Throughout the globe, weirs and dams, often combined with hydropower use, form barriers for migratory fish species (Gehrke et al., 2002; Dugan et al., 2010; Pelicice et al., 2015). The resulting negative effects of blocked migration routes, for example, on the accessibility of suitable spawning sites, annual recruitment success or genetic exchange, are well-known (e.g., Antonio et al., 2007; Fukushima et al., 2007; Liermann et al., 2012).

Generally, a distinction needs to me made between upstream and downstream migration of fish since both follow different principles (Geist, 2021). For decades, intensive efforts have been made to make migration barriers passable for upstream migrating fish species through provisioning alternative corridors, e.g., various types of fish ladders (Banks, 1969; Evans and Johnston, 1980). In contrast, the great importance of suitable bypasses that enable downstream moving fish to pass a barrier without being seriously injured was underestimated for a long time (Larinier and Travade, 2002).

Particularly the passage of conventional hydropower turbines (e.g., Francis, Kaplan turbines) poses a high injury and mortality risk for fish (Pracheil et al., 2016; Algera et al., 2020; Mueller et al., 2022). As fish primarily follow the main current during their downstream migration to save energy (Williams et al., 2012; Anderson et al., 2015) and the bulk of discharge typically flows through the turbines at most hydropower plants (HPPs), a large proportion of active or passive downstream moving fish also follows this route (Fjeldstad et al., 2018). To prevent fish from passing through the turbines and to direct them to alternative downstream corridors, so-called “bypasses”, various types of physical and behavioral barriers are used that are commonly installed upstream of the turbine inlet. Typical physical barriers are bar or mesh screens designed to block fish from swimming through whose body dimensions are larger than the clearance between the openings (Ebel, 2013; Schwevers and Adam, 2020). In contrast, behavioral barriers are intended to induce an avoidance or scaring reaction in fish through external stimuli such as electricity, light, sound or air bubbles (e.g., Schilt, 2007; Noatch and Suski, 2012). Recently, so-called “hybrid” barriers have also been developed, where physical barriers (e.g., bar screens) are combined with electric fields to provide an additional behavioral barrier (Tutzer et al., 2021, 2022; Haug et al., 2022).

Corridors specifically built for downstream migrating fish, such as surface, bottom or full-depth bypasses, should provide an alternative route to the turbine corridor and transport fish unharmed into the tailwater. In addition, structures to divert flotsam or floodwater to the tailwater such as flushing channels, spillways, periodically opened flap gates and sluices as well as fish passes that were actually built to enable upstream migration can also be used by downstream moving fish (e.g., Scruton et al., 2007; Travade et al., 2010; Egg et al., 2017). Recent empirical evidence suggests that larger fish than expected from modelling approaches can pass through physical screens at hydropower inlets (Knott et al., 2023), emphasizing the necessity of better understanding the roles of alternative corridors for their safe downstream passage.

However, there are also HPPs where fish protection measures and bypasses for downstream migrating fish are deliberately not in place [e.g., at innovative Archimedes screw turbines or very low-head (VLH) turbines], as it is assumed that the installed turbine technology should cause less mortality and injuries than conventional turbine types and therefore fish passage via the turbine corridor is intended (Anderson et al., 2015; Simmons and Lubitz, 2021).

In general, bypass efficiency depends not only on the available discharge, but also on bypass dimensioning, spatial proximity to fish guiding structures, location in the water column (top, middle, bottom) and prevailing hydraulic conditions (Larinier and Travade, 2002; Katopodis, 2005). Since these variables can differ considerably among sites, even the efficiency of state-of-the-art bypass systems can vary strongly depending on the local conditions.

