Aquaculture 262 (2007) 65 – 72

www.elsevier.com/locate/aqua-online

Rearing of African catfish (Clarias gariepinus) and vundu catfish

(Heterobranchus longifilis) in traditional fish ponds (whedos): Effect
of stocking density on growth, production and body composition
Imorou Toko a,c , Emile D. Fiogbe a , Bruno Koukpode b , Patrick Kestemont c,⁎
a
Research Unit in Wet Land, URZH/FAST/UAC, 01BP526 Cotonou, Benin
b
Centre Régional pour la Promotion de l’Agriculture, CeRPA-Ouémé, Porto-Novo, Benin
c
Research Unit in Organismic Biology, URBO/FUNDP, 61 rue de Bruxelles, 5000 Namur, Belgium

Received 26 January 2006; received in revised form 24 August 2006; accepted 29 August 2006

Abstract

African catfish (Clarias gariepinus) (initial body weight: 34.8 ± 4.8 g) and vundu catfish (Heterobranchus longifilis) (initial
body weight: 39.1 ± 8.2 g) fingerlings were stocked at densities of 4, 6 or 8 fish m− 3 in traditional fish ponds (whedos) constructed
in the floodplain of the Oueme River (South Benin, West Africa), for 70 days from March to June 2005. Fish were fed twice a day
with 34% crude protein feed formulated with locally available ingredients. The effects of stocking density were evaluated in growth
responses, gross production and body composition. Water quality variables were similar ( p N 0.05) in all compartments.
Temperature and pH were at the optimum level for fish. Dissolved oxygen ranged from 0.9 to 1.2 mg l− 1 during the experiment and
secchi disc transparency was low (b14 cm). In both species, growth responses increased with the increasing density, significantly in
African catfish stocked at density of 8 fish m− 3 compared to the other densities (4 and 6 fish m− 3) but not significantly in vundu
catfish. Production data ranged from 3.1 ± 0.5 to 22.8 ± 4.5 t ha− 1 year− 1 in African catfish and from 6.1 ± 1.2 to 15.1 ± 3.1 t ha− 1
year− 1 in vundu catfish. Production increased with increasing stocking densities but only significantly ( p b 0.05) between the
density of 8 fish m− 3 and the other densities. In both species, carcass fat increased with increasing density ( p b 0.05) while carcass
protein and moisture decreased ( p N 0.05). These results are important because they indicate that, as far as growth rate and
production are concerned, African catfish is more profitable than vundu catfish for culture at high density in whedo.
© 2006 Elsevier B.V. All rights reserved.

Keywords: African catfish; Vundu catfish; Traditional fish pond; Whedo; Stocking density

1. Introduction (1100 ha according to Welcomme, 1971) which were

constructed during medieval times in the floodplains of
The extensive catfish culture in south-eastern Benin rivers and lakes and commonly known as ‘whedo’. This
(West Africa, Fig. 1) is generally based on many ponds practice improved with the passing years and especially
developed in the valley of the Oueme River (Fig. 1) where
one counts more than 1000 whedos, each longer than
⁎ Corresponding author. Tel.: +32 81724363; fax: +32 81724362. 120 m (Noutaï, 1987). Their primary objective was, and
E-mail address: Patrick.kestemont@fundp.ac.be (P. Kestemont). commonly still is, to store water during the dry season
0044-8486/$ - see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.aquaculture.2006.08.054
66 I. Toko et al. / Aquaculture 262 (2007) 65–72

Fig. 1. Map of south Benin showing the study area.

