Aquatic macroinvertebrate community responses to wetland mitigation in the Greater Yellowstone Ecosystem

Wetlands are critical components of freshwater biodiversity and provide ecosystem services, but human activities have resulted in large‐scale loss of these habitats across the globe. To offset this loss, mitigation wetlands are frequently constructed, but their ability to replicate the functions of natural wetlands remains uncertain. Further, monitoring of mitigation wetlands is limited and often focuses exclusively on vegetation and physical characteristics. Wetland fauna are assumed to be present if suitable habitat restoration is achieved, but this assumption is rarely tested. We used the macroinvertebrate community as a proxy for wetland function to compare recently created mitigation wetlands, natural wetlands impacted but not destroyed by road construction activity, and unimpacted reference wetlands along a highway corridor in the Greater Yellowstone Ecosystem. Unlike most other studies of invertebrate communities in created wetlands which have occurred in warm climates, our study area has a cold temperate climate with short growing seasons. We estimated macroinvertebrate taxonomic richness and used linear models to test for effects of wetland design features (wetland age, isolation, depth, vegetation, size, and pH) on invertebrate richness. We also used non‐metric multidimensional scaling to visualise differences in community composition among wetland types and used indicator species analysis to determine which taxa were causing observed differences. Taxonomic richness of macroinvertebrates was lower in created wetlands than impacted or reference wetlands, whereas richness was similar in impacted and reference wetlands. Wetland age was positively correlated with taxonomic richness. The amount of aquatic vegetation in wetlands had the greatest influence on taxonomic richness, so that recently created wetlands with little vegetation had the simplest invertebrate communities. Community composition of invertebrates in created wetlands also differed from community composition in reference and impacted wetlands. Most notably, created wetlands lacked some passive dispersers that were common in other wetland types, although we found no relationship between taxonomic richness and wetland isolation. Overall, constructed wetlands had diminished and altered macroinvertebrate communities relative to reference and impacted wetlands, suggesting that periods in excess of 5 years may be required for wetland mitigation projects in cold temperate climates to attain fu