Measurements of riverine dissolved inorganic carbon (DIC), total alkalinity (AT), pH, and the partial pressure of carbon dioxide (pCO2) can provide insights into the biogeochemical function of rivers including the processes that control biological production, chemical speciation, and air-water CO2 fluxes. The complexity created by these combined processes dictates that studies of inorganic carbon be made over broad spatial and temporal scales. Time-series data like these are relatively rare, however, because sampling and measurements are labor intensive and, for some variables, good measurement quality is difficult to achieve (e.g., pH). In this study, spectrophotometric pH and total alkalinity (AT) were quantified with high precision and accuracy at biweekly to monthly intervals over a four year period (2018-2021) along 216 km of the Upper Clark Fork River (UCFR) in the northern Rocky Mountains, USA. We use these and other time-series data to provide insights into the processes that control river inorganic carbon, with a focus on pCO2 and air-water CO2 fluxes. We found that seasonal snowmelt runoff increased pCO2 and that expected increase and decrease of pCO2due to seasonal heating and cooling were likely offset by an increase and loss of algal biomass, respectively. Overall, the UCFR was a small net source (0.08 ± 0.14 mol m-2 d-1) of CO2 to the atmosphere for all four years of our study with highly variable annual averages. The highly dynamic, seasonally correlated, offsetting mechanisms highlight the challenges in predicting pCO2 and air-water CO2 fluxes in rivers.

Freshwater ecosystems are globally significant sources of greenhouse gases (GHG) to the atmosphere. Generally, we assume that in-situ production of GHG in streams is limited by turbulent reaeration and high dissolved oxygen concentrations, so stream GHG flux is highest in headwater streams that are connected to their watersheds and serve as conduits for the release of terrestrially derived GHG. Low-gradient streams contain pool structures with longer residence times conducive to the in-situ production of GHG, but these streams, and the longitudinal heterogeneity therein, are seldom studied. We measured continuous ecosystem metabolism alongside concentrations and fluxes of carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O) from autumn to the following spring along an eight kilometer segment of a low-gradient third order stream in the North Carolina Piedmont. We characterized spatial and temporal patterns of GHG in the context of channel geomorphology, hydrology, and ecosystem metabolic rates using linear mixed effects models. We found that stream metabolic cycling was responsible for most of the CO2 flux over this period, and that in-channel aerobic metabolism was a primary driver of both CH4 and N2O fluxes as well. Long water residence times, limited reaeration, and substantial organic matter from terrestrial inputs foster conditions conducive to the in-stream accumulation of CO2 and CH4 from