Successes and failures of conservation actions to halt global river biodiversity loss

People have settled near rivers for thousands of years to benefit from the ecosystem services they provide, such as clean water, transport, energy, food and recreation¹. However, human activities have impaired biodiversity in approximately half of the world’s rivers, particularly in densely populated regions in East Asia, Europe and North America, and in arid and equatorial climate zones². These activities include wastewater discharge, fertilizer and pesticide runoff from adjacent farmland, water removal, channel and flow regime alteration, and the introduction of invasive species.

Since 1950, existing river stressors have intensified and new stressors have emerged, including neonicotinoid pesticides, pharmaceuticals and personal-care products, worsening climate change and the development of global hydropower³. New stressors often co-occur with historical persisting stressors, potentially exacerbating their combined effects on river biodiversity. To mitigate these stressors, countries and jurisdictions have introduced legislation aiming to reduce pollution, protect species and habitats, and restore ecosystems and river biodiversity, such as the US Clean Water Act, the European Union (EU) Urban Wastewater Treatment Directive and the Convention on Biological Diversity’s Kunming–Montreal Global Biodiversity Framework (Fig. 1). In response, a broad range of conservation actions have been implemented worldwide that seek to maintain or improve biodiversity⁴, including through restoration, protection and management. Examples include wastewater treatment plants to reduce water pollution from households and industry; hydromorphological restoration to improve physical shape and flow characteristics, such as by re-meandering channelized rivers and reconnecting rivers to their adjacent floodplains; and national or international efforts to reduce fertilizer, pesticide and sediment inputs from agricultural areas.

Policies are classified as either focusing on addressing the drivers of biodiversity loss or on promoting conservation actions. ECE LRTAP, Economic Commission for Europe: Convention on Long-range Transboundary Air Pollution; EU, European Union; IUCN, International Union for Conservation of Nature; UN, United Nations.

Full size image

Although river conservation actions have been implemented for decades^5,6,7, the outcomes of these actions for maintaining or improving river biodiversity remain poorly known. Furthermore, conservation actions and their effects on biodiversity are rarely adequately monitored^8,9,10, which requires data collected across relevant spatial and temporal scales, comparisons with reference ecosystems and consistent monitoring of target biota. In addition, actions that are monitored often exhibit wide variability in biodiversity outcomes^11,12. Consequently, evidence on the effectiveness of conservation actions is scarce and many actions are implemented because they are assumed to be effective. Assessing the effectiveness of conservation actions in improving river biodiversity is therefore an essential first step in examining these assumptions and guiding further research and river conservation.

In this Review, we identify and define nine categories of river conservation actions and how they aim to ameliorate the effects of anthropogenic stressors on river biodiversity. We assess evidence for their effectiveness in protecting or improving river biodiversity, focusing in particular on the principal factors determining success and failure, and identify research and implementation gaps. On the basis of this assessment, we provide recommendations to inform river conservation research and improve the outcomes of future actions.

Freshwater vertebrate populations are declining at higher rates than those in terrestrial and marine ecosystems¹³. These declines are reflected in the threatened status of 62% of freshwater turtles and 43% of freshwater mammals^5,14, and by an average population abundance decline of 88% in monitored freshwater megafauna between 1970 and 2012 (ref. ¹⁵), including river dolphins, crocodiles, large turtles, sturgeons and giant salamanders. River biodiversity decline is similarly illustrated by declines of 81% in monitored populations of global freshwater migratory fish (averaged across species) between 1970 and 2020 (ref. ¹⁶), and by the impaired status of 50% of river macroinvertebrate communities, which include dragonflies, mussels and crayfish².

This global trend of river biodiversity loss is predominantly driven by five broad driver groups, ordered by declining importance: land-use change, pollution, direct exploitation, invasive species and climate change^17,18 (Box 1). All five groups encompass multiple stressors that often co-occur in many river catchments; however, substantial variation exists in the severity, extent and combination of drivers within and among catchments and the taxonomic groups affected. Additionally, despite the reported lower importance of invasive species¹⁹ and climate change, the importance of these drivers is expected to increase, resulting in often unpredictable interactions with the other drivers³.

To address river biodiversity loss and its drivers, conservation actions have been implemented globally. On the basis of commonly discussed initiatives to improve river biodiversity^4,5 these actions can be grouped into nine broad categories that aim to maintain or improve river biodiversity: restore habitats and connectivity; implement environmental flows; manage point-source pollution; manage diffuse pollution; manage species and substrate exploitation; prevent and control invasive species; address climate change effects; protect habitat; and protect species (Table 1, Supplementary Box S1). Further, individual actions undertaken across these categories are, as a whole, often referred to as ‘nature-based solutions’^20,21, which aim to protect, restore and manage ecosystems to provide benefits to both human wellbeing and biodiversity²². Seven of the nine conservation action categories generally address effects from only one of the five driver groups, whereas two action categories (‘protect habitat’ and ‘protect species’) address all driver groups. Some conservation actions can also be assigned to multiple action categories. For example, restoring vegetation in river riparian zones is a commonly applied habitat restoration action, but might also help to mitigate diffuse pollution by capturing runoff or to mitigate climate warming by providing shade.

