Impact of habitat type on the conservation status of tamarind (Tamarindus indica L.) populations in the W National Park of Benin

Abstract

The conservation status of many wild fruit tree species that support rural people in Africa remains poorly documented despite its importance for their management. We compared the viability of tamarind (Tamarindus indica) populations, a dry land species that has nutritional, medicinal and cultural importance for rural communities, under different human-pressure levels.
Materials and methods. The data relative to the tree diameter and height as well as the number of adults and stems were collected in plots of inventory and made it possible to calculate the dendrometric parameters for each targeted population, and to establish their diameter distribution. The dendrometric characteristics were analyzed by using nonparametric tests and the diameter distribution was adjusted to a truncated normal distribution.
Results and discussion. Numbers of mature tamarind trees per hectare and regeneration (expressed as stem.ha(-1)) were relatively low, suggesting tamarind populations may not be self-rejuvenating. Nonetheless, significant variation occurred between habitat types (P < 0.001). Mature tree density in gallery forests [(18.2 +/- 10.1) trees.ha(-1)] was three to eight times higher than that of savannah woodlands [(5 +/- 4.5) trees.ha(-1)] and farmlands [(2.5 +/- 0.4) trees.ha(-1)]. Young plants followed the same trend, with (11.2 +/- 9.3) plants.ha(-1), (1.1 +/- 0.6) plants.ha(-1), and 0.00 plants.ha(-1), respectively. Diameter size class distributions departed from normality (P < 0.0001) and coefficient of skewness was positive irrespective of habitat type, indicating declining populations. However, median diameter values would suggest the species' populations in farmlands and savannah woodlands to be more vulnerable than those occurring in gallery forests. These findings would suggest that gallery forests best suit tamarind in situ conservation. The observed severe reduction of trees and juveniles in farmlands and woodlands may negatively impact the long-term viability of tamarind populations. Juveniles' introduction into farmlands may be needed to ensure conservation in agroforestry systems.

Full text

Original article
Fruits, vol. 65 (1) 1
Impact of habitat type on the conservation status of tamarind (Tamarindus indica L.)
populations in the W National Park of Benin.
Abstract – Introduction. The conservation status of many wild fruit tree species that support rural peo-
ple in Africa remains poorly documented despite its importance for their management. We compared the
viability of tamarind (Tamarindus indica) populations, a dry land species that has nutritional, medicinal
and cultural importance for rural communities, under different human-pressure levels. Materials and
methods. The data relative to the tree diameter and height as well as the number of adults and stems
were collected in plots of inventory and made it possible to calculate the dendrometric parameters for
each targeted population, and to establish their structure in diameter. The dendrometric characteristics
were analyzed by using nonparametric tests and the diameter distribution was adjusted to a truncated
normal distribution. Results and discussion. Numbers of mature tamarind trees per hectare and regene-
ration (expressed as stem·ha–1) were relatively low, suggesting tamarind populations may not be self-
rejuvenating. Nonetheless, significant variation occurred between habitat types (P< 0.001). Mature tree
density in gallery forests [(18.2 ± 10.1) trees·ha–1] was three to eight times higher than that of savannah
woodlands [(5 ± 4.5) trees·ha–1] and farmlands [(2.5 ± 0.4) trees·ha–1]. Young plants followed the same
trend, with [(11.2 ± 9.3) plants·ha–1], [(1.1 ± 0.6) plants·ha–1], and [0.00 plants·ha–1], respectively. Diame-
ter size class distributions departed from normality (P< 0.0001) and coefficient of skewness was positive
irrespective of habitat type, indicating declining populations. However, median diameter values would
suggest the species’ populations in farmlands and savannah woodlands to be more vulnerable than those
occurring in gallery forests. These findings would suggest that gallery forests best suit tamarind in situ
conservation. The observed severe reduction of trees and juveniles in farmlands and woodlands may
negatively impact the long-term viability of tamarind populations. Juveniles’ introduction into farmlands
may be needed to ensure conservation in agroforestry systems.
Benin / Tamarindus indica / resource conservation / habitats / stand
characteristics / forest inventories / anthropic influence
Impact du type d'habitat sur la conservation des populations du tamarinier
(Tamarindus indica L.) dans le parc national du W au Bénin.
