Editorial: Green or red: Challenges for fish and freshwater biodiversity conservation related to hydropower

Abstract

The global sustainable development goals of the United Nations (United Nations, 2015) and the associated necessary transformations, such as energy decarbonization (Sachs et al., 2019), have resulted in increased hydropower development around the world (Zarfl et al., 2015). Advocates of hydropower point to the important contributions of this form of energy to the reduction of CO2 emissions, its low cost, and its use as a stable source of energy, justifying it as a ‘green’ energy that makes an important contribution to electricity supply compared with other forms of regenerative energy. For the year 2018, the International Energy Agency (IEA) lists a global energy supply from hydropower of 4,325,111 GWh, compared with 1,273,409 GWh for wind energy and 554,382 GWh for solar/photovoltaic energy (IEA, 2020). In contrast to wind or solar power, hydropower production is available at any time of day and is less dependent on current weather conditions. According to the Collins Dictionary (www.collinsdictionary.com, accessed 14 February 2021), ‘green energy’ is defined as ‘power that comes from sources that do not harm the environment …’. The debate about how ‘green’ the green energies really are is controversial (Gibson, Wilman & Laurance, 2017). Opponents of hydropower production often refer to it as a ‘red’ energy because of the mortalities and injuries that fish face when passing turbines (Mueller, Pander & Geist, 2017; Mueller et al., 2020b), as well as other ecological harm to free-flowing rivers associated with habitat fragmentation and alteration. Hydropower is often referred to as one of multiple impacts on river ecosystems that has potentially synergistic effects with other stressors (Ormerod et al., 2010; Mueller et al., 2020a; van der Lee & Verdonschot, 2020). In many cases hydropower plants were built after other structural modifications of river systems, such as damming and straightening (some of them irreversible) that had already taken place (Auerswald et al., 2019). Still, the general global consensus remains that hydropower can significantly alter river systems. Several international targets in the conservation of natural habitats and wild fauna and flora (Habitats Directive, Council of the European Communities, 1992) conflict with further hydropower development. This includes achieving ‘good ecological status’ of water bodies (or ‘good ecological potential’ in the case of ‘heavily modified water bodies’) as required by the European Water Framework Directive (Council of the European Communities, 2000). Such policies result in enormous efforts (financial and physical) dedicated towards improving freshwater conservation and restoration (Geist, 2015; Geist & Hawkins, 2016). If applied as intended these would be expected to hinder any further expansion of hydropower development, and would also support an argument for dam removal if conservation targets are not met. Despite this paradox, global hydropower capacity is projected to approximately double from the 2010 installed capacity, requiring a dramatic increase in the number of hydropower dams in river basins around the world (Opperman, Grill & Hartmann, 2015). It would be highly ineffective – also from a financial point of view – if both the expansion of hydropower and the restoration of streams and rivers received support from public funds in spite of conflicting targets. It is thus important to bridge the gap between hydropower use and the conservation of aquatic biodiversity, ecosystems, and services that they provide to human society. This is clearly a difficult task and many experts question the feasibility of this ambitious target. From the perspective of conservation biology, the expansion of hydropower is of particular concern in rivers that are considered to be biodiversity hotspots, such as the Amazon (Latrubesse et al., 2017; Latrubesse et al., in press), the Congo and the Mekong (Winemiller et al., 2016), the São Francisco (Gomes et al., 2020), or regions in south-east Europe (Hudek, Zganec & Pusch, 2020). From a general perspective, aquatic conservation in the age of sustainable development goals and ecological assessments in connection with the ‘shifting baseline syndrome’ have already been broadly explored (Irvine, 2018). The intention of this editorial is to provide more specific guidelines that help to assess comprehensively the effects of hydropower in relation to aquatic biodiversity conservation. This requires the better integration of ecological effects, technological properties, and socio-economic factors, as well as conservation prioritization, and a transition from assessing the effects of single facilities on single species to an assessment of cumulative effects, followed by an open, fact- and evidence-based discussion that involves all stakeholder groups before proceeding with political decision making (Figure 1). At present, baseline information is rarely complete even within specific fields. For example, ecological information rarely includes effects on food webs, complete life cycles, or populations, and instead mostly focuses on the organismic damage of a few fish species. Whereas the conservation of fishes is typically prioritized in the context of