Evaluating the effectiveness of conservation actions and their geographic distribution, coverage of relevant organism groups and alignment with regional drivers should help to improve ongoing and future conservation. Qualitative assessment of case studies provides further insight into the overall outcomes of each action category, their geographic patterns of use, and whether they match the principal drivers in different geographic regions (Supplementary Information 1 and 3).

Defining and describing conservation action categories

Each of the nine conservation action categories has the potential to improve river biodiversity, but overarching challenges hinder their effectiveness.

Restore habitats and connectivity

The degradation and fragmentation of habitats as a result of land-use change has been identified as the most important driver of freshwater biodiversity decline¹⁷. In rivers, land-use change typically occurs across entire catchments, but conservation actions addressing this driver tend to be local in scale. Examples include enhancing habitat by changing river channel form (increasing sinuosity, or levee removal and/or setback²³) or improving habitat structure and complexity (by adding gravel, boulders or woody debris²⁴). Habitat fragmentation, such as from dams and weirs^23,24, can be addressed either by restoring longitudinal connectivity via barrier removal or by the installation of fishways^25,26. Restoration actions also include altering river banks or bed elevation to reduce sediment runoff from agriculture or to lower erosive flows from impervious surfaces²⁷. These actions are all usually referred to as river restoration.

Despite the large number of habitat improvement projects, few are empirically evaluated^27,28 and projects disproportionately focus on macroinvertebrates or fish over other organism groups, which are less frequently (or never) assessed. Of the studies that report outcomes, either no apparent change or negative (that is, adverse) changes are as frequent as positive changes^9,28,29. A lack of biodiversity improvement generally arises when actions primarily focus on restoring local habitat heterogeneity, but other factors are present that can limit recovery. For example, even if hydromorphological restoration is complete, species recovery might be constrained by pollutants^9,30,31 or because source populations for new colonists are too distant^6,32,33.

Dam removal projects often focus on fish and macroinvertebrates, and outcomes of connectivity restoration are almost entirely assessed only for fish. Less consideration is given to how these actions affect other taxa, such as macrophytes and plankton. Dam removal sometimes leads to considerable biodiversity improvement³⁴, but responses are highly variable among sites³⁵ and communities do not necessarily return to their pre-impact state³⁶, perhaps owing to other unaddressed stressors. Outcomes of connectivity restoration using fishways are also mixed³³. Fishways are ineffective when fish are unable to navigate the passage, or when post-passage issues prevent the recovery of populations, such as when fish that have passed through are unable to access critical habitats for reproduction^37,38,39.

Implement environmental flows

Environmental flows are implemented for various purposes, such as increasing water availability in downstream river sections and riparian areas, triggering life-history events of river species, enhancing connectivity within the river network, prolonging floodplain inundation, improving sediment transport and flushing salt from riparian soil^40,41. Conservation actions to implement environmental flows include managing water abstraction^5,42, limiting alterations to the natural flow regime to protect biodiversity and ecosystem condition, or simulating a natural flow regime to support ecosystem functions and services in the context of multiple competing demands⁴³. For example, water might be released from dams to mimic the magnitude and timing of natural flow to meet the ecological requirements of river species in downstream reaches, although dams often still hinder downstream natural sediment transport.

Despite the increasingly recognized importance of environmental flows for river species, these actions have been implemented only in a small proportion of hydrologically altered rivers⁴⁴. Most studies that have evaluated the effectiveness of environmental flows for river biodiversity have focused on fish and macroinvertebrates, and generally report positive outcomes, including enhanced recruitment of fish⁴⁵, increased abundance of macroinvertebrates and fish⁴⁶, reduced abundance of invasive species⁴⁷ or facilitated movements of fish and aquatic mammals^40,48. Failure can occur when the magnitude and frequency of environmental flows are not well aligned with the ecological needs of river species⁴⁹, resulting in limited benefits for biodiversity. This highlights the need for adaptive management strategies in which flow quantity and timing are continuously updated as information about the ecological requirements of target species is gathered⁴⁰.

Manage point-source pollution

Managing point-source pollutants in rivers typically involves reducing the concentration of pollutants via collection and/or treatment prior to their release into the environment, such as in wastewater treatment plants (WWTPs), oil-spill containment booms or tailings ponds for waste materials from mining. Actions to manage industrial waste and spills or to remediate accidental pulse pollution events can be very effective if applied rapidly, and recovery of river biodiversity can occur within a short time (for example, within 1–4 years after exposure^50,51). By contrast, remediation of press pollution events, in which rivers are continually inundated with contaminated effluent from tailings ponds or mines, are mostly ineffective in improving river biodiversity^52,53 if the source of pollution is ultimately not addressed⁵⁴.