Résumé. – Introduction. Le statut de conservation de nombreuses espèces fruitières sauvages dont
dépendent les populations rurales en Afrique demeure mal documenté en dépit de son importance pour
la gestion de leurs populations. Nous avons comparé la viabilité des populations de tamariniers (Tama-
rindus indica), espèce ayant une importance alimentaire, médicinale et culturelle pour les communautés
rurales, sous différents degré de pression humaine. Matériel et méthodes. Les données relatives au dia-
mètre et hauteur des arbres ainsi que le nombre d’adultes et de juvéniles ont été collectées dans des pla-
cettes d’inventaire et ont permis de calculer pour chaque population ciblée, les paramètres
dendrométriques et d’établir leur structure en diamètre. Les caractéristiques dendrométriques ont été ana-
lysées en utilisant des tests non paramétriques et la distribution des diamètres a été ajustée à une distribu-
tion normale tronquée. Résultats et discussion. Le nombre de tamariniers adultes et la régénération
(exprimée en nombre de tiges·ha–1) ont été relativement faibles, suggérant que les populations de tama-
riniers ont une faible capacité à se régénérer. Néanmoins, une différence significative est apparue en
fonction du type d'habitat (P< 0,001). La densité des arbres adultes dans les forêts de galerie
[(18,2 ± 10,1) arbres·ha–1] a été trois à huit fois plus élevée que dans les savanes boisées [(5 ± 4,5)
arbres·ha–1] et dans les terres cultivées [(2,5 ± 0,4) arbres·ha–1]. Les jeunes plants ont exprimé la même
tendance avec [(11,2 ± 9,3) plants·ha–1], [(1,1 ± 0,6) plants.ha–1] et 0 plant·ha–1, respectivement. La distri-
bution des diamètres s’est écartée de la distribution normale (P< 0,0001) et le coefficient d’asymétrie a
été positif quel soit le type d'habitat, indiquant des populations en déclin. Cependant, la valeur médiane
des diamètres suggèrerait que les populations de T. indica dans les terres cultivées et dans les savanes
boisées sont plus vulnérables que celle des forêts de galerie. Ces résultats indiquent que les forêts galerie
conviendraient le mieux à la conservation in situ du tamarinier. La diminution critique des arbres et jeu-
nes plants observée dans les terres cultivée et les savanes boisées pourrait influencer négativement la via-
bilité à long terme des populations de tamariniers. L'introduction de jeunes plants dans les zones de
culture pourrait être nécessaire pour assurer la conservation de T. indica dans des systèmes d’agroforesterie.
Bénin / Tamari nd us in dic a / conservation des ressources / habitat /
caractéristique du peuplement / inventaire forestier / influence anthropique
1Lab. Appl. Ecol. Fac. Agron.
Sci., Univ. Abomey-Calavi, 01
BP 526, Cotonou, Benin
Belarmain.fandohan@laposte.
net
2Lab. Trop. Subtrop. Agric.
Ethnobot., Dep. Plant Prod.,
Fac. Biosci. Eng., Univ. Ghent,
Coupure Links 653, B-9000,
Ghent, Belgium
Impact of habitat type on the conservation status of tamarind
(Tamarindus indica L.) populations in the W National Park of
Benin
Adandé Belarmain FANDOHAN1*, Achille Ephrem ASSOGBADJO1, Romain Lucas GLELE KAKAÏ 1, Brice SINSIN1, Patrick VAN DAMME2
* Correspondence and reprints
Received 18 December 2008
Accepted 10 June 2009
Fruits,
2010, vol. 65, p. 1–9
© 2010 Cirad/EDP Sciences
All rights reserved
DOI: 10.1051/fruits/2009037
www.fruits-journal.org
RESUMEN ESPAÑOL, p. 9

2Fruits, vol. 65 (1)
A.B. Fadohan et al.
1. Introduction
Many tropical fruit trees are not the objects
of large world markets, but still have con-
siderable importance in local and national
economies, being harvested by rural popu-
lations for local consumption and commer-
cialization on a small scale [1]. Interest in
these species has reached a peak with new
initiatives in agroforestry. These initiatives
seek to integrate, into tropical farming sys-
tems, indigenous trees whose products have
traditionally been gathered from natural for-
ests [2], in order to provide marketable prod-
ucts from farms that will generate cash for
resource-poor rural households. To sustain
this goal, many studies have addressed mor-
phometric, genetic variabilities among and
within populations and fruiting potential
(for example, baobab [3, 4], Sclerocarya [5],
Vitellaria [6], Irvengia [7], Detarium [8]).
However, efforts in this direction are limited
by lack of information about the conserva-
tion status of populations of most of these
important multipurpose trees.
Like most other tropical countries, Benin
ecosystems are full of multipurpose tree
species [9]. Tamarindus indica L. is a semi-
evergreen multipurpose tree species of the
Fabaceae family. The species occurs in most
tropical countries in a wide range of habitats
ranging from natural stands to anthropo-
genic areas. It is a tree species that has nutri-
tional, medicinal and cultural importance
for rural communities [10, 11]. To date it has
shown a high economic value and has been
chosen as a model species to be domesti-
cated in Sub-Saharan Africa [12]. Studies on
the species in Asia and Africa have docu-
mented its domestication potential in terms
of socio-cultural and nutritional values, and
aptitude for seed and vegetative propaga-
tion [11]. The species’ distribution in Africa
has been mapped [13]. Although not
exhaustive, it has given evidence of the spe-
cies’ plasticity under tropical areas.
