1. For a period of 8 weeks, rainbow trout (Salmo gairdneri), mean initial weight 21 g, were given either a low-magnesium or control diet containing 0·03 and 0·58 g Mg/kg diet respectively. Both groups of trout were then given the control diet for a further 11 weeks.
2. Weight gains over the initial 8-week period were lowest in the Mg-deficient trout. Feeding the deficient fish the control diet rapidly improved growth rate until it was the same as that of the control trout.
3. Plasma Mg was significantly lower in the Mg-deficient trout at week 8. Feeding with the control diet for 11 weeks did not increase plasma Mg. Few changes were observed in the plasma concentrations of the other electrolytes.
4. Renal calcium concentrations were unaffected by dietary Mg levels. Similarly, the renal levels of phosphorus, sodium and potassium all fell within the range found in normal rainbow trout.
5. Muscle Mg concentrations were reduced in those trout given the Mg-deficient diet. Feeding with the control diet for a further 11 weeks increased muscle Mg but the level was still significantly lower than that found in trout given the control diet for 19 weeks.
6. The bone ash Mg concentration was significantly lower, and the Ca higher, in the deficient fish at week 8, when compared with the control group.
7. When compared with the value at the start of the experiment, total bone Mg fell slightly in the deficient trout over the initial 8-week period, but increased in the control group of fish. Feeding with the control diet for a further 11 weeks increased total bone Mg in both Mg-deficient trout and control trout.
8. The results show that the Mg deficiency imposed on the rainbow trout was of limited severity and almost complete recovery was obtained when the control diet was fed.