Nile × blue tilapia hybrid (Oreochromis niloticus × O. aureus) has become an important food fish in intensive freshwater aquaculture. Recently, the parasite Myxobolus bejeranoi (Cnidaria: Myxozoa) was found to infect hybrid tilapia gills at high prevalence, causing immune suppression and high mortality. Here, we explored additional characteristics of M. bejeranoi–tilapia interaction, which enable efficient proliferation of this parasite inside its specific host. Highly sensitive quantitative polymerase chain reaction (qPCR) and in situ hybridization analyses of fry collected from fertilization ponds provided evidence to an early-life infection of fish by a myxozoan parasite, occurring less than 3 weeks post-fertilization. Because Myxobolus species are highly host-specific, we next compared infection rates in hybrid tilapia and in both its parental species following a 1-week exposure to infectious pond water. Analysis by qPCR and histological sections showed that while blue tilapia was as susceptible to M. bejeranoi as the hybrid, Nile tilapia appeared to be resistant. This is the first report of differential susceptibility of a hybrid fish vs its parental purebreds to a myxozoan parasite. These findings advance our understanding of the relationship between M. bejeranoi and tilapia fish and raise important questions regarding the mechanisms that allow the parasite to distinguish between very closely related species and to infect a specific organ at very early-life stages.

This study aimed to determine the effects of supplementing the diet of adult Nile tilapia Oreochromis niloticus with phosphatidylcholine (PC) on growth performance, body composition, fatty acid composition and gene expression. Genetically Improved Farmed Tilapia fish with an initial body weight of 83·1 (sd 2·9) g were divided into six groups. Each group was hand-fed a semi-purified diet containing 1·7 (control diet), 4·0, 6·5, 11·5, 21·3 or 41·0 g PC/kg diet for 68 d. Supplemental PC improved the feed efficiency rate, which was highest in the 11·5 g PC/kg diet. Weight gain and specific growth rate were unaffected. Dietary PC increased PC content in the liver and decreased crude fat content in the liver, viscera and body. SFA and MUFA increased and PUFA decreased in muscle with increasing dietary PC. Cytoplasmic phospholipase A2 and secreted phospholipase A2 mRNA expression were up-regulated in the brain and heart in PC-supplemented fish. PC reduced fatty acid synthase mRNA expression in the liver and visceral tissue but increased expression in muscle. Hormone-sensitive lipase and lipoprotein lipase expression increased in the liver with increasing dietary PC. Growth hormone mRNA expression was reduced in the brain and insulin-like growth factor-1 mRNA expression in liver reduced with PC above 6·5 g/kg. Our results demonstrate that dietary supplementation with PC improves feed efficiency and reduces liver fat in adult Nile tilapia, without increasing weight gain, representing a novel dietary approach to reduce feed requirements and improve the health of Nile tilapia.

The present study evaluated the effects of dietary microbial phytase on the growth and gut health of hybrid tilapia (Oreochromis niloticus ♀×Oreochromis aureus ♂), focusing on the effect on intestinal histology, adhesive microbiota and expression of immune-related cytokine genes. Tilapia were fed either control diet or diet supplemented with microbial phytase (1000 U/kg). Each diet was randomly assigned to four groups of fish reared in cages (3×3×2 m). After 12 weeks of feeding, weight gain and feed conversion ratio of tilapia were not significantly improved by dietary microbial phytase supplementation. However, significantly higher level of P content in the scales, tighter and more regular intestinal mucosa folds were observed in the microbial phytase group and the microvilli density was significantly increased. The adhesive gut bacterial communities were strikingly altered by microbial phytase supplementation (0·41<similarity coefficient<0·54). Stimulated intestinal inflammation and stress status were observed in the fish fed diet supplemented with microbial phytase, as indicated by the up-regulated intestinal expressions of the cytokine genes (tnf-α and tgf-β) and hsp70. In addition, the gut microvilli height was significantly decreased in the phytase group. These results indicate that dietary microbial phytase may exert mixed effects on hybrid tilapia, and can guide our future selection of phytases as aquafeed additives – that is, eliminating those that can stimulate intestinal inflammation.