For economically important fish species such as salmon and eel, there is already a large number of studies on the effectiveness of fish guiding structures and bypasses for downstream migration, mostly relying on telemetry or sonar detection (e.g., Gosset et al., 2005; Havn et al., 2017; Nyqvist et al., 2018). In contrast, research on corridor usage during downstream passage at hydropower facilities of fish species with little or no perceived economic value (particularly potamodromous species) is severely underrepresented. However, knowledge on the downstream movement behavior of the whole fish community is important, as many potamodromous fish species that are not known to be migratory species also move downstream and inevitably have to pass HPPs (Katopodis, 2005; Knott et al., 2020). Since there can be large differences between fish species in terms of morphology (e.g., body size, body shape) and behavior (e.g., habitat use, swimming behavior), it is often not known which corridors the various fish species preferentially use to pass barriers such as HPPs. Consequently, studies are needed that assess the corridor usage for a broad range of species in order to identify possible deficits and to improve bypass efficiency. This would help to better protect not only economically important species such as the European eel (Anguilla anguilla L.) from negative ecological effects of hydropower, but also other species such as the small-bodied bullhead (Cottus gobio L.), which is also threatened and protected by law (European Commission, 1992).

In this study, we investigated the corridor usage of downstream moving fish at four hydropower sites by a net-based monitoring during different seasons. For this purpose, downstream moving fish were captured and examined after turbine passage and the passage of alternative downstream corridors such as surface and bottom bypasses, flushing channels, flap gates and fish passes.

Specifically, we hypothesized that (i) the majority of downstream moving fish are guided towards alternative bypasses due to the physical and behavioral barrier function of the fine screens installed in front of the turbine inlets and that (ii) the sampled downstream corridors at a specific site differ in terms of number, size and species composition of downstream moving fish due to differing fish communities and species-specific behavior.

During the study period, 6,646 downstream moving fish from 42 species were caught after passing the different corridors of the investigated HPPs. The most frequently caught species were spirlin (Alburnoides bipunctatus L.), bleak (Alburnus alburnus L.) and roach (Rutilus rutilus L.), summing up to 42% of the total catch of individuals (Supplementary Table S3). TLs of all captured fish ranged from 2 cm (Rudd Scardinius erythrophthalmus L., Tench Tinca tinca L., Topmouth gudgeon Pseudorasbora parva Temminck and Schlegel) to 85 cm (European eel) (a.m.±s.d.: 10.2 ± 5.6 cm). Small-bodied species and juvenile fish <20 cm TL accounted for 96% of the total number of captured fish, 3% were between 20 cm and 40 cm TL and 1% were >40 cm (mostly European eels).

For an objective evaluation of the impacts of HPPs on fish welfare, not only fish species-specific mortalities and injury rates, but also the efficiency of fish guiding structures and the actual use of the different downstream migration corridors need to be considered. In the present study, the corridor usage of downstream moving fish was investigated at four small-scale HPPs with different concepts to prevent turbine entrainment and to bypass fish. The results indicate that despite existing bypass systems and fine screens with 15 mm and 20 mm bar spacing installed upstream of the turbine inlets, a large proportion of fish still passed the turbines, albeit at a slightly lower ratio than expected from the proportions of discharge through turbines versus alternative corridors. Physical exclusion effects of the bar screens and behavioral effects due to the arrangement of the fine screens (angled, inclined or horizontal at the river bottom), which are supposed to prevent passage and guide fish to bypasses, appear to be mostly low at the assessed HPPs. However, standardizing the number of downstream moving fish to a specific discharge revealed that the discharge-related downstream movement rates via the bypasses were mainly higher than via the turbine corridor. This suggests that increasing the bypass discharge, e.g., by enlarging the openings, could increase the actual bypass efficiency. Moreover, recent studies suggest that a combination of physical barriers such as bar screens with behavioral barriers (e.g., electric fields) could increase fish protection from turbine entrainment and the guiding effect towards bypasses (cf. Tutzer et al., 2021, 2022; Haug et al., 2022). However, since fish downstream movements can also take place at very low water temperatures in late autumn/winter and reduced metabolic rates potentially increase the proportion of fish drifting passively with the main current (Wiesner et al., 2004), it would be worthwhile for future studies to investigate whether the turbine entrainment risk is increased under these conditions.

