A laboratory study of the effects of shelter availability and invasive crayfish on the growth of native stream fish

1. The alteration of stream habitats by urbanisation reduces the availability of shelter. Reduced shelter availability may increase both predation risks and metabolic costs, negatively affecting the growth performance of stream fish. Although urbanised streams often allow the establishment of invasive species, the additive or interactive effects of shelter availability and invasive species are rarely explored. The invasive red swamp crayfish (Procambarus clarkii) is a strong shelter competitor and predator in streams. 2. We conducted a stream‐channel experiment to investigate how shelter availability and this invasive crayfish species affect the growth performance (measured as the loss in body mass under starved conditions) of two native benthic species, Japanese crucian carp (Carassius auratus complex) and an endangered cobitid fish (Lefua echigonia). 3. We found that both the absence of shelter and the presence of crayfish increased body mass loss in the crucian carp. However, the growth performance of the crucian carp was improved by the presence of shelter except in the presence of crayfish, in which case the shelter available did not ameliorate the negative effect of the crayfish on growth performance (i.e. an interactive effect). This result suggests that crayfish may affect the growth performance of the crucian carp through chemical cues as well as through shelter occupancy. 4. In contrast, shelter availability is the primary factor affecting the body mass loss of the cobitid fish. However, the growth performance of these fish was unaffected by the presence of crayfish. The change in the growth performance of the cobitid fish in response to the absence of shelter was greater than that of the crucian carp. This finding suggests that the cobitid fish may be more vulnerable than the crucian carp to a reduction in shelter availability. 5. Our study demonstrates that reduced shelter availability and/or crayfish invasion can have significant, indirect negative effects on the growth performance of native fish, but whether those effects are interactive might vary depending on the species. Our findings have clear implications for stream restoration and habitat assessment.