Evaluation of Poultry By-product Meal in Commercial Diets for Hybrid Striped Bass, Morone chrysops ♀×Morone saxatilis ♂, in Pond Production

The efficacy of replacing fish meal with petfood-grade poultry by-product meal (PBM) on an ideal protein basis in commercial diets for hybrid striped bass (HSB) was evaluated under production conditions in pond culture. A generic production diet (GEN) for HSB was formulated to contain 45% protein, 12% lipid, and 3.7 kcal/kg. Protein in the generic diet was supplied by a mix of animal and plant sources typically used by the industry that included more than 20% select menhaden fish meal and less than 10% PBM. A positive control diet (GEN + AA) was formulated by supplementing the generic diet with feedgrade Met and Lys to match the level of those amino acids in HSB muscle at 40% digestible protein. Substitution diets were formulated by replacing 35, 70, or 100% of fish meal in the GEN diet with PBM on a digestible protein basis and then supplementing with Met and Lys (designated 35PBM, 70PBM, and 100PBM, respectively) as needed to maintain concentrations equal to those in the GEN + AA diet. Diet formulation and extrusion were conducted by a commercial mill, and all diets met or exceeded known nutritional requirements for HSB. Twenty 0.10-ha ponds (4 ponds/diet) were randomly stocked with juvenile HSB (76 6 10 g; mean 6 SD) at a density of 7400/ha and fed for 600 d (October 2004 to May 2006). Diets were fed once daily to apparent satiation to a maximum of 95 kg feed/ha. Total weight and number of fish in each pond were determined at harvest. Weight distributions in each pond were estimated by selecting every 15th fish during harvest. Subsets of ten fish from each of these samples were selected randomly for the determination of body composition and nutrient and energy retention. The availability of indispensable amino acids as well as ammonia production from the commercial test diets were determined in separate tank trials. Most production characteristics were not statistically different (P . 0.10) among dietary treatments. Distributions of individual fish weights from each of the ponds were not affected by poultry by-product level in the diet. Multivariate analysis of body compositional indices grouped diets into two clusters composed of GEN, GEN + AA, 35PBM vs. 70PBM, and 100PBM mainly because fish fed the 70PBM and 100PBM diets had greater (P 5 0.001) body fat (visceral somatic indices) than fish fed the other diets. Ammonia production in tanks was not different among diets and peaked 6–8 h after feeding when fish were fed at 1.5% of body weight; ammonia-N excretion ranged from 197 to 212 mg/kg/d and 18.5–21.5% of nitrogen intake. Some imbalances in the levels and ratios of selected amino acids to Lys were found in the diets containing higher amounts of PBM and were attributed to a lack of accurate availability coefficients during formulation for some dietary proteins. These imbalances in essential amino acids may have been the predominant factor in the somewhat fattier fish observed fed diets containing the two highest levels of PBM. Nevertheless, these results from fish stocked at commercial densities and raised to market size in ponds suggest that formulating diets on an available amino acid basis for all protein 1 Corresponding author. JOURNAL OF THE WORLD AQUACULTURE SOCIETY Vol. 40, No. 2