ecological assessments of hydropower, other aspects relevant to biodiversity conservation deserve inclusion, such as managing the hydromorphological changes to flow regimes, thermal regimes, and altered substrates, which can all affect a range of processes and organisms, not just fishes. The focus of assessing the ecological impacts of hydropower has traditionally been on the direct consequences of the entrainment of fishes. This comprises mortality or injuries from contact with structures, such as screens and bars, collision with turbine blades, shear stress, and barotrauma arising from abrupt pressure changes during turbine passage. Direct and delayed mortality (Ferguson et al., 2006) as well as types of external and internal injury (Mueller, Pander & Geist, 2017; Mueller et al., 2020b) are affected by the physical and hydraulic forces encountered by the fishes and can be predicted to some extent (Deng et al., 2007; Boys et al., 2018), although strong species-specific differences related to anatomical properties prevail. Even less is known about how fish behaviour affects mortality and differences in injury, related to different species, personality types, and diurnal activity patterns, as well as route choice at hydropower facilities (Baumgartner et al., 2014b). Egg et al. (2017) demonstrated that an eel bypass system that was effective in laboratory experiments was not used by eels in a field setting during the silver eel migration. Knott et al. (2019) and Knott et al. (2020) showed pronounced diurnal differences in the downstream movement of different fish species as well as their route choice when approaching hydropower plants, and the resulting injury patterns. A major limitation is that many assessments of the direct impacts of hydropower facilities are based only on one or two fish species (often eel or salmon smolts), without considering the typically much more diverse fish community at its full size range in the specific habitat. In addition, there is little to no integration of information on all life stages of species (for instance, eggs or larvae), nor on other taxonomic groups such as primary producers or macroinvertebrates, which are key elements in the functioning of aquatic food webs. Another difficulty in relation to conservation is that it is often difficult or unethical to experiment with species that are already highly endangered and only present in low numbers. The subjective perception of the mortality or injury of individual fish, which is relevant from the perspective of animal welfare, is different from an assessment of population-level effects, which is more relevant from the perspective of conservation. As the quality of modelling effects at the population level depends strongly on the quality of input information (i.e. on the sensitivity of different life stages as well as the interactions among species), large data gaps concerning the autecology of many species of conservation concern still have to be filled as a first step (Smialek et al., 2019). For example, there is still little knowledge on the safe passage of drifting fish eggs and larvae when passing hydropower turbines. The published results point in different directions as they depend heavily on local conditions, hydraulic mechanisms, and the species studied (Boys et al., 2016; Navarro et al., 2019). Another difficulty is the general lack of studies investigating the effects of turbine passage on taxonomic groups other than fishes. Globally, hydropower development is contributing to one of the largest expansions of dams seen in history (Opperman, Grill & Hartmann, 2015). This is problematic in many ways: rivers are naturally four-dimensional systems that depend on a high degree of connectivity (Ward, 1989; Auerswald et al., 2019), and there are hardly any free-flowing rivers in the world any more (Grill et al., 2019). Practically all hydropower development depends on the introduction of concrete structures into rivers, resulting in habitat fragmentation; indeed, Cooke et al. (2020) have proposed that the concrete conquest of aquatic ecosystems should cease. The creation of impoundments, urbanization, and catchment land use have been identified as the most important factors affecting fish community composition in streams (Bierschenk et al., 2019a; Mueller et al., 2020a). It should be noted that the impacts of hydropower on habitat are not primarily caused by the generation of hydropower itself, but mostly result from the dams and weirs that reduce or fully block connectivity and migration, and alter flow regimes. The importance of free movement and migration is well understood for fishes (Vasconcelos et al., 2020), particularly for diadromous species that depend on migrations between oceans and fresh water to complete their life cycles (Smialek et al., 2019). The provision of fish passage for charismatic flagship species is often proposed as a solution (Silva et al., 2018a). In this respect, a clear distinction must be made between upstream and downstream movements of fishes, as these movements follow different principles (Calles & Greenberg, 2009). Although there is a wealth of scientific studies related to technical and nature-like solutions for facilitating the upstream movement of fishes, their safe downstream passage remains poorly understood (Knott et al., 