Translating bumblebee dialect…
Summary Mosses in streams attach primarily to stable substrata subject to low hydrological perturbation, serving as shelter and as substratum for epiphytic algae, which are a food source for many grazing macro‐invertebrates. Given the changes in the hydrodynamic conditions occurring at substrata surfaces and the increasing habitat complexity, predator–prey interactions may be altered in streams dominated by moss, compared with those dominated by diatom‐covered rocks. However, little is known about the relative effects of the food and refuge functions of moss and how they influence grazers in streams. Results from a field survey indicated that streams with greater moss cover supported higher densities of less mobile, armoured, predator‐resistant but disturbance‐vulnerable caddisfly grazers than streams where moss was rare. However, the abundance of more mobile, disturbance‐resistant but predator‐vulnerable mayfly grazers was unrelated to the extent of moss cover. To distinguish potential mechanisms explaining those observed relationships, we performed a microcosm experiment comparing the responses of two grazer species, one caddisfly (Allomyia gnathos) and one mayfly (Baetis bicaudatus), to variation in substrata and risk of predation. We estimated grazer growth rates, algal accrual rates and the per capita effect of grazers on algal biomass in circular flow‐through chambers that differed in substratum (artificial moss‐covered tiles versus diatom‐covered tiles) and predation risk (non‐feeding predatory stoneflies present versus absent). Results showed that, although artificial moss‐covered tiles provided more food (epiphytic algae biomass – chl‐a), only the mayflies grew faster with higher food resources available on the artificial moss compared with those feeding on diatom‐covered tiles. Furthermore, algal resource depression was greater on diatom‐covered tiles, which suggests that diatoms on bare tiles were more vulnerable to grazing than epiphytes on artificial moss. Although non‐feeding predators did not affect the biomass of either grazer species, algal accrual rates were higher in chambers where mayflies were exposed to predation risk, which is consistent with a behavioural trophic cascade. Mayflies also had a higher per capita impact on algal biomass in the absence of predatory stoneflies. We conclude that armoured, less mobile and disturbance‐vulnerable caddisfly grazers primarily colonise mossy substrata for protection, because we found no evidence that they benefit from the higher epiphytic food supply on moss. However, although disturbance‐resistant mayflies can exploit increased food resources provided by moss, in nature they do not aggregate on mossy substrata in preference to diatom‐covered mineral substrata. Therefore, classical trade‐offs between food acquisition and safety do not explain the observed distributions of caddisflies, which prefer habitats that both are safe and have the best food resources, nor of mayflies, which prefer neither the safest nor the most resource‐rich habitats. These results demonstrate that mayflies and caddisflies have evolved different solutions to the problems of foraging under risks of predation and disturbance.