when the rivers are dropped out of the floodplains. These considered after wild fish are captured (from February
traditional ponds are also used to trap the wild fish when to June).
the water level drops. In spite of low oxygen concentra- The expansion of aquaculture, especially the culture of
tion, these whedo shelters many wild fish species such as African catfish (C. gariepinus) and vundu catfish
Clarias gariepinus (Burchell 1822), C. ebriensis (Pelle- (Heterobranchus longifilis, Valenciennes 1840) is due to
grin 1920), Protopterus annectens (Owen 1839), Para- their tolerance to a wide range of temperatures, as well as
channa obscura (Güther 1861), Polypterus senegalus to low oxygen and high salinity levels (Hecht et al., 1996;
(Cuvier 1829), etc. (Welcomme, 1971). Such fish farming Otémé et al., 1996). Moreover, they have high nutritive
is predominantly rural, and orientated principally to the value, good taste, and few bones. Because of their high
immediate needs of the farmers and their families. fecundity and growth rate, they become an important
According to Nonfon (1988), the output of a whedo is commercial species worldwide (Africa, Europe, China
higher than 1.57 t ha− 1 year− 1 and 58.3% of the fishermen and Brazil) (Hecht et al., 1996). A preliminary experiment
have 23% of their income who depend of their unique conducted in whedo showed the possibility to increase the
exploitation per year. The production of the whedos, catfish production by artificial feeding. However, stock-
averaging 1719 t in 1970 (Welcomme, 1971), decreased ing density is a significant factor in determining the fish
from year to year to 1500 t only in 1986 (Nonfon, 1988). production in these structures where oxygen level is
From December to June (dry season), the wild fish trap- already lower than 1 mg l− 1. Growth and survival of
ped in these whedos are captured using traditional fishing African and vundu catfish fry are known to be influenced
methods (cast net, hoop net, small seine net, basket, by stocking density (Hecht, 1982; Hecht and Appelbaum,
harpoon, etc.), often in February. They are then invaded 1988; Appelbaum and Van Damme, 1988; Haylor, 1991,
progressively by aquatic vegetation (Ipomea sp, Lemna 1992; Kaiser et al., 1995a; Baras et al., 1998; Ewa-Oboho
sp, Salvinia sp, Azolla spp, Echinochloa sp, Leersia sp, and Enyenihi, 1999). In some species, increasing stocking
etc.) until the next flood of the Oueme River, generally at density usually results in stress (aggressive behaviour,
the end of June. The management patterns of these fish dominance) which leads to enhanced energy requirements
production as part of an extensive fish farming system causing reduced growth, food utilization and net yield
have remained virtually unchanged in these traditional (Grant, 1997; Begout-Anras and Lagardère, 2004).
ponds. It is to improve this management that the However, catfishes showed some interesting deviation
development of semi-intensive catfish cultured is from this typical trend. Juvenile African catfish are
 I. Toko et al. / Aquaculture 262 (2007) 65–72 67

normally highly aggressive when confined in small 8.2 g, n = 140) fingerlings were obtained from male and
numbers in a large volume of water. Aggressiveness, female brood fish reared in the pond of the Research
territorial defense and development of hierarchies and Unit of Wet Land, University of Abomey-Calavi, Benin.
individual dominance are often reduced at high stocking Breeding was carried out using Ovaprim® (0.5 ml/kg
density (Hecht and Appelbaum, 1988; Kaiser et al., body weight) as inducing agent, following the proce-
1995b; Hecht and Uys, 1997) and fish may start to be dures of Viveen et al. (1985). Larvae were fed Artemia
stressed once certain threshold densities are attained nauplii (INVE Aquaculture, Belgium) for 8 days and