Worldwide, WWTPs are a frequently applied river conservation action to reduce inputs of nitrogen, phosphorus and organic matter to rivers; 63% of domestic and municipal wastewater is collected and 52% is treated in WWTPs⁵⁵. Despite their widespread use, few studies have investigated the effects of WWTPs on river biodiversity. In Europe^56,57 and North America^58,59, considerable improvements in river biodiversity have been reported after the implementation of WWTPs, as well as negative effects⁶⁰. Negative outcomes can be reported if treated sites are compared to sites that receive no wastewater rather than to sites that receive untreated wastewater⁶⁰. Additionally, many regions do not sufficiently treat wastewater⁶¹ and the increasing number and use of chemicals, including pharmaceuticals, have driven a rise in micropollutants that remain largely untreated in most WWTPs, resulting in negative effects on river biodiversity^62,63. Advances in WWTPs, such as a fourth purification stage, have the potential to remove many of these micropollutants^64,65, but wider implementation is hampered by their high cost and energy consumption⁶⁴. A further challenge is that wastewater and rainwater are often collected in the same sewer and transported to WWTPs. During heavy rain events, this mixed wastewater and rainwater can be directly discharged into receiving rivers via sewer overflow systems when the total volume of water is too high to allow any treatment⁶⁶.

Manage diffuse pollution

Diffuse pollution is a particularly challenging stressor to mitigate in river ecosystems because pollutants are distributed across large spatial extents from multiple sources. Common conservation actions typically aim to reduce pollutant emission or runoff by, for example, adding emission filters, reducing the use and application of heavy metals and chemicals⁶⁷, and installing or maintaining riparian buffer strips to capture pollutants before they enter rivers^68,69.

Legislative efforts to control emissions, such as the United Nations Economic Commission for Europe’s Convention on Long-range Transboundary Air Pollution Sulphur Protocols and the US Clean Air Act, have exhibited broad success in reducing diffuse pollutants. The resulting measures led to substantial declines in nutrient aerosol inputs (for example, nitrogen) and acidification in many affected European and North American rivers^70,71,72. These reductions are probably the result of emission control measures that prevent or strongly reduce emissions from vehicles or industrial facilities. Although directly quantifying the benefits of these regulations to river biodiversity is difficult because their effects encompass large spatiotemporal scales and co-occur with numerous other stressors and restoration efforts, substantial recovery is thought to have followed the implementation of these and other large-scale environmental regulations^{73,74,75,76,77}.

By contrast, the effectiveness of actions to mitigate polluted runoff, which includes efforts to reduce inputs of chemicals such as nutrients and pesticides, is more ambiguous. The positive effects of reducing fertilizer use in agriculture are often discussed or modelled⁷⁸, but empirical studies investigating the realized effects of these practices on river biota are lacking. Similarly, although microplastics have been detected in many rivers exposed to urban runoff⁷⁹, evidence of potential ecological effects of microplastics on river biota is scarce and little to no effective conservation actions have been implemented to mitigate microplastics in rivers⁸⁰. Maintaining, revegetating or expanding riparian buffers is one of the most studied actions to address runoff, and considerable⁸¹ or at least slight improvements⁸² in river biodiversity have been reported. However, most studies examine macroinvertebrate responses, whereas little is known of the effects of these actions on other organism groups. Additionally, although riparian buffers can reduce diffuse pollution, they are often not implemented explicitly for this purpose but to restore habitat. Consequently, it is difficult to determine whether improvements arose from better management of runoff quality or from habitat restoration and other benefits of riparian vegetation, such as reducing bank erosion or mitigating climate warming effects by increased shading. Further work is needed to identify mechanisms and to evaluate the efficacy of different actions in mitigating the effects of diffuse pollutants on different organism groups, particularly those that reduce the use of chemicals such as fertilizers. The benefits of riparian buffer strips need to be linked specifically to the mitigation of diffuse pollution.

Manage species and substrate exploitation

The harvesting of species and river substrates for human use are both considered exploitation. However, whereas substrate exploitation, such as sand and gravel mining for construction, has global effects on river biodiversity⁸³, actions are rarely undertaken specifically to address this stressor, resulting in a lack of knowledge in this area. Species exploitation primarily occurs through commercial and recreational fishing. Actions to mitigate fishing include legislation and enforcement of fishery closures (which can be seasonal or permanent), efforts to reduce exploitation (for example, via catch quotas, gear restrictions, catch-and-release, and so on), or compensating species losses by stocking from artificial propagation and rearing programmes^84,85.

