An in vivo digestibility trial and a feeding trial were conducted to determine the extent to which heat treatment, nutritional quality, and palatability of hexane-extracted, defatted soy flour (SF) influenced growth of fingerling coho salmon. Heat treatment of SF (autoclave, 1.7 atm, 121 degrees C) lowered trypsin units inhibited (TUI) from 181 to 1.8 after 20 min and lowered protein solubility from 98% to 70%. Longer heating periods further reduced SF protein solubility, but did not substantially reduce TUI. The apparent digestibility coefficient (ADC) of protein from heated SF (90.8) was significantly higher than the ADC of unheated SF (74.3), but significantly lower than the ADC of the wheat gluten basal diet (98.2). In the feeding trial, fish weight gain was reduced in a step-wise fashion at each level (15, 20, or 25%) of unheated SF inclusion in the diet. Fish fed diets containing heat-treated SF gained more weight at each SF inclusion level compared to their corresponding unheated treatment groups, but inclusion of heat-treated SF greater than 15% substitution of herring meal protein resulted in reduced weight gain compared to fish fed the control diet. Krill supplementation (5%, combination of dried and frozen) increased weight gain in dfish fed diets containing 25% heated or unheated SF, primarily by restoring feed intake. In fish fed the heated 25% SF diet with krill, feed intake and average weight gains were equivalent to control diet levels. Fish fed the unheated SF diet with added krill had higher feed intake and weight gains than fish fed the equivalent diet without added krill, but gains were significantly lower than controls. Calculated indices of fish performance, e.g., specific growth rates, feed conversion ratios, protein retention ratios, and apparent net protein utilization (percentage protein retention) displayed patterns similar to weight gain results. In this study, a 20 min heat treatment of SF reduced trypsin inhibitor activity to physiologically insignificant levels, increased protein digestibility, and also allowed a successful 15% substitution of SF for herring meal protein. It appears that reduced feed intake associated with unheated SF was responsible for approximately one-half of the observed reduction in weight gain. Lower nutritional value of the unheated or insufficiently heated SF was responsible for the remaining reduction in fish performance. Therefore, a combination of proper heat treatment and supplementation with palatability-enhancing feed ingredients can overcome these problems, even in young Pacific salmon which are known to be extremely sensitive to soybean meal in their feeds. (C) 1999 Elsevier Science B.V. All rights reserved.