(Hecht and Uys, 1997). Indeed, aggression can result in then weaned over an 8-day period by substituting Ar-
stock losses, reduced food conversion efficiency and temia with a commercial Catfish diet (CatCo Crumble
slower growth. So, there must be a density from which Excellent, 0.3–0.5 mm, Coppens-International, Hel-
both growth and production decreased with increasing mond, Netherlands). After weaning, fish were trans-
stocking density. Evaluation of the critical stocking ferred into a circular concrete tank for 60 days. During
density in whedo may provide commercially important the first 2 weeks the fish were fed CatCo Crumble
information in designing an efficient culture system for Excellent (0.5–0.8 mm, Coppens-International, Hel-
catfish. mond, Netherlands). They were then fed on the CatCo
In this study, the effects of stocking density were Grower-12 EF (2 mm, Coppens-International, Helmond,
assessed to determine the relationships between stocking Netherlands) for 56 days. Fingerlings were kept in a
density and production in African and vundu catfishes concrete pond for an adaptation period of 3 months
cultured in whedo. Other related objectives were to before being allotted to the experiment.
determine the effect of stocking density on growth, food
utilization, body composition and finally the feasibility 2.3. Food and feeding
of this clariid catfish as a species suitable for raising in
whedo. The 34% crude protein feed (ADCP, 1983) was
formulated using locally available ingredients (Table 1).
2. Material and methods Feed was prepared by mixing the dry ingredients with
the addition of oil and water until a desirable paste-like
2.1. Location and experimental design consistency was reached. This paste was sun-dried at
about 30–40 °C and preserved in airtight containers.
The study was carried out from March to June 2005 Finally, the feed was manually crumbled to pellet size
(70 days) in three whedos located at Gangban village (2 cm) before served. Fish were hand fed daily at 5% of
(06°39′48″N 02°27′66″E), 108 km northeast from body weight, divided into two meals delivered at 8:00 a.
Cotonou (Fig. 1). m and 5:00 p.m. After 2 weeks, the ration was adjusted
In each whedo longer than 100 m six compartments according to Hogendoorn and Koops (1983) after esti-
of 10 m2 were made. The level of water ranged from mation of body weight using a growth curve reported by
0.6 m at the onset of experiment to 0.4 m only at the end Legendre et al. (1992) for African and vundu catfishes.
due to evaporation and percolation. Barriers between
compartments were achieved locally with bamboo
Table 1
sticks. Three treatments using three stocking densities Formulation, proximate composition and cost of experimental feed
have been tested: 4 fish m− 3, 6 fish m− 3 and 8 fish m− 3.
Ingredients a (g kg− 1) Protein level (%)
These densities, although lower than those usually used
for African and vundu catfishes in intensive culture Fish meal 480 51
Maize meal 200 8.5
system, were chosen with regard to water quality and the
Cottonseed cake 250 37
fact that whedos are used by limited resource populace. Palm oil 60 0
Each treatment was applied in triplicate. To prevent NaCl 10 0
warm water contamination due to density treatment, the
replicates of the same density were disposed by species, Proximate composition (%)
Crude protein 33.9 ± 0.7
in three successive compartments of the same whedo.
Crude lipid 16.3 ± 0.4
Ash 7.0 ± 0.1
2.2. Fish Cost of feed per kg (Fcfa) b 213
a
Locally made, supplied by Ovograin Feeds Depot, Abomey-
African catfish (initial body weight: 34.8 ± 4.8 g, Calavi, Benin.
b
n = 140) and vundu catfish (initial body weight: 39.1 ± 1 Euro = 655.957 Fcfa.
68 I. Toko et al. / Aquaculture 262 (2007) 65–72

