In the tank experiment, the δ13C and δ15N values of the assimilated fish diets (from Eq. 2) in the diet-switching treatment closely matched those of the provided foods (chironomid larvae during weeks 1–5 and feed during weeks 6–10; Fig. 3A). In the pond experiment, the assimilated fish diet δ13C and δ15N values in the low and high-feeding treatments during weeks 1–3 did not differ from those values for fish in the no-feeding treatment (one-way ANOVA, both P N 0.21; Fig. 3B).
However, fish in the low and high-feeding treatments had higher δ13C values during weeks 4–7 (derived from feed assimilation) as compared to weeks 1–3 (two-sample t-tests, both P b 0.025), whereas only the fish in the low-feeding treatment had higher δ15N values during weeks 4–7 as compared to weeks 1–3 (two-sample t-test, P= 0.012).