Climate frameworks must better account for river ecosystem responses and feedbacks

There has been considerable debate on the contribution of rivers to the global carbon cycle, particularly regarding how climate change alters riverine carbon processing and greenhouse gas (GHG) emissions. Rivers are increasingly recognized as active components of the global carbon cycle¹, yet river ecosystem deterioration and its reciprocal interactions with climate change remain insufficiently represented in Earth system and climate assessments. Rivers actively regulate GHG emissions, carbon and nutrient cycling, and ecosystem resilience through interconnected hydrological and biogeochemical processes. However, current climate assessment and mitigation frameworks, including IPCC assessments, national GHG inventories, and UNFCCC reporting, typically evaluate rivers through water resource potential, carbon fluxes, pollution profiles, and GHG emissions independently rather than as interacting components of climate feedback processes. Consequently, they may underestimate how climate change and anthropogenic disturbances collectively degrade river functioning and, in turn, influence regional and global carbon budgets. Further, the impacts of climate change can be more severe in ecosystems whose hydrological and ecological attributes have already been modified by human activities. This is because anthropogenic perturbations have already altered riverine carbon, oxygen and water cycles across the land-ocean aquatic continuum^2,3,4. These disturbances alter not only the magnitude of climate responses but also the underlying ecohydrological and biogeochemical mechanisms that regulate them. We therefore argue that future climate action frameworks should explicitly integrate river ecosystem functioning and changing biogeochemistry into climate assessment and mitigation strategies. We conceptualize these interactions as a Riverine Climate Feedback Domain (RCFD) (Fig. 1), linking climatic and anthropogenic drivers to river ecosystem responses and climate feedbacks.

Among the mechanisms most overlooked by climate assessment frameworks are hydroclimatic disturbances, which fundamentally regulate river biogeochemistry and ecosystem resilience. Climate warming and hydroclimatic extremes are already shifting river systems towards reduced ecological resilience, increasing their susceptibility to abrupt biogeochemical and ecological state changes⁵. This results from the compounded effects of warming, hydroclimatic variability, nutrient enrichment, habitat fragmentation, and contaminant redistribution whose combined effects are often nonlinear rather than additive. Dry events reduce the dilution capacity of river water, concentrating nutrients and pollutants, and intensifying thermal stress on ecosystems, while flooding events mobilize sediments, nutrients, organic matter, and legacy contaminants. Consequently, rapidly changing drying-rewetting cycles may destabilize carbon pools from floodplains, wetlands, and sediments, increasing their susceptibility to hydrological export during subsequent rainfall events. Intense precipitation and runoff events can rapidly transfer this mobilized carbon to rivers in dissolved and particulate forms, altering microbial metabolism, GHG emissions, and downstream carbon fate. Because these hydrological shifts regulate when and how rivers transition between carbon storage and GHG production, they represent critical processes that should be represented in climate assessments.

Furthermore, hydrological fragmentation represents one of the most apparent changes through which human activities amplify climate-driven changes in river functioning. For instance, damming, channel modification, and excessive extraction of water and sediments fundamentally alter river flow regimes by disrupting hydrological connectivity, sediment transport, nutrient exchanges, and ecological dispersal pathways across catchments. These physical disruptions have important implications for long-term carbon storage. Riverine sediments are important for carbon sequestration, with terrestrial particulate organic carbon (POC) burial efficiencies reaching up to 36%, substantially higher than those reported for marine systems (~0.25%)⁶. However, only about 23% of large rivers remain free-flowing globally, while sediment delivery to the ocean has declined by nearly 49%, because of human interventions such as damming and sediment extraction, resulting in reduced flow. These shifts reduce opportunities for recolonization, dispersal, and functional compensation of riverine communities following disturbance, thereby weakening ecosystem resilience at larger spatial scales. Also, they promote depositional environments favorable for methanogenic activity, increasing the likelihoods of localized GHG hotspots⁷. Consequently, evaluating climate mitigation approaches without accounting for hydrological and sediment continuity may underestimate important ecological trade-offs and downstream carbon cycling.

Beyond hydrological fragmentation, anthropogenic pressures further deteriorate river water quality, reinforcing climate change-driven shifts in ecosystem functioning, by enhancing the influx of nutrients, contaminants, and organic matter through urban and industrial wastewater discharge along with surface runoff. Urbanization further amplifies river warming through increased thermal loading associated with expanding impervious surfaces, reduced riparian vegetation, and anthropogenic heat emissions, leading to elevated temperatures, commonly discussed as the urban heat island effect⁸. Together, these interacting climatic and anthropogenic pressures may precondition river ecosystems to climate warming by shifting their biogeochemical states before additional climate-driven perturbations occur, thereby increasing their susceptibility to nonlinear ecological responses. Since dissolved oxygen (DO) regulates microbial respiration, GHG production, and nutrient cycling, its continued decline represents a critical mechanistic link that climate assessments must consider alongside conventional carbon-emission metrics.