2020). Even the upstream solutions are typically directed towards one or a few target species. If the target species are strong swimmers, such as the Atlantic salmon (Salmo salar), then a functional fish pass for the target species may still hamper the movements of other species that co-occur in the same habitat. In Europe, this is particularly true for small and weak swimmers that are hardly considered in the planning of fish passes, such as Cottus gobio, Gobio gobio, and Barbatula barbatula. They are often referred to as ‘stationary species’ yet have been found to demonstrate significant movement in fish passes (Pander, Mueller & Geist, 2013). A similar problem arises for downstream migration, as illustrated for the European eel (Anguilla anguilla). This target species for conservation receives great attention and has a unique life cycle in which adults metamorphose into silver eel that migrate down the rivers on their long spawning migration into the Sargasso Sea. Silver eel have a long, elongated body shape and thus are at high risk of getting injured when passing hydropower facilities in a downstream direction (Egg et al., 2017; Mueller, Pander & Geist, 2017; Mueller et al., 2020b). For this reason, there have been various proposals for adjusting turbine operations in relation to eel activity patterns, or for trapping eels, moving them around the barriers, and then releasing them near the ocean (‘trap and truck’). Although the usefulness of such measures that require continuous human management action remains controversial, there is no doubt that other, less prominent species, such as lampreys, are left behind. Compared with the many challenges in facilitating connectivity for fishes, much less is known on how other taxonomic groups are affected by reduced connectivity, although it appears likely that dispersal ability connected with the life cycle strongly determines potential impacts (Pander, Mueller & Geist, 2018). For instance, organisms such as waterborne insects that can fly over barriers in their adult stages are probably less affected by habitat fragmentation than species that are only dispersed passively on hosts (e.g. the glochidia larvae of freshwater mussels), which themselves are strongly impaired in their dispersal ability by barriers. In addition to connectivity, habitat alteration by barriers also needs to be considered. Areas upstream of dams and weirs are typically characterized by increased water depth, lower flow velocity, increased sedimentation, and altered water temperatures, including stratification. Stream systems with a large number of dams and weirs ultimately change from a fluvial system into a stagnant one. Depending on their functional traits, species can be affected differently by these habitat alterations with winners (typically generalists of low conservation value) and losers (typically riverine specialists of high conservation value). The resulting shifts in the biological community occur at all trophic levels from primary producers to fishes (Mueller, Pander & Geist, 2011), with eurytopic species being favoured (Jumani et al., 2018), whereas the life cycles of rheophilic species often cannot be completed (Pander & Geist, 2016) because of the impairment of resources, recruitment, or refugia (van Looy et al., 2019). Sometimes, proponents prepare an argument that establishing hydropower use in natural river stretches would increase biodiversity because it helps to create diverse habitats, such as stagnant flow areas above the dams and high-flow areas directly at the turbine outlets. From a conservation perspective these benefits are false, as they ignore the many damaging impacts on the system as well as the fragmentation of key habitats for the life stages of rheophilic species. Even if rheophilic target species of conservation significance, such as salmonids, manage to spawn in the swift currents downstream of a dam, their larvae typically do not find shallow water habitats in highly modified tailwaters (Pander & Geist, 2016) or manage to use the stagnant flow area upstream of the dam as a foraging habitat. Finding solutions to the challenges ahead requires finding answers to a diversity of questions, such as: should priority be given to fish welfare or habitat conservation; how can the effects of hydropower be assessed and addressed considering the cumulative effects over entire river systems; which principles can guide prioritization at the level of species, ecological processes, and sites; should the effects of mortality and fish injury related to fish passage through hydropower facilities be minimized, or should the entry of fishes into the facilities and moving them around be minimized; to what extent can innovative hydropower technologies provide solutions; and how can decision makers find an acceptable balance between the ‘green’ and the ‘red’ aspects of hydropower? Assessing the impacts of hydropower on aquatic biodiversity conservation is complicated by several factors. Even for individual hydropower facilities, data are rarely available on all taxonomic groups and all life stages that are crucially important for the thorough modelling of population-level effects. From a hydropower licensing perspective, the population-level effects of fishes are a primary concern, often requiring expert assessments based on