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

Abstract

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

How sampling influences the statistical power to detect changes in abundance: an application to river restoration

“…The number of significant changes varied strongly across reaches. For example, when compared with fish taxa with reduced abundance at Pierre-B enite according to our tests (Barbatula barbatula, Lepomis gibbosus), all taxa with increased abundance (Alburnus alburnus, C. nasus, Perca fluviatilis, Alburnoides bipunctatus; Table 1) were also expected to be favoured by the restoration according to the habitat model of Lamouroux & Olivier (2015, their Fig. At Pierre-B enite where more significant changes were observed, the total number of sampling points was lower (N = 611) than in other reaches (N between 1203 and 2083, cf.…”

“…We analyse here the microhabitat abundance variations of 13 fish taxa considered by the hydraulic habitat models of Lamouroux & Olivier (2015) (Table 1; Telestes souffia was excluded due to very low abundance) and of all 35 invertebrate taxa considered by the habitat model of M erigoux et al (2015) ( Table 2). We used the same data set as in these articles but removed invertebrate dredge samples made in the channel centre at Pierre-B enite that were made at a much larger spatial scale than the other samples.…”

“…The 5th and 95th percentiles ofŝ were 0.04 and 15.3. Observed changes in abundance cannot be directly compared with the results of M erigoux et al (2015) and Lamouroux & Olivier (2015), because their hydraulic habitat model deals with average fish densities across microhabitats and not abundance itself, due to technical constraints of their hydraulic preference models. For fish, the number of significant changes was generally related to the magnitude of the flow restoration, because 6 of 13 taxa had significant abundance changes at Pierre-B enite (where the target minimum flow was multiplied by 10 after restoration), 5 of 13 had significant changes at Chautagne (target minimum flow multiplied by 5), 0 of 13 had significant changes at Belley (target minimum flow multiplied by 2.4) and 1 of 13 had significant changes at Br egnier-Cordon (target minimum flow unchanged).…”

“…Gutreuter et al, 2010;Potts et al, 2012), notably in systems such as large rivers where only a small proportion of the total available surface area can be sampled (e.g.~0.1% of the total reach area was sampled on each fish survey in the Rhône; Lamouroux & Olivier, 2015). Gutreuter et al, 2010;Potts et al, 2012), notably in systems such as large rivers where only a small proportion of the total available surface area can be sampled (e.g.~0.1% of the total reach area was sampled on each fish survey in the Rhône; Lamouroux & Olivier, 2015).…”

Realised and predicted changes in the invertebrate benthos after restoration of connectivity to the floodplain of a large river

“…Following M erigoux et al (2015) and Lamouroux & Olivier (2015), predicted changes (PC) were expressed relative to the observed pre-restoration density for a given taxon: Following M erigoux et al (2015) and Lamouroux & Olivier (2015), predicted changes (PC) were expressed relative to the observed pre-restoration density for a given taxon:…”