Despite the variety of actions available to reduce fish exploitation and their long history of use in rivers, evaluations of their efficacy are generally lacking^86,87 and reports on efficacy often show no improvement in target species^86,88. Reasons for failure are varied, but can generally be attributed to one of four problems. First, even if fishing is restricted, populations cannot recover if other stressors constrain recovery. This problem can be addressed through holistic management approaches such as integrated river management⁸⁹, in which actions to address exploitation are combined with actions such as establishing fishways, restoring habitat and controlling pollution that address other, relevant stressors. Second, conventional fishery management actions can increase the incidence of illegal or unsustainable fishing when they do not integrate local residents into management and decision-making^90,91, resulting in little overall change in exploitation. Third, rivers often span multiple jurisdictions, which presents problems for managing migratory fishes because actions can be compromised by poor conditions and continued exploitation in other locations (for example, marine fishing^88,92). Finally, fish stocking may fail owing to mortality of artificially reared individuals^93,94 and if stocking increases exploitation of wild populations when fishing effort increases to take advantage of the augmented population⁹⁵. These issues require improvements in the quality of hatchery-reared fish, such as by improving genetics and disease management⁸⁷, and restrictions to ensure fishing effort remains consistent after stocking.

In summary, actions to address species exploitation in rivers have been most successful where fishing is the dominant stressor, people readily comply with restrictions, the managed population occupies a restricted spatial area, and/or focal species are amenable to captive breeding and hatcheries produce viable individuals. Beyond these circumstances, successfully addressing exploitation is difficult and further research, including on outcomes of conservation actions that address the effects of substrate exploitation on river biodiversity, is needed to develop comprehensive, inclusive and adaptive management practices.

Prevent and control invasive species

Actions to address invasive river species can be divided into two groups: those that focus on preventing new invaders from arriving and establishing, and those that focus on controlling already established invasive populations^7,96. Prevention involves identifying key vectors that introduce river invaders (for example, the online aquarium trade⁹⁷), enacting regulations to prevent or reduce introductions from these vectors, and implementing early detection and rapid response to ensure that introduced species are quickly identified and removed. Control includes the eradication or reduction of established invasive species populations using biological, chemical or physical means⁹⁸. Additionally, installing dispersal barriers within rivers to prevent invasive species from further spread is an increasingly considered control action, although its application is restricted primarily to North America, Australia and Europe^99,100. However, these barriers can limit movements of native species, potentially fragmenting populations and causing further risks for river biodiversity⁹⁹.

Studies evaluating the effectiveness of actions to address invasive species are limited in number and among countries. Prevention is typically more successful in terms of cost and effectiveness than control actions^101,102, but is the focus of fewer studies. Control can be very successful, even at the scale of entire countries (for example, the removal of muskrat and coypu from Britain¹⁰³). However, success is often determined on the basis of elimination of the target species and not of the subsequent recovery of native river biodiversity. Where biodiversity outcomes are reported, projects to control higher trophic-level invaders, such as fish or mustelids, are often successful^104,105,106 and can lead to the recovery of native biota^107,108,109 and positive ecosystem changes¹¹⁰. By contrast, projects to control lower-trophic level invaders, such as macrophytes or invertebrates, exhibit mixed outcomes^111,112,113. Differences in outcomes among taxa are probably caused by the greater ease of finding and removing higher-trophic-level organisms such as fish and mammals, given that lower-trophic-level organisms can be smaller and are therefore difficult to find, isolate and remove. Lower-trophic-level taxa can also have shorter lives and more varied reproductive strategies, such as asexual reproduction, which facilitate rapid population recovery. In addition to differences among taxa, the nature of river ecosystems means that many invasive species are hidden underwater and are easily transported via river networks, making them inherently difficult to control¹¹⁴, particularly across large spatial extents. These difficulties highlight why more work is needed on prevention, because control might be successful only for particular species or in restricted areas and environments, such as small streams or river reaches during the early phases of invasion.

Address climate change effects

Proposed measures to reduce climate change effects in rivers aim to improve ecosystem or species adaptability, or to address precipitation and temperature changes. These actions include planting trees in riparian zones to increase shading, reducing water extraction and backwaters, establishing cold-water refugia¹¹⁵, and improving flow management¹¹⁶. Increased flood and drought risks caused by shifts in precipitation and temperature can be mitigated by increasing water infiltration and retention capacity in the catchment, for example, by reducing impervious areas or increasing green infrastructure for flood management^117,118. Other proposed measures include reducing heat discharge into rivers from power plants to partially compensate for warming rivers, stocking or relocating vulnerable species to suitable regions¹¹⁹ and locating protected areas based on future climate change scenarios¹²⁰.

