Content uploaded by Kasso Daïnou
Author content
All content in this area was uploaded by Kasso Daïnou on Dec 11, 2015
Content may be subject to copyright.
Tropical Ecology 55(3): 327-338, 2014 ISSN 0564-3295
© International Society for Tropical Ecology
www.tropecol.com
Reproductive ecology of Coula edulis Baill., source of a valuable non-
timber forest product
CHRISTIAN MOUPELA1,2,1*, JEAN-LOUIS DOUCET2 , KASSO DAÏNOU2, YVES BROSTAUX3,
ADELINE FAYOLLE4 & CÉDRIC VERMEULEN2
1University of Science and Technology of Masuku, B.P. 941 Franceville, Gabon
2University of Liege, Gembloux Agro-Bio Tech, Unit of Forest and Nature Management,
Laboratory of Tropical and Subtropical Forestry, Passage des Déportés 2, 5030
Gembloux, Belgium
3University of Liege, Gembloux Agro-Bio Tech, Unit of Applied Statistics, Computer Sciences and
Mathematics, Gembloux Agro-Bio Tech, University of Liege, Passage des
Déportés 2, 5030 Gembloux, Belgium
4University of Liege, Gembloux Agro-Bio Tech, Unit of Forest and Nature Management, Passage
des Déportés 2, 5030 Gembloux, Belgium
Abstract: The reproductive ecology of Coula edulis (Olacaceae), source of a highly valuable
non-timber product for African populations, has been investigated on a large tree sample size in
a Gabonese forest for a 3-year period, in order to propose options for its domestication.
Reproduction is annual and the minimum tree diameter for flowering was 10.6 cm while the
diameter for regular fructification was 23 cm. The annual diameter increment (ADI; mean =
0.22 cm year-1) was affected by crown exposure, but not by initial tree diameter or fruit
production. Tree diameter influenced fruiting frequency and fruit production. There was a
strong correlation between fruit production of 2011 and 2012 (Pearson's r = 0.85; P < 0.001),
suggesting a high heritability of that trait. Phenotypical selection should be performed prior to
any propagation activity. Moreover, as most of C. edulis fruits were sterile (64 %), we propose
that sylvicultural strategies should be based on vegetative propagation.
Resumen: La ecología reproductiva de Coula edulis (Olacaceae), fuente de un producto no
maderable con gran valor para poblaciones africanas, fue investigada en una muestra grande de
este árbol en un bosque de Gabón durante un período de tres años, con el fin de proponer
opciones para su domesticación. La reproducción es anual y el diámetro mínimo del árbol para
la floración fue de 10.6 cm, mientras que el diámetro para la fructificación regular fue de 23 cm.
El incremento anual en diámetro (IAD, con una media = 0.22 cm año-1) se vio afectado por la
exposición de la copa, pero no por el diámetro inicial del árbol o de la producción de frutos. El
diámetro del árbol tuvo un efecto en la frecuencia de fructificación y la producción de frutos.
Hubo una correlación fuerte entre la producción de frutos de 2011 y 2012 (r de Pearson = 0.85,
P < 0.001), lo que sugiere una alta heredabilidad de ese rasgo. La selección fenotípica debe ser
realizada antes de cualquier actividad de propagación. Además, como la mayoría de los frutos de
C. edulis fueron estériles (64 %), proponemos que las estrategias silvícolas deben basarse en la
propagación vegetativa.
Resumo: A ecologia reprodutiva de Coula edulis (Olacaceae), fonte de um produto não-
*Corresponding Author; e-mail: cmoupela@student.ulg.ac.be / cmoupela@yahoo.fr
328 REPRODUCTIVE ECOLOGY OF COULA EDULIS
lenhoso de alto valor para as populações africanas, foi investigada numa grande amostra de
árvoresnuma floresta do Gabão, por um período de 3 anos, a fim de propor alternativas para a
sua domesticação. A reprodução é anual eo diâmetro mínimo da árvore para a floração foi 10,6
cm, enquanto o diâmetro para frutificação normal era de 23 cm. O incremento anual de
diâmetro (ADI; média = 0,22 cm ano-1) foi afetado pela exposição da copa, mas não pelo diâmetro
inicial da árvore ou pela produção de frutos. O diâmetro da árvore influenciou a frequência da
frutificação e a produção de frutos. Encontrou-se uma forte correlação entre a produção de frutos
de 2011 e 2012 (r de Pearson = 0,85, P < 0,001), o que sugere uma alta heritabilidade dessa
característica. A seleção fenotípica deve ser realizada antes de qualquer atividade de
propagação. Além disso, como a maioria dos frutos da C. edulis foram estéreis (64 %), propõe-se
que as estratégias silvícolas devem ser baseadas na propagação vegetativa.
Key words: Diameter growth, fruit production, phenology, NTFP, tropical rainforest.
Introduction
For centuries, non-timber forest products
(NTFPs) have played a considerable role for rural
people inhabiting tropical rainforest zones, by
providing them with food and trade products (FAO
2010; Hill et al. 2007; Moupela et al. 2011). Despite
their importance, NTFPs have received little
attention by ecologists as they are economically less
important than timber products (Aiyelaagbe et al.
1998; Anegbeh et al. 2003; Atangana et al. 2001;
Debroux 1998; Guariguata et al. 2010; Guedje et al.
2003). As a result, there is a real risk of loss of
certain NTFPs due to unsustainable harvesting,
especially taxa with a small or restricted spatial
range. Domestication initiatives can prevent such a
loss as they aim to integrate NTFPs into farming
systems; thus reducing the pressure on natural
forests (Atangana et al. 2001; Kanten & Beer 2005;
Leakey & Simons 1998). While providing food and a
diverse range of other products to farmers,
agroforestry practices can also generate substantial
cash for households. As domestication requires a
perfect command of the propagation mechanisms of
the target species, prior knowledge of its natural
reproductive system is essential (Simons & Leakey
2004). If efficient, the natural regeneration process
may simply be encouraged for establishing artificial
plantations in crop lands.
In the African rainforest, plant taxa with
NTFP value that have been studied are generally
those with a high commercial potential
internationally, such as Baillonella toxisperma
Pierre, Garcinia lucida Vesque, Dacryodes edulis
Lam., Irvingia gabonensis Baill. and Prunus
africana Hook. In contrast, the African walnut,
Coula edulis Baill., may appear less interesting
since its harvestable non-timber products are
generally marketed only at a national scale, and it
has a limited distribution range. It occurs only in
the Guineo-Congolian evergreen rainforests in
West and Central Africa (Tchiegang et al. 1998).
However, C. edulis is undoubtedly very important
for Central and West African populations, as the
seeds are much appreciated by local people
(Moupela et al. 2011).