2.4. Parameters measured and chemical analyses data were normalized by arc-sine transformation prior
analyses.
Water quality parameters such as dissolved oxygen,
temperature, pH and transparency (using temperature 3. Results
and dissolved oxygen meter WTW Oxi 197; pHmeter
WTW pH324 ; secchi disk, respectively) were monitored 3.1. Survival, growth, condition and production
in each compartment throughout the experimental
period, twice a month at 10:30–11:00 a.m. Water Survival of fish under experimental conditions
samples (1.5 l) were also collected at the same time in ranged from 96 to 100% without any significant
each compartment for nitrite nitrogen (NO2−, colorimet- difference between treatments ( p N 0.05). Growth per-
ric analysis using NitriVer® 3 Nitrite reagent) and formances of African and vundu catfishes differed
chlorophyll a for phytoplankton abundance (Pechar, significantly ( p b 0.05) in terms of final mean weight
1987). Each parameter was monitored at a depth of 10– and growth rate (Table 2). Overall, growth responses in
20 cm. The water quality parameters ranged from 28.5 to both species increased with increasing density, but not
29.8 °C, 0.9 to 1.2 mg l− 1, 6.2 to 6.5, 12.4 to 14.0 cm, all differences were significant. In African catfish, there
0.03 to 0.15 mg l− 1 and 0.5 to 1.3 μg l− 1, for tem- were significant differences ( p b 0.05) in Wt and SGR
perature, dissolved oxygen, pH, secchi disk transparen- between a density of 8 fish m− 3 and the other two lower
cy, nitrite and chlorophyll a, respectively. There was no densities (4 fish m− 3 and 6 fish m− 3) which were similar
significant difference ( p N 0.05) of the water quality while in vundu, growth was not significantly affected
variables among compartments and treatments. ( p N 0.05) by stocking density between 4 and 8 fish m− 3,
After 70 days, the total number and biomass (kg) of although the higher densities supported the best growth.
survivors in each compartment were recorded. Six fish In both species, production values differed with stock-
randomly taken from each compartment (n = 18 fish/ ing density. Total production increased with increasing
treatment) were stored in polyethylene bags, and frozen densities (Table 2) but differences were significant
(− 20 °C) for body analysis. Fish samples were analysed ( p b 0.05) only between the density of 8 fish m− 3 and
by standard methods for moisture (oven drying at the others. The CV weight ranged between 19 and
105 °C for 24 h), crude protein (N-Kjeldahl x6.25) and
ash (oven incineration at 550 °C). Carcass total lipids
were extracted according to Bligh and Dyer (1959). Table 2
Growth (final mean body weight: Wt, specific growth Survival, final weight, specific growth rate (SGR), production data,
coefficient of variation of final weight (CV) and condition factor (K ) of
rate: SGR) and apparent feed conversion rate (AFCR) were African and vundu catfishes reared at different socking densities in
calculated as follows: Wt (g) =Bt N− 1, where Bt = final fish whedo
biomass (g) per compartment and N = number of survivors; Parameters Stocking density
SGR (% day− 1) =100 (LnWt − LnW0) Δt− 1, where W0 = −3
Fish m 4 6 8
initial mean body weight (g), Δt=time (days); AFCR =
(weight of feed used) (fish weight gain)− 1. Survival (%)
African catfish 100 (0.0) 100 (0.0) 100 (0.0)
The coefficient of variation of final weight (CV =
Vundu catfish 100 (0.0) 96.7 (5.8) 98.3 (2.9)
100 SD mean− 1) and the condition factor (K = 100 Wt Final weight (g)
L− 3, where Wt is in g and L: fish longer in cm) were also African catfish 61.7 (4.0)a 82.9 (11.9)a 149.8 (24.2)b
calculated. Vundu catfish 89.9 (14.0) 93.2 (13.9) 121.3 (14.3)
Production (P) was calculated using extrapolated SGR (% day− 1)
African catfish 0.82 (0.09)a 1.23 (0.21)b 2.07 (0.24)c
value of final biomass per hectare (ha): P (t ha− 1
Vundu catfish 1.18 (0.22) 1.23 (0.21) 1.61 (0.18)
year− 1) = (Bt − Bo) Δt− 1 S− 1, where Bt and Bo = final Production (t ha− 1 year− 1)
and initial biomass (ton, t) respectively, Δt = time (year) African catfish 3.1 (0.5)a 7.9 (2.0)a 22.8 (4.5)b
and S = area of production (ha). Vundu catfish 6.1 (1.2)a 8.8 (2.6)a 15.1 (3.1)b
CV weight (%)
African catfish 19.0 (7.2) 27.0 (7.7) 55.6 (20.6)
2.5. Statistical analysis
Vundu catfish 56.3 (8.6) 40.0 (3.4) 46.1 (6.0)
Condition factor (K )
The effects of density were compared using one-way African catfish 0.9 (0.1)a 0.8 (0.1)a 1.2 (0.1)b
analysis of variance (ANOVA). Significant differences Vundu catfish 1.0 (0.2) 0.9 (0.1) 1.1 (0.1)
between treatments means ( p b 0.05) were determined Data are means (±SD) of three replicates. In each line, means with
using a Fisher's PLSD test (Saville, 1990). Percentage different letters as superscripts are significantly different ( p b 0.05).
 I. Toko et al. / Aquaculture 262 (2007) 65–72 69