Despite the clear need for actions to reduce the effects of climate change in rivers¹²¹, little evidence exists that action has been taken specifically to maintain or improve river biodiversity¹²². Other conservation actions can include climate change mitigation as a possible outcome. For example, habitat restoration that increases riparian shading might also mitigate river warming. Similarly, restoring river connectivity and implementing environmental flows might help to mitigate the effects of heatwaves on river species¹¹⁶ or help species to disperse to habitats with suitable temperatures¹²³. Addressing climate change is rarely the focus of these actions but should become a higher priority in the near future. Therefore, whether these actions have helped to provide microclimate buffers—particularly given that they often fail to achieve their primary, non-climate-related outcomes—remains unknown. The general lack of climate-focused conservation actions is one of the largest and most critical knowledge gaps in mitigating river biodiversity loss.

Protect habitat

Protecting river habitat is ideally accomplished by designating allowable, permitted and prohibited activities on and around rivers, and by assigning and managing protected areas that encompass key habitats, species and a large proportion of a river’s catchment¹²⁴. However, this ideal is rarely realized. Instead, most inland protected areas typically target terrestrial flora, fauna and habitats^125,126, although some river-focused protected areas exist, such as the Vindelälven Ramsar Site in Sweden or Special Areas of Conservation along the Garonne river in France. That inland protected areas usually focus on terrestrial species means they generally lack goals or conservation strategies for rivers^125,127, often under the assumption that rivers will be indirectly protected through better land management, which usually does not occur^128,129. Consequently, protected areas fail to maintain or improve river biodiversity approximately as often as they succeed¹³⁰. Failure occurs when protected areas do not encompass the ecosystem components vital to maintaining river species, such as key ecological and hydrological processes, and spatial and temporal interactions in river networks^131,132,133. Furthermore, terrestrial-focused protected areas do not necessarily address key river stressors that originate from beyond protected area boundaries, such as upstream pollutants¹³⁴ or downstream fishing⁹². Many protected areas are also not strictly protected, because the International Union for Conservation of Nature (IUCN) recognizes different categories that range from strict nature reserves (category Ia) to those that allow multiple human uses (such as categories V or VI)¹³⁵. Multi-use protected areas can fail to halt river biodiversity loss if they prioritize socioeconomic needs, such as land development¹³⁶, over ecological goals.

Improving the success of protected areas for river biodiversity requires explicit consideration of how best to protect river habitats and species. For example, rather than limiting protections to certain river sections (such as in the EU Habitats Directive¹³⁷), extending protections to the broader catchment or to large parts of the catchment can better counteract both local and broader-scale stressors^132,138. Consistently setting catchment-specific targets and management actions can also greatly improve biodiversity¹²⁷. These targets and actions could include: identifying priority catchments for protection; improving coordination between land and water managers¹³⁹; and setting catchment-specific goals for the global protected area coverage target of 30% by 2030 set by the Kunming–Montreal Global Biodiversity Framework, which should be counted separately for terrestrial and inland waters¹⁴⁰, including rivers. These measures would better ensure that river protection is directly considered and is not just an assumed byproduct of terrestrial protected areas. Furthermore, for rivers and their catchments subject to human use, engagement with residents and other stakeholders is critical for success. This engagement can include collaborative management between agencies and resource users, such as the restoration of fisheries in the Mekong River in Cambodia¹⁴¹, or entirely community-based management, as has been successful for fish communities in Thailand’s Salween catchment¹⁴². Such efforts prioritize both river biodiversity and the needs of local residents^142,143 and reduce unsustainable or illegal resource use¹⁴⁴.

Protect species

Species-specific conservation actions include creating Red Lists¹⁴⁵ to assign species a conservation status, identifying flagship or umbrella species, listing species to be monitored, protected or managed, and protecting habitats on which targeted species depend (for example, the EU Habitat Directive¹⁴⁶ and the US Endangered Species Act¹⁴⁷). Protecting species can also involve translocations, which include moving species to new habitats. Examples include moving freshwater mussels to Gulf Coast USA rivers¹⁴⁸, reintroducing species to habitats they once occupied, such as bullhead fish (Ameiurus nebulosus) to Belgian headwaters¹⁴⁹, and supplementing wild populations with captive-bred individuals, such as golden perch (Macquaria ambigua) and Murray cod (Maccullochella peelii) in the Australian Murray–Darling Basin¹⁵⁰.

Red-listing and other listing actions, such as in the Convention on Migratory Species (CMS) and the Convention on International Trade in Endangered Species of Wild Fauna and Flora (CITES), have known benefits for developing policy, setting conservation goals, allocating funding and raising awareness. However, evaluations of the effects of listing on specific species are generally lacking¹⁵¹. Many freshwater status assessments are also far from complete; for instance, according to the global Red List of the IUCN, 26% of the freshwater fishes, 30% of the decapods and 16% of the Odonata (dragonflies and damselflies) are threatened¹⁵², but we do not know the status of most invertebrates (such as molluscs^153,154) and plants. Consequently, even if listing is generally beneficial for river biota, most species have not been assessed and, therefore, cannot be listed. Research into the effects of listing on river species is sorely needed, as are assessments and actions for underrepresented taxonomic groups.