Despite the regional importance, C. edulis is
still considered as a wild tree species, and only
scarce information is available about its
reproductive ecology and population dynamics
(reviewed by Moupela et al. 2011). Specifically,
documented topics comprise botanical descriptions
(Adam 1971; Adriaens 1951; Aubréville 1959;
Louis & Léonard 1948; Villiers 1973; Vivien &
Faure 2011), some propagation methods (Bonnéhin
2000; De La Mensbruge 1966; Miquel 1987) and its
use by local populations (Ekop & Eddy 2005; Fort
et al. 2000; Van Neer & Clist 1991). The topics that
have not received attention so far regarding this
taxon and relevant aspects for domestication
processes include characterization of its natural
reproductive ecology, its fruit production potential
and the factors, which determine fruiting
characteristics.
The present work aimed to characterize the
natural reproductive patterns and fruit production
potential in C. edulis in an evergreen forest of
Gabon. Specifically, the study aimed to: (1)
describe the reproductive phenology of the focal
species, (2) record fruit and seed characteristics
and fruit production over two years, and (3) assess
relationships between dendrometric tree traits
(tree growth and crown exposure), reproductive
phenology and fruit production.
MOUPELA et al. 329
Material and methods
Study species
Coula edulis (Olacaceae) is commonly called
African walnut or Gabon nut. Its distribution
extends from Sierra Leone to the Democratic
Republic of Congo (Vivien & Faure 2011). It
displays a dependence on the evergreen forest zone
(Vivien & Faure 2011), prefers clayey soils (Sosef et
al. 2004) and can be found in the forest understorey
as well as in the top canopy, with an average adult
diameter of 100 cm and a height of 30 m.
The African walnut plays an important role in
the local economy (Bonnéhin 2000; Bukola & Kola
2008; Johnson & Johnson 1976; Schnell 1957): its
wood is renowned for resistance to termites and is
used locally in construction; its seeds contain a
sweet and odorless yellow oil with a high nutritive
value (Adraiens 1951; Busson 1965; Louis &
Léonard 1948; Tchiegang et al. 1998); and its bark
is used in traditional medicine as a purgative and
for treatment of back pain (Tamokou et al. 2011;
Walker & Sillans 1995). Cirad (2008) suggested
that the wood of C. edulis could be of economic
interest in the future.
Study site
The study was conducted in the forest
concession managed by the logging company
Precious Woods Gabon, located near Lastourville,
southeastern Gabon, between 0° 30´-1° 00´ S and
12° 30´-14° 0´ E (Fig. 1). The vegetation is a typical
evergreen forest dominated by Caesalpiniaceae
and Burseraceae (White 1986). The climate is
equatorial, with two high-rain seasons (March-
May and September-December; thereafter called
“wet” seasons) and two seasons with less rainfall
(June-August and January/February; hereafter
called “dry” seasons). Annual rainfall is 1,700 mm
and average annual temperature is around 26 °C,
with peaks from February to April and minimum
temperatures in July/August.
The dominant soil-type is a yellow lateritic
derived from a hydrolyzed substratum in which
most of the bases are exported. These soils display
a low chemical level of fundamental components,
but physical properties are suitable when the clay
content is high enough (Martin et al. 1981).
Phenological monitoring and assessment of
diameter growth
A total of 150 trees with a diameter at breast
height (dbh) ≥ 5 cm were monitored. Phenological
observations were recorded once per month from
March 2009 to July 2012 (40 months). Each
month, a team of three local technicians used a
semi-quantitative method to estimate the intensity
of phenophases, using binoculars to observe the
percentage of the crown covered by each organ
type: leaves, flowers and fruits (e.g. Bentos et al.
2008; Bourland et al. 2012; Engel & Martins 2005;
Tesfaye et al. 2011). For each organ, values
provided by the three observers were averaged to
obtain the estimate of phenophase intensity. In
addition, to minimize inter-observer variance, the
same team of technicians conducted all pheno-
logical observations throughout the 3-year study
period.
Annual diameter measurements were obtained
for a subsample of 130 trees (Table 1). We
excluded 20 trees from the 150 individuals
sampled for phenological observations since their
trunks had major defects. Diameter was consis-
tently measured by the same technician, using the
same diameter tape. The measurement line on the
tree trunk was delimited by two painted bands.
The annual increment in stem growth was
estimated by the difference in diameter of the last
and the first measurements (cm) divided by time
(years). In order to assess the impact of crown
position (sensu Dawkins 1958) on growth, the
method suggested by Moravie et al. (1999) was
used. This method is based on the codification
proposed by Dawkins (1958) and Synnott (1979).
The value assigned to a tree depends on its
exposure to sunlight. Moravie et al. (1999) and
Gourlet-Fleury (1998) demonstrated the reliability
of this codification (analogous to crown illumi-
nation index) in estimating the quantity of light
received by a target tree crown.
Fruits of Coula edulis and their characteristics
The total fruit biomass of C. edulis individuals
was assessed for two consecutive fruiting seasons
(2011 and 2012), from December to April (corres-
ponding to the yearly period of fruit production),
based on a sample of 25 adult trees of different
diameters. These met all the following criteria: (i)
the trees were isolated and selected from 130
individuals monitored for diametric growth and (ii)
the adult trees did not occur in areas accessible to
local populations. Each of these sampled trees was
far enough away from other conspecific trees
(minimum 30 m) that fruits found beneath a target
crown were sure to have come from that focal tree.
A plastic circular collector as large as the crown
330 REPRODUCTIVE ECOLOGY OF COULA EDULIS
Fig. 1. Location of the study site in a logging concession, Gabon.
Table 1. Diameter structure of sampled trees and parameters of mean fruit production (standard error).
Diameter
class
Number of trees (N)
Number of
mother trees
sampled
Fruit mass (kg tree-1) Number of fruits tree-1
Phenology Growth
diameter 2011 2012 2011 2012 2011 2012
[5-10 cm] 23 22 0 0 0 0 0 0
[10-20 cm] 29 28 6 6 1.8 (0.8) 4.3 (1.6) 62 (31) 135 (48)
[20-30 cm] 27 26 7 7 19.2 (4.7) 20.6 (4.9) 578 (160) 600 (161)
[30-40 cm] 25 21 6 6 60.1 (13.4) 48 (7.3) 2048 (492) 1544 (292)
≥ 40 cm 46 33 6 6 92.4 (51.2) 67.4 (31.1) 3207 (1808) 2810 (1209)
Total 150 130 25 25 42.2 (13.3) 36 (8.8) 1438 (470) 1246 (335)
surface was installed about 1 m above the ground,
to avoid predation from terrestrial animals. The
following data were gathered every week for each
collector: (1) the total number and weight of all
fruits, (2) the total weight of fresh and intact fruits,
(3) the proportion of fruits without kernel (sterile
fruits), based on a random subsample of 50
fruits/tree (a total of 1,250 fruits), and (4) the pro-
portion of fruits displaying insect damage, based on
a random subsample of 10 fruits/tree (250 fruits in
total). Points (3) and (4) were conducted only in
2011.
