Differential controls on CO2 and CH4 emissions from the free-flowing Neretva River, Bosnia and Herzegovina

Abstract

Streams and rivers emit methane (CH4) and carbon dioxide (CO2), two greenhouse gasses contributing to global warming. Estimates for diffusive gas emissions can be obtained by multiplying the concentration gradient between water and atmosphere with the gas transfer velocity. The latter is purely physically constrained, yet spatially highly variable. And - in a flowing water ecosystem - the local concentration gradient is the result of a dynamic balance between upstream evasion and resupply. The collection of representative emission data is thus challenging and emissions of river ecosystems are rarely assessed considering temporal variability and spatial dependence at network scale. In this study, we uncover spatial heterogeneity and controls of concentrations and emission fluxes of the two greenhouse gasses, CH4 and CO2, along a 50 km length of a pristine river system, the Neretva River in Bosnia and Herzegovina. This remote river network has so far remained barely influenced by human activities and the hydromorphological status is to date not altered. The Neretva can therefore serve as a reference for similar systems in the region. This seems to be particularly important as rivers in the Western Balkans, including the Neretva, are currently experiencing a surge in hydropower development and damming, which is known to strongly affect riverine greenhouse gas emissions. We found high emissions as a result of co-occurrence of high concentration with high exchange velocity, but we identified different underlying mechanistic processes driving the evasion of the two gasses. CH4 was strongly supply-limited: elevated concentrations were exclusively measured in a large pool (0.84 µmol L-1 compared to a median concentration of 0.005 µmol L-1 in the entire study section). This resulted in CH4 evasion being four orders of magnitude higher in the turbulent reach following the pool (22 mmol m-2 d-1) compared to the median evasion at network scale (0.06 mmol m-2 d-1). In contrast, CO2 evasion was more variable in time and equally dependent on CO2 and gas exchange velocity. The construction of dams intended in this area would lead to reservoirs of slowly flowing or standing water with similar habitat conditions as the observed CH4-hotspot. The concomitant increase in residence time and higher retention of organic material will lead to an increase of CH4 production replacing aerobic respiration. Consequently, CH4 emissions can be expected to drastically increase by orders of magnitude. This greenhouse gas footprint of hydropower generation may counteract the promised climate benefits in terms of renewable energy production.