It is widely assumed that fish can only pass bar-structured barriers if the body width or height is smaller than the bar spacing (Ebel, 2013; Schwevers and Adam, 2020). However, recent studies have shown that even much larger fish than expected by modelling or rules of thumb can pass through such bar screens (cf. Knott et al., 2023). This is probably due to the fact that fish can also squeeze through barriers and damage to the screen caused by debris or manufacturer-related deviations in bar spacing, which can be found in regular operation under field conditions, allow larger fish than expected to pass. In addition, the poor barrier function of the assessed fine screens may also be explained by the fact that the majority of the fish population in the study rivers consists of small-bodied species and juveniles (cf. Knott et al., 2020), for which the fine screens did not act as a physical barrier. This is also supported by the finding that 96% of the captured fish over all corridors in this study were <20 cm in TL.

In addition to the primary protective effect as a physical barrier, fine screens can also lead to an avoidance reaction or a guiding effect through optical or hydraulic stimuli, whereby physically permeable barriers can also be effective (Ebel, 2013; Harbicht et al., 2022). However, this behavior-influencing effect no longer works at high approach velocities at the screen (≥0.5 m/s according to DWA, 2005). High current velocities at the fine screen can be a reason for the high number of screen and turbine passages, especially in Baiersdorf-Wellerstadt, where the threshold value of 0.5 m/s was permanently exceeded. As a result, behavior-influencing effects that could also keep smaller individuals from passing the screen or guide them towards bypasses no longer apply. Moreover, fish can be injured or killed if the current velocities at the screen are too high, as they can no longer actively avoid the screen, but are pressed against or between the bars (Deng et al., 2010; Brown et al., 2014; Amaral et al., 2018).

The low proportion of fish that moved downstream via the bypasses located in the immediate vicinity of the fine screens in Lindesmuehle (fish slide and flap gate) and Baiersdorf-Wellerstadt (flushing channel) suggests that the guiding effect towards these corridors was low.

In general, bypass efficiency is, besides proper location and hydraulic conditions at the entrance, very much dependent on the available discharge. The higher the discharge, the more fish typically enter a bypass (Turnpenny et al., 1998) which is also supported by the findings of this study. Commonly, a minimum of 2% of turbine discharge is recommended for “good” bypass efficiency (Klopries et al., 2018: 2%–35%; Larinier and Travade, 2002: 2%–10%; Odeh and Orvis, 1998: 2%–5%). Bypass discharge via the fish slide and the flap gate in Lindesmuehle each amounted to 0.1% of the turbine discharge and via the flushing channel in Baiersdorf-Wellerstadt to 0.4% of the turbine discharge and was thus well below these recommendations. However, “good” bypass efficiency can only be achieved at low discharges if guiding structures, location and hydraulic conditions are ideal (Klopries et al., 2018). In addition to insufficient discharge, the fish slide and the flap gate in Lindesmuehle probably also have deficits in terms of detectability for fish. The fish slide was predominantly used by bottom-oriented species such as bullhead, ruffe and European eel, indicating that the bottom entrances were preferentially used over the near-surface opening. Results from Egg et al. (2017) also suggest that the functionality of the bottom entrances is reduced particularly in late autumn, as they can become clogged by leaf debris that is mobilized during increased flows. The low bypass efficiency of the flap gate in Lindesmuehle can further be explained by a potentially reduced detectability for fish due to the periodical operation compared to permanently opened corridors. In addition, the periodic bypass opening causes an immediate change in hydraulic conditions (e.g., during the screen cleaning and flushing process; cf. Williams et al., 2012), which can scare fish away (Schwevers and Adam, 2020). Moreover, an unfavorable bypass location may not only reduce detectability, but also lead to an increased predation or impingement risk (Turnpenny et al., 1998).