modelling and fish testing. Any fish testing under realistic field settings can help to understand the factors that govern fish mortality and injuries, and consequently reduce the impact of hydropower stations on biota. Therefore, comparative studies with harmonized study designs can help to identify the least harmful technologies under specific conditions. In addition, such studies can help to optimize operational modes so that harm to aquatic fauna is minimized. This includes technological and construction aspects, such as adjusting the design of turbine blades (Deng et al., 2007), as well as shutting down turbines and/or opening alternative corridors at peak times of migration. On the other hand, such studies also come with several ethical concerns, especially if fishes need to be released upstream of hydropower facilities in order to assess their damage following passage. Some of those stocked fishes may escape, resulting in the unintended introduction of diseases or, where using genetically distant lineages that interbreed with local fishes, in introgression and reduced adaptation in the local population. From the viewpoint of conservation, the use of non-native species in such experiments (such as the European eel in the Danube system) can also be problematic, especially if the river was not stocked with these species in the past. The use of species that are already endangered is of special concern: their use in such experiments may result in serious population declines or – in the case of hatchery-bred fishes – result in competition with supplementary stocking related to existing conservation programmes. This holds true, for example, for the European grayling (Thymallus thymallus), a species that has severely declined in Central Europe (Mueller, Pander & Geist, 2018). Innovative approaches of using buoyant sensors that record strike, sheer, acceleration, and pressure changes may provide perspectives for mitigating this challenge in the future (Deng et al., 2014), yet they cannot fully incorporate anatomical, physiological, and behavioural differences among species that result in diverse injury patterns among different species passing through the same facility (Mueller, Pander & Geist, 2017; Mueller et al., 2020b). At the community level, interactions among species are difficult to include and uncritical translation of results from laboratory experiments into field situations can sometimes be problematic (Egg et al., 2017). Monitoring abiotic habitat changes requires systemic and long-term datasets that are rarely included within hydropower planning. The variables measured should primarily target an assessment of the serial discontinuity at hydropower facilities related to the mean and variability of, for example, water depth, current velocity, temperature, and substrate composition, as these are commonly used when assessing the effects of dams and weirs on habitat quality (Mueller, Pander & Geist, 2011). In the context of hydropower, additional properties such as water-level changes that may result in stranding of larval fishes as well as in the mortality of macroinvertebrates under drying-out conditions, also need to be considered. It is important to recognize that extreme conditions are often more relevant than average conditions because these can have far greater impacts on aquatic life. It is surprising that critical processes such as flow regime-dependent sediment transport, habitat quality, and lateral connectivity are rarely used as indicators, yet should be important conservation targets. Animal welfare organizations tend to focus in particular on publishing shocking pictures of dead or amputated and bleeding animals following hydropower passage to increase public awareness of hydropower as a ‘red’ energy, whereas a consideration of habitats is likely to be more relevant for overall biodiversity conservation. Even when the effects of individual hydropower stations on certain taxonomic groups and habitats are well understood, the interaction with other stressors such as climate change and land use are difficult to disassociate. Moreover, regulated rivers often have a series of hydropower stations along their length and not just single developments. An assessment of such cumulative effects is difficult, yet this is crucial for situations where tipping points are exceeded only when considering multiple stations. For instance, Latrubesse et al. (in press) presented evidence that the cumulative effects of planned dams represent a major threat to Amazonian biodiversity. Yet, environmental impact assessments and licensing typically only consider individual plants. A solution to this challenge could be to adjust the licensing and operation of individual hydropower plants for entire river systems on the basis of holistic, river-wide thresholds related to minimum viable population sizes of key species of conservation concern, as well as to the habitat quality and quantity on which they depend. In countries like Switzerland, a national fund is available for ecological improvements at hydropower plants; such a system is likely to facilitate shifting the focus from a consideration of individual sites to entire river systems. It is well known that the impacts of hydropower facilities and pumping stations on fishes can be reduced by adjusting operational