Integrated Multi-Trophic Aquaculture takes advantage of the mutualism between some detritivorous fish and phytoplankton. The fish recycle nutrients by consuming live (and dead) algae and provide the inorganic carbon to fuel the growth of live algae. In the meanwhile, algae purify the water and generate the oxygen required by fishes. Such mechanism stabilizes the functioning of an artificially recycling ecosystem, as exemplified by combining the euryhaline tilapia Sarotherodon melanotheron heudelotii and the unicellular alga Chlorella sp. Feed addition in this ecosystem results in faster fish growth but also in an increase in phytoplankton biomass, which must be limited. In the prototype described here, the algal population control is exerted by herbivorous zooplankton growing in a separate pond connected in parallel to the fish–algae ecosystem. The zooplankton production is then consumed by tilapia, particularly by the fry and juveniles, when water is returned to the main circuit. Chlorella sp. and Brachionus plicatilis are two planktonic species that have spontaneously colonized the brackish water of the prototype, which was set-up in Senegal along the Atlantic Ocean shoreline. In our system, water was entirely recycled and only evaporation was compensated (1.5% volume/day). Sediment, which accumulated in the zooplankton pond, was the only trophic cul-de-sac. The system was temporarily destabilized following an accidental rotifer invasion in the main circuit. This caused Chlorella disappearance and replacement by opportunist algae, not consumed by Brachionus. Following the entire consumption of the Brachionus population by tilapias, Chlorella predominated again. Our artificial ecosystem combining S. m. heudelotii, Chlorella and B. plicatilis thus appeared to be resilient. This farming system was operated over one year with a fish productivity of 1.85 kg/m2 per year during the cold season (January to April).

Tilapia species are able to hybridize in captivity or in natural conditions following introduction of alien species. Here, we report a case of hybridization between two (and perhaps three) autochthonous species, Tilapia zillii and T. guineensis, resulting from a modification of their environment due to human activities. The man-made dam Ayamé (Côte d'Ivoire) transformed the River Bia into a lake (Lake Ayamé, area: 197 km2). In this new environment, both tilapia species have hybridized. In order to study hybridization along the Bia System, 25 allozyme loci have been screened in 470 specimens from ten locations. Two loci LDH-1 and PGI-2 were diagnostic between the two different species and can be used to identify hybrids in each sampling location of the River Bia basin. The percentage of pure T. zillii individuals in Lake Ayamé varied from 18.1% in Ebikro to 4.2% in Kétesso while percentage of pure T. guineensis varied from 81.3% in Yaou to 50.0% in Ebikro. The proportions of hybrids varied from 8.3% in Yaou to 31.9% in Ebikro. The existence of private alleles observed in tilapia from Lake Ayamé compared to what has been observed in both reference populations, T. zillii and T. guineensis led to the hypothesis that a third species could have been engaged in this hybridization. This third species is likely to be T. busumana, formerly present in the lake but now rarely observed. If this natural hybridization is due to the dam construction and the correlated ecological changes, then other hybridizations could result since many dams have been constructed within the range of these species in Côte d'Ivoire.

Two 8-week feeding trials were conducted to determine the dietary Na requirement for juvenile hybrid tilapia (Oreochromis niloticus × O. aureus) reared in fresh water and seawater. In each experiment, NaCl was added to the basal diet at 0, 0·5, 1, 2, 3, 5, or 7g Na/kg diet (fresh water) and at 0, 0·2, 0·5, 0·8, 1·2, 1·5, 2, or 3g Na/kg diet (seawater). Each diet was fed to three replicate groups of fish, individual fish initially weighing 0·69 (se 0·01) g, in a closed, recirculating rearing system. In fresh water, the tilapia fed the diet supplemented with 2g Na/kg diet had significantly (P<0·05) greater weight gain than the fish fed the diets supplemented with ≥3 and ≤0·5g Na/kg diet. Feed efficiency (FE) in fish generally followed the weight-gain pattern. Gill Na+–K+ ATPase activity was highest in the fish fed the diets supplemented with 1–3g Na/kg diet, followed by the fish fed the diet with 7g Na/kg diet and lowest in the fish fed the unsupplemented control diet. In seawater, the weight gain, FE and gill Na+–K+ ATPase activity in fish were not affected by the dietary treatment. Analysis by polynomial regression of weight gain, by broken-line regression of gill Na+–K+ ATPase activity and by linear regression of whole-body Na retention of the fish reared in fresh water, indicated that the adequate dietary Na concentration for tilapia is about 1·5g/kg diet. The present study also suggests that no dietary Na is required for tilapia reared in seawater.

