Predictors of chironomid abundance following dam removal

Dam removal is increasingly used as a river restoration measure, though the effects of dam removal on river ecosystems are often poorly understood. The goal of this study was to examine the changes in chironomid abundance in a lowland river undergoing dam removal. Sampling was undertaken in the River Drzewiczka (Poland) from multiple habitats downstream: pool, standing water, macrophytes, bank, and riffle, for a full year before (DI), during (DII) and following (DIII) dam removal. Larval chironomid abundance was quantified from direct counts in samples, along with biotic and abiotic habitat variables. Using an information theory approach, we compared a series of a priori alternative models to identify the biotic and abiotic predictors of chironomid abundance. The best-fitting a priori model to predict the density of Chironomini was one that comprised an interaction between inorganic substrate index (SI) and dam removal. The model showed a significant positive association between Chironomini density and SI in DIII in comparison with phases DI and DII. The best-fitting a priori model to predict the density of Orthocladiini was a model that comprised interactions between inorganic substrate index and dam removal, and benthic particulate organic matter (BPOM) and dam removal. The model demonstrated a highly significant difference in the relationship between SI and Orthocladiini density in phase DIII. Following restoration (DIII), this relationship was strongly positive, in contrast to phases DI and DII, when the relationship was weakly negative. Moreover, there was a significant negative relationship between Orthocladiini density and BPOM over the three phases of restoration. These findings indicate that dam removal can significantly impact the biota of river ecosystems and that these effects should be considered when planning dam removal projects.