Testing predictions of changes in fish abundance and community structure after flow restoration in four reaches of a large river (French Rhône)

摘要: 1. Projects that restore river flows can be considered as in situ experiments and should used to test predictions of the effects flow changes on fish populations communities. However, restoration projects often lack appropriate monitoring replication. The Rho^ne project has included repeated changes, four bypassed reaches river, where increase minimum daily varied from minimal change a tenfold increase. 2. Fish communities (>55 000 individuals 36 species) were electrofished at nine sites main channels reaches, for 2–9 years before 5–10 after restora- tion. An instream hydraulic habitat model, published based observations microhabitat preferences independent was applied predict density 14 species accounted 94% total abundance. 3. In two considerably increased (fivefold tenfold), abundance preferring fast-flowing deep microhabitats by factors 1.9 2.4, respectively, whereas other strongly decreased. Predicted made using model these agreed with several involved species. contrast, less, observed weak less related predictions. 4. Hydraulic models predicted also explained community responses changed flows. Ten first restoration, our results suggest generated perennial structure, reversing patterns prior restoration.