modes (Bierschenk et al., 2019b), by the provision of corridors (Knott et al., 2019; Knott et al., 2020), and by the use of behavioural barriers, such as electrified fences (Egg et al., 2019), with species-specific differences in effectiveness. The same holds true for different types of hydropower plants that differ in their species-specific mortalities depending on the local conditions. Thus, from a conservation viewpoint, priorities need to be set for differential management among rivers and among the designs of hydropower facilities. The question of which species should be prioritized is often determined by a few flagship species for conservation that are popular or economically important, but not necessarily the most endangered or ecologically important species (Geist, 2011). At the species level, giving priority to long-distance migratory species (e.g. eel and salmon) as well as to particularly sensitive, endangered, and endemic species with limited geographical distribution (e.g. the Danube salmon, Hucho hucho, in the upper Danube system) appears most promising. If such a selection can be combined with popularity or ecosystem services such as food provision, then this can facilitate action. Hydropower technologies are typically long-term investments (30–50 years, with renewal). This means that any decision made today will continue to have a persistent impact in the future, and that conservation, mitigation, and management strategies and decisions should consider long-term impacts, but without disregarding the positive short-term effects of conservation decisions such as shutting turbines during the peak migration of sensitive key species. Given the high number of hydropower facilities in developed parts of the world, it is unlikely that we will be able to return to near-natural river systems in the near future, despite the targets set, for example, by the European Water Framework Directive (Council of the European Communities, 2000). A balanced approach can be accomplished by the processes of prioritizing (especially for new hydropower developments as well as dam removal) and minimizing adverse impacts (mostly in the case of existing facilities). Concerning prioritization, I had previously proposed a step-wise approach in freshwater conservation that must set clear priorities and objectives (Geist, 2015). Translating this to the field of hydropower means that the few remaining intact river and stream systems should receive the greatest conservation priority, and that any form of hydropower development on these systems should be avoided, in line with the suggestions on ´hydropower by design´ (Opperman, Grill & Hartmann, 2015). In addition to the few large rivers in which the hydropower potential has not yet been fully explored (e.g. the Amazon, Congo, and Mekong rivers), this also relates to many of the headwater areas in European streams where further expansion of ‘small’ hydropower technology is often only profitable economically if subsidized or compromised in terms of environmental mitigation measures. On the other hand, whether investment in costly ecological improvements of already highly degraded or artificial river systems should have first priority can be questioned, unless these systems are essential in providing access and connectivity to important areas upstream. In channelized river systems, three different conservation strategies are of particular importance in enabling complete life cycles of rheophilic species: creating and managing bypass channels as substitute habitat; connecting free-flowing tributaries as recruitment areas; and restoring environmental flows. Areas directly downstream of hydropower plants and fish bypass channels primarily constructed to facilitate fish migration are often the only refuge areas with steep enough hydraulic gradients to provide a spatially limited, high-quality refuge habitat for specialized rheophilic species of conservation concern (Pander, Mueller & Geist, 2013; Pander & Geist, 2018). Thus, in addition to their role as migration corridors, the substitute habitat function of bypass channels must be considered equally, even though their limited space cannot fully compensate for deficits in the main stem (Pander & Geist, 2013). Especially in large, dammed rivers, free-flowing tributaries are important for the recruitment of rheophilic fishes (Silva et al., 2018b; Vasconcelos et al., 2020). There is also increasing focus on restoring environmental flow (e-flow) regimes in regulated rivers: i.e. flows that mimic the natural hydrographs (Auerswald et al., 2019; Pander et al., 2019) or reproduce fish-related cues (Baumgartner et al., 2014a), as well as other conservation approaches that take into account the natural dynamics of stream systems (Geist, 2011; Geist, 2015). As a matter of fact, existing hydropower stations are designed to control flow releases and thus – at least to a certain extent – can be used to help realize designed e-flow regimes for conservation purposes. Such approaches can also mitigate the impacts of hydropower on the biotic disturbance of floodplains and estuaries. Innovative hydropower technologies with the aim of minimizing environmental impacts can provide a chance for future improvements. Several new technologies have been developed such as the shaft hydropower plant concept