Functional redundancy of stream macroconsumers despite differences in catchment land use

1. The conversion of forested landscapes to agriculture and, increasingly, to suburban and urban development significantly affects the structure and function of both terrestrial and aquatic ecosystems. While a growing body of research is examining how biotic communities change in response to human alteration of landscapes, less is known about how these changes in community structure affect biotic interactions. 2. The objective of this study was to examine top-down control by macroconsumers (fish and crayfish) across a human-impacted landscape. We predicted that changes in stream macroconsumers and physicochemical characteristics associated with increased catchment development (e.g. decreased abundance of fish that are obligate benthic invertivores, increased sedimentation) would diminish top-down control of benthic insects. We expected that effects on algal assemblages would be more variable, with increased top-down control at sites dominated by algivorous fish and diminished control elsewhere. To test these predictions, we experimentally excluded fish and crayfish from areas of the bed of five streams whose catchments ranged from 100% to < 50% forested, and examined the effects of exclusion on benthic insects and algae. 3. Despite cross-site differences in physical, chemical and biological characteristics, the outcome of our experiments was consistent across five sites representing a range of catchment development. Across all sites, macroconsumers reduced total insect biomass, largely due to decreases in Chironomidae and Hydropsychidae larvae. Macroconsumers also affected algal assemblages, reducing chlorophyll-a and the proportion of upright and filamentous diatoms (e.g. Melosira, Cymbella) but increasing the proportion of adnate diatoms (e.g. Achnanthes) across all sites. 4. We expected that differences in factors such as macroconsumer assemblage composition, nutrient and light availability and sedimentation would result in variable responses to macroconsumer exclusion in the five streams. Contrary to these expectations, only one response variable (ash-free dry mass) showed a statistically significant interaction (i.e. site x exclusion) effect. Most responses to exclusion were relatively consistent, suggesting functional redundancy in assemblages of macroconsumers among the sites despite differences in catchment land use.