In intensive culture, the two species of brackish water tilapia, T. guineensis and S. melanotheron, have proved disappointing. Even under most favourable conditions, i. e. the monosex male culture of S. melanotheron, a poor growth rate (about O. 5g. d-') and a mediocre feed conversion do not presage a profitable aquaculture exploitation. On the other hand, extensive culture trials using the original "acadja-enclosure" technique have given very encouraging resu1ts.A nnual fish yields of more than 8 t/ha, including 70 to Sol ;/, of S. rnelanotheron of marketable size, are possible without any artificial feeding.

Accepté le 12 avril 1995. Abstract RésuméLuquct P., Z. J. Otémé, A. Cissé. Aquat. Living Resour., 1995, 8, 389-394. Evidence for ct>rnpensutov growth and its urility in the culture cf Heterobranchu~ longifilis.Three trophic levels wcre tested with the African catfi\h IIvterobranchus 1ongiJili.s. In thc 1st trial, fish with a rriean weight of 60 g (12 batchcs of 8 fish), were subjcctcd to severely restrictcd feeding (1% of body weight) for three weeks, then incrcased rations for thrcc weeks either to ad libitum feeding or to 3% of body wcight. Growth sates wcrc compared to thosc of fish that had becn continuously fed eithcr ud libitum or at 3% of body wcight. Although the final live weights of fish on restricted diets were lower than those of fi5h on unrestrictcd diets, the results indicate that a restrictive feeding rate ultimately induces a significantly higher specific growth rate, and a better food conversion ratio. This compensatory growth is caused both by hypcrphagy and by an improved metabolism. A positive effect of hyperphagy was still noted during the following two weeks.During the 2nd experirnent, silurids of an initial mean weight of 141 g (6 batchcs of 18 fish) werc fed for 10 weeks either at a constant rate of 4% body weight, or fed altemately on a diet restrictcd to 2% of body weight for one wcck then increased to 4 or 5% of body weight for the next week. At the end of the trial (10 weeks), mcan weights were not significantly difLerent. The best food convcrsion ratio (1.8) was found among the fish on restricted dicts alternated with increased rations equal to that of the control (4%). In the 3rd experiment, silurids of a mean weight of 340 g (6 batches of 16 fishes) were fed for five months at a constant rate of 3% of body weight and those of fish fed altcmately at 2% then 4% using 15-day intervals following lunar cycles were compared. At the end of the experiment, final mean wcights varied from 977 to 1 127 g, while the food conversion ratios varied from 2.1 to 2.5. The best performances, considcring al1 critena, were those of group of fish fed restricted diets of 2% of body weight during the waning moon and then incrcased to 4% during the waxing moon. Conversely, the worst results were obtained with those fish on a restricted diet during the waxing moon. To fully utilize the phenomenon of compensatory gain as illustrated in the silurid, dietary restriction, even a mild one for a one wcek period can be recommended. However, there appears to be very little profit in attempting to benefit from hyperphagy after restricted diets for two weeks.Keywurds: African catfish, Heterobrunchus longifilis, diets, growth, lunar cycles.Trois niveaux trophiques ont été expérimentalernent étudiés chez le silure africain Heterobranchus longijîlis. Au cours d'un 1" essai, des poissons d'un poids moyen initial de 60 g ont été rationnés (1 % du poids vif) pendant trois semaines puis alimentés pendant 3 semaines soit ad libitum soit à 3 % du poids vif. Leur croissance a été comparée à celle de poissons nourris en perm...

Use of chrmical or hiologicaljsh silugr as feed for Chrysichthys nigrodigitatus (Bugridue).These studies were undertaken to test the nutritional value of marine fish offal products prescrvcd by different silage nlethods and used as an ingredient in african catfish (Clzry.sichthys nigrodigitutus) feeds. The first experiment concerned cooked tuna wastc silage preparcd by addition of 3% formic acid. After incubation for 16 days, a desirablc and stable pH of 4.03 was obtained. The hydrolysate, containing 66.7% of crude protein (N x 6.25) on a dry matter (DM) basis, was incorporated with a basal mixture to formulate feeds containing 38% crude protein which were fed to Chrysichthys nigrodigitatus. A commercial fish meal was uscd as control on an iso-nitrogenous basis. Diets were distributed to 6 groups of 50 catfish of an initial weight of 34 g for 56 days at 3/70 (dry) of fish body wcight. The zootcchnical performances obtaincd with fish fed with the expcrimental diet (CNE) were bcttcr in terms of the daily weight gain (0.86 g.d-'), food conversion ratio (FCR: 1.79) and protein cfficiency ratio (PEU: 1.46), compared to those of the control diet (no significant differencc). During the second experiment, fish silage was preparcd by fermentation with 65% (wct weight, wlw) cooked and ground small sardines, 25% (wlw) corn flour, 5% sugar and 5% (wlw) fcrmented cassava used as fermentation startcr. After 35 days, a stable pH of 4.1 8 was attained and the source of Sermentation substrate did not affect the nitrogen content of the hydrolysate aftcr 15 days storage. The liquid silage (cosilage) containing 35.65% crude protein (DM) was also added to a mixture and tested against control fish meal on an iso-nitrogenous basis in experimental conditions identical to those described above. The incorporation of cosilage in the diet (CE) resulted in an increase in the daily weight gain (0.94 g.d-'), FCR (1.77) and PER (1.61) compared to control diet.In both experiments, the increase in growth rate and the presence of greater amount of lipids in the body of fishcs receiving silage are partially due to greater food ingestion, owing probably to a different texture of the pellets or the possible presence of food attractants. The acceptability of fish silage as an ingredient in catfish feed, as shown in these studies, constitutes an alternative to fish meal in utilizing marine fish waste or low value fish.Keywords: Chrysichthys nigrodigitatu.~, catfish, feeding, growth rate, tropical aquaculture, Côte d'Ivoire.Deux expériences ont été réalisées pour tester la valeur alimentaire, pour le mâchoiron (Chrysichthys nigrodigitatus), de dérivés de poissons de mer conservés par la méthode de l'ensilage. Le premier essai a concerné des déchets de conserveries de thons conservés par la méthode de l'ensilage acide (3 % d'acide formique). L'hydrolysat, renfermant 66,7 % de protéines brutes (N x 6,25) sur la base de la matière sèche (MS) a été incorporé dans un aliment expérimental en substitution à la farine de poisson sur une base iso-azotée. Les al...

The feeding experiment was designed on the basis of food allowance rather than on feed formulation. Food allowances are fixed in an attempt to provide the desired amount of nutrients per body mass of fish per day (g -$ kg-' ' > d-' or J ' > kg-' ' > d-'). Therefore, the initial determination of optimum feeding ration is not required, and utilization of neutral feed complement (e.g. cellulose) is unnecessary. Data interpretation for such a design stud is presented by a sample experiment for protein-energy re uirements of a tilapia species, Sarothero&z melanotheron. The interrelationships of commonly used inlices for growth and nutritional efficiency are discussed.