Methane (CH₄) is a potent greenhouse gas, contributing significantly to atmospheric radiative forcing. Inland waters are identified as the largest natural source of CH₄, with estimates reaching 398.1 (±79.4) TgCH₄ yr⁻¹. However, these estimates primarily rely on data from solute-poor waters, neglecting the substantial contribution of salt-rich inland waters, which comprise roughly half of the global inland water volume and a fifth of the surface area. Evidence suggests that salinity, particularly sulfate (SO₄²⁻), inhibits CH₄ production through various mechanisms, potentially leading to lower CH₄ emissions from these systems. The lack of empirical data from salt-rich inland waters raises concerns about the accuracy of current understanding of aquatic CH₄ regulation and global emission estimates. Salinity significantly influences aquatic microbial communities, affecting the abundance and distribution of methanogens and methanotrophs. Methanogenesis is an energy-inefficient process, and the presence of various ions favors more efficient reactions. Sulfate and iron reducers can outcompete methanogens for labile carbon substrates. Furthermore, the reduction of SO₄²⁻, nitrate (NO₃⁻), Fe³⁺, and other ions can be coupled to anaerobic CH₄ oxidation, further suppressing CH₄ concentrations at high salinity. Salinity's impact on CH₄ is multifaceted, interacting with organic carbon availability, nutrient availability, and sediment sorption. While salinity's inhibitory effect on CH₄ has been observed in coastal and salt-rich wetlands, the impact on regional and continental CH₄ emissions remains largely unknown. This study aims to address this knowledge gap by investigating the role of salinity in regulating CH₄ emissions from inland waters in the Canadian Prairies, a region characterized by a high density of salt-rich inland water systems. The findings will have implications for refining global CH₄ emission estimates and predicting future emissions under changing environmental conditions.

Previous research has established the importance of salinity in influencing methane (CH₄) cycling in various aquatic ecosystems. Studies have demonstrated that salinity, particularly sulfate (SO₄²⁻), inhibits CH₄ production through multiple mechanisms. These mechanisms include the outcompetition of methanogens by sulfate and iron reducers for labile carbon substrates, and the coupling of the reduction of various ions to anaerobic CH₄ oxidation. The availability of organic carbon also plays a crucial role, influencing the extent of SO₄²⁻ inhibition of methanogenesis. In organic-rich environments, the inhibitory effect of salinity may be less pronounced due to the abundance of substrates for methanogenesis. Nutrient availability, affected by salinity through sediment sorption, further complicates the interplay of factors regulating CH₄ production. Existing studies have shown the impact of salinity on CH₄ emissions in coastal and salt-rich wetlands, but there is a scarcity of data from inland salt-rich waters. This knowledge gap prompted the current research, aiming to quantify the impact of salinity on CH₄ emissions across various aquatic ecosystems and improve the accuracy of global CH₄ emission estimates.

This study employed a comprehensive approach combining extensive field surveys with high-resolution eddy covariance measurements. The primary survey encompassed 193 aquatic ecosystems in the Canadian Prairies, including rivers, lakes, wetlands, and agricultural ponds, spanning a wide range of morphometry, hydrology, chemistry, and trophic status. These sites were sampled once during the summer months (June-August 2011-2021). A peripheral survey included 48 wetland ponds across the three Prairie Provinces, sampled repeatedly throughout the 2021 ice-free season to assess spatio-temporal patterns. Two wetland sites were equipped with eddy covariance flux towers for continuous, ecosystem-scale measurements of CH₄ flux over a full year (May 20, 2021 to May 19, 2022). At each primary survey site, various limnological parameters were measured, including water temperature, dissolved oxygen, pH, specific conductivity, salinity, dissolved organic carbon (DOC), total phosphorus (TP), and total nitrogen (TN). In a subset of sites, sulfate (SO₄²⁻) concentrations were also measured. CH₄ concentrations were determined using the headspace technique, with variations in gas sampling methods among the independently collected datasets. CH₄ emissions were calculated using the diffusive flux equation (F=k(Ceq-Cw)), with the gas transfer velocity (k) estimated from previous studies. Ebullitive CH₄ fluxes were measured at 10 sites using inverse funnel traps. Eddy covariance measurements provided continuous, ecosystem-scale data for CH₄ fluxes. Statistical analyses, including multiple linear regressions and marginal effect analysis, were used to assess the relationships between CH₄ and other variables. The study also compared measured CH₄ emissions to predictions from existing empirical models to quantify the impact of salinity on model accuracy. Spatial upscaling of CH₄ emissions was conducted for the Canadian Prairies using conservative estimates of the area covered by small lentic systems and simulations based on two salinity levels (0.5 ppt vs. 0.1 ppt). Similar scaling exercises were performed globally to assess the potential impact of salinity on global CH₄ emission estimates. Temporal trends in SO₄²⁻ concentrations were analyzed using long-term monitoring data from southern Saskatchewan.