Recent work based on neutral molecular-
genetic markers [14] has highlighted a high
genetic diversity in West African popula-
tions, indicating no immediate risk of
genetic erosion in this region, as long as eco-
logical conditions continue to permit regen-
eration. In West Africa, the maintenance of
regeneration in tamarind populations has
been reported as being not assured at all [1]
but, to date, very few studies have directly
addressed tamarind populations’ structure
nor the causes underlying lack of regener-
ation using dendrometric characteristics and
diameter size distribution. Thus, the popu-
lation status of T. indica remains poorly
documented and understood, as for many
wild fruit trees that support rural popula-
tions in Africa [15].There is therefore a need
to document T. indica current populations’
conservation status, in order to generate
information that would allow guiding the
design of comprehensive intervention strat-
egies for the species’ conservation and
enhanced utilization.
Our study was therefore initiated with the
aim of examining the impact of different
habitat types; farmlands, savannah wood-
lands and gallery forests with different land
use models (that is, human pressure), on the
conservation status of T. indica popula-
tions, using dendrometric characteristics
and diameter size class distributions of the
trees as indicators of populations’ viability.
The research was undertaken posing the fol-
lowing questions: is the population struc-
ture of T. indica independent of habitat type
and human pressure? Which habitat type
may best suit T. indica in situ conservation?
2. Materials and methods
2.1. Study area
This study was carried out on a number of
tamarind populations established within the
W national park of Benin and around, where
the species has been documented to occur
in different habitats [16]. The sampled region
is located in the sudanian climatic zone
between lat. 11° 40’ N to 12° 23’ N and long.
2° 2’ E to 3° 2’ E, coinciding with up to 80%
of the W National Park of Benin (figure 1).
The mean annual rainfall is about 950 mm
and soils are mainly ferruginous [16].
The study was undertaken in three dif-
ferent vegetation types; gallery forests,
savannah woodlands and farmlands. In
both gallery forest and savannah woodland

Conservation of T. indica populations in Benin
Fruits, vol. 65 (1) 3
habitats, T. indica is associated with Grewia
bicolor Juss., Cissus quadrangularis L., Cap-
paris sepiara L., Feretia apodanthera Del.,
Anogeissus leiocarpa (D.C.) Guill. et Perr.
and Balanites aegyptiaca (L.) Del., while in
farmland, it is commonly seen with Vitel-
laria paradoxa Gaertn. f., Adansonia digi-
tata L., Sclerocarya birrea (A. Rich.) Hochst.
and Balanites aegyptiaca L. [16].
Gallery forests are characterized by a
longer period of soil water availability due
to the proximity of water courses. In these
forests, the species faces lower drought
stress conditions compared with the situa-
tion in farmlands and savannah woodlands.
This also results in more litter on the soil and
thus offers better conditions for seed germi-
nation and seedling survival even during
drought periods (personal field observa-
tions). Other important characteristics of
tamarind habitats are the termite mounds
and ant hills that occur under mature tam-
arind individuals, most frequently in savan-
nah woodlands and farmlands. The
localization of most of the surveyed gallery
forests (well-protected zones of the W
national Park), the density of their canopies,
topographic conditions along water courses
and the presence of some feared snakes
such as Python sebea Gmelin limit and make
tamarind resource harvesting and other
human disturbances difficult (field survey).
In contrast, the savannah woodlands as well
as farmlands are more open, close to sur-
rounding villages, and as a result subject to
intensive extractive uses of tamarind [16]. In
the two latter habitats, tamarind mature trees
and juveniles are subject to seasonal fire. On
top of this, farmlands are characterized by
land clearing and tree removal for agricul-
tural purposes. Regarding the aforemen-
tioned characteristics the investigated
habitats were categorized as follows: gallery
forests (less human pressure), savannah
woodlands (medium human pressure) and
farmlands (high human pressure).
2.2. Inventory approach and data
collection
The indicators of viability are: mature tree
density, young plants (expressed as
stems·ha–1), diameter size class distribution
and its coefficient of skewness (g1), and
median diameter (Dm). A s a re su lt o f t he di f-
ferences in the vegetation structure, man-
agement systems and size in the different
habitat types, plot outlines and layout were
fulfilled differently to ensure easy data col-
lection and enhance accuracy in the estima-
tion of the variables under investigation.
Finally, in each natural habitat type (gallery
forests and savannah woodlands), 17 circu-
lar plots of 5 000 m2 were laid out along four
azimuthal transects. In farmlands, the entire
farm of a given household was taken as a
Figure 1.
Area prospected for studying
the conservation status of
tamarind (Tamarindus indica L.)
populations in the W National
Park of Benin and location of
sampled sites [16].

