The Effects of Improved Water Quality on Fish Assemblages in a Heavily Modified Large River System

A long history of human alterations has affected the hydrology, physical habitat and water quality of most large river ecosystems. For more than a century, the Illinois River Waterway has been subject to channelization, damming, dredging, agricultural runoff and industrial and municipal effluents. This study evaluates how subsequent improvements in water quality have influenced long-term changes in fish assemblages (1983–2010). We used five metrics to characterize the changes in fish assemblages. These metrics depicted shifts in the abundance and biomass of predatory and native fishes and species richness. Random forests (RF) and multiple linear regressions (MLRs) were used to relate the fish metrics to individual water quality and weather variables, with weather primarily used to account for inter-annual variation. Model performances varied spatially and among fish metrics (0 ≤ pseudo-R² ≤ 0.73 for RF; 0.10 ≤ adjR² ≤ 0.88 for MLR), but dissolved oxygen, un-ionized ammonia and water clarity were often the best predictors. As the distance downstream of major pollutant sources increased, water quality became less important for explaining the changes in fish metrics and weather more important. These results indicate that water quality improvement largely accounts for fish assemblage recovery in the river system, although within some reaches we examined, weather had substantial compounding effects. The results could be used to prioritize water quality variables for long-term monitoring and aid in predicting fish assemblage responses to future changes in water quality and climate.