Assessing macroinvertebrate community response to restoration of Big Spring Run: Expanded analysis of before‐after‐control‐impact sampling designs

Stream restoration projects utilize a variety of approaches to improve conditions for aquatic organisms or enhance ecosystem function. Alterations to abiotic conditions to enhance certain ecosystems services may not lead to concurrent changes in the benthic macroinvertebrate community indicative of improved stream health. Big Spring Run was the location of a novel restoration project to recreate an anabranching “wet meadow” habitat typical of precolonization conditions without the primary goal of restoring a macroinvertebrate community characteristic of single‐channel lotic systems. We examined the effect of the restoration on the macroinvertebrate community using a multivariate analysis of assemblage composition, a before‐after‐control‐impact (BACI) approach, and an assessment of potential aerial migrants. We also examined subsets of the data using a BACI approach that represented restricted sampling designs often employed in stream restoration projects. Benthic macroinvertebrates were collected in the Spring twice prior to restoration (2010 and 2011) and 3 years after restoration (2012–2014). Adult stream insects were collected in 2014. Analyses of benthic macroinvertebrates and adult insects using the full dataset suggested that restoration had no effect on the macroinvertebrate community due to poor in‐stream conditions likely from sediment deposition following restoration. Dispersal barriers are likely acting as a secondary constraint on recolonization. Analyses using subsets of the data demonstrated that reference site quality and sampling extent can alter conclusions from a BACI approach. We found that a holistic approach using multiple lines of evidence required a nuanced approach to interpreting the data but was also informative for assessing project success. Robust monitoring protocols are likely the best approach for producing convincing results through a single line of evidence. The additional BACI analyses performed for this study, however, allowed the modest sampling regime employed to generate a broad narrative demonstrating that the macroinvertebrate assemblage did not respond to this type of restoration. Thus, we believe the holistic approach we employed can strengthen assessments of stream restoration projects when resources for monitoring are limited.