Finally, we selected 10 intact fruits from each
of 14 mother trees for a detailed characterization
of fruit physical traits. For each fruit and its nut
and kernel (the latter being dried before characte-
rization), we recorded: (1) mass, using a precision
scale, (2) length and width (median diameter),
measured with a caliper.
Data analysis
Following the description of each phenological
phase, the different diameter thresholds for the
onset of fertility (sexual maturity of trees) and
fructification were determined. The minimum
diameters for fertility (MDFe) and fructification
(MDFr) correspond to the minimum stem diameters
for which flowers and fruits were observed,
respectively. The relationship between stem dia-
meter and probability of fruiting, at least, once over
a three-year period was established using a binary
logistic regression. The relationship between
diameter and frequency of fruiting was tested with
a Kruskal-Wallis non-parametric analysis of
variance (ANOVA), to compare the average
diameter of four expected productive groups (trees
with dbh > MDFr): (1) those that did not bear fruit,
T0; (2) those that produced fruit once in three years,
MOUPELA et al. 331
T1; (3) those that bore fruit twice in three years, T2;
and finally (4) those that bore fruit three times in
three years, T3. A pairwise comparison of groups
was then performed using the Mann-Whitney test.
For each tree, the mean annual diameter
increment (ADI), over three years, was calculated,
and the influence of initial diameter and crown
exposure on ADI was tested by performing a linear
regression and analysis of covariance (ANCOVA),
respectively, assuming that diameter is the cova-
riate of the crown exposure parameter for the latter.
Variation in ADI between years was tested using an
ANCOVA with repeated measures. To verify a
possible impact of fruiting frequency on stem
growth, ADI of the different fruiting groups (T0, T1,
T2 and T3) was compared. For that, we performed
an ANCOVA and specified tree diameter as the
covariate of fruiting frequency. Along the same
lines, we performed a linear regression between
individual ADI and fruit production for testing the
hypothesis that with more resources allocated to
reproduction tree growth will be less.
A biometric characterization of the diaspore
traits (fruits, kernels and nuts) was carried out.
We then presented a correlation matrix illus-
trating at which extent a given organ can be
estimated from the others.
The statistical analyses were carried out with
the STATISTICA version 6.0 program. Mean
values of different parameters were provided with
their standard errors.
Results
Phenological spectrum and reproductive
demography
As the African walnut is an evergreen species,
leaves are abundantly present in the crown
regardless of climatic season. The renewal of older
leaves is continuous, though more pronounced
during the long dry season (June-August). Leaf
cover peaks during the wet seasons.
Flowering and fruiting are regular and annual
in C. edulis (Fig. 2). At individual level, these two
reproductive phases last on average 46 ± 34 days
and 76 ± 47 days, respectively. Flowering peaks in
November or December, but total flowering time
can encompass a 6-month period, from July to
January. The end of the dry season from February
to March corresponds with the period of greatest
fruit production.
Flowering was observed in individuals with a
minimum diameter of 10.5 cm (MDFe). However,
mature fruits were observed only on individuals of
at least 12.3 cm diameter (MDFr). The probability
of an individual being able to produce ripe fruits
can be estimated from Fig. 3; this suggests dbh ≈
23 cm is the most probable minimum size for
bearing ripe fruits (with a probability > 95 %), at
least once during three years (Fig. 3). Therefore,
dbh of 23 cm may be considered a reliable estimate
of the diameter for regular fructification.
Fruiting frequency was affected by tree
diameter (ANOVA K-W, H = 30.24, P < 0.0001). A
pairwise comparison showed that there is no
significant difference between the three first cate-
gories (T0, T1 and T2). Trees that produced fruit
every year had significantly greater diameters
than the other trees: dbh = 38.9 ± 15.2 cm versus
19.9 ± 7.0 cm.
Diameter growth and interaction with
other tree traits
Overall, the mean ADI (130 individuals, 3-year
measurement period) was 0.22 ± 0.2 cm2 year-1.
Pearson’s correlation revealed no significant
relationship between initial tree diameter and ADI
(r = 0.149 and P = 0.097). However, crown exposure
influenced diameter growth (ANCOVA, F5,117 =
3.63, P = 0.004). The pairwise comparison
highlighted a significant difference between cate-
gories 2c (least exposed) and 2b (grouping of highly
exposed individuals) (Fig. 4). ADI significantly
varied from one year to another (repeated ANOVA;
F3,120 = 55.78 and P < 0.001). Repeated Student's t
test showed that growth in year 2 was significantly
different from the two other years. Finally, fruiting
frequency did not affect annual diameter
increment (ANCOVA; F3,77 = 0.96 and P = 0.415).
Fruit production and interaction with tree size
parameters
Mean fruit cover ranged from 21 % in 2012 to
38 % in 2011. As expected, fruit production followed
the same trend, with 36.0 ± 8.8 kg tree-1 in 2012
versus 42.0 ± 13.3 kg tree-1 in 2011. The yearly
number of fruits per tree also varied from 1246 to
1438 on average (Table 1). Fruit production signi-
ficantly correlated with tree diameter, with
Pearson's correlation coefficients ranging from 0.70
to 0.75 (Fig. 5). In addition, there was a significant
correlation between yearly fruit production at an
individual level: trees that gave the lowest amount
of fruit in 2011 did similarly in 2012, while the
best producers in 2011 maintained their status in
2012 (Pearson's r = 0.85; P < 0.001). But there was
3
F
2
F
d
e
x
n
a
(
a
4
k
l
T
7
1
c
l
o
3
32
F
ig. 2. Ph
e
2
= Februar
y
F
ig. 3. Pro
b
d
iameter. L
o
e
-8.62 + (0.54x
);
w
x
= diameter
n
o relation
s
a
verage fr
u
(
Pearson's r
Thorou
g
a
mean ma
s
4
.0 ± 0.3 c
k
ernel wei
g
l
ength of 3.
6
T
he nuts p
r
7
.4 ± 1.5 g
w
1
.8 ± 0.12
c
c
haracters
s
l
ated (an e
o
rgans was
g
e
nology dyn
a
y
; etc.
b
ability of
o
gistic regre
s
w
here y = pr
o
of the tree (
c
s
hip betwee
u
it produc
t
= 0.34; P =
g
h measure
m
s
s of 38.5 ±
7
m and dia
m
g
hed 21.3
±
6
± 0.2 cm
a
r
esented a
w
ith length
c
m. Simple
s
howed tha
t
xample of
g
iven in Ta
b
REP
R
a
mics. Mont
h
fructificatio
n
s
sion: y = e
o
bability of f
r
c
m).
n diameter
t
ion over
t
0.098).
m
ents of 14
0
7
.7 g with
a
m
eter 3.8
±
±
3.7 g on
a
nd diamet
e
mean indi
v
2.0 ± 0.13 c
m
correlatio
n
t
all are sig
n
the mass
o
b
le
2).
