Tracing sources of riverine CO2 evasion in a densely populated and dammed river system
Recent syntheses of riverine CO2 evasion have suggested that global inland water systems might emit much larger amounts of carbon than previously estimated. Anthropogenic perturbations could significantly alter riverine carbon fluxes, but their contributions to riverine organic carbon transformations and CO2 evasion remain poorly constrained. To investigate human-induced alterations of natural riverine C fluxes, organic carbon characterization was combined with cross-site comparison of the partial pressure of CO2 (pCO2) across the Han River draining a densely populated and dammed region in Korea. The study consisted of (1) a monsoon season sampling at representative land use components of the river system and (2) a sensor-based continuous monitoring in an upper tidal reach of the estuary along the megacity Seoul. The pCO2 at 20-cm depth was determined using a diffusion-type non-dispersive infrared sensor sealed by a gas-permeable polytetrafluoroethylene membrane. Water temperature, pH, conductivity, dissolved oxygen (DO), air temperature, and barometric pressure were also determined in situ. Water samples were collected for determining dissolved and particulate organic carbon (DOC and POC), UV absorbance, fluorescence excitation emission matrices (EEMs), alkalinity, and major anions and cations.
The pCO2 varied with the type of inland waters and the distance from the river mouth. Higher levels of DOC and its optical measurements concurred with downstream increases in the pCO2 along an urban tributary and the estuary compared to the headwater streams, reservoirs, and upper main-stem reaches. Increased algal photosynthesis in the summer presumably lowered the pCO2 to a level of undersaturation in reservoirs. High-frequency, in-situ measurements of pCO2 in the estuary exhibited large seasonal and diurnal variations. The pCO2 decreased by 38% with transition from summer to winter and fluctuated diurnally, with day-to-night shifts ranging from 1.1–17.6 times. Both spatial and temporal variations of pCO2 were strongly correlated with changes in pH and DO. The 'microbial' humic-like fluorescence better explained the spatial variation of pCO2 than DOC and the 'terrestrial' humic- or protein-like fluorescence. The results from the initial phase of the multi-year project suggest a potential linkage between the spatiotemporal variations of pCO2 and organic carbon characteristics that might be explained by the interplay between hydro- biogeochemical processes-and anthropogenic pollution in thisintensely modified river system.