4Fruits, vol. 65 (1)
A.B. Fadohan et al.
plot. This was done because T. indica's
niche on a farm is farmer-controlled and not
random in distribution as is always the case
for natural habitats. Randomization of natu-
ral habitat plots was achieved by laying out
plots at 1-km intervals, while farm plots were
laid out at 500-m intervals. To make the on-
farm data representative, two to three plots
were laid out in each of the 18 villages sur-
rounding the W National Park, which is a
total of 39 plots. Parameters recorded in each
plot were: all individuals above 25 cm
height, diameter at 130 cm height (D130)
using a forest tape, number of mature trees
(trees with D130 5 cm), and their height
using a clinometer. We considered as young
plants all individuals simultaneously fulfill-
ing the following conditions: height above
25 cm and D130 inferior to 5 cm.
2.3. Computation of dendrometric
parameters
2.3.1. Stem density
Stem density (N), that is, the average
number of tamarind individuals per hectare
of a given stage (here, mature trees and
young plants), was expressed as individu-
als·ha–1: N=n/s, where n is the overall
number of individuals of a given stage in the
plot and s is the plot surface (ha). Compar-
ison between habitats was made using the
Kruskal-Wallis test for k independent sam-
ples and the Mann-Whitney test.
2.3.2. Stem diameter class distribution
Stem diameter structures were established
by grouping D130 data into size classes of
10 cm in width. A histogram showing tree
density according to diameter classes was
established for each habitat type, using the
Minitab statistical package. The obtained
structures were subsequently adjusted to a
truncated normal distribution. The limit con-
sidered for the truncated distributions was
5 cm and was related to the inventory
design, that only took into account trees of
at least 5 cm in D130. Log-linear analysis was
then performed with SASv9 software to
check models’ adequacy.
To discuss population trends within hab-
itats, the coefficient of skewness (g1) was
used. This coefficient, which describes
the evenness of truncated distribution, is
a measurement of the relative proportion
of small versus large stems in a population
[17] and is defined as follows:
, where
is the number of stems; xi and are the
log(D130) of stem i and the mean of xi,
respectively; s is the standard deviation of
xi; g1>0 for size distributions with rela-
tively few small stems and many large stems
(that is, declining population); and g1<0
for distributions with relatively few large
stems and many smaller sized stems (that is,
increasing population) [17].
The median value of the diameter
size class distribution (Dm) was computed
to discuss the population trends
between habitats and is defined as
Dm = L1 + [((n/2) – (Σf )1)/fn] × c, where Dm
is the median value of the diameter size class
distribution; L1 is the lowest class boundary
of the median class (that is, the class con-
taining the median); n is the number of
stems; (Σf )1 is the sum of frequencies of all
classes lower than the median class; fm is the
frequency of the median class; and c = size
of median class interval.
For a given site, a population with a lower
median diameter (that is, relative to smaller
stems) tends to increase in abundance more
rapidly than at other sites [18].
3. Results
3.1. Impact of habitat type
on dendrometric characteristics
The findings of this study indicated signifi-
cant variations in the number of mature trees
per hectare between habitats (H = 43.54;
P< 0.001). It was three to eight times higher
in gallery forests than, respectively, savan-
nah woodlands and farmlands (table I).
Comparing the regeneration, the density
obtained in gallery forests was above
10 times higher than that of savannah wood-
lands (H = 54.82; P<0.001). No regeneration
was found in farmlands. The ratio [mature
≥
g1
–n xixi
–()
3
i
∑ / n1–()n2–()s3
=
x

Conservation of T. indica populations in Benin
Fruits, vol. 65 (1) 5
trees / regeneration] also differed with hab-
itat type and increased with higher human
pressure.
The mean diameter significantly differed
between habitats, ranging from 47.9 cm in
farmlands to 28.9 cm in gallery forests
(H = 45.41; P< 0.001). In contrast, differ-
ences among values of the mean height
were found to be not significant, ranging
from 12.8 m in farmlands to 10.5 m in savan-
nah woodland (H = 4.86; P= 0.088).
3.2. Impact of habitat type on stem
diameter class distributions
In savannah woodlands and farmlands, the
size structure of trees was bell-shaped with
a small proportion of young stems, but
departed from the normal distribution, that
is, the characteristic distribution of the stem-
diameter structure of a population, and is
skewed to the right (respectively,
P< 0.0001; g1= 0.55; Dm= 28.2 and
P< 0.0001; g1= 2.18; Dm= 43.7). In gallery
forests, in contrast, the size structure was
more similar to the inverse J-shape with an
apparently high proportion of small stems,
and departed from the normal distribution,
but is also skewed to the right (P< 0.0001;
g1=0.34; Dm= 25.2). The values obtained
of the coefficient of skewness (g1) are all
positive, meaning that the size distributions
showed relatively few small stems and many
large stems irrespective of habitat type.
However, the median diameter (Dm) value
was lower in gallery forests than that in
savannah woodlands, farmlands having the
highest value (figure 2).