We
a
R
ODUCTIVE
h
s: 1 = Jan
u
n
accordin
g
(-8.62 + 0.54x)
/
r
uctification,
increment
t
he two
y
0
fruits rev
e
a
mean leng
t
±
0.2 cm.
E
average wi
t
e
r 3.0 ± 0.2
v
idual weig
h
m
and dia
m
n
between t
h
n
ificantly c
o
o
f the diff
e
a
lso noted
ECOLOGY
O
u
ary;
g
to
/
(1+
and
and
y
ears
e
aled
t
h of
E
ach
t
h a
cm.
h
t of
m
eter
h
ese
o
rre-
e
rent
that
Fi
g
fro
m
cod
e
su
n
tre
e
du
a
att
a
in
ent
i
no
occ
u
P
=
not
64
be
por
t
in
t
0.4
2
Co
u
spe
c
po
p
dia
m
to
l
int
o
ter
m
rai
n
of
pre
per
i
the
r
O
F COULA E
D
g
. 4. Growth
m
Moravie
e
e
s follows a
light.
e
diameter
d
a
l mass of fr
u
A
notable
a
cked by in
s
the kernel
.
i
rely destro
y
significant
u
rrence of
d
=
0.339). We
appear to
b
% of non-d
a
sterile, as
t
ion of "em
p
t
rees with
h
2
and P = 0.
0
This is th
u
la edulis,
a
c
ies for
W
p
ulations.
m
eter grow
t
l
ong-term o
b
o
account i
n
m
investig
a
n
forest con
d
studies ar
e
s
ent work,
b
i
od and co
r
efore, high
l
D
ULIS
following t
h
e
t al. 1999).
gradient of
e
d
id not infl
u
u
it, kernel
a
proportion
s
ects (56 %),
.
In some
y
ed, and b
e
correlation
d
amaged fr
u
also obser
v
b
e efficient
i
a
maged fru
i
the kernel
p
ty" fruits
w
h
igher valu
e
0
37).
Discu
s
e first des
c
a
n importa
n
W
est and
C
Ideally, s
t
h or fruit
b
servation
p
n
fluences of
a
tions are
d
itions, and
e
able to
p
b
ased on da
t
mprising a
l
y valuable.
h
e Dawkins
c
The arran
g
e
xposure of
t
u
ence the a
v
a
nd nut (Ta
b
of Coula
resulting i
n
cases, the
e
came meal
y
between t
r
u
its (Pearso
n
v
ed that pol
l
in the stud
y
i
ts sampled
was lacki
n
w
as signific
a
e
s of dbh (
P
s
sion
c
ription of
n
t NTFP-p
r
C
entral Afr
i
tudies of
production
p
eriods in
o
climate. H
o
challengin
g
only a sm
a
p
rovide suc
h
t
a collected
o
large sa
m
c
ode (adapt
e
g
ement of t
h
t
he crowns
t
v
erage indiv
i
b
le 3).
fruits we
r
n
many hol
e
kernel w
a
y
. There w
a
r
ee size a
n
n
's r = -0.1
9
l
ination do
e
y
populatio
n
appeared
t
n
g. The pr
o
a
ntly great
e
P
earson's r
its type f
o
r
oducing tr
e
i
can huma
phenolog
y
require mi
d
o
rder to ta
k
o
wever, lon
g
g
in tropic
a
a
ll proportio
h
data. T
h
o
ver a 3-ye
a
m
ple size, i
s
e
d
h
e
t
o
i
-
r
e
e
s
a
s
a
s
n
d
9
;
e
s
n
:
t
o
o
-
e
r
=
o
r
e
e
n
y
,
d
-
k
e
g
-
a
l
n
h
e
a
r
s
,
F
p
y
2
t
G
d
p
s
s
s
d
u
A
f
o
i
p
F
ig. 5. Rel
a
p
roduced a
n
y
ear 2011 (
r
2
012 (r² = 0.
5
Heckets
w
t
he phenolo
g
G
abon. His
d
bh ≥ 19 c
m
p
artial leaf-
s
eason (Jun
e
s
tudy confi
r
s
tudies usi
n
d
iameter
u
nderstand
i
A
frican wal
n
f
ruits from
2
o
ccur in s
m
i
nformation
p
lanting th
i
a
tionship b
e
n
d diameter
r
² = 0.49; P
=
5
8; P < 0.00
1
P
henolog
y
w
eiler (199
2
g
y of C. ed
u
results we
r
m
. Hecketsw
e
shedding o
c
e
to August
)
r
m this, an
n
g small
s
threshold
i
ng the re
p
n
ut. Coula
2
3 cm of d
b
m
aller tree
s
is essenti
a
i
s tree spec
e
tween bi
o
of the adul
t
=
0.001) an
d
1
).
y
of
C. edu
l
2
) was the
f
u
lis in the I
m
r
e based on
e
iler (1992)
c
curs durin
g
)
. The resul
t
d also sho
w
s
ample size
were i
n
p
roductive
edulis regu
b
h, but frui
t
s
, to dbh
o
a
l for farme
r
ies in crop
M
o
mass of
t
tree (a) i
n
d
(b) in the
l
is
f
irst to des
c
m
passa fore
15 trees w
i
observed t
h
g
the longes
t
t
s of the pre
w
that pre
v
s with a
h
n
adequate
ecology of
larly bears
t
production
o
f 12 cm.
T
r
s intereste
land. One
M
OUPELA et
a
fruit
n
the
year
c
ribe
st of
i
th a
h
at a
t
dry
sent
v
ious
h
igh
for
the
ripe
can
T
his
d in
may
not
e
eve
r
fru
i
ind
u
tra
n
flo
w
dry
A
n
infl
an
d
In
G
71
rep
r
the
(19
9
pre
wa
s
Ko
u
Da
ï
an
d
ex
a
ph
e
et
a
te
m
fru
i
Do
n
Sti
l
rel
a
ma
y
flo
w
dis
e
bec
a
of t
h
an
a
int
e
mi
n
bia
s
cas
e
cue
ini
t
dat
a
are
wo
u
we
e
ph
e
ph
e
im
p
str
a
an
d
las
t
dis
p
(e.
g
a
l.
e
that alth
o
r
green for
e
i
ting appea
r
u
ctions bec
a
n
sition per
i
w
ering (Nov
season for
t
n
umber of a
u
uence of v
a
d
fruiting r
a
G
abon, Hec
% of tree
r
oductive ti
m
southern
9
9) also not
e
s
ents two p
e
s
confirmed
u
adio (200
9
ï
nou et al. (
2
Reproducti
v
d
intrinsic
ct climatic
e
nomena re
m
a
l. (1999)
m
perature t
o
i
ting. Othe
r
n
ati 2005;
D
l
es 2009;
W
a
tive humi
d
y
constitu
t
w
ering and
f
e
ntangling
a
use most o
f
h
e potentia
l
a
lyses. In
a
e
r-correlate,
n
ants of re
p
s
ed and po
w
e
, we were
, which
m
t
iation in
C
a
for the st
u
annual in
C
u
ld be nece
s
e
n change
s
e
nophases.