Translocations often have considerable benefits for targeted river species, especially those with narrow endemic distribution ranges^155,156. However, few projects include genetic management of translocated populations to ensure their long-term viability and fitness^157,158,159. Translocation failure is typically attributed to the effects of unaddressed stressors (such as habitat degradation^160,161) or to logistical difficulties, such as poor support for the long-term monitoring necessary to track population success and to respond to problems that arise during establishment^162,163.

Action outcomes and their global distribution

Reviewing and evaluating the outcomes of 436 published studies examining the effects of conservation actions on river biodiversity provides further insight into the factors determining action success and failure (Supplementary Tables S2 to S4 and Supplementary Data S1). These studies encompass 7,195 different projects across 26 global regions, including different continents, countries and/or regions, or individual or grouped river catchments. Study outcomes were categorized qualitatively according to the conservation action, geographic region, taxonomic group and outcome of the action as reported by the study authors. Approximately two-thirds of the projects investigated the responses of fish and the remaining third investigated macroinvertebrates; other organism groups such as macrophytes, diatoms, plankton, reptiles, amphibians or mammals were only occasionally the focus of study. The action categories ‘Protect habitat’ and ‘Manage species and substrate exploitation’ were each examined by about 25% of projects, and ‘Restore habitat and connectivity’ and ‘Manage diffuse pollution’ were each examined by 15–20%. All other action categories typically comprised between 0 and 5% of projects (Supplementary Table S4).

Based on the aim of the actions as reported in each published study, outcomes were scored for each taxonomic group assessed in each study using three categories. Outcomes that were negative or neutral were categorized as ‘no desired change’ (scored as 0). Studies in which authors reported only partial recovery in certain taxa or metrics were categorized as ‘slight improvement’ (scored as 0.5), whereas studies reporting substantial increases in biota, or a return to a community similar to reference conditions, were categorized as ‘considerable improvement’ (scored as 1). These scores were then weighted by the number of projects to ensure that outcomes supported by more data carried more weight. This qualitative approach enables outcomes to be compared across a diversity of studies, including those that do not include comprehensive statistical testing or quantitative information per project.

Generally poor outcomes

Across all regions, action categories and organism groups, river conservation actions generally elicited little improvement in biodiversity (based on an average outcome score of 0.15; Fig. 2). However, this overall outcome must be interpreted with caution owing to an unequal distribution of projects across action categories (categories examined by more projects tended to score lower) and an overrepresentation of projects from certain regions, particularly Europe and North America. Additionally, the underrepresentation of certain actions might be due to differences in interest or responsibilities of stakeholders such as farmers or authorities, and the poor availability of published research on action effectiveness for river biota. For example, broad-scale legislation for better wastewater management, such as the US Clean Water Act, probably drove substantial recovery of river biodiversity. However, the average outcome of ‘Manage point-source pollution’ is low (Fig. 2). This disconnect occurs because little published research exists directly linking effects of WWTPs to river biota. Instead, improvements are typically inferred either from chemical water quality or from broad-scale analyses showing recovery coinciding with the timing of legislation^74,164. As a result, the effects of broad-scale legislation are difficult to capture in a qualitative review of individual project outcomes.

Outcomes are scored from no desired change (0.00, red) to considerable improvement (1.00, blue) (Supplementary Information). Globally and for each continent, average outcomes for each action (colours), the number of projects implementing each action (ring size), and overall average outcome across all conservation actions (numbers below diagrams) are provided. We caution against overinterpretation of these outcomes, given the differences in project numbers between continents and action categories, which might partially drive differences in outcomes.

Full size image

Despite these caveats, the overall poor outcome of the reviewed river conservation actions matches the frequency of action failure reported across the literature on the nine conservation action categories, as well as reports of generally mixed outcomes from other global-level or continental-level reviews and meta-analyses, including on river restoration^29,54, protected areas¹³⁰ and overexploitation⁹². Furthermore, many examples exist of river species that remain threatened today, despite being historically listed as threatened, illegal to exploit and with designated recovery plans¹⁴⁷.

These overall poor action outcomes for river biodiversity contrast with outcomes for terrestrial and marine ecosystems, in which conservation is reported to be more successful (for example, in two-thirds of cases⁴). The lower rate of success of river conservation actions might occur because rivers are the lowest point in the landscape and are therefore exposed to the combined effects of multiple stressors acting across broad geographic areas, including from urban centres, agriculture and forestry.

Geographic differences

The projects examined were overwhelmingly from high-income countries (about 95%), and half of the remainder were from South America, particularly the Amazon Basin. These socioeconomic biases have been reported elsewhere^{12,27,113,165}, highlighting a potential lack of support for river conservation and associated research in many regions, in addition to potential biases in the publication process. Furthermore, a clear socioeconomic difference exists in the implementation of different actions. Projects in high-income countries often encompass a variety of actions that target specific drivers, typically a combination of habitat restoration, exploitation management and pollution management. By contrast, projects in low-income or middle-income regions primarily focus on general actions of protecting habitat and species, suggesting a lack of both political will and financial support for implementing varied and targeted actions (Fig. 3).

