diameter were also collected. The results show similar total tree heights (7.0 m-9.0 m) but significantly (p < 0.05) smaller tree dbh in woodlands (mean ± SD: 20.5 ± 0.49 cm) and fallows (29.3 ± 0.64 cm) than in farmlands (32.8 ± 0.15 cm). Adult tree density (trees/ha) was 1.3 and 2.7 times higher in woodlands (82.37 ± 6.57) than in fallows (62.00 ± 3.98) and farmlands (30.02 ± 1.63), respectively. The density of regeneration in woodlands was 28 and 6 times higher than in fallows and farmlands, respectively. The majority (> 50%) of regenerating plants were suckers and no seedling regeneration was found in farmlands. The distribution of trees in diameter classes was J-shaped in woodlands, bell-shaped in farmlands and positive asymmetric in fallows, indicating recruitment bottlenecks. We found that 80.18% of individuals encountered were unhealthy. Intensive debarking and cutting were the main threats to the species and no conservation strategy was in place in the study region. We suggest measures to reduce intensive debarking and cutting, which should contribute to better management of the species.
Full-text available
A wildlife survey was carried out in Pendjari National Park of Benin in April 2000. The park covers an area of 2,660 km(2). Larger mammals were censused along 97 parallel line transects. The transects lay 1 km apart and were 15 km long on the average. The total length of strips (effort) was 1,455 km. Count data were analysed with the "Distance" programme. Twenty species were recorded during the survey, including most of the larger mammals of West Africa, in particular bovids. The most abundant species were olive baboons (Papio anubis), western buffalo (Syncerus caffer brachyceros) and kob (Kobus kob), with respective densities of 3.06, 1.0 and 0.98 animals/km(2). The total biomass of larger mammals was 0.63 t/km(2) (elephants: Loxodonta africana excluded) and 1.12 t/km(2) (elephants included). The carrying capacity for herbivores was estimated at 2.8 t/km(2). Except for buffalo, roan antelope (Hippotragus equinus) and hartebeest (Alcelaphus buselaphus major), both species richness and abundance were lower than in a previous survey ten years earlier, and species such as topi (Damaliscus lunatus korrigum) and leopard (Panthera pardus) were no longer detected. The results signify the need to revise and improve current wildlife conservation and management strategies to assure long-term protection of larger mammals in Pendjari National Park.
Full-text available
Book dealing with : diameter measurement (instruments, types of errors), height measurements (instruments, types of errors), age, bark, crown, stump merasurements, form and volume of trees - determination, mass and biomass, volume tables (construction), forest stands (characterization), stand measurements (areas,volumes), growth of stands and trees, forest productivity (estimation methods), yield tables, growth models (construction and application), forest inventories (sampling methods : random sampling, stratified, systematic, point sampling, sampling with partial replacement, multiphase sampling,..
Question: Is the failure of establishment of rare flood-meadow species in habitat restoration primarily due to seed or microsite limitation? How do sown species respond to artificially created gaps and added litter at two neighbouring sites with similar physical conditions but contrasting vegetation matrix (young arable fallow field vs species-poor meadow sward)? Location: Upper Rhine valley, southwestern Germany, 85 m a.s.l. Methods: Seeds of six typical flood-meadow species were sown in four treatment combinations of the factors gap creation and litter addition. Seedling recruitment was monitored for three years. Results: Five of the six species established successfully at both sites largely irrespective of treatments, indicating seed limitation. Only in the small-seeded Arabis nemorensis, which was revealed to be strictly gap-dependent at the meadow site, could an obvious microsite limitation be shown. The non-significance of gap treatments in all other species at the relatively high productive meadow site is probably due to biomass removal by mowing in early summer. Only at the extremes of the seed size spectrum did the results meet predictions of plant ecological theory, such as the strict gap dependence of small-seeded species in closed swards or the positive to neutral response of large-seeded species to litter layers. Conclusions: Species identity was revealed to be the major factor influencing differences in recruitment. Due to the lack of a general trend in the response towards treatments the results support conceptual models that describe the interplay of facilitation and interference as a highly dynamic equilibrium, driven by variable abiotic and biotic marginal conditions.
Light gap disturbances have been postulated to play a major role in maintaining tree diversity in species-rich tropical forests. This hypothesis was tested in more than 1200 gaps in a tropical forest in Panama over a 13-year period. Gaps increased seedling establishment and sapling densities, but this effect was nonspecific and broad-spectrum, and species richness per stem was identical in gaps and in nongap control sites. Spatial and temporal variation in the gap disturbance regime did not explain variation in species richness. The species composition of gaps was unpredictable even for pioneer tree species. Strong recruitment limitation appears to decouple the gap disturbance regime from control of tree diversity in this tropical forest.
One of the most dramatic plant and animal relationships in the West African ecosystem is that which takes place between such frugivorus bats as Epomophorus gambianus and Eidolon helvum and the introduced neem tree, Azadirachta indica. While earlier studies on bat behavior were carried out by personal observation and ordinary photographic means, recently developed night vision equipment has allowed the author to make nocturnal observations previously not possible. New observations, using this equipment, have now been made on bats and their interaction with trees of the following species: Mangifera indica, Anacardium occidentale, Ficus umbellata, Psidium guajava, Carica papaya, Kigelia africana, Spathodea campanulata, Parkia clappertoniana, Ceiba pentandra, and Adansonia digitata. Observations are reported on the roosting characteristics of bats of the genus Epomophorus. The effect of feeding by bats on the fruits of various trees and the resulting dispersal of seeds on the coastal savannah-grassland, (Accra Plains) of Ghana, are discussed. The effect of introduced plants on the bat population and disadvantages to the West African ecosystem which are attributable to the plant and bat interactions are summarized.