Diets incorporating three different sources of extracted cottonseed meal (CM), soybean meal and an animal protein mixture were evaluated for juvenile rainbow trout. Fish averaging 0.96 g were divided into groups of 30; 3 groups per treatment, and each group was fed one of four diets for a 16-week period. Fish meal (FM) was replaced on a 25% protein basis by each of three different sources of CM from California (CA), Tennessee (TN), and Arkansas (AR), U.S.A. In the three CM-containing diets another 25% soybean meal protein and 50% animal protein mixture were also incorporated to completely replace FM protein. The results of growth rate and feed utilization showed that FM could be entirely replaced by a mixture of plant proteins (CM and soybean meal) and animal by-product proteins. Hematocrit levels were significantly lower in the group fed CM-containing diets than in the control. The findings suggest that CM can be used as a good protein source by the incorporation of at least 15% in diets (25% of fish meal protein replacement), and that the nutritive values of CM in juvenile trout can be different due to their different origin. Significantly higher concentrations of total gossypol were found in faeces of CM-TN (5.8 +/- 0.4 micromol/g) and CM-AR (5.6 +/- 0.6) groups than in that of CM-CA (3.7 +/- 0.4) group. It was documented that gossypol enantiomers, present in an equal proportion in diets, selectively accumulated in liver and bile, whereas equal proportions of (+)- and (-)-enantiomers were found in whole-body and faeces. Depending on CM source, fish can absorb approximately 35-50% of dietary gossypol, and the majority of the absorbed gossypol seemed to be excreted.

Rainbow trout (Oncorhynchus mykiss) males were fed, during a 9‐month period, five experimental diets where fishmeal proteins were gradually replaced with cottonseed meal (CS) proteins (0, 25, 50, 75 and 100%; diets 1–5, respectively). This study was carried out to evaluate the action and tissue concentrations of gossypol. Growth performance of fish was not affected with the increasing levels of CS in the diets. Haemoglobin concentration and haematocrit significantly decreased in fish fed 100% CS compared with other dietary treatments. Gonadosomatic index, plasma sex steroids (testosterone, 11‐ketotestosterone and 17,20β‐dihydroxy‐4‐pregnen‐3‐one) and sperm characteristics (concentration, motility, protein concentration and lactate dehydrogenase activity) were not negatively affected by increasing levels of CS. For the first time, a comprehensive study of gossypol concentrations, total and (+) and (−)‐enantiomers, in several tissues (liver, blood plasma, spermatozoa and seminal plasma) was performed. The concentrations of both (+) and (−)‐enantiomers significantly increased with increasing levels of CS in the diet. The highest concentrations were found in the liver of fish fed diet 5 (185 ± 18 μg g−1). In blood plasma, the concentrations of total gossypol were 10 times lower than the one found in the liver, but 10 and 100 times higher than the concentrations in the spermatozoa and seminal plasma, respectively. In all tissues studied, the concentration of (+)‐enantiomer was higher than the (−)‐enantiomer. The ratio (−)‐enantiomer/total gossypol did not change significantly with the increasing levels of CS in the diet. The results of the present study indicate that CS can be used over a period of 9 months to replace fishmeal proteins completely without compromising growth and reproduction of rainbow trout males.