MIV sampling
The benthic community was sampled during base flow with Hess samplers (sampling area 0.045 m2, mesh size 500 µm,) 2 – 5 m upstream of each drift net. At each sampling site, three benthic samples were collected transect-wise (from the bank margin to the in-stream area) along a flow velocity gradient to include taxa of different current preferences (Figure 2a). Benthic samples in the HP reaches were collected four times a day (before each HP scenario and at the end of the day; Figure 2b), whereas they were taken twice a day (beginning and end of the day) in the RF reaches.
Drifting MIV were collected at each sampling site with drift nets (frame size 53 x 27 cm, mesh size 500 µm) fixed with thick rebars on both sides. The long side of each net was placed on the riverbed. Samples in the HP reaches were taken separately (nets exchanged) during base flow (exposure time 60 min) and during each HP phase: up-ramping (UR), first and second part of the peak (P1, P2; each 30 min) and down-ramping (DR; Figure 2b). In the RF reaches, MIV drift was sampled for approximately 60 min, simultaneously with the induced peak flows in the HP reaches.
Stranding MIV were collected according to Tanno et al. (2016). Four flat nets (net size 50 x 50 cm, mesh size 500 µm) were installed at each sampling site nearby the drift nets on the dewatering area (temporally wetted area, dry at base flow and wetted during peak flow; Figure 2a): two nets directly at the waterline and two nets 1 m away from the waterline and 0.5 m offset from the other nets. Ten stones of the dominant grain size were placed on each flat net to increase roughness and thus better mimic the substrate condition of the dewatering area. Stranded MIV on the nets and on the stones were collected after each HP scenario (Figure 2b). In the RF reaches, stranding was not assessed, due to stable flow conditions.