Abstract
The effects of roach (Rutilus rutilus) density of perch (Perca fluviatilis) and ruffe (Gymnocephalus cernua) were examined experimentally in ten enclosures in a shallow pond in southern Sweden. The growth and diets of the fish species as well as the biomass of zooplankton and benthic macroinvertebrates in the enclosures were followed for two months during the summer. The biomass of cladocerans decreased with increasing roach density in September but not in October, which was probably due to the lower biomass of cladocerans in October. The biomass of copepods was not related to roach density. The growth of perch and roach decreased with increasing roach density, whereas the growth of ruffe did not change with roach density. Roach density did not affect the diet of perch, whereas roach consumed a higher proportion of plant material at high roach densities. The diet of ruffe was not affected by roach density. A high density of a planktivore (roach) should be able to force perch to shift from zooplankton to benthic macroinvertebrates. To explain the lack of a diet shift in my experiment I developed a hypothesis about how the simultaneous effects of two competitors (roach and ruffle) on a third species (perch) influence the expected diet shift of perch.