Knowledge
e
nophases a
n
p
ortant for
a
tegies, in
o
d
to plan cr
o
t
up to six
m
p
ersed pla
n
g
. see Debro
u
o
ugh C. ed
u
e
sts, the p
r
to be se
n
a
use they
m
i
ods: the e
n
ember/Dece
t
he fruiting
u
thors have
a
rious clim
a
a
tes in the
t
ketsweiler
(
species in
m
ing simil
a
part of C
a
e
d that frui
t
e
aks, with
o
in other sp
e
9
) for
M
a
n
2
012) for
M
i
l
v
e phenop
h
physiologic
factors un
m
ain obscu
r
suggested
o
be the
m
r
s (Anderso
n
D
avid et a
l
W
hite 199
4
d
ity, day le
n
t
e the tri
g
f
ruiting. Th
the role
o
f
these stud
l
influencin
g
a
ddition, se
v
meaning
t
p
roductive
w
erful met
h
not able t
o
m
ay explai
n
C
. edulis.
W
u
dy area, b
u
C
. edulis,
m
s
sary for t
e
s
in rainf
a
of the d
u
n
d associat
e
the devel
o
o
rder to sch
e
o
p rotation.
F
m
onths; simi
l
n
t species
u
x 1998). Th
e
u
lis is a ty
p
eaks of fl
o
n
sitive to s
o
m
atch climat
n
d of a we
mber), and
peak (Febr
u
documente
d
a
tic factors
t
ropical eq
u
(
1992) repo
r
our study
a
r to that of
a
meroon, M
t
ripening o
f
o
ne in Marc
h
e
cies-specifi
c
n
sonia alti
s
l
icia excelsa
h
ases depen
d
al cues.
H
n
der the o
n
r
e (Debroux
changes i
n
m
ain factor
s
n
et al. 20
0
l
. 2012; Ro
j
4
) have p
r
n
gth and so
l
g
gers of
i
e difficultie
s
o
f each fa
c
ies only con
g
factors for
v
eral clima
that testin
g
phenology
r
h
ods of an
a
o
determine
n
flowering
W
e had mo
n
u
t because
m
any years
o
e
sting relat
i
a
ll and t
h
u
ration of
e
d plant re
q
o
pment of
e
dule diasp
o
F
ruiting in
l
ar to other
from the
s
e
oretically,
t
3
3
p
ical plant
o
o
wering a
n
o
me climat
i
ic seasons
o
t season f
o
the end of
u
ary/March
)
d
the possib
l
on floweri
n
u
atorial zon
e
r
ted that 6
3
region ha
v
C. edulis. I
barga et a
f
tree speci
e
h
. This tre
n
c
studies [e.
g
s
sima Che
v
Berg].
d
on clima
t
H
owever, t
h
n
set of tho
s
1998). Kha
n
rainfall
o
s
influenci
n
0
5; Bollen
&
j
as-Robles
&
r
oposed th
a
l
ar radiatio
i
nitiation
o
s
incurred i
c
tor may
b
sidered so
m
their specif
i
t
ic variabl
e
g
the dete
r
r
equires u
n
a
lysis. In o
u
the climat
i
or fruiti
n
n
thly rainfa
l
phenophas
e
o
f monitori
n
i
onships be
t
h
e start
o
reproducti
v
q
uirements
i
agroforest
r
o
re collectio
C. edulis ca
animal see
d
s
ame regio
t
he selectiv
e
3
3
o
f
n
d
i
c
o
r
o
r
a
)
.
l
e
n
g
e
.
3
-
v
e
n
l.
e
s
n
d
g
.
v
.,
t
e
h
e
s
e
n
o
r
n
g
&
&
a
t
n
o
f
n
b
e
m
e
i
c
e
s
r
-
n
-
u
r
i
c
n
g
l
l
e
s
n
g
t
-
o
f
v
e
i
s
r
y
n
n
d
-
n
e
334 REPRODUCTIVE ECOLOGY OF COULA EDULIS
Table 2. Correlation coefficients (r) between three
fruit components: mass of the entire fruit, the kernel
and the nut ** and *** indicate significant correlations
(Pearson's r).
Fruit
mass
Kernel
mass Nut mass
Fruit mass - 0.516*** 0.321***
Kernel mass - - 0.245**
Table 3. Correlation coefficient (r) between tree
diameter and mean weight of the entire fruit, the
kernel and the nut.
Parameters r P-value
Fruit mass -0.006 0.945
Kernel mass 0.017 0.841
Nut mass -0.086 0.327
pressure against individuals to reproduce at peak
moments has been proposed as an explanation for
long flowering and fruiting periods in plants
(Ollerton & Lack 1992). This theory could apply to
the case of C. edulis: although we did not record
the temporal dynamics of predation incidence, it is
likely that fruit predation or dispersal will mostly
occur at the peak of fruiting, as is the case for the
majority of animal-dispersed plant species (Aizen
2003; Howe & Smallwood 1982; Tadwalkar et al.
2012). As most frugivorous animals interested in
C. edulis fruits act as predators rather than
dispersers (Moupela, submitted), diaspores
produced at fruiting peak may not be selectively
favored, leading to the persistence of a large
variance in flowering and fruiting times, as
described for other species (Nyiramana et al. 2011;
Beaune et al. 2012) .
Annual diameter increment (ADI):
determinants and interactions with tree
characteristics
The annual diameter increment of C. edulis
was low (0.22 cm year-1), as is that of Garcinia
lucida (0.1 cm year-1, Guedje et al. 2003), which is
also an understorey NTFP-producing tree species
of tropical forests. In general, shade-bearing woody
plants have a slower growth rate than canopy tree
species (mostly light-demanders). In the case of C.
edulis, diameter growth was not influenced by
either initial tree diameter or fruit production. The
former result concurs with the findings of Gourlet-
Fleury (1998) who demonstrated that in general,
initial diameter be used to predict diameter incre-
ment. However, the latter result was surprising: as
Fenner (1998) showed, efficient reproduction
generally occurs to the detriment of diameter
growth. It should be emphasized that for the
purpose of the present study, we considered fruit
production to reflect reproductive effort, and stress
that flowering effort should also be included. Even
in cases where flower cover could represent
reproduction, in the global equation to quantify
individual reproductive effort, the relative
importance of flowering and fruiting is unknown.
