• brook trout;
  • mottled sculpin;
  • predation;
  • top-down control;
  • trophic cascade


1. In some situations fish have strong top-down effects in stream communities while in others they seem to be relatively unimportant. Differences in the impact of fish may depend on a variety of factors including the foraging mode of the fish, interactions among fish species and temporal variation in environmental conditions and species interactions.

2. We investigated the effect of brook trout (Salvelinus fontinalis) and mottled sculpin (Cottus bairdi) on lower trophic levels in Appalachian streams and whether or not interactions between these fish changed their influence. Mesocosms were placed in a headwater stream in a randomized complete block design. Within blocks, mesocosms were randomly assigned to one of the following treatments: (i) no fish; (ii) sculpin only; (iii) trout only and (iv) both sculpin and trout. Fish biomass was the same in all three fish treatments. Invertebrate density and algal biomass in mesocosms were determined after 3 weeks. We repeated the experiment in the autumn, spring and summer to test for seasonality of fish effects.

3. The effect of fish on invertebrate assemblages was seasonal and depended on prey identity. Sculpin strongly suppressed grazer abundance in spring while trout had little effect on grazers in any season. The influence of both fish on insect predators was similar and relatively constant across seasons. We found little evidence of an interaction between sculpin and trout that strongly influenced their effect on prey across seasons.

4. None of the fish treatments influenced algal biomass during any of the seasons. Algal growth was also seasonal, with a two- to four-fold increase in algal biomass in spring compared to autumn and summer.

5. Our results indicate that benthic and drift feeding fish differ in their effects on some, but not all prey. Furthermore, fish effects on prey were strongly seasonal for some, but not all prey types. While the temporal context is not commonly considered, our results indicate seasonality can be an important component of predator–prey interactions in streams.