and very low head turbines, as well as new Archimedes screw-based designs. All of these developments require robust testing under realistic conditions as this is the only way to identify technologies for the future that are less harmful to aquatic systems than existing ones. Yet, the key question – of whether it is more reasonable to minimize the adverse effects of turbine passage by investments in more ‘fish-friendly hydropower solutions’ than to minimize fish entry or maximize fish passage mitigation – is not an easy one to answer. Although protecting large fishes by using protection screens and bars appears feasible, my own observations from various projects in European streams and rivers clearly suggest that the majority (typically >90%) of all fishes are less than 15 cm in total length. Thus, there are limited possibilities to prevent their entrainment owing to technical limitations of minimum screen and bar sizes under field conditions. For larval fishes and other aquatic organisms, it is likely that as yet there are no effective ways of excluding, or even significantly reducing, their entrainment. A better understanding of population-level effects is urgently needed, yet it appears logical that the absence of effective options to mitigate the damage linked with downstream passage makes it mandatory to focus on the type of technology as well as on technical and operational adjustments, such as temporal turbine management to allow safe fish passage (Figure 1). The discussion on hydropower and conservation is highly controversial and often results in conflicts. Identifying and realizing solutions for the challenges ahead can only be successful if an honest discussion based on scientific evidence is used as a basis for policy and management decisions (Figure 1). Also, integrative approaches that translate aspects such as ecosystem functioning and technical detail, which may be abstract to practitioners, into clear operational guidelines, based on uses and services directly relevant to humans, tend to be most successful (Geist, 2011). From an ecological point of view, a shift must be made from considering individual organisms and species to a consideration of population-level effects, communities, and habitats, ideally focusing more on incorporating ecological functions into conservation decision making (Decker et al., 2017). For instance, a recent study by Gerke et al. (2021) demonstrated that top-down effects of nase (Chondrostoma nasus) and chub (Squalius cephalus) can mitigate the adverse effects of eutrophication and increase oxygen availability to the hyporheic zone, which are problems that typically arise in the headwater areas of dams. As populations of both species can be severely affected by hydropower dams, the consequences for other species, such as macroinvertebrates that depend on oxygenated stream beds, need to be better emphasized and considered in decision making. Such thinking needs to be applied more broadly to other projects. On a more global level, decision making in both hydropower development and conservation management is closely tied to societal and political processes. It thus lies in the nature of things that a natural scientist’s or conservationist’s perspective as well as an engineer’s perspective alone will not be able to address the challenges ahead. Instead, natural scientists and engineers should provide important information from their disciplines as baseline information that then requires the integration of socio-economic and legal aspects as well as the consideration of different technological and operational options (Figure 1). Ideally, such information should not only be assessed at the level of individual hydropower facilities, but instead include cumulative effects for entire river systems. Such a process requires the transparent incorporation of stakeholder interest and cost trade-offs: for example, between hydropower and fish-passage mitigation (Venus et al., 2020), as well as among biodiversity, food security, and hydropower (Ziv et al., 2012). Especially for large river systems, such as the Mekong and the Amazon, food security for people and power provision for economic development are key considerations that need to be integrated with conservation targets. The more that the needs of different stakeholders become part of decision making, the easier it will be to achieve a balanced compromise that mediates the controversy about hydropower as green or red energy. I would like to thank Phil Boon for inviting me to write this editorial, which was inspired by ongoing research projects on the ecological effects of innovative hydropower technologies (partly funded by the European Union’s Horizon 2020 research and innovation programme under grant agreement no. 727830, ‘Fish-friendly innovative technologies for hydropower (FIThydro)’, and a comparative monitoring project on the effects of conventional and innovative hydropower technologies funded by the Bavarian State Ministry of Environmental and Consumer Protection: OelB-0270-45821/2014) conducted at the Aquatic Systems Biology Unit of TUM. I would also like to thank Lee Baumgartner, Josef Knott, Birgit Lohmeyer, Martin Mörtl, and Joachim Pander for proof-reading earlier drafts of the article and our entire hydropower team at TUM for their great work. None to declare.