Fish community dynamics (1985–2010) in multiple reaches of a large river subjected to flow restoration and other environmental changes

Abstract

Summary To assess how climate warming, flow regulation and flow restoration affect community dynamics, we analysed long‐term (15–25 years) series of fish community data from four restored and two unrestored reaches of the Rhône River. Environmental variables (low and high flow hydraulics, temperature, sediment flushing operations) were measured in all reaches. Five of the reaches were bypassed by artificial channels, downstream from diversion dams whose construction finished in 1984. Minimum flows in four of these reaches were restored (i.e. increased by factors up to 10) between 2000 and 2006. We examined hypotheses concerning the inter‐annual response of fish community guilds and size structure to flow restoration, climate warming and dam completions. We also performed a principal component analysis (PCA) to explore observed community changes at the river scale (all reaches pooled). We used mixed‐effects linear models to infer potential environmental effects not included in our hypotheses. Our results indicated an inter‐annual effect of flow restoration in the two reaches where minimum flows were most modified (higher percentages of ‘midstream’ species preferring deep and fast water). They also revealed effects of climate warming in the two warmed reaches (higher percentages of southern and small individuals and higher total density). These effects were consistent with our hypotheses. In contrast, dam construction had no consistent effects across reaches. The first PCA axis indicated that the longitudinal organisation of fish communities at the river scale was unchanged, but independent inter‐annual effects of flow restoration and climate change were apparent along the second and third axes. Observed annual variations within sites were weakly related to annual variation in environmental variables. This may be due to the presence of long‐term or time‐lagged environmental effects and the difficulty of estimating community structure of large river fishes accurately enough for this kind of analysis. Fish community responses to flow restoration and other environmental changes may be consistent across multiple reaches when the degree of environmental change is taken into account. Our results support the effectiveness of flow restoration on communities of large rivers subjected to multiple environmental changes.