We acknowledge and stress that fruiting effort is
not always analogous to reproductive effort. On
the other hand, we found that ADI was influenced
by crown exposure of the tree: the least exposed
trees displayed the lowest growth values. While
this result was in accordance with several other
studies (e.g. Dyer et al. 2010; King et al. 2005), it
highlighted uncertainties regarding light require-
ments of C. edulis reported by Doucet (2003): a
shade-bearing tree species typical of mature
evergreen forest should present greater growth at
understory stage than at canopy level.
Based on an ADI of 0.22 cm year-1, it would
take approximately 104 years for an individual to
reach the regular diameter for fructification (23
cm) in natural conditions. Assuming that this
natural growth pattern would not significantly
change in a farming environment, vegetative
propagation techniques (stem cutting, grafting,
layering, etc.) for domestication are recommended,
as they generally accelerate both vegetative and
sexual maturity processes (Meunier et al. 2008).
Fruit production: perspectives for agroforestry
This study was the first to extensively
characterize fruit production and fruit traits of C.
edulis.
Fruit production significantly varied between
individuals and between years. We note that our
estimates could have been biased by fruit
consumption from arboreal rodents before
diaspores fell beneath mother trees. Fruit
production is also under genetic influence, and
more than flowering for example (Nanson 2004),
and this is evident here as the correlation
coefficient between years was high and significant
(Pearson's r = 0.85). This assertion requires further
investigation, in order to determine the heri-
tability of the trait; but we assume fruiting para-
meters to be highly heritable in C. edulis as they
are in most of plant species (McCarthy & Quinn
1990). Therefore, for agroforestry purposes, trees
MOUPELA et al. 335
should be selected on the basis of favorable
fruiting characteristics compared to other indivi-
duals from the same population. Ideally, candidate
trees should have similar diameters to prevent a
biased selection work due to the significant
correlation between tree diameter and fruit
production (see also Anderson et al. 2005;
Takenoshita et al. 2008; Zardo & Henriques 2011)
(otherwise, the majority of selected individuals
would be simply the larger ones).
Unlike the observations of Hecketsweiler
(1992) and Loung (1996), C. edulis seems to be a
highly productive species with a regular fruiting
rate in Gabon. Nevertheless, a significant pro-
portion of diaspores are subject to insect attack
which can reduce their economic value. Moreover,
more than 60 % of the fruits appeared to be sterile.
This is likely to be the main difficulty for mass pro-
duction of C. edulis by sowing in agroforestry
systems. Undoubtedly, vegetative propagation is
necessary to domesticate the African walnut.
The average sizes of fruits, kernels and nuts
observed in Gabon were similar to those indicated
for other African countries (Aubréville 1959;
Villiers 1973; Vivien & Faure 2011). Although tree
diameter determined the total fruit production per
individual, it did not influence the average mass of
each fruit. Similar observations have been made
for Irvingia gabonensis Baill. in Cameroon
(Atangana et al. 2001).
Conclusions
The African walnut is an important resource
for people in Central and West Africa, making it a
valuable candidate for domestication. Our study
constituted the first overview of the reproductive
ecology of C. edulis, with a special emphasis on
aspects pertaining to agroforestry. We revealed
that, as with most rainforest tree species, C. edulis
can become sexually mature at small diameters.
But this might be due to a low annual diameter
growth, meaning that the youngest flowering C.
edulis trees are as old as those of other taxa, which
display greater diameter increments. We also
found interactions between phenology, tree size
parameters, diameter growth and fruit production.
For example, it should be borne in mind that the
number of fruits produced per tree seems to be
highly heritable, but this suggestion requires
further investigation. For agroforestry purposes,
vegetative propagation methods should be tested
in the future, since sowing is not likely to provide
good results and the majority of seeds are
naturally sterile. In addition, vegetative
propagation maintains tree characteristics while
accelerating flowering and fruit production.
Acknowledgements
The authors would like to thank the state of
Gabon through PAI-DRH for its financial support
for the doctoral work of M. Moupela. Our thanks
also go to the partner structures (Institut National
Supérieur d’Agronomie et de Biotechnologies/
University of Science and Technology of Masuku,
Precious Woods Gabon, Millet, Nature Plus and
Gembloux Agro-Bio Tech/University of Liège) for
the technical and logistical support and to the
anonymous readers for their criticisms and
suggestions. Finally, we would like to thank M. C.
Decurtins, P. Geffroy, N. Tagg, P. Lejeune, M.
Federspiel, J. Phillipart, A. G. Boubady, A. Assame,
R. Milemba, J. B. Bokomba, and J. F. Toka for
their collaboration.
References
Adam, J. G. 1971. Flore Descriptive des Monts Nimba.
Muséum national d’histoire naturelle, Paris.
Adriaens, E. L. 1951. Les Oléagineux du Congo Belge.
Ministère des Colonies, Bruxelles.
Aiyelaagbe, I. O. O., A. O. Adeola, L. Popoola & K. O.
Obisesan. 1998. Agroforestry potential of Dacryodes
edulis in the oil palm-cassava belt of southeastern of
Nigeria. Agroforestry Systems 40: 263-274.
Aizen, M. A. 2003. Influences of animal pollination and
seed dispersal on winter flowering in a temperate
mistletoe. Ecology 84: 2613-2627.
Anderson, D. P., E. V. Nordheim, T. C. Moermond, Z. B.
Gone Bi & C. Boesch. 2005. Factors influencing tree
phenology in the Taï National Park, Ivory Coast.
Biotropica 37: 631-640.
Anegbeh, P. O., C. Usoro, V. Ukafor, Z. Tchoundjeu, R.
R. B. Leakey & K. Schreckenberg. 2003. Dome-
stication of Irvingia gabonensis: 3. Phenotypic
variation of fruits and kernels in a Nigerian village.
Agroforestry Systems 58: 213-218.
Atangana, A. R., Z. Tchoundjeu, J. M. Fondoun, E.
Asaah, M. Ndoumbe & R. R. B. Leakey. 2001.
Domestication of Irvingia gabonensis: 1. Phenotypic
variation in fruits and kernels in two populations
from Cameroon. Agroforestry Systems 53: 55-64.
Aubréville, A. 1959. La Flore Forestière de la Côte
D’ivoire. Tome 1. CTFT, Nogent-sur-Marne.
Beaune, D., L. Bollache, B. Fruth & F. Bretagnolle.
2012. Bush pig (Potamochoerus porcus) seed pre-
dation of bush mango (Irvingia gabonensis) and
336 REPRODUCTIVE ECOLOGY OF COULA EDULIS
other plant species in Democratic Republic Congo.
African Journal of Ecology 50: 509-512.
Bentos, T. V., R. C. G. Mesquita & G. B. Williamson.
2008. Reproductive phenology of Central Amazon
pioneer trees. Tropical Conservation Science 1: 186-
203.
Bollen, A. & G. Donati. 2005. Phenology of littoral forest
of Sainte Luce, southeastern Madagascar. Bio-
tropica 37: 32-43.
