Integrative analysis of stressor gradients reveals multiple discrete trait-defined axes underlie community assembly

The generalizable functional attributes of organisms (traits) relate strongly to their environment across multiple levels of biological organization, making trait-based approaches a powerful mechanistic framework to understand species distributions and community composition in relation to environmental change. To investigate how a wide range of stressor types shape stream macroinvertebrate communities, we conducted an integrative analysis using community and taxon trait information across drying, flooding, eutrophication, fine sediment, and acid mine drainage (AMD) gradients. Each gradient spanned relatively unimpacted to severely impacted sites. To characterize community change in response to stressors, we incorporated abundance-based trait information from all stressor gradients in a single trait-based ordination (nonmetric multidimensional scaling), defining the trait space within which each stressor gradient acted. We hypothesized that different stressors would apply different environmental filters, moving communities along distinct axes in trait space and resulting in communities with definable trait combinations. Particularly strong relationships were associated with anthropogenically derived stressors (fine sediment, eutrophication, and AMD) compared to natural stressors (drying and flooding). Anthropogenic stressors instigated significant movement of communities along multiple axes in trait space, likely driven by limited adaptation to these novel stressors. We demonstrate that trait-based analysis of communities across multiple stressor gradients can support a more comprehensive understanding of how community composition changes than taxonomic methods or investigation of a single stressor type, and could underpin community-focused management actions.