The study revealed a clear link between salinity and CH₄ emissions, varying across ecosystem types. In rivers and larger lakes, the influence of salinity on CH₄ was non-significant, likely due to the co-linearity of salinity, DOC, and nutrients. However, in small lentic waterbodies (wetlands and agricultural ponds), salinity was a significant predictor of CH₄ emissions, particularly via ebullition. Multiple linear regression analysis showed that salinity negatively impacted pCH₄ in these systems, while the ratio of DOC to salinity positively correlated with pCH₄, suggesting that organic-rich systems can partially offset the inhibitory effect of salinity. This relationship was validated by an independent survey of wetland ponds. The negative correlation between salinity and both diffusive and ebullitive CH₄ fluxes was observed, with ebullition showing a stronger sensitivity to salinity. Eddy covariance measurements at two wetland sites with contrasting salinity levels confirmed the inverse relationship between salinity and ecosystem-scale CH₄ emissions over annual scales. Meta-analysis of data from other global regions (Tibet, Mexico, and India) supported the global applicability of this inverse relationship. Existing empirical models, developed primarily for freshwater systems, underestimated CH₄ emissions in salt-rich systems. Salinity significantly improved model predictions, accounting for a substantial portion of the deviation from the observed data, particularly for ebullitive fluxes. Upscaling analyses revealed that ignoring salinity resulted in a significant overestimation of CH₄ emissions in the Canadian Prairies (at least 81% overestimation, equivalent to 0.97 Tg CO₂ equivalents) and potentially globally (a potential overestimation of 1.7 TgCH₄ yr⁻¹, or 4% of the global lentic CH₄ emissions). Finally, analysis of long-term monitoring data showed a significant increase in SO₄²⁻ concentrations in Saskatchewan lakes over the past 30 years, highlighting the potential for future salinization to further reduce aquatic CH₄ emissions.

This study's findings significantly advance our understanding of CH₄ cycling in salt-rich inland waters and highlight the need to incorporate salinity into global CH₄ emission models. The strong negative correlation between salinity and CH₄ emissions, particularly in small lentic systems, underscores the importance of considering salinity's inhibitory effects on methanogenesis. The interaction between salinity and organic matter availability further emphasizes the complex interplay of factors regulating CH₄ dynamics. The substantial overestimation of CH₄ emissions by existing models, which largely ignore salinity, has significant implications for national and global greenhouse gas inventories. The inclusion of salinity as a key variable in future emission models is crucial for improving accuracy and informing effective climate change mitigation strategies. The observed increase in SO₄²⁻ concentrations in Saskatchewan lakes suggests that future salinization may further modulate CH₄ emissions, warranting further investigation into the long-term implications of this phenomenon.

This study demonstrates the significant role of salinity in restricting methane emissions from small inland waters, particularly in the Canadian Prairies. The findings highlight a substantial overestimation of CH₄ emissions in salt-rich regions when using models developed for freshwater systems. Incorporating salinity as a crucial variable in CH₄ emission models is necessary for accurate estimations and effective climate change mitigation efforts. Future research should focus on expanding the study to other regions, investigating the long-term impacts of salinization on CH₄ emissions, and exploring finer-scale mechanisms underlying salinity's influence on microbial communities and CH₄ cycling. Further investigation into the relationship between agricultural practices (e.g., sulfur-based fertilizer use) and increasing SO₄²⁻ levels in prairie waters is warranted.

While this study provides a significant advancement in understanding the influence of salinity on CH₄ emissions, some limitations should be acknowledged. The primary survey involved single summer sampling of most sites, limiting the assessment of seasonal and interannual variability. The headspace technique for CH₄ measurement varied across datasets, although this was deemed to introduce a minor error relative to the observed range in CH₄ concentrations. The spatial upscaling calculations relied on conservative estimates of the area covered by small lentic systems, and uncertainties might exist in extrapolating regional findings to global scales. Future studies could address these limitations by incorporating more extensive spatiotemporal data and standardizing CH₄ measurement techniques.