Table 3 (1.7 ± 0.3) but differences with other densities were not
Feed utilization and body composition data for African and vundu significant. In both species, proximate carcass fat
catfishes reared at different stocking densities in whedo
increased with increasing density ( p b 0.05) while carcass
Parameters African catfish Vundu catfish protein level and moisture decreased but not significantly
−3
Fish m 4 6 8 4 6 8 ( p N 0.05) (Table 3).
AFCR 3.4 2.2 1.1 2.1 2.2 1.7
(0.7)a (0.6)b (0.2)c (0.5) (0.6) (0.3) 3.3. Economics
I. protein 12.2 12.2 12.2 12.9 12.9 12.9
F. protein 14.7 14.2 13.6 15.2 14.9 14.5
Net profits were directly related to stocking density
(0.5) (1.1) (0.2) (1.2) (0.5) (0.2)
Crude fat 6.8 7.4 10.5 4.5 5.5 7.2 in both species. So, the highest density provided the
(0.2)a (0.1)a (0.4)b (0.2)a (0.1)a (0.2)b highest profit (Table 4).
Moisture 72.5 71.9 69.8 75.4 74.2 72.7
(1.2) (1.0) (1.1) (0.5) (1.3) (0.3) 4. Discussion
Ash 4.2 4.0 4.3 3.3 3.5 3.4
(0.1) (1.2) (0.1) (0.0) (0.1) (0.1)
The interpretation of the effects of stocking density
Data are means (±SD) of three replicates. In each line and by species,
on fish growth and body composition is complex as the
means with different letters as superscripts are significantly different
( p b 0.05). I. protein=initial protein and F. protein=final protein.
results are affected by many interrelated factors such as
water quality, composition of feed and size of the ration
(Hepher, 1988). In our study, water quality parameters
55.6% in African catfish and between 40 and 56.3% in were found to be similar in all compartments. The same
vundu and did not vary significantly ( p N 0.05) between feed was used and the ration was fixed similarly in all
stocking densities. Stocking density did not affect the compartments. Results from this study showed that both
condition factor of vundu while K of African catfish growth rate and production of African catfish (2.07%
stocked at 8 fish m− 3 was significantly ( p b 0.05) higher day− 1 and 22.8 t ha− 1 year− 1, respectively) and vundu
than those of fish stocked at lower density. catfish (1.61 % day− 1 and 15.1 t ha− 1 year− 1, respec-
tively) were found better over stocking at density of
3.2. Food utilization and body composition 8 fish m− 3. With both species, apparent food conversion
rate (AFCR) was lower at high (1.1 and 1.7 with African
In African catfish, AFCR was affected by stocking and vundu catfishes, respectively) than at low stocking
density, decreasing significantly ( p b 0.05) with increas- density (3.4 and 2.1 with African and vundu catfishes,
ing stocking density (Table 3). In vundu, the best AFCR respectively). The highest body fat content was also
was achieved with the stocking density of 8 fish m− 3 observed at density of 8 fish m− 3 while protein and