The climatic significance of these interacting stressors depends on how they alter river biogeochemical processes and associated climate feedbacks. One of the clearest signatures of these changes is the decline in DO. Evidence from North American and European rivers⁹ indicates that approximately 87% of rivers are warming and 70% are undergoing deoxygenation. Water temperature alone explains up to 95% of DO variability in some major systems¹⁰, suggesting that continued warming will intensify oxygen stress across river systems. Under oxygen-rich conditions, carbon processing is dominated by aerobic respiration, but oxygen depletion progressively shifts microbial pathways toward anaerobic metabolism, enhancing denitrification, nitrous oxide emissions, and methane production. Such transitions also alter food-web structure, microbial community composition, and nutrient release, often through nonlinear responses. Positive feedbacks between hypoxia, nutrient release, microbial respiration, and oxygen demand therefore emerge which progressively reduce ecosystem recovery, causing disturbed river reaches to remain trapped in degraded states¹¹. Warming further worsens these oxygen depletion-mediated changes. Under warmer and oxygen-limited conditions, microbial communities increasingly allocate energy to cellular maintenance and stress tolerance¹² rather than biomass production, thereby increasing carbon loss as CO₂. These conditions may favor methane-producing and denitrifying bacteria while reducing the dominance of oxygen-sensitive taxa. Hypoxia may therefore function not only as a symptom of ecosystem degradation but also as a driver of further biogeochemical change, shifting sediment functioning from long-term sinks to active sources of GHGs.

In addition to these stressors, climate change is modifying atmospheric inputs that regulate watershed biogeochemistry and carbon mobilization. Rivers are exposed to increasing atmospheric inputs of anthropogenic contaminants originating from dust transport, wildfire emissions, industrial aerosols, intensive agriculture, and precipitation. Climate change is altering their delivery and fate by modifying rainfall intensity, wildfire frequency, atmospheric circulation, glacier melt, and dryland expansion. More importantly, these changes destabilize long-stored pools of terrestrial carbon within river catchments. Atmospheric nutrient deposition can stimulate soil microbial activity and organic matter decomposition within catchments, accelerating the transformation of soil organic matter into dissolved organic carbon (DOC)¹³. Enhanced mobilization of these carbon pools by rainfall and surface runoff subsequently increases carbon delivery to riverine mainstems, thereby influencing overall metabolism and downstream carbon export to coastal oceans. This is particularly relevant because rivers integrate multiple carbon pools, including surface litter, peatlands, permafrost, and vegetation. Although runoff preferentially exports young surface carbon, a significant proportion of ancient carbon is being released by climate change-driven permafrost degradation¹⁴. Emerging evidence¹⁵ suggests that hydrological processes strongly regulate the routing of carbon from land to atmosphere through rivers, obscuring the bulk organic carbon ages. Millennial-aged carbon contributes about 59 ± 17% of global river CO₂ emissions¹⁶, with sedimentary lithology and catchment size as crucial drivers. Hence, additional uncertainties arise because riverine emissions derive from both contemporary and aged carbon pools, complicating efforts to constrain global carbon budgets. These findings suggest that rapidly changing carbon processing is disrupting the riverine carbon sequestering capacity. Incorporating watershed carbon mobilization and riverine carbon cycling into climate assessments would therefore reduce important uncertainties in carbon accounting.

These interactions are particularly important because rivers already contribute substantially to atmospheric carbon fluxes. Rivers and streams emit approximately 1.8 Pg C yr⁻¹ as CO₂¹⁷ and 30.5 ± 17.1 Tg CH₄ yr⁻¹ as methane¹⁸ to the global emissions. They also emit approximately 291 Gg N₂O-N yr⁻¹, a flux growing three times faster than terrestrial N₂O emissions¹⁹, driven primarily by anthropogenic nitrogen loading. Together, these emissions determine the overall CO₂-equivalent climatic impact of river GHG fluxes and therefore should be considered collectively. In addition to atmospheric emissions, rivers transport nearly 1.02 Pg C yr⁻¹ of terrestrial carbon to the oceans²⁰, alongside a global riverine discharge of approximately 39,000 km³ yr⁻¹ as freshwater⁶. Consequently, degradation of river ecosystems may increasingly affect the quantity, quality, and fate of carbon and nutrients exported to downstream waters and coasts, thereby influencing carbon emission/sequestration and oceanic nutrient cycling. The climatic significance of rivers, hence, depends not only on the magnitude of these fluxes but also on the cross-domain carbon linkages between terrestrial, freshwater, and coastal systems. Because ecosystem degradation and resulting biogeochemical responses regulate the transitions between carbon storage and release¹, river ecosystems are increasingly exposed to compound disturbances whose cumulative ecological effects may exceed those of individual stressors²¹. These interactions demonstrate why climate assessments must evaluate river ecosystem functioning alongside GHG fluxes rather than considering emissions as independent accounting variables.