4. Discussion
4.1. Habitat type impact on
T. ind i ca
populations
Measuring conservation status is straightfor-
ward [19]. But investigating anthropogenic
influence on plants requires long-term mon-
itoring data on population trends, which is
extremely complicated when addressing
long-living slow-growing trees such as tam-
arind. In the absence of long-term data, a
single survey of size class distribution is a
rapid and useful approach to assess popu-
lation structure [20]. Factors that may affect
the size class distribution of a plant species
are light tolerance, life form [21], habitat type
[22], climate, anthropogenic activities [23],
agroforestry uses [24] and individual species
survival strategies [25]. In the present study,
the investigated populations experience the
same conditions except for habitat type,
anthropogenic activity pressure (fire and
land clearing for agricultural purposes) and
agroforestry uses (harvesting pressure).
The findings of our study clearly show
that tamarind mature trees and regeneration
decline from gallery forests to farmlands and
tend to be critically reduced in habitats of
higher human pressure and drought stress,
that is, farmlands. The differences may be
due to the different behavior of regeneration
shown by the species according to its hab-
itats: seed- and root sucker-originated seed-
lings were observed. Root suckers were
observed in gallery forests only and are
probably one of the reasons underlying the
gregarious spatial repartition shown by the
Tab l e I .
Mean and standard deviation of dendrometric characteristics of Tamarindus indica in three habitat types of the
W National Park of Benin.
Habitat types Mature trees
(trees·ha–1)
Regeneration
(stems·ha–1)
Ratio
[mature trees
/ regeneration]
Diameter at breast height
or D130
(cm)
Total height
(m)
Gallery forest 18.2 ± 10.06 a 11.2 ± 9.3 a 1.63 28.9 ± 1.4 b 10.7 ± 0.5 a
Tree savannah 5.0 ± 0.9 b 1.1 ± 0.6 b 4.64 29.0 ± 2.4 b 10.5 ± 0.8 a
Farmland 2.5 ± 0.4 c 0 c –47.9 ± 1.9 a 12.8 ± 0.4 a
Means in a column followed by the same letter are not significantly different.

6Fruits, vol. 65 (1)
A.B. Fadohan et al.
species in that habitat type. The relatively
high aptitude of tamarind trees to make root
suckers has been documented even if it is
not explicit it depends on habitat type [26].
Contribution of human harvesting pressure
to the low rate of regeneration is compli-
cated and still not well documented [27].
However, a recent study on Pentadesma
butyracea Sabine has suggested over-
extraction of the reproductive structures
(fruits, seeds, flowers and stamens) to be a
partial consequence of low regeneration
[28]. Thus, the harvesting of seeds or fruits
needs to be undertaken with caution, since
heavy harvesting could lead to long-term
detrimental effects on recruitment of new
individuals [29]. In the specific case of tam-
arind, if not extracted, fruits may remain on
the trees for one to two years after ripening
and generally most of the seeds are
destroyed and eaten by insects before the
pods fall (field observation). Seedlings that
originated from seeds might then be natu-
rally very few and, thus, harvesting pressure
influence on the regeneration might be of
less importance. In contrast, land clearing,
fire and tree density reduction for agricul-
tural purposes [30] may significantly affect
seedling establishment and threaten the
species. Little attention has been given,
despite the need for it, to the evaluation of
T. indica in its habitats [31]. In Kenya, it was
reported that T. indica population structure
showed low regeneration, while mature
trees revealed a higher density along water
courses than that of farmlands and savannah
woodlands [32]. It is required ecologically
that an adequately high abundance of indi-
viduals of a given species occurs in the dif-
ferent reproductive stages (seedlings,
juveniles and mature trees) in habitats, for
continuity [33]. Species occurring in low
densities in populations are vulnerable to
extinction under human use pressure and
other internal and external disturbances
[34]. In this regard, reduction of mature tree
density linked with scarcity of regeneration
of T. indica, especially in farmlands, might
indicate the absence of any process of sys-
tematic sparing or recruitment favoring as is
the case in the Sahel zone [1], and thus indi-
cates nonself-rejuvenating populations.
The average diameter is significantly
lower in natural habitats compared with that
in farmlands. This may be attributed to
lower growing rates in the former due to a
more intense interspecific competition,
which is naturally lower in woodlands and
artificially eliminated in farmlands. Such
results have been mentioned for other spe-
cies, especially Sclerocarya birrea in South
Africa [35] and Quercus ilex in Spain [24]. In
contrast to our expectations, the average
height showed the same gradient between
habitats. This may be due to an abundance
of relatively young trees in gallery forests.
However, the tallest trees are found in
gallery forests, a possible result of higher
competition for light.
Figure 2.
Diameter size class distribution
of Tamarindus indica trees in
three habitat types (savannah
woodlands, gallery forests and
farmlands) of the W National
Park of Benin.