Table 4
Benefit–cost analysis for African and vundu catfishes reared at different stocking densities in whedo (per year)
Components African catfish Vundu catfish
−3
Fish m 4 6 8 4 6 8
Expenditure (× 1000 Fcfa a, per ha)
Whedo preparation 15 15 15 15 15 15
Fingerlings cost b 1500 2250 3000 2000 3000 4000
Feed cost c 434.3 656.9 1004.5 432.6 712.9 1133.6
Maintenance and harvest 30 30 30 30 30 30
Total expenditure 1979.3 2951.9 4049.5 2477.6 3757.9 5178.6

Income (per ha)

Final biomass (kg) 1235 2488 5990 1798 2795 4853
Total income d (× 1000 Fcfa) 1482 2985.6 7188 2697 4192.5 7279.5
Net benefit (Fcfa, per ha) − 497.3 33.7 3138.5 219.4 434.6 2100.9
Benefit–cost ratio (BCR) − 0.25 0.01 0.78 0.09 0.12 0.41
a
1 Euro = 655.957 Fcfa.
b
75 Fcfa per African catfish and 100 Fcfa per vundu catfish.
c
1 kg feed = 213 Fcfa.
d
1 kg African catfish = 1200 Fcfa and 1 kg vundu catfish = 1500 Fcfa.
70 I. Toko et al. / Aquaculture 262 (2007) 65–72

moisture did not show any significant difference among breathers, have a relatively high tolerance to poor water
treatments. quality conditions and display unusual behavioural
The growth performances obtained for African responses under high density conditions.
catfish and vundu in the present study figure well in Indeed, Hecht and Uys (1997) found that the level of
the range of those reported in literature (Hogendoorn aggression decreased dramatically with increasing den-
and Koops, 1983; Viveen et al., 1985; Legendre et al., sity. This is in accordance with Kaiser et al. (1995b)
1992; de Graaf et al., 1996). Under field conditions, reporting that agonistic behaviour in African catfish was
Hogendoorn and Koops (1983) found that African highest at low stocking densities. In vundu catfish,
catfish reached 300 g in 22 weeks. During the same Kerdchuen and Legendre (1992) also reported a de-
period, but in fertilized ponds and without supplemental creasing level of aggression with increasing density.
food, Bok and Jongbloed (1984) obtained only 135 g for Aggressiveness, territorial defense and development of
African catfish. de Graaf et al. (1996) reported 200– hierarchies and individual dominance in fish cost a lot of
300 g for African catfish after 6 months in polyculture energy which otherwise would be used for growth
with Oreochromis niloticus in pond. Viveen et al. (Hecht and Uys, 1997).
(1985) in monoculture rearing reported lower final body Although vundu catfish are known to grow faster
weight (62 g) of African catfish in 13 weeks. than African catfish as reported by previous studies
Growth and production of African catfish in this study (Legendre et al., 1992; Hecht et al., 1996; Otémé et al.,
was directly density dependent, i.e. fish grown at the 1996), our results showed a deviation from this trend
highest density achieved the highest final mean weight with increasing stocking density. Does this mean that
and production value. This is in agreement with the vundu is less efficient than African catfish under high
results reported for African catfish raised in ponds stocking density? Our results and those reported by
(Hogendoorn and Koops, 1983), in cages (Hengsawat Kerdchuen and Legendre (1992) can support this hypo-
et al., 1997) or in tanks (Machiels and Van Dam, 1987; thesis. Vundu catfish is probably less adapted to tra-
Almazán-Rueda, 2004). Hogendoorn and Koops (1983) ditional whedo environment than African catfish.
found that the highest biomass was achieved at the Despite the fact that vundu is present in the Oueme
highest stocking density for African catfish cultured in River, this species had never been caught in whedos
pond. Hengsawat et al. (1997) reported increasing daily (Welcomme, 1971).
instantaneous growth rate and total production of In the present study, AFCR was significantly influ-
juvenile African catfish (initial mean weight 32 g) reared enced by increasing stocking density of African catfish.
when densities increased from 33 to 133 fish m− 3. This was in agreement with other reports in African
Similarly, Almazán-Rueda (2004) found that juvenile catfish as well as in other species (Cruz and Ridha, 1989;
African catfish kept at high stocking density (2500 fish Watanabe et al., 1990; Almazán-Rueda, 2004). The re-
m− 3 vs. 250) were heavier and had a higher SGR than duction of food conversion rate with increasing density
those kept at low stocking density. In vundu catfish, no can be attributed to the feeding behaviour of catfish
significant density treatment effect on growth was (Hecht and Uys, 1997; Almazán-Rueda, 2004). Hecht
observed although growth rate increased with increasing et al. (1997) reported that at high density, African catfish
stocking density. A similar scenario was observed by react faster to the presence of food and consume a meal
Kerdchuen and Legendre (1992) in vundu catfish (initial much faster than at low density. Hecht and Uys (1997)
mean weight 110 g) reared in tank under densities also found that the time it took for fish to respond to food
ranging from 56 to 555 fish m− 3. The relationships and to consume a meal decreased with increasing density.
between stocking density and growth and production In earthen pond such as whedo, food conversion may be
obtained in the present study with both species agree influenced by the amount of natural food. However, at
with many reports on other fish species such as sutchi high stocking density, natural production, although not
catfish (Pangasius sutchi) in cage (Rahman et al., 2006), analysed in this study, was probably low because of high
channel catfish (Ictalurus punctatus) raised in raceway grazing, and fish were more closely dependent on the
(Woiwode and Adelman, 1989) and Nile tilapia (O. formulated diet than at low density. Formulated food is
nioticus) in cages (Daungsawasdi et al., 1986) or in open more energetic than plankton and macroinvertebrates,
pond (Yi et al., 1996). The underlying mechanisms which were probably more consumed by fish at low
causing increasing growth and production with increas- density. In this case, ammonia from uneaten feed which is
ing density are still poorly understood. However, not checked during the present study, could also explain
according to Hecht et al. (1997), African catfish can be the high AFCR observed with slower growing fish which
grown at very high densities because they are air- have the potential for excess feed.
 I. Toko et al. / Aquaculture 262 (2007) 65–72 71