Climate change is also reorganizing biological community structures that directly impact river carbon and nutrient cycling. Warming increasingly favors smaller-bodied organisms and faster biological turnover, changing microbial pathways, and reducing the persistence of longer-lived biomass pools that contribute to ecosystem carbon storage. For instance, climate-driven drying, droughts, and heatwaves can act as strong environmental filters, restructuring communities across functional and taxonomic scales²². Also, climate change coupled with anthropogenic eutrophication may shift primary producer communities from diatom-dominated assemblages toward cyanobacterial blooms²³, intensifying further hypoxia and GHG outgassing. These ecological changes are further exacerbated by ongoing biodiversity loss driven by human activities. Current rates of biodiversity loss are already estimated to be 100–1000 times greater than pre-human background rates²⁴, highlighting an alarming loss of ecological buffering capacity across freshwater ecosystems globally. Biodiversity loss, therefore, affects more than species-richness alone; it can weaken the ecological functioning under stress²⁵. This makes biodiversity loss a greater concern because biological communities regulate the pathways through which carbon and nutrients are retained, transformed or emitted. Consequently, disruptions to these biological controls may propagate downstream, influencing estuarine and coastal biogeochemistry. Conserving biodiversity should hence be considered an integral component of climate change mitigation approaches because biological communities regulate many of the ecosystem processes that underpin river-climate feedbacks.

As ecological and biogeochemical changes accumulate, river ecosystems may progressively lose resilience and approach critical ecological thresholds. Beyond these thresholds, relatively small additional disturbances may trigger disproportionate ecological changes. These coupled processes increase the likelihoods of shifting towards ecological tipping points, whereby gradual environmental change pushes river ecosystems beyond a critical threshold, resulting in abrupt and potentially irreversible shifts to alternative ecological states. River ecosystems often appear resilient to gradual environmental change because compensatory biogeochemical mechanisms buffer disturbances. However, resilience may decline progressively before visible ecosystem collapse occurs due to repeated warming, eutrophication, and low-flow conditions. For instance, cyanobacterial dominance, once established through warming and nutrient enrichment, represents a major state shift that reorganizes carbon and nutrient cycling in ways that are difficult to reverse over management-relevant timescales. This suggests that understanding how close river ecosystems are to tipping points may be as important as understanding how rapidly climate change is progressing. Therefore, future climate frameworks must explicitly incorporate river ecosystem resilience, hydrological continuum, DO dynamics, biodiversity, and carbon processing pathways for early detection of ecological tipping points and to guide effective mitigation strategies. Critically, conventional emissions-focused climate frameworks, which do not monitor ecosystem state and response variables, struggle to detect these thresholds before they are crossed.

Current climate mitigation strategies do not always account for the river ecosystem responses they generate. Hydropower expansion can increase reservoir residence times and promote oxygen-depleted conditions that favor methane production, partially offsetting anticipated climate benefits. Similarly, unplanned afforestation and large-scale land-use interventions may alter catchment hydrology and seasonal flow regimes, thereby modifying downstream carbon and nutrient transport. Likewise, bioenergy crop cultivation, particularly where intensive fertilizer use and poor land management occur, can enhance nutrient runoff, accelerating eutrophication and associated oxygen depletion in receiving rivers. These river responses are rarely quantified within carbon accounting frameworks that evaluate mitigation effectiveness, meaning that the net climate benefit of such strategies may be unresolved. Practical implementation pathways already exist for incorporating these considerations into climate action. River-sensitive climate assessments could combine indicators of hydrological integrity (environmental flows and connectivity), geomorphic stability (sediment continuity), biogeochemical functioning (dissolved oxygen, nutrient cycling, and GHG fluxes), and ecological resilience (biodiversity and ecosystem recovery). Integrating these indicator-based approaches into national GHG inventories, Nationally Determined Contributions (NDCs), National Adaptation Plans (NAPs), and reporting of the Global Stocktake and the Intergovernmental Platform on Biodiversity and Ecosystem Services (IPBES) would enable mitigation strategies to be evaluated not only for atmospheric carbon reductions but also for their consequences for river ecosystem functioning and long-term climate resilience. A practical framework linking river ecosystem indicators to climate assessment and policy mechanisms is presented in Table 1.

Current climate frameworks remain focused primarily on GHG emissions and carbon accounting. While these approaches are undoubtedly indispensable for climate monitoring and governance, they may overlook the ecohydrological and biogeochemical processes that regulate emissions in the first place. Explicit incorporation of river ecosystem changes could reduce major uncertainties in global carbon budget constraints²⁶ and improve the representation of river-climate feedbacks. We therefore suggest that future assessments and governance frameworks, including those being developed by the IPCC, UNFCCC-COP, the Global Stocktake, and the Sendai Framework, explicitly incorporate river health variables including ecosystem responses, hydrological stability, and biogeochemical functioning, as central components of resilience and carbon assessment. Concretely, this would require standardized riverine monitoring indicators within national GHG inventories and cross-framework coordination between climate and biodiversity governance. Such approaches would also strengthen implementation of UN Sustainable Development Goals (SDGs), particularly SDGs 6, 13, 14, and 15. Without such integration, current approaches may underestimate the role of rivers in shaping carbon fluxes and over-simplify their contribution to climate resilience. As climate pressures intensify, it is a time to evaluate progress not only by emissions reduced, but by the ecosystem processes that ultimately regulate them.