For each region, averaged outcomes per region range from ‘no desired change’ (0.00) to ‘considerable improvement’ (1.00). The total number of projects (represented by the size of the pie charts) and proportion of projects enacting each conservation action are also displayed for each region. For information on how we assigned studies to regions, see Supplementary Information 3.

Full size image

Mismatch between actions and drivers

The number and type of conservation actions should ideally reflect the intensity and type of stressor and the underlying drivers in each river catchment. To determine whether this alignment has occurred, information on the intensity of the five driver groups of biodiversity loss globally¹⁶⁶ (Supplementary Information 1) can be compared to the geographic distribution of conservation actions. In general, most regions are affected by medium to high stress, particularly from land-use change, direct exploitation and pollution (Fig. 4). However, only in high-income regions, particularly the USA and Europe, does higher stress correspond to a higher number of conservation projects (Fig. 4). Frequently, higher river stress combined with a comparatively low number of conservation actions indicates that global anthropogenic effects in rivers are probably insufficiently addressed in most regions (Fig. 3).

For each region, total stress level (Supplementary Information 1) and the relative contribution of each driver (bar colours on the left) are shown alongside the total number of projects implementing conservation actions and the relative proportion of projects for each conservation action (bar colours on the right). To simplify the links between actions and drivers, some conservation action categories were merged according to the groupings shown in Table 1, resulting in six action categories rather than nine. Five of the six action categories have a primary matching driver group (as shown in the key), but the sixth action category (‘Protect habitat and species’) encompasses all drivers (as in Table 1). Ratios shown on the right indicate the number of drivers and the number of actions per region. For information on how we calculated stress levels, see Supplementary Information 1. For information on how we assigned studies to regions, see Supplementary Information 3.

Full size image

The overall poor outcomes from river conservation actions result from three principal causes that consistently feature across the literature and in our assessment of the 436 studies: failure to address multiple stressors; failure to act at the relevant spatial scale; and failure to monitor actions appropriately. Numerous other causes indirectly result in failure of conservation actions, such as a lack of political support, funding constraints or complex socioeconomic, political and cultural reasons. Such causes of failure are beyond the scope of this Review but are often the ultimate reason for one or all of the three causes of failure listed above.

Multiple stressor challenges

Multiple stressors co-occur in many rivers and often interact in unpredictable ways (for example, warming can augment nutrient pollution effects by altering the food web¹⁶⁷). However, in practice, only one or two conservation actions are typically implemented (Fig. 3) and might not address the most limiting stressor, thus hindering recovery⁶. Examples include instances of ineffective river habitat restoration owing to the presence of stressors such as pollution or erosive flows, which had greater effects on biodiversity than habitat loss^54,168,169. Mitigating or reversing multiple stressor effects requires a combination of conservation actions tailored to the most relevant stressors. This process is complicated by changes in stressors between rivers or among different river sections, and by logistical and funding restrictions that complicate the disentangling of complex stressor interactions. An ideal approach would be to identify the suite of relevant stressors and corresponding conservation actions using the best available information, and then to prioritize actions that address multiple stressors simultaneously. For example, riparian restoration or maintaining a buffer strip of vegetation can be an affordable action that addresses several stressors at once, including riparian habitat loss, diffuse pollution resulting from agricultural and urban runoff, and climate change. However, such actions can fail¹⁷⁰ and further research is necessary to identify approaches that improve conservation action effectiveness for multiple stressors.

Spatial challenges

Success in river conservation also depends on the spatial extent, location and overlap of both actions and stressors. River species often have broad ranges⁶ and complex habitat requirements, such as migratory fish or river insects that move longitudinally throughout their lifespan. Localized actions covering only a few hundred metres or several kilometres of river length, or actions enacted in only one segment of a degraded river, might not extend across the length of habitat these species use (including whole catchments, floodplains and riparian zones), reducing the likelihood of action success. Additionally, the connected nature of river environments means that stressors can operate at different spatial scales. For example, localized actions do not account for stressors transported from upstream aquatic and terrestrial areas¹⁷¹. Furthermore, nearby source populations of the targeted species may be necessary for recovery success³².

This combination of spatial scaling issues in species habitat requirements, stressor effects and action implementation means that the spatial extent of the problems species experience are often mismatched to the spatial scale addressed by the actions¹⁷². Such scaling problems are not uncommon given difficulties in coordinating cross-boundary management (for example, in river ecosystems spanning multiple administrative units) and logistical constraints that limit the spatial scale of action, including limitations in financial resources or land availability.