Hitherto, meagre information is available on body ment — Interuniversity council of French Speaking Uni-
composition of African and vundu catfishes according to versities of Belgium) and the FUNDP (Namur
stocking conditions. Data from the present study showed University). We wish to thank other colleagues, research
that increasing stocking density influenced body com- assistants and fishermen from Gangban village who
position resulting in increased fat level and decreased provided help for successfully conducting the experiment.
protein and moisture levels at high stocking density. It
cannot be concluded whether this body composition is
References
directly influenced by stocking conditions, or indirectly
due to differences in feeding (higher part of natural food
ADCP: Aquaculture Development and Coordination Programme,
at low density), growth rate or fish size at the harvest. In 1983. Fish feeds and feeding in developing countries. An interim
channel catfish, Robinson et al. (2003) reported that fast- report on the ADCP feed development programme. FAO-ADCP/
growing fish are generally high in body fat and low in REP/83 Rome, Italy, pp. 18–97.
moisture and protein concentrations compared with Adelman, I.R., 1977. Effects of bovine growth hormone on growth of
slow-growing fish. Percent moisture and percent fat of carp (Cyprinus carpio) and the influences of temperature and
photoperiod. J. Fish. Res. Board Can. 34, 509–515.
whole body are generally negatively associated, as Adelman, I.R., 1978. Influence of temperature on growth promotion
shown by Wilson et al. (1988) in Channel catfish. In and body composition of carp (Cyprinus carpio) due to bovine
other teleosts, a weight increase was also accompanied growth hormone. Trans. Am. Fish. Soc. 107, 747–750.
by body fatness increases as well (Cyprinus carpio, Almazán-Rueda, P., 2004. Towards assessment of welfare in African
Adelman, 1977, 1978; Ictalurus melas, Kayes, 1977). catfish, Clarias gariepinus: the first step. PhD thesis, Wageningen
University, The Netherlands. 151pp.
Although all fishes harvested after 70 days rearing did Appelbaum, S., Van Damme, P., 1988. The feasibility of using
not reach their optimum size with regard to market exclusively artificial dry feed for the rearing of Israeli Clarias
(N250 g), production economics revealed that high gariepinus (Burchell 1822) larvae and fry. J. Appl. Ichthyol. 4,
stocking density (8 fish m− 3) gave the highest net benefit 105–110.
Baras, E., Tissier, F., Westerloppe, L., Mélard, C., Philippart, J.C.,
and best benefit–cost ratio (BCR). Moreover, at high
1998. Feeding in darkness alleviates density-dependent growth of
densities raising of African catfish is more profitable juvenile vundu catfish Heterobranchus longifilis (Clariidae).
than vundu in whedo. Aquat. Living Resour. 11 (5), 335–340.
Begout-Anras, M.L., Lagardère, J.P., 2004. Domestication et compor-
5. Conclusion tement chez les poisons téléostéens. INRA Prod. Anim. 17 (3),
211–215.
Bligh, E.G., Dyer, W.J., 1959. A rapid method for total lipid extraction
This study has shown that the two African catfishes and purification. Can. J. Biochem. Physiol. 37, 911–917.
(C. gariepinus and H. longifilis) can be efficiently grown Bok, A.H., Jongbloed, H., 1984. Growth and production of sharptooth
in whedo and can generate 3–5 t per hectare in only catfish, Clarias gariepinus (Pisces: Clariidae) in organically
2.5 months. So, raising in whedo is a viable alternative to fertilized ponds in the Cape Province, South Africa. Aquaculture
increase catfishes production in Benin. On the other 36, 141–155.
Cruz, E.M., Ridha, M., 1989. Preliminary study on the production of
hand, this study shows that African catfish are more able tilapia, Oreochromis spilurus (Günther), cultured in seawater
to support high stocking densities in whedo than vundu. cages. Aquac. Fish. Manage. 20, 381–388.
While growth and production values increased with Daungsawasdi, S., Chomchei, C., Yamorbsin, R., Kertkomut, B.,
increasing stocking density, there must be a density from 1986. Net Cage Culture of Tilapia and Puntius in Klong Praew
Irrigation Tank. Tech. Pap. NIFI, No. 64, Bangkok, Thailand.
which mortality increased with simultaneous reduction
10 pp.
in growth and production. For African catfish this critical de Graaf, G.J., Galemoni, F., Banzoussi, B., 1996. Recruitment control
density was not reached in this study even at the highest of Nile tilapia, Oreochromis niloticus, by the African catfish,
stocking density of 8 fish m− 3 while vundu catfish is Clarias gariepinus (Burchell 1822) and, the African snakehead,
probably less suitable for rearing over 8 fish m− 3 in Ophiocephalus obscuris. I. A biological analysis. Aquaculture
whedo. Further research of optimal stocking density is 146, 85–100.
Ewa-Oboho, I.O., Enyenihi, U.K., 1999. Aquaculture implications of
needed since it may make the culture of these species in growth variation in the African catfish: Heterobranchus longifilis
whedo more attractive to fish farmers in Benin. (Val.) reared under controlled conditions. J. Appl. Ichthyol. 15,
111–115.
Acknowledgements Grant, J.W.A., 1997. Territoriality. In: Godin, J.G.J. (Ed.), Beha-
vioural Ecology of Teleost Fishes. Oxford Univ. Press, Oxford,
pp. 81–103.
This research was funded by the CBRST (Centre Haylor, G.S., 1991. Controlled hatchery production of Clarias
Béninois de Recherches Scientifiques et Techniques), the gariepinus (Burchell 1822): growth and survival of fry at high
CUD-CIUF (University Co-operation towards Develop- stocking density. Aquacult. Fish. Manage. 22, 405–442.
72 I. Toko et al. / Aquaculture 262 (2007) 65–72

