1. In stream ecosystems, Pacific salmon deliver subsidies of marine-derived nutrients and disturb the stream bed during spawning. The net effect of this nutrient subsidy and physical disturbance on biological communities can be hard to predict and is likely to be mediated by environmental conditions. For periphyton, empirical studies have revealed that the magnitude and direction of the response to salmon varies from one location to the next. Salmon appear to increase periphyton biomass and/or production in some contexts (a positive response), but decrease them in others (a negative response).

2. To reconcile these seemingly conflicting results, we constructed a system dynamics model that links periphyton biomass and production to salmon spawning. We used this model to explore how environmental conditions influence the periphyton response to salmon.

3. Our simulations suggest that the periphyton response to salmon is strongly mediated by both background nutrient concentrations and the proportion of the stream bed suitable for spawning. Positive periphyton responses occurred when both background nutrient concentrations were low (nutrient limiting conditions) and when little of the stream bed was suitable for spawning (because the substratum is too coarse). In contrast, negative responses occurred when nutrient concentrations were higher or a larger proportion of the bed was suitable for spawning.

4. Although periphyton biomass generally remained above or below background conditions for several months following spawning, periphyton production returned quickly to background values shortly afterwards. As a result, based upon our simulations, salmon did not greatly increase or decrease overall annual periphyton production. This suggests that any increase in production by fish or invertebrates in response to returning salmon is more likely to occur via direct consumption of salmon carcasses and/or eggs, rather than the indirect effects of greater periphyton production.

5. Overall, our simulations suggest that environmental factors need to be taken into account when considering the effects of spawning salmon on aquatic ecosystems. Our model offers researchers a framework for testing periphyton response to salmon across a range of conditions, which can be used to generate hypotheses, plan field experiments and guide data collection.