Monitoring challenges

The successful implementation of conservation actions requires adequate pre-action and post-action monitoring of biodiversity and its environmental drivers to assess their effectiveness and plan future actions. However, many actions implemented across the globe are based on national or international legislation that does not mandate biodiversity monitoring (such as the US Clean Water Act and the EU Urban Wastewater Treatment Directive). Furthermore, actions that require monitoring often face limitations of time, personnel and funding, resulting in no or very limited monitoring. This lack of adequate monitoring could create a biased perspective of outcomes, as the subset of projects that conduct monitoring might not represent action successes and failures worldwide.

Ideally, monitoring should encompass both biodiversity and environmental stressors, and should comprise three elements. First, appropriate monitoring sites must be identified and tracked, including sites where actions have been applied and appropriate control sites. These control sites should include both unaffected locations (to assess the degree of progress towards a less-affected state¹⁷³) and affected sites not subject to conservation action (to track the degree of departure from the degraded state). Multiple control sites should be monitored pre- and post-action for at least a decade¹⁷⁴ to assess variability in both space and time at scales appropriate for focal biota and stressors¹⁷⁵. Second, post-action monitoring should assess biotic and abiotic responses across the initial (pioneer) and long-term succession phases, again at appropriate temporal scales¹⁷⁶. Third, monitoring should identify the species and communities of most concern (that is, those particularly vulnerable to the specific stressors affecting the river) and their nearby source populations, and should track changes in these target taxa. Decisions about which species and communities to monitor should be guided by policy emphasis, ecological relevance and societal interest, and not necessarily on which species have received historical research interest such as fish or macroinvertebrates.

River conservation actions have generally poor outcomes, indicating that further efforts are required for maintaining or improving river biodiversity. Poor outcomes often result from the complexity of river ecosystems, inundation by multiple stressors, difficulties in matching actions to the spatial scale of effects, and insufficient monitoring. Many localized river conservation actions do not fully address these challenges but still successfully mitigate targeted stressors. This might not lead to full recovery of river biodiversity but provides progress towards this goal and contributes to our understanding of action effectiveness. Similarly, broader efforts to address stressors that span entire countries and continents can be successful, such as the implementation of wastewater treatment plants that have successfully reduced effects on river biodiversity. Generally poor outcomes might, therefore, indicate an insufficient overall effort to implement multiple actions and monitoring rather than a failure of applied conservation actions.

Research on improving river conservation outcomes has identified the need to move away from focusing on disturbed river sections or target species and towards broader ‘catchment-scale management’^{30,124,177,178}. This approach considers the interconnectedness of the entire river network and linked terrestrial ecosystems. Catchment-scale management can operate in various ways and under different names, such as integrated river management¹⁷⁷, integrated water resources management¹⁷⁹, river management plans¹⁸⁰ or ecosystem-based management¹⁸¹. However, these approaches all aim to coordinate management across multiple ecosystem components, including within the river and at land–water interactions throughout the catchment. To do so, first, stressors and the scale at which they operate are typically identified and prioritized for targeted actions, particularly stressors that dominate over others¹⁶⁷. Second, target biota and any nearby source populations are defined, followed by selecting and monitoring appropriate indicators of change in these species. Third, once stressors and indicators have been selected, targeted actions must be identified, prioritized and implemented. This process should incorporate previously implemented actions to coordinate information and provide a comprehensive perspective of progress. Fourth, progress on stressors and target biota must be monitored to quantify the effectiveness of implemented actions and to identify causes of success and failure. Finally, actions are adjusted according to outcomes; for example, switching from unsuccessful species-focused management to social-ecological management⁹¹. This approach is often referred to as ‘adaptive management’.

A key element of successful catchment-scale management is ‘co-production’, that is, joint contributions from multiple relevant stakeholders. Co-production acknowledges that river management will probably fail if ecological needs are not balanced with socioeconomic, political and cultural needs, such as for water security, sanitation and hygiene (WASH). Achieving this balance requires moving beyond traditional management approaches, which are usually distributed among various authorities without sufficient coordination. Instead, co-productive river management should comprise consulting, coordinating and incorporating the needs of multiple agencies and other stakeholders (such as regional authorities, industry, farmers and Indigenous peoples). The effectiveness of co-production is supported by conservation research^12,182,183 and its incorporation in national and international science organizations and funding agencies. Co-production helps stakeholders to navigate complex interactions among multiple stressors, biodiversity and human needs for land and water by promoting cooperation and compliance towards achieving management goals and resolving conflicts, thus avoiding undesirable outcomes. Additionally, international legislation and organizations (such as the Kunming–Montreal Global Biodiversity Framework and the EU Nature Restoration Law¹⁸⁴) provide vital mandates for conservation and funding opportunities, and promote cooperation across different agencies and stakeholders. In summary, catchment-scale management and co-production are the principal ways forward for improving river biodiversity and promoting interdisciplinary and transdisciplinary cooperation to achieve transformative river management.