An 8-week feeding trial was conducted to determine the dietary copper (Cu) requirement of juvenile hybrid tilapia, Oreochromis niloticus 5 O. aureus. Purified diets with eight levels (0, 1, 2, 3, 4, 8, 12 or 20 mg Cu per kg diet) of supplemental Cu were given to tilapia (mean initial weight 0·79±0·03 g). Each diet was given to three replicate groups of fish. The rearing water contained 0·98 µg Cu per l. Weight gain was highest in fish given diets supplemented with 2 mg Cu per kg diet, followed by the group given 1 mg Cu per kg diet, then the unsupplemented control group, and was lowest in the 20 mg Cu per kg diet group. The differences between each of these four groups were significant (P < 0·05). Food efficiency (FE) of fish showed a similar trend to that of weight gain. Blood haemoglobin (Hb) and haematocrit (Hct) concentrations were lowest in fish given ≥ 12 mg Cu per kg diet, followed by fish given the unsupplemented control diet and highest in fish given 2 mg Cu per kg diet (Hb) and 2 and 3 mg Cu per kg diet (Hct). Plasma ceruloplasmin activity was higher in fish given 1 and 2 mg Cu per kg diet, followed by 3 to 12 mg Cu per kg diet and the control group, and lowest in fish given 20 mg Cu per kg diet. The body Cu content in fish generally increased as the dietary Cu supplementation level increased. Analysis by broken-line regression of weight gain percent and by linear regression of the whole-body Cu retention of fish indicated that the adequate dietary Cu concentration in growing tilapia is about 4 mg Cu per kg diet.

17α-ethynylestradiol sex-reversed males of Oreochromis aureus (pseudofemales, ΔF, genotype ZZ) are used in aquaculture to produce a male monosex population by crossing with ZZ homogametic normal males. When placed with males (M) and females (F) in the same spawning tank, the spawning rate of F is higher than for ΔF. In order to understand this phenomenon, comparisons were made between the behaviour of 18 F (446 ± 96 mm) and 18 ΔF (401 ± 59 mm). ΔF showed a more aggressive behaviour and were significantly more dominant than normal F in fighting pair experiments (F × ΔF) or in fighting group experiments in four different stocking densities (8, 12, 16 and 83 fish 10–3 l with F/ΔF RATIO = 1). ΔF were also more aggressive towards males than F were. The results support the idea that behavioural differences exist between F and ΔF. These differences are probably due to the effect of the sexual genotype on behaviour.

A 10-week feeding trial was conducted to re-evaluate the level of dietary vitamin E (DL- α-tocopheryl acetate) that was adequate for juvenile tilapia Oreochromis niloticus ✕ O. aureus given diets containing two dietary lipid concentrations. Purified diets with eight levels of vitamin E (0, 25, 50, 75, 100, 150, 200, 400 mg/kg diet) at either 50 or 120 g lipid per kg were each given to three replicate groups of tilapia (mean weight: 0·69 (s.e.0·02) g) reared in a closed, recirculating system. Food efficiency and protein deposition were significantly (P < 0·05) higher in fish given 50 mg vitamin E per kg diet and 75 mg/kg diet in the 50 and 120 g lipid per kg groups respectively, compared with fish given the unsupplemented control diet. Mortality of fish was not affected by dietary treatment. Weight gain and liver microsomal ascorbic acid-stimulated lipid peroxidation data analysed by broken-line regression indicated that the optimum dietary vitamin E requirements in juvenile tilapia are 42 to 44 mg vitamin E per kg and 60 to 66 mg vitamin E per kg in 50 and 120 g lipid per kg diets, respectively.

The reproductive characteristics of Oreochromis niloticus, a multiple spawner, have been compared between populations submitted experimentally to three types of spatial structure. Fish from a same original stock were divided and reared for 2 months in: 1) concrete tanks with a bare bottom, 2) concrete tanks with a sandy bottom, or 3) concrete tanks equipped with an artificial reef. The females were then killed and their gonads examined for sexual stage and fecundity determination. Fecundity was inversely correlated to the spatial complexity of the environment whereas oocyte size tented to increase together with spatial complexity. Females placed in tanks with artificial reefs presented a lower fecundity and reproductive investment and bigger oocytes than those maintained in the bare bottom tanks. Fish reared in tanks with a sandy bottom had intermediate reproductive traits. The response of females to changes in the spatial structure of their environment was observed after only one or two reproductive cycles, providing the first experimental evidence of such rapid phenotypic adaptation to environmental complexity. The results also indicate that the perception of living space, arising from the availability of refuges and spawning sites, is more likely to influence the reproduction of tilapias than substrate quality.

