Introduction

High-head storage hydropower plants often have detrimental ecological impacts on downstream river ecosystems because of their high-volatile electricity production, which cause severe daily and sub-daily downstream fluctuations in discharge and water levels due to so-called ‘hydropeaking’ (hereafter, HP; Bruder et al., 2016; Young, Cech, & Thompson, 2011). Those hydropower operations influence the flow regime not only during peak flow. Due to the temporal water storage in the reservoir, the river segment located between storage and hydropower plant experiences relatively constant base-flow conditions, so-called residual flow (hereafter, RF). An increase of 1.2 TWh is forecasted by 2050 in the energy storage capability of the Swiss hydropower reservoirs, corresponding to 20% of today’s energy storage capacity (Boes et al., 2021). Consequently, further ecological impacts within riverine ecosystems are expected with an increase in HP operation.
In alpine rivers, HP often occurs in high frequency and its intensity can be characterized by several hydrological parameters (Greimel et al., 2016; Li & Pasternack, 2021). HP operation in the up-ramping phase often leads to abrupt increase in discharge and related hydraulic forces (e.g., flow velocity, bed shear stress). This can promote unintentional and mechanical dislodgement of aquatic macroinvertebrates (hereafter, MIV) from the substrate and their downstream transport in the water column until passive (e.g., in hydraulic dead zones) or active (behavioral or through morphological adaptations) exit of the drift (Ciborowski, 1987; Naman, Rosenfeld, & Richardson, 2016). The causes for MIV drift by HP operation are manifold and complex and often strongly taxon- and trait-specific (e.g., Kjærstad, Arnekleiv, Speed, & Herland, 2018; Leitner, Hauer, & Graf, 2017). The following parameters are of major importance regarding HP-induced passive MIV drift: up-ramping rate, flow magnitude and amplitude, flow velocity, and to a lesser extent also Froude number and bed shear stress (Bruno, Cashman, Maiolini, Biffi, & Zolezzi, 2016; Bruno, Maiolini, Carolli, & Silveri, 2010; Gibbins, Vericat, Batalla, & Buendia, 2016; Imbert & Perry, 2000; Miller & Judson, 2014; Schülting, Feld, Zeiringer, Huđek, & Graf, 2018a; Timusk, Smokorowski, & Jones, 2016). Moreover, sudden variations in water temperature (i.e., thermopeaking) may induce behavioral MIV drift (Bruno, Siviglia, Carolli, & Maiolini, 2013; Schülting, Feld, & Graf, 2016), and re-suspension of fine sediments increases turbidity and clogging risk (Hauer, Holzapfel, Tonolla, Habersack, & Zolezzi, 2019) further affecting MIV (Bo, Fenoglio, Malacarne, Pessino, & Sgariboldi, 2007; Crosa, Castelli, Gentili, & Espa, 2010; Jones et al., 2012). In addition to drift, stranding is a possible consequence of HP operations. MIV stranding has been documented, also if not unequivocally, to be related to down-ramping rate, flow magnitude, amplitude and rate, flow velocity and the extension of the dewatering area (Kroger, 1973; Perry & Perry, 1986; Tanno, Wächter, & Gerber, 2021; Tanno, Wächter, & Schmidlin, 2016). Tanno et al. (2021) further found a positive correlation between MIV drift and stranding. In a similar vein as for drift, water temperature and turbidity are probably likely to influence stranding. For example, water temperature has been shown to influence stranding risk of fish (e.g., Halleraker et al., 2003). Besides the short-term effects of MIV drift and stranding, long-term effects of altered hydromorphological habitat conditions in HP rivers are reported to affect colonization patterns of benthic populations (Bretschko & Moog, 1990; Cushman, 1985; Kjærstad et al., 2018). Morphological heterogeneity, for instance, is known to be crucial in providing diverse habitats and refuges for MIV communities under HP conditions (Hauer, Holzapfel, Leitner, & Graf, 2017). The complex interplay of drift, stranding and altered hydromorphological habitat conditions caused by HP operation, most likely contribute to reductions in MIV abundance and biomasses (Céréghino, Cugny, & Lavandier, 2002; Céréghino & Lavandier, 1998; Elgueta et al., 2021; Leitner et al., 2017; Moog, 1993). Another possible consequence is the alteration of the MIV community composition due to the evolution of specific behavioral (e.g., mobility, ability to regain a foothold, sinking postures) and morphological (e.g., body shape, hooks) traits as well as life history strategies. HP may, for example, contributes to a selection of rheobiont and rheophilic taxa (Bretschko & Moog, 1990; Cushman, 1985; Ruhi, Dong, McDaniel, Batzer, & Sabo, 2018) against limnophilic taxa or of taxa associated with lentic and substrate surface areas (Graf et al., 2013; Leitner et al., 2017; Ruhi et al., 2018; Schülting et al., 2018a; Schülting, Dossi, Graf, & Tonolla, 2021). Community changes due to HP may consequently affect the local food web structure and the ecological functioning of river systems (e.g., Holzapfel, Leitner, Habersack, Graf, & Hauer, 2017; Pearce et al., 2019).
The complex interactions between MIV drift, stranding and established benthic community remain poorly understood, and field studies are rare (but see for example Miller & Judson, 2014; Tanno et al., 2021; Timusk et al., 2016). Such a comprehensive evaluation is yet of importance for an exhaustive understanding of HP-related impacts on MIV and finally for implementation of associated mitigation measures and sustainable operation of HP power plants (Bruder et al., 2016; Moreira et al., 2019; Tonolla, Bruder, & Schweizer, 2017).
The main goal of our field experimental study – conducted in three HP-regulated Swiss rivers – was therefore to quantify the short-term effects of three HP scenarios with different intensity (simulated through an increase in flow amplitude and up-ramping rate) on MIV drift and stranding in addition to the long-term HP effects on the established community composition. As comparison, MIV drift and community composition was quantified in HP-unaffected RF reaches upstream. We hypothesize that: (i) drift increases during the HP scenarios compared to base flow and to drift in the RF reaches as well as a positive relationship between drift and stranding; (ii) HP-intensity and associated hydraulic forces, as summarized by pre-selected environmental variables, can explain any observed differences in drift and stranding propensity; (iii) compared to the RF reaches with almost constant hydrological conditions, flow fluctuations due to HP operation lead to lower benthic densities and a different MIV community composition, linked to drift and stranding. Finally, we expected taxa-, trait- and river-specific patterns.