Evaluation of protein quality in fish meals by chemical and biological assays

Fish meals (herring, menhaden, and anchovy), commercially available in Atlantic Canada and a Norwegian fish meal (Norse-LT94Reg.) formulated to contain 0, 16, 28 or 40% protein were evaluated for protein quality by in vitro assays and by growth studies with Atlantic salmon fingerlings. Pepsin digestibility and multienzyme digestibility were useful in predicting protein quality. Assays for total volatile basic-nitrogen, available lysine, sulphydryl groups, and disulphide bonds, particularly when used alone, were of limited value. Atlantic salmon fingerlings (initial weight 7.65 +or- 0.14 g) fed on diets containing 2 steam-dried herring meals and Norse-LT94Reg. for 70 days were evaluated on the basis of weight gain, protein efficiency ratio, net protein ratio, net protein utilization, and slope assay. Fish fed Norse-LT94Reg. gained 11-60% more weight than fish fed on the Canadian meals. The protein quality of the Canadian fish meals was lower than that of Norse-LT94Reg.. Pepsin digestibility ranked the fish meals in the same order as the biological tests, but the biological testing of feedstuffs to determine protein quality is recommended as the final method of comparison. Biological evaluation is the preferred method of measuring the overall quality of a feedstuff, since in vitro tests may miss-rank the quality of some feedstuffs..