An 8 week feeding trial was conducted to determine the dietary K requirement for juvenile hybrid tilapia (Oreochromis niloticus × O. aureus). Purified diets with eight levels (0, 1, 2, 3, 4, 5, 7, 10 g/kg diet) of supplemental K were fed to tilapia. Each diet was fed to three replicate groups of fish initially weighing a mean value of 0.77 (SE 0.01) g/fish in a closed, recirculating rearing system. Weight gain was higher (P<0.05) in fish fed the diets supplemented with 2, 3 and 4 g K/kg diet than in fish fed diet with 10 g K/kg diet and the unsupplemented control diet. Gill Na+-K+ ATPase activity was highest in fish fed the diets supplemented with 1–3 g K/kg diet, followed by fish fed the diet with 5 g K/kg diet and lowest in fish fed the diet with 10 g K/kg diet. Whole-body K content in fish were generally increased as the dietary K supplementation level increased. Analysis by polynomial regression of weight gain and gill Na+-K+ ATPase activity and by linear regression of whole-body K retention of the fish indicated that the adequate dietary K concentration for tilapia is about 2–3 g/kg diet.

Large differences in reproductive traits were observed between populations of Oreochromis niloticus sampled in 9 manmade lakes of Ivory Coast. In order to assess whether these variations in reproductive characteristics resulted from short term adaptation or from a longer evolutionary process, living specimens were caught in the two most differentiated populations in term of life history traits, and placed in a common environment in culture conditions. Genetic analysis of fish from these two populations were performed using four microsatellite markers and revealed that both descended originally from the same strain (Bouaké station) which was constituted from broodfish initially caught in the Nile and Volta basins. Fish from the two populations were subjected to a common environment (pond and aquariums) for five months. Then, their reproductive characteristics were analysed and no significant differences were found in fecundity, egg size and spawning frequency. These results indicate that reproductive differences between the two populations, originally observed in the two reservoirs, mostly reflect the phenotypic plasticity of the species in facing different environmental conditions.

L'inversion hormonale du sexe chez Oreochromis aureus (génotype mâle ZZ et génotype femelle ZW) avec un œstrogène (1717α-éthynylœstradiol) conduit à l'obtention de pseudofemelles, caractérisées par un genotype mâle (ZZ). Le croisement de ces pseudofemelles avec des mâles (ZZ) donne, en théorie, des populations 100 % mâles. Nous avons déterminé et comparé les principaux paramètres caractéristiques de la reproduction au sein de trois groupes de géniteurs O. aureus. Le premier groupe (groupe 1) est composé de femelles, le second (groupe 2) de pseudofemelles et le troisième (groupe 3) de femelles et de pseudofemelles. La fécondité et le poids des œufs des femelles et des pseudofemelles ne sont pas significativement différents. Dans le groupe 3, le taux de reproduction des pseudofemelles (0,6 pontes) est plus faible que celui des femelles (6,4 pontes). Dans ce groupe 60 % des pseudo-femelles ne se sont jamais reproduites. Dans les groupes 1 et 2, il n'existe pas de différence entre les femelles (6,9 pontes) et les pseudofemelles (4,7 pontes). Le plus faible taux de reproduction des pseudofemelles pourrait résulter de l'expression du génotype sexuel mâle chez ces individus.

Surgery procedures were used to implant telemetry transmitters into the body cavity of adult (574–1 033 g) tilapias Oreochromis aureus in aquaculture tanks (4 m2, 1.5 m3, 26.5 ± 0.5 °C, ≥ 5.0 mg O2 l−1) and their effect on fish survival, growth and behaviour was evaluated. Only one out of 35 implanted fish died. With one exception, all fish consistently retained their transmitter until the end of the study (up to 30 months). Healing was faster when the incision was sutured with polyamide monofilament (5–14 days) than with other suture materials, due to tunnelling with atraumatic needles for catgut or fouling of braided silk. In all 10 fish sacrificed after 30 and 50 days, the transmitter had become encapsulated by connective tissue. No infection or damage to the viscera was observed.

The activity of four tilapias (903–1 033 g) equipped with motion sensitive transmitters was telemetered during the recovery from anaesthesia and surgical procedures. All four fish maintained a normal diurnal activity rhythm pattern throughout the study but had low levels of activity during the first 12–24 h. Based on the evolution of their resting posture after surgery, it is suggested that tilapias need 3 to 4 days to completely compensate the negative buoyancy resulting from anaesthesia and tagging.

