The reduction of phosphorous waste from farmed rainbow trout (Oncorhynchus mykiss) was attempted by feeding diets of reduced P content. The protein source was fishmeal in which the bone fraction had been removed. This was achieved either by solubilizing the fish material (press cake) by enzymatic treatment and subsequent filtering, or by removing the fish bones by a skin/bone separator. Two fish species were used, sand eel (Ammodytes marinus) and sprat (Sprattus sprattus). Due to the small size of the fish, only bones were retained in the skin/bone separator. Traditional fishmeal from the same species were used for control experiments. The fishmeal samples were used together with other ingredients in a traditional extruded dry feed formulation (Ecolife 19).

The diets were used in full scale experiments in a fresh water fish farm in Denmark, as well as in small scale experiments in aquaria. Direct calculations on the feed conversion ratio and the specific growth rate on both experiments showed that the enzyme treated codes performed less well than the other codes. However, when compensating for the different energy content of the diets, the enzymatic treatment did not give rise to a higher feed conversion ratio, nor did it give rise to a lower average growth rate than the other codes. It was not possible to make any final conclusions with respect to the effect of the enzymatic treated fishmeal on the pollution risk from P. This was due to large variation on the P determinations.