Conservation of T. indica populations in Benin
Fruits, vol. 65 (1) 7
In diameter size class distribution, all the
obtained curves deviated from the normal
distribution. However, the distribution was
closer to a bell-shaped type both in savan-
nah woodlands and farmlands, while it was
similar to an inverse J-shaped type in gallery
forests. Such distributions would suggest
that gallery forest populations are extending
while savannah woodlands’ and farmlands’
are declining [36]. Although studies have
often relied on the assumption that standing
distribution reflects future population
changes, studies to directly test this hypoth-
esis have proven that diameter size distribu-
tion is not a good predictor of population
changes [18, 37]. However, the coefficient of
skewness (g1) of this distribution is able to
predict the direction of population change,
according to the same authors. In our study,
all the distributions show a positive coeffi-
cient of skewness (g1> 0), indicating a dis-
tribution with relatively few small stems
(that is, juveniles) and many large stems
(that is, mature trees) and, thus, vulnerable
populations irrespective of habitat type.
However, a population with a lower median
diameter at a given site tends to increase in
abundance more rapidly than at other sites
[18]. In this regard, T. indica’s population
might increase more rapidly in gallery for-
ests than in savannah woodlands and farm-
lands. Thus, gallery forest tamarind
populations might be less vulnerable than
those of savannah woodlands and farm-
lands.
4.2. Implication for conservation
and sustainable management
Tamarind is often quoted as a dry ecosystem
species but our findings would suggest that
the species’ populations seem to be more
viable in gallery forests and thus under less
drought conditions. This habitat type might
therefore best suit the species' in situ con-
servation. However, trends in the species
population showed no evidence of stability
or long-term viability irrespective of habitat
type. In farmlands where regeneration fail-
ure can clearly be linked to agricultural
activities, the low rhythm of expansion
noticed may be attributed to anthropogenic
interference. Thus, the domestication of the
species by means of young plant production
and introduction into rural agricultural sys-
tems could help ensure rejuvenation and
thus long-term circa situ conservation and
utilization of the species. In contrast, in the
natural habitats of the species, particularly
in gallery forests, interpreting the ecological
status of the species and building effective
in situ conservation strategies will need bet-
ter understanding of factors and processes
that impact the population dynamics of the
species. Whatever the case, silvicultural
method building, identification of the best
propagation methods and drought tolerance
strengthening with regard to specific envi-
ronments will be necessary for an effective
management plan for T. indica.
Acknowledgements
Our study was carried out with the financial
support of the European Union through the
Project of Domestication and Development
of Baobab and Tamarind (DADOBAT).
References
[1] Diallo B.O., Mckey D., Chevallier M.-H., Joly
H.I., Hossaert-Mckey M., Breeding system
and pollination biology of the semi-domesti-
cated fruit tree, Tamarindus indica L. (Legu-
minosae: Caesalpinioideae): Implications for
fruit production, selective breeding, and
conservation of genetic resources, Afr. J.
Biotechnol. 7 (22) (2008) 4068–4075.
[2] Leakey R.R.B., Simons A.J., The domestica-
tion and commercialization of indigenous
trees in agroforestry for the alleviation of
poverty, Agrofor. Syst. 38 (1998) 165–176.
[3] Assogbadjo A.E., Kyndt T., Chadare F.J.,
Sinsin B., Gheysen G., Eyog-Matig O., Van
Damme P., Genetic fingerprinting using
AFLP cannot distinguish traditionally classi-
fied baobab morphotypes, Agrofor. Syst.
(2008) DOI: 10.1007/s10457–008–9157-y.
[4] Assogbadjo A.E., Sinsin B., Van Damme P.,
Caractères morphologiques et production
des capsules de baobab (Adansonia digitata
L.) au Bénin, Fruits 60 (5) (2005) 327–340.
[5] Leakey R.R.B., Shackleton S., du Plessis P.,
Domestication potential of marula (Sclerocarya

8Fruits, vol. 65 (1)
A.B. Fadohan et al.
birrea sbsp. caffra) in South Africa and
Namibia: 1. Phenotypic variation in fruit
traits, Agrofor. Syst. 64 (2005) 25–35.
[6] Kelly B.A., Hardy O.J., Bouvet J.-M., Tempo-
ral and spatial genetic structure in Vitellaria
paradoxa (shea tree) in an agroforestry sys-
tem in Mali, Mol. Ecol. 13 (2004) 1231–1240.
[7] Anegbeh P.O., Usoro C., Ukafor V., Tchoundjeu
Z., Leakey R.R.B., Schreckenberg K.,
Domestication of Irvengia gabonensis: 3.
Phenetopic variation in fruits and kernels in
Nigeria, Agrofor. Syst. 58 (2003) 213–218.
[8] Kouyaté A.M., Van Damme P., Caractères
morphologiques de Detarium microcarpum
Guill. et Perr. au sud du Mali, Fruits 57 (2002)
231–238.
[9] Codjia J.T.C., Assogbadjo A.E., Mensah
M.R., Diversité et valorisation au niveau local
des ressources forestières alimentaires du
Bénin, Cah. Agric. 12 (2003) 321–331.