Morphological features of the cap enameloid and dental epithelial cells were investigated by light and transmission electron microscopy during the various stages of enameloid mineralisation and early maturation in the tilapia. The pattern of mineralisation along collagen fibrils in the enameloid differed from that in the dentine. Many matrix vesicles were found in the predentine and in the enameloid, suggesting that they may be involved in the initial mineralisation in both regions. Most of the organic matrix disappeared from the cap enameloid during mineralisation and maturation. The disappearance of the organic matrix could be divided into 2 stages. Initially a fine network-like matrix, which probably consisted of glycosaminoglycans and extended between collagen fibrils, began to disappear. At the same time, fine crystallites and electron-dense, fine granular material covered the collagen fibrils as mineralisation of the enameloid began. In the second stage, the maturation of the enameloid, the collagen fibrils degenerated completely and disappeared from the cap enameloid, being replaced by large numbers of large crystals. At the mineralisation stage, the numbers of lysosomal bodies tended to increase in the inner dental epithelial (IDE) cells, which contained a well developed Golgi apparatus and rough endoplasmic reticulum (rER). At the early stage of maturation, a ruffled border was noted at the distal ends of the IDE cells, which contained many mitochondria and lysosomal bodies, but less rER. These features suggest that the cells actively absorb the organic matrix, which includes collagen fibrils, in the cap enameloid. The outer dental epithelial (ODE) cells were translucent cells that contained well developed labyrinthine canalicular spaces from the onset of the mineralisation stage to the middle stage of maturation. The IDE and ODE cells were clearly involved in the mineralisation of the cap enameloid at the mineralisation and maturation stages.

The experiment was conducted for 122 days in central Thailand to test rearing of hybrid catfish in cages in earthen ponds, where tipalia (Oreochromis niloticus) was stocked to utilize the waste products derived from intensively reared catfish. Catfish fingerlings (13-17 g size) was stocked in plastic net cages at a density of 275 fish/m3. Six earthen ponds (250 m2-110 m3 each) were used to suspend the cages in two loading densities as experimental treatments: 880 and 1760 catfish/pond. Each loading density was replicated in 3 ponds. Each pond was stocked with 440 sex-reversed male tilapia of 6-7 g size in open water, giving the catfish and tilapia stocking ratios of 2:l and 4:1, respectively. Catfish was fed twice daily at 3-10% body weight per day with commercial floating pellets containing 25-30% crude protein. Wate quality was analyzed bi-weekly for concentration of dissolved oxygen, ammonia and chlorphyll a.

Results showed that there was no significant difference (p < 0.05) in catfish growth rate and survival between the two loading densities. The mean weight gain of harvested catfish was 259.5 ± 34.5 g and 255.2 ± 8.4 glfish, giving the total net yield of 218.0 ± 26.8 kg and 391.5 ± 88.0 kg per pond in low and high loading density, respectively. Tilapia was harvested with mean weight of 172.3 ± 37.8 g and 297.5 f 32.0 glfish, and the net yield of 68.1 ±13.6 and 86.0 ± 27.2 kg/pond for low and high loading catfish density treatment, respectively. Chlorophyll a and total ammonia concentrations differed greatly between the two treatments. Early morning dissolved oxygen (DO) declined steadily both treatments over the culture period.

This work describes the application of a homologous radioimmunoassay (RIA) in the measurement ofplasma and pituitary growth hormone levels of tilapia maintained at 20 and 26 °C, using recombinant tilapia growth hormone (rtiGH). RIA sensitivity (ED90) was 0.08 ng/ml and ED50 was 0.62 ± 0.02 ng/ml. Intra- and inter-assay coefficients of variation were 3.3 and 10.6% respectively, for a plasma sample of 1 ng/ml of tiGH. Mammalian or salmonids growth hormones, thyrotropins or tilapia prolactins did not show cross-reactivity with rtiGH antiserum. Plasma and pituitary extract of Oreochromis niloticus gave inhibition curves parallel to the rtiGH standard curve without interaction with serial dilutions either of tissue extracts or plasma from hypophysectomized tilapia. Pituitary extracts from other cichlid fish showed parallel competitive binding curves. Pituitary extracts of perciform fish exhibited a lower affinity to anti-rtiGH serum, while extracts of species from more distant families showed only negligible cross-reactivity or none at all. The effects of water temperature on plasma concentrations and pituitary contents of GH were investigated in tilapia acclimatized to 26 and to 20 °C. Measurements have been performed at the beginning and at the end of a two-week experiment. Whereas in fish from colder water plasma GH mean values were decreased, only a similar tendency was observed in pituitary GH contents. These results emphasize the hypothesis wich reports that the increase of growth with temperature is, at least in part, controlled by GH.