Bonnéhin, L. 2000. Domestication Paysanne des Arbres
Fruitiers Forestiers. Cas de Coula edulis Baill.
(Olacacée) et de Tieghemella heckelii Pierre
(Sapotacée) Autour du Parc National du Taï, Côte
d’Ivoire. Ph.D. Thesis, Wageningen Agricultural
University.
Bourland, N., L. Kouadio Yao, P. Lejeune, B. Sonké, J.
Philippart, K. Daïnou, F. Fétéké & J. L. Doucet.
2012. Ecology of Pericopsis elata (Fabaceae), a
timber species considered as endangered, in
southeastern Cameroon. Biotropica 44: 840-847.
Bukola, C. A. T. & A. Kola 2008. Antimicrobial activities
of Coula edulis. Research Journal of Medicinal
Plant 2: 86-91.
Busson, F. 1965. Les Plantes Alimentaires de l’Ouest
Africain. Etude Botanique, Biologique et Chimique.
Leconte, Marseille.
Cirad. 2008. Fiche Technique de Coula edulis. Fiche
n°197. Tropix 6.0. Caractéristiques Technologiques
de 245 Essences Tropicales. Cirad, Montpellier.
Daïnou, K., E. Laurenty, G. Mahy, O. J. Hardy, Y.
Brostaux, N. Tagg & J. L. Doucet. 2012. Pheno-
logical patterns in a natural population of a tropical
timber tree species, Milicia excelsa (Moraceae):
Evidence of isolation by time and its interaction
with feeding strategies of dispersers. American
Journal of Botany 99: 1-11.
David, J. P., B. S. Murugan & R. Manakadan. 2012.
Seasonality in fruiting of fig and non-fig species in a
tropical dry evergreen forests in Sriharikota Island,
southern India. Tropical Ecology 53: 1-13.
Dawkins, H. C. 1958. The Productivity of Tropical High-
forest Trees and the Reaction to Controllable
Environment. Commonwealth Forestry Institute,
Oxford.
Debroux, L. 1998. L’aménagement des Forêts Tropicales
Fondé sur la Gestion des Populations D’arbres:
l’exemple du Moabi (Baillonella toxisperma Pierre)
Dans la Forêt du Dja, Cameroun. Ph.D. Thesis,
Faculté Universitaire Des Sciences Agronomiques
de Gembloux, Gembloux.
De La Mensbruge, G. 1966. La Fermination et les
Plantules des Essences Arborées de la Forêt Dense
Humide de la Côte D’Ivoire. CTFT, Nogent-sur-
Marne.
Doucet, J. L. 2003. L’Alliance Délicate de la Gestion
Forestière et de la Biodiversité Dans les Forêts du
Centre du Gabon. Ph.D. Thesis, Faculté Univer-
sitaire des Sciences Agronomiques de Gembloux,
Gembloux.
Dyer, L. A., T. R. Walla, H. F. Greeney, J. O. Stireman &
R. F. Hazen. 2010. Diversity of interactions: A metric
for studies of biodiversity. Biotropica 42: 281-289.
Ekop, A. S. & N. O. Eddy. 2005. Comparative studies of
levels of toxicant in the seeds of Terminalia catappa
(Indian Almond) and Coula edulis (African Walnut).
Chem Class Journal 2: 74-76.
Engel, V. L. & F. R. Martins. 2005. Reproductive
phenology of Atlantic forest tree species in Brazil:
an eleven year study. Tropical Ecology 46: 1-16.
FAO. 2010. Ressources Phytogénétiques. Ne pas les
Utiliser, c’est les Perdre. http://www.fao.org/nr/cgrfa
(Accessed on 23 September 2011).
Fenner, M. 1998. The phenology of growth and repro-
duction in plants. Perspectives in Plant Ecology,
Evolution and Systematics 1:78-91.
Fort, D. M., S. R. King, T. J. Carlson & S. T. Nelson.
2000. Minquartynoic acid from Coula edulis.
Biochemical Systematics and Ecology 28: 489-490.
Gourlet-Fleury, S. 1998. Indices de compétition en forêt
dense tropicale humide: étude de cas sur le
dispositif sylvicole expérimental de Paracou
(Guyane Française). Annales des Science Forestières
55: 623-654.
Guariguata, M. R., C. García-Fernández, D. Sheil, R.
Nasi, C. Herrero-Jáuregui, P. Cronkleton & V.
Ingram. 2010. Compatibility of timber and non
timber forest product management in natural
tropical forests: perspectives, challenges, and
opportunities. Forest Ecology and Management 259:
237-245.
Guedje, N. M., J. Lejoly, B. A. Nkongmeneck & W. B. J.
Jonkers. 2003. Population dynamics of Garcinia
lucida (Clusiaceae) in Cameroonian Atlantic forests.
Forest Ecology and Management 177: 231-241.
Hecketsweiler, P. 1992. Phénologie et Saisonnalité en
Forêt Gabonaise: l’Exemple de Quelques Espèces
Ligneuses. Ph.D. Thesis, Université de Montpellier
II, Montpellier.
Hill, T., Y. Ouedraogo & L. Conditamde. 2007. L’en-
treprise villageoise d’exploitation des arbres au
Burkina Faso-appuyer le développement des petites
entreprises dont l’activité est fondée sur les produits
forestiers non ligneux. Unasylva 228: 32-37.
Howe, H. F. & J. Smallwood. 1982. Ecology of seed
dispersal. Annual Review of Ecology and Syste-
matics 13: 201-228.
Johnson, E. J. & T. J. Johnson. 1976. Economic plants in a
rural Nigerian market. Economic Botany 30: 375-381.
MOUPELA et al. 337
Kanten, R. V. & J. Beer. 2005. Production and phenology
of the fruit shrub Eugenia stipitata in agroforestry
systems in Costa Rica. Agroforestry Systems 64:
203-209.
Khan, M. L., P. Bhuyan, U. Shankar & N. P. Todaria.
1999. Seed germination and seedling fitness in
Mesua ferrea L. in relation to fruit size and seed
number per fruit. Acta Oecologica 20: 599-606.
King, D. A., S. J. Davies, M. N. Nur Supardi & S. Tan.
2005. Tree growth is related to light interception
and wood density in two mixed dipterocarp forests
of Malaysia. Functional Ecology 19: 445-453.
Kouadio, Y. L. 2009. Mesures Sylvicoles en vue
D’améliorer la Gestion des Populations D’essences
Commerciales de l’Est du Cameroun. Ph.D. Thesis,
Gembloux Agro-Bio Tech/ULg, Gembloux.
Leakey, R. R. B. & A. J. Simons.1998. The domestication
and commercialization of indigenous trees in agro-
forestry for the alleviation of poverty. Agroforestry
Systems 38: 165-176.
