Can biological traits of stream invertebrates help disentangle the effects of multiple stressors in an agricultural catchment?

Biological traits, which may give insights into the mechanisms driving the distribution of organisms along gradients of stressor intensities, have been proposed as a tool for disentangling the effects of multiple stressors acting simultaneously on scales ranging from climatic region to river basins, valleys, reaches and microhabitats. However, the combined effects of farming intensity and flow reduction on biological traits of stream invertebrates remain to be studied. We assessed the benthic invertebrate community and physicochemistry at 43 stream sites along gradients of farming intensity (0-95% of the catchment in intensively managed grassland) and water abstraction (0-92% streamflow reduction). Using general linear models and an information-theoretic approach, we studied individual and combined effects of agricultural stressors on invertebrate traits and community composition. Traits often followed predictable patterns along stressor gradients, and non-additive interactions between paired stressors were common. Farming intensity was more frequently related to life-history, resistance and resilience traits, whereas water abstraction was correlated more often with general biological traits such as feeding habits, dietary preference and respiration. Further, traits and traditional measures of community structure, such as taxon relative abundances and community indices, offered a similar level of distinction along the gradients of stressor intensities. Our findings indicate that invertebrate traits can differentiate the effects of multiple stressors and provide insights into potential mechanisms. At the landscape scale, farming intensity exerted stronger effects via invertebrate habitat quality and water abstraction via food availability. At the reach scale, both fine sediment and nutrients affected habitat quality, whereas nutrients showed more marked effects via food availability. Finally, we propose a suite of traits that may provide the strongest differentiation of stressor intensities.