Le régime alimentaire de Tilapia, Oreochromis niloticus, principale espèce exploitée commercialement au Lac Muhazi, a été étudié de décembre 1989 à septembre 1990, en fonction des saisons, des stations et de la taille des individus capturés dans les filets de pêche. Tous les indices considérés montrent que O. niloticus du Lac Muhazi consomme bien le phytoplancton et que les Cyanobactéries sont les plus représentées dans les contenus stomacaux comme c'est le cas dans le lac. Pour les 118 contenus stomacaux analysés, l'indice d'occurence arrive à 100 % pour certaines espèces de Cyanobactéries (Merismopedia tenuissima). Pour ce groupe dominant, l'indice d'abondance est en moyenne de 64,3 %, l'indice volumétrique de 83 % et l'indice de dominance de Lauzanne de 58,1. Deux espèces, Microcystis aeruginosa et Ceratium hirundinella représentent à elles seules plus de 90 % de la biomasse contenue dans les estomacs des poissons dont la taille est supérieure à 90 mm quelles que soient la station et la saison de prélèvement. Pour la période d'étude, nous avons constaté que O. niloticus a une activité essentiellement diurne (il se nourrit entre 6 h et 21 h 30), avec une ration journalière élevée (26,5 g.j-1). Ceci explique sans doute la bonne croissance de cette espèce dans les milieux où le phytoplancton est abondant. Cependant, au Lac Muhazi, la population d'O. niloticus n'atteint qu'une biomasse relativement faible et ne consomme qu'une faible fraction de la production primaire nette (6 %).

Les lagunes côtières doivent leur forte productivité aux échanges qu'elles entretiennent avec les eaux marines et continentales. Au Togo, en période de fermeture du cordon littoral (situation observée en 1994), la salinité est faible: 1 psu à Agbodrafo, ainsi que la diversité spécifique du peuplement étudié (Ish = 2,13). A l'inverse, après une période de communication avec la mer comme ce fut le cas en 1989, la salinité remonte sur l'ensemble du plan d'eau (7 psu à Agbodrafo) et les peuplements de poissons et de crustacés sont plus variés (Ish = 2.74) et à tendance marine ou marine-estuarienne (Liza, Caranx, Penaeus, Gerres...). Dans ce cas, les rendements élevés (300 à 400 kg/ha/an) s'expliquent par un enrichissement du milieu lagunaire dû aux apports marins et à l'introduction d'espèces marines. En situation d'isolement, les peuplements sont essentiellement estuariens et largement dominés par les Cichlidae et plus particulièrement Sarotherodon melanotheron qui représente à lui seul 50 % des captures totales. Dans un contexte de pêche particulièrement active, l'expansion de cette espèce est certainement due au fait qu'elle présente une reproduction continue. Néanmoins, les valeurs des rendements totaux observés (160 kg/h/dan), des tailles maximales (250 mm), des tailles moyennes de capture (120 mm) et des tailles de première maturité sexuelle (85 mm), inférieures aux valeurs rencontrées dans les autres lagunes de la région (Nigeria et Côte-d'Ivoire), semblent traduire une croissance de Sarotherodon plus faible au Togo. L'absence de communication avec la mer et la pression de pêche intensive seraient à l'origine d'une prolifération de cette espèce dont les fortes abondances entraîneraient à leur tour un ralentissement de la croissance.

Dry diets containing varying levels of dried fermented fish silage and soybean meal blend (FS: SBM) as replacement for 25,50 or 75% of fish meal protein were fed to juvenile Oreochromis niloticus (8.2 ± 0.1 g) and Clarias gariepinus (10.8 ± 0.3 g) at 5% body weight per day for 70 days. Diets containing co-dried FS: SBM supported weight gains and growth rates similar to those in the control treatment without significant differences (p > 0.05). Feed conversion, protein efficiency ratio and protein productive values followed a similar trend with growth performance values. Apparent protein digestibility decreased in both fish species with increasing dietary level of co-dried FS: SBM Carcass composition of experimental fish were not affected by the diets and morphological defects were not observed. Results indicate that co-dried FS: SBM can partially replace fish meal protein as well as establish the potential of utilizing waste/undersized tilapias as a protein feedstuff in dry aquaculture diets.