Louis, J. & J. Léonard. 1948. Olacaceae. pp. 250-253. In:
W. Robyns & P. Starner (eds.) Flore du Congo belge
et du Ruanda-Urundi. INEAC, Bruxelles.
Loung, J. F. 1996. Les pygmées camerounais face à
l’insuffisance des produits alimentaires végétaux
dans la forêt tropicale. pp. 325-336. In: C. M. Hladik,
A. Hladik, H. Pagezy, O. F. Linares, G. J. A.
Koppert & A. Froment (eds.) L’alimentation en Forêt
Tropicale: Interactions Bioculturelles et Perspectives
de Développement. Unesco, Paris.
Martin, D., Y. Chatelin, J. Collinet, E. Guichard & G.
Sala. 1981. Les Sols du Gabon: Pédogénèse,
Répartition et Aptitudes. Orstom, Paris.
Mbarga Bibani, R., W. B. J. Jonkers & J. Essama
Etoundi. 1999. Phénologie de 86 essences
productrices de bois d’œuvre de la forêt dense humide
sempervirente du Sud-Cameroun - Résultats
préliminaires. In: R. I. Nasi, Amsallem & S.
Drouineau (eds.) La Gestion des Forêts Denses
Africaines Aujourd'hui: Actes du Séminaire Forafri
de Libreville, Gabon. Cirad-forêt, Montpellier.
McCarthy, B. C. & A. J. Quinn. 1990. Reproductive
ecology of Carya (Jugandaceae): Phenology,
pollination, and breeding system of two sympatric
tree species. American Journal of Botany 77: 261-
273.
Meunier, Q., R. Bellefontaine & O. Monteuuis. 2008. La
multiplication végétative d’arbres et arbustes
médicinaux au bénéfice des communautés rurales
d’Ouganda. Bois et Forêts des Tropiques 295: 71-82.
Miquel, S. 1987. Morphologie fonctionnelle de plantules
d'espèces forestières du Gabon. Adansonia 9:
101-121.
Moravie, M. A., M. Durand & F. Houllier. 1999.
Ecological meaning and predictive ability of social
status, vigor and competition indices in a tropical
rain forest (India). Forest Ecology and Management
117: 221-240.
Moupela, C., C. Vermeulen, K. Daïnou & J. L. Doucet.
2011. Le noisetier d’Afrique (Coula edulis Baill.). Un
produit forestier non ligneux méconnu. Bio-
technologie, Agronomie, Société et Environnement
15: 451-461.
Nanson, A. 2004. Génétique et Amélioration des Arbres
Forestiers. Presses Agronomiques de Gembloux,
Gembloux.
Nyiramana, A., I. Mendoza, B. A. Kaplin & P.-M. Forget.
2011. Evidence for seed dispersal by rodents in
Tropical Montane Forest in Africa. Biotropica 43:
654-657.
Ollerton, J. & A. J. Lack. 1992. Flowering phenology -
An example of relaxation of natural selection?
Trends in Ecology & Evolution 7: 274-276.
Rojas-Robles, R. & F. G. Stiles. 2009. Analysis of a
supra-annual cycle: reproductive phenology of the
palm Oenocarpus bataua in a forest of the
Colombian Andes. Journal of Tropical Ecology 25:
41-51.
Schnell, R. 1957. Plantes Alimentaires et vie Agricole de
l’Afrique Noire: Essai de Phytogéographique
Alimentaire. Larose, Paris.
Simons, A. J. & R. R. B. Leakey. 2004. Tree dome-
stication in tropical agroforestry. Agroforestry
Systems 61: 167-181.
Sosef, M. S. M., Y. Issembe, H. P. Bourobou & W. J. M.
Koopman. 2004. Botanical diversity of the
Pleistocene forest refuge Monts Doudou. pp. 17-92.
In: B. Fischer (ed.) Monts Doudou, Gabon: a Floral
and Faunal Inventory with Reference to Elevational
Variation. California Academy of Sciences.
Synnott, T. J. 1979. A Manual of Permanent Plot
Procedure for Tropical Rainforests. Commonwealth
Forestry Institute, Oxford.
Tadwalkar, M. D., A. M. Joglekar, M. Mhaskar, R. B.
Kanade, B. Chavan, A. V. Watve, K. Ganeshaiah &
A. A. Patwardhan. 2012. Dispersal modes of woody
species from the northern Western Ghats, India.
Tropical Ecology 53: 53-67.
Takenoshita, Y., C. Ando, Y. Iwata & J. Yamagiwa.
2008. Fruit phenology of the great ape habitat in
Moukalaba-Doudou National Park, Gabon. African
Study Monographs 39: 29-39.
Tamokou, J. D. D., J. R. Kuiate, D. Gatsing, A. P. Nkeng
Efouet & A. J. Njouendou. 2011. Antidermatophytic
and toxicological evaluations of dichloromethane-
methanol extract, fractions and compounds isolated
from Coula edulis. Iranian Journal of Medical
Sciences 36: 111-121.
338 REPRODUCTIVE ECOLOGY OF COULA EDULIS
Tchiegang C., C. Kapseu & M. Parmentier. 1998.
Chemical composition of oil from Coula edulis (Bail.)
nuts. Journal of Food Lipids 5: 103-111.
Tesfaye, G., D. Teketay, M. Fetene & E. Beck. 2011.
Phenology of seven indigenous tree species in a dry
Afromontane forest, southern Ethiopia. Tropical
Ecology 52: 229-241.
Van Neer, W. & B. Clist. 1991. Le site de l’Age du Fer
Ancien d’Oveng (Province de l’Estuaire, Gabon),
analyse de la faune et de son importance pour la
problématique de l’expansion des locuteurs bantu en
Afrique Centrale. Académie des Sciences 32: 105-
110.
Villiers, J. F. 1973. Icacinacées, Olacacées, Pentadi-
plandracées, Opiliacées, Octoknémacées. pp. 144-
146. In: A. Aubréville & J. F. Leroy (eds.) La flore
du Gabon. n°20. Muséum National d’Histoire
Naturel, Paris.
Vivien, J. & J. J. Faure. 2011. Arbres des Forêts Denses
d’Afrique Centrale. Nguila-Kerou, Clohars Carnoet.
Walker, R. A. & R. Sillans. 1995. Les Plantes Utiles du
Gabon. Sépia, Libreville.
White, F. 1986. La Végétation de l’Afrique. Orstom-
Unesco, Paris.
White, L. J. T. 1994. Patterns of fruit-fall phenology in
the Lopé Reserve, Gabon. Journal of Tropical
Ecology 10: 289-312.
Zardo, R. N. & R. P. B. Henriques. 2011. Growth and
fruit production of the tree Caryocar brasiliense in
the Cerrado of central Brazil. Agroforestry Systems
82: 15-23.
(Received on 07.12.2012 and accepted after revisions, on 01.04.2013)