Haylor, G.S., 1992. Controlled hatchery production of Clarias Machiels, M.A.M., Van Dam, A., 1987. A dynamic simulation model
gariepinus (Burchell): growth and survival of larvae at high for growth of tha African catfish, Clarias gariepinus (Burchell
stocking density. Aquac. Fish. Manage. 23, 303–314. 1822). III. Effect of body composition on growth and feed
Hecht, T., Appelbaum, S., 1988. Observations on intraspecific utilization. Aquaculture 60, 55–71.
aggression and coeval sibling cannibalism by larval and juvenile Nonfon, M., 1988. Données préliminaires sur l'écologie et la
Clarias gariepinus (Clariidae; Pisces) under controlled conditions. production halieutique des « trous à poissons » de la rive gauche
J. Zool., Lond. 214, 21–44. de la basse vallée de l'Ouémé. Mémoire d'Ingénieur Agronome,
Hecht, T., Oellermann, L., Verheust, L., 1996. Perpectives on clariid FSA/UNB. 152 pp.
catfish culture in Africa. Aquat. Living Resour. 9, 197–206 (Hors Noutaï, J., 1987. Etude sur les possibilités d’amélioration des « trous à
série). poissons » dans la vallée de l'Ouémé. Mémoire de DEAT, CPA2,
Hecht, T., 1982. Intensive rearing of Clarias lazera larvae (Claridae: Sékou. 52 pp.
Pisces). S. Afr. J. Wildl. Res. 12, 101–105. Otémé, Z.L., Hem, S., Legendre, M., 1996. Nouvelles espèces de
Hecht, T., Uys, W., 1997. Effect of density on the feeding and poissons-chats pour le développement de la pisciculture africaine.
aggressive behaviour in juvenile African catfish, Clarias gariepi- Aquat. Living Resour. 9, 207–217 (hors série).
nus. S. Afr. J. Sci. 93, 537–541. Pechar, L., 1987. Use of an acetone:methanol mixture for the
Hecht, T., Sorgeloos, P., Verreth, J., Ollevier, F., 1997. Guide to the extraction and spectrophotometric determination of chlorophyll a
biology and culture of the African catfish. Website project funded in phytoplankton. Arch. Hydrobiol., Suppl. 78, 99–117.
by the Flemish Ministry of Education, the University of Ghent, Rahman, M.M., Islam, M.S., Halder, G.C., Tanaka, M., 2006. Cage
Rhodes University and Liberty Life Educational Foundation culture of sutchi catfish, Pangasius sutchi (Fowler 1937): effects
Johannesburg. http//cdserver2.ru.ac.za/cd/catfish. of stocking density on growth, survival, yield and farm profit-
Hengsawat, K., Ward, F.J., Jaruratjamorn, P., 1997. The effect of ability. Aquac. Res. 37, 33–39.
stocking density on yield, growth and mortality of African catfish Robinson, E.H., Li, M.H., Manning, B.B., 2003. How a catfish reaches
(Clarias gariepinus Burchell 1822) cultured in cages. Aquaculture commercial weight is key to composition. Catfish J. 18 (3), 12.
152, 67–76. Saville, D.J., 1990. Multiple comparison procedures: the practical
Hepher, B., 1988. Nutrition of pond fishes. Cambridge Univ. Press, solution. Am. Stat. 44 (2), 174–180.
New York, USA. 388 pp. Viveen, W.J.A.R., Richter, C.J.J., van Oordt, P.G.W.J., Janssen, J.A.L.,
Hogendoorn, H., Koops, W.J., 1983. Growth and production of Huisman, E.A., 1985. Practical manual for the culture of the
African catfish, Clarias lazera (C. and V.). I. Effects of stocking African catfish (Clarias gariepinus). The Netherlands Ministry for
density, pond size and mixed culture with tilapia (S. niloticus L.) Development Cooperation, Section for Research and Technology.
under extensive field conditions. Aquaculture 34, 253–263. 128 pp.
Kaiser, H., Weyl, O., Hecht, T., 1995a. Observations on agonistic Watanabe, W.O., Clark, J.H., Dunham, J.B., Wicklund, R.I., Olla, B.I.,
behaviour of Clarias gariepinus larvae and juveniles under 1990. Production of fingerling Florida red tilapia (Tilapia
different densities and feeding frequencies in a controlled hornorum × T. mossambica) in floating marine cages. Prog.
environment. J. Appl. Ichthyol. 11, 25–36. Fish-Cult. 52, 158–161.
Kaiser, H., Weyl, O., Hecht, T., 1995b. The effect of stocking density Welcomme, R.L., 1971. Evaluation de la pêche intérieur, son état
on growth, survival and agonistic behaviour of African catfish. actuel et ses possibilités. Rome, FAO AT 2938. 95 pp.
Aquac. Int. 3, 217–225. Wilson, R.P., Poe, W.E., Nemetz, T.G., MacMillan, J.R., 1988. Effect
Kayes, T., 1977. Effects of temperature on hypophyseal (growth of recombinant bovine growth hormone administration on growth
hormone) regulation of length, weight, and allometric growth and and body composition of channel catfish. Aquaculture 73,
total lipid and water concentrations in the black bullhead (Ictalurus 229–236.
melas). Gen. Comp. Endocrinol. 33, 382–393. Woiwode, J.G., Adelman, I.R., 1989. Influence of density and
Kerdchuen, N., Legendre, M., 1992. Effect favorable des fortes multipass water use on channel catfish performance in raceways.
densités pour l'adaptation d'un silure africain, Heterobranchus Prog. Fish-Cult. 51, 183–188.
longifilis (Pisces, Clariidae) en bacs de petit volume. Rev. Yi, Y., Kwei Lin, C., Diana, J.S., 1996. Influence of Nile tilapia
Hydrobiol. Trop. 25, 63–67. (Oreochromis niloticus) stocking density in cages on their growth
Legendre, M., Teugels, G.G., Cauty, C., Jalabert, B., 1992. A and yield in cages and in ponds containing the cages. Aquaculture
comparative study on morphology, growth rate and reproduction 146, 205–215.
of Clarias gariepinus, Heterobranchus longifilis and their
reciprocal hybrids (Pisces, Clariidae). J. Fish Biol. 40, 59–79.