Deficits for fish to enter the flushing channel in Baiersdorf-Wellerstadt seemed to be the low overflow height of the downstream end of the vertically inclined fine screen (<5 cm) during the study period, which made it difficult for larger fish to get into the flushing channel. Nevertheless, considerably larger fish moved downstream via the flushing channel at this site than via the turbines and the fish pass. It was observed that larger individuals of common carp and Prussian carp were impinged at the 15 mm fine screen and transported into the flushing channel by the screen cleaner. These individuals were physically unable to pass the fine screen and probably could not escape the high approach velocities at the screen (0.97 m/s on average).

It is remarkable that in Lindesmuehle and Baiersdorf-Wellerstadt more fish moved downstream via the fish passes than via other bypasses. This can be explained by the considerably higher discharge through the fish passes compared to the other bypasses, whereby the upstream entrance of the fish pass in Lindesmuehle was about 70 m away from the HPP on the opposite bank (Figure 1) and its entrance was therefore expected to not be easily discovered by downstream moving fish. However, due to the applied fish catching method, it was not possible to differentiate whether the captured fish moved downstream from the headwater or used the fish passes as habitat before moving downstream. Particularly for small-bodied fish species such as spirlin, bullhead and European minnow and juvenile fish, fish passes not only serve as a migration corridor, but also as a suitable habitat (Pander et al., 2018).

Großweil was the only site where more fish moved downstream via the surface- and bottom-near bypasses than via the turbine corridor. In addition to a considerably higher share of bypass discharge of 9.9% of the turbine discharge compared to Lindesmuehle (0.2%) and Baiersdorf-Wellerstadt (0.4%), site-specific characteristics also seem to have affected the corridor usage of downstream moving fish. For example, this finding was influenced by a dominance of European grayling (68% of all captured fish at this site with TLs mostly ranging from 10 to 13 cm), which is an open water-oriented species that prefers surface-near corridors for downstream migration (Knott et al., 2019).

At Hoellthal, similar to Lindesmuehle and Baiersdorf-Wellerstadt, most of the captured fish moved downstream via the Kaplan turbine corridor. The lack of permanently open bypasses in spatial proximity to the HPP and approach velocities >0.5 m/s may have influenced the high number of screen passages. About one-third of the captured fish at this site passed the Archimedes screw turbines, where fish protection measures upstream of the turbine inlet are deliberately not in place. Although Archimedes screw turbines are often referred to as “fish-friendly” due to low runner speeds and pressure changes (Nuernbergk and Rorres, 2013; Simmons and Lubitz, 2021), recent studies have shown that this technique can indeed cause severe injuries and mortality, depending on the fish species in question and the site-specific conditions (e.g., Pauwels et al., 2020; Mueller et al., 2022).

Despite existing fish protection screens and bypasses, a large proportion of fish passed the turbine corridor during their downstream migration. Physical exclusion screens with 15 mm and 20 mm bar spacing, which are also supposed to have a behavior-influencing effect through various inclinations and guide fish to bypasses, could not prevent the majority of fish from passing the hazardous turbine corridor. The mainly poor effectiveness of the investigated bypasses was probably due to the fact that the discharge was too low and the location or detectability of the bypasses was unfavorable, reflecting the real operational conditions at these sites. The different bypass types were used by a different range of fish species and sizes which was also strongly site-dependent. This highlights the need to check existing bypasses not only for compliance with current requirements in terms of design, dimensioning and spatial location to the HPP, but also in relation to the site-specific fish community. Bypasses should ideally function for the widest possible range of fish species present in a system. This could be achieved through either less selective bypass types such as optimally positioned full-depth bypasses, or a combination of different bypass systems. Moreover, less harmful turbine technologies must be developed and identified in order to enhance fish protection from HPP-related injuries and mortality during downstream passage, since fine screens with 15 mm and 20 mm bar spacing cannot prevent small-bodied fish species and juvenile fish <20 cm from turbine entrainment.