[10] El-Siddig K., Gunasena H.P.M., Prasad B.A.,
Pushpakumara D.K.N.G., Ramana K.V.R.,
Vijayanand P., Williams J.T., Tamarind, Tam a -
rindus indica L., Southampt. Cent. Underutil.
Crop., Southampt., UK, 2006.
[11] Jama B.A., Mohamed A.M., Mulatya J., Njui
A.N., Comparing the “Big Five”: A framework
for the sustainable management of indige-
nous fruit trees in the drylands of East and
Central Africa, Ecol. Indic. (2007) DOI:
10.1016/j.ecolind.2006.11.009.
[12] Eyog Matig O., Gaoué O.G., Dossou B.,
Réseaux « Espèces Ligneuses Alimentaire ».
C. R. Prem. Réun. Réseau tenue du 11–13
déc. 2000 au CNSF, Ouagadougou, Burkina
Faso, Inst. Int. Ressour. Phytogénét., 2002.
[13] Salim A., Simons A., Waruhin A., Orwa C.,
Agroforestry database: A tree species refer-
ence and selection guide and tree seed sup-
pliers directory, Int. Counc. Res. Agrofor.,
Nairobi, Kenya, 1998.
[14] Diallo B.O., Joly H.I., Mckey D., Hossaertmckey
M., Chevallier M.H., Genetic diversity of
Tamarindus indica populations: Any clues on
the origin from its current distribution? Afr. J.
Biotechnol. 6 (7) (2007) 853–860.
[15] Leakey R.R.B., Potential for novel food prod-
ucts from agroforestry trees: a review, Food
Chem. 66 (1999) 1–14.
[16] Fandohan A.B., Structure des populations et
importance socioculturelle du tamarinier
(Tamatindus indica L.) dans la commune de
Karimama (Bénin), Fac. Sci. Agron., Univ.
Abomey-Calavi, Bénin, Mém. DEA, 2007,
55 p.
[17] Bendel R.B., Higgins S.S., Teberg J.E., Pyke
D.A., Comparison of skewness coefficient,
coefficient of variation, and Gini coefficient
as inequality measures within populations,
Oecol. 78 (1989) 394–400.
[18] Feeley J.K., Davies S.J., Nur Supardi Noor
Md., Kassim A.R., Tan S., Do current stem
size distributions predict future population
changes? An empirical test of intraspecific
patterns in tropical trees at two spatial
scales, J. Trop. Ecol. 23 (2007) 191–198.
[19] Lykke A.M., Assessment of species compo-
sition change in savannah vegetation by
means of woody plant’s size class distribu-
tions and local information, Biodivers. Con-
serv. 7 (1998) 1261–1275.
[20] Shackleton S.E., The significance of local
level trade in natural resource products for
livelihoods and poverty alleviation in South
Africa, Rhodes Univ., PhD thesis, Grahamst.,
S. Afr., 2005.
[21] Tesfaye G., Teketay D., Fetene M., Regener-
ation of fourteen tree species in Harenna for-
est, southeastern Ethiopia, Flora 197 (2002)
461–474.
[22] Wadt L.H.O., Kainer K.A., Gomes-Silva
D.A.P., Population structure and nut yield of
Bertholletia excelsa stand in southwestern
Amazonia, For. Ecol. Manage. 211 (2005)
371–384.
[23] Omeja P., Obua J., Cunningham A.B.,
Demand and supply of wood for drum mak-
ing in Central Uganda, Int. For. Review 7 (1)
(2005) 21–26.
[24] Pulido F.J., Diaz M., Hidalgo de Trucios S.J.,
Size structure and regeneration of Spanish
holm oak Quercus ilex forest and dehesas:
Effects of agroforestry use on their long term
sustainability, For. Ecol. Manage. 146 (2001)
1–13.
[25] Su J.C., Debinski D.M., Jajubauskas M.E.,
Kindscher K., Beyond species richness:
community similarity as a measure of cross-
taxon congruence for coarse-filter conserva-
tion, Conserv. Biol. 18 (1) (2004) 167–175.
[26] Bellefontaine R., Pour de nombreux ligneux,
la reproduction sexuée n’est pas la seule
voie : analyse de 875 cas, Sécher. 16 (4)
(2005) 315–317.
[27] Murali K.S., Shankar U., Shaanker R.U.,
Ganeshaiah K.N., Bawa K.S., Extraction of
non-timber forest products in the forests of
Biligiri Rangan Hills, India. 2. Impact of NTFT
extraction on regeneration, population struc-
ture, and species composition, Econ. Bot.
50 (3) (1996) 252–269.

Conservation of T. indica populations in Benin
Fruits, vol. 65 (1) 9
[28] Avocèvou-Ayisso C., Sinsin B., Adégbidi A.,
Dossou G., Van Damme P., Sustainable use
of non-timber forest products: Impact of fruit
harvesting on Pentadesma butyracea regen-
eration and financial analysis of its products
trade in Benin, For. Ecol. Manag. 257 (2009)
1930–1938.
[29] Shackleton C.M., Guthrie G., Main R., Esti-
mating the potential role of commercial over-
harvesting in resource viability: a case study
of five useful tree species in South Africa,
Land Degrad. Dev. 16 (2005) 273–286.
[30] Sinsin B., Sinadouwirou Th., Valorisation
socio-économique et pérennité de Penta-
desma butyracea en galeries forestières au
Bénin, Cah. Agric. 12 (2003) 1–5.
[31] Gunasena G., Hughes A., Ta marindus indic a,
Int. Cent. Underutil. Crop., Southampt., UK,
2000.
[32] Nyadoi P., Population structure and socio-
economic importance of Ta m a rind u s i n d i ca
in Tharaka District, Eastern Kenya, Makarere
Univ., Thesis, Uganda, 2004, 110 p.
[33] Denslow J.S., Disturbance and diversity in
tropical rain forests: the density effect, Ecol.
Appl. 5 (1995) 962–968.
[34] Cunningham A.B., Mbenkum F.T., Sustaina-
bility of harvesting Prunus africana bark in
Cameroun: a medicinal plant in international
trade, Div. Ecol. Sci., UNESCO, Paris,
France, People Plants Work. Pap. 2 (1993)
11–46.
[35] Shackleton C.M., Botha J., Emanuel P.L.,
Productivity and abundance of Sclerocarya
birrea Caffra in and around rural settlements
and protected area of the bushbuckridge
lowed, South Africa, For. Trees Livelihoods
13 (2003) 217–232.
[36] Sokpon N., Biaou H.S., The use of diameter
distribution in sustained-use management of
remnant forests in Benin: case of Bassila for-
est reserve in North Benin, For. Ecol. Man-
age. 161 (2002) 13–25.
[37] Condit R., Sukumar R., Hubbell S.P., Foster
R.B., Predicting population trends from size
distribution: Direct test in tropical tree com-
munity, Am. Nat. 152 (1998) 495–509.
Impacto del tipo de hábitat sobre la conservación de las poblaciones del
tamarindo (Tamarindus indica L.) en el parque nacional del W en Benín.
Resumen.– Introducción. El estado de conservación de numerosas especies fruteras silvestres,
de las cuales dependen las poblaciones rurales en África, permanece mal documentada, a pesar de
su importancia para la gestión de sus poblaciones. Comparamos la viabilidad de las poblaciones
de tamarindos (Tamarindus indica), especie con una importancia alimenticia, medicinal y cultural
para las comunidades rurales, bajo diferentes niveles de presión humana. Material y métodos.
Los datos relativos al diámetro y a la altura de los árboles, así como el número de adultos y de
juveniles, se recopilaron en las parcelas de ensayo y permitieron calcular, para cada población
determinada, los parámetros dendrométricos así como establecer su estructura en diámetro. Las
características dendrométricas se analizaron mediante el empleo de pruebas no paramétricas; y, la
distribución de los diámetros se ajustó a una distribución normal truncada. Resultados y discu-
sión. El número de tamarindos adultos así como la regeneración (expresada en número de
ramas·ha–1) fueron relativamente débiles, lo que sugirió que las poblaciones de tamarindos poseen
una capacidad de regenerarse débil. No obstante, una diferencia significativa apareció en función
del tipo de hábitat (P< 0.001). La densidad de los árboles adultos en los bosques de galería
[(18.2 ± 10.1) árboles·ha–1] fue de tres a ocho veces más elevada que en las sabanas arboladas
[(5 ± 4.5) árboles·ha–1] que en tierras cultivadas [(2.5 ± 0.4) árboles·ha–1]. Los jóvenes plantones
expresaron la misma tendencia con [(11.2 ± 9.3) plantones·ha–1], [(1.1 ± 0.6) plantones .ha–1] y 0
plantón·ha–1, respectivamente. La distribución de los diámetros se separó de la distribución normal
(P< 0.0001) y el coeficiente de asimetría fue positivo, indiferentemente del tipo de hábitat, lo que
indicó así las poblaciones en decadencia. Sin embargo el valor medio de los diámetros sugirió que
las poblaciones de T. indica en las tierras cultivadas y en las sabanas arboladas son más vulnera-
bles que aquellas de bosque de galería. Dichos resultados indican que los bosques de galería con-
vendrían mejor para la conservación in situ del tamarindo. La disminución crítica de los árboles y
de los jóvenes plantones, observada en las tierras cultivadas y en las sabanas arboladas podría
influenciar negativamente la viabilidad a largo plazo de las poblaciones de tamarindos. La intro-
ducción de jóvenes plantones en las zonas de cultivo podría ser necesaria para garantizar la conser-
vación de T. indica en sistemas de agroforestería.
Benin / Tamarindus indica conservación de los recursos / habitat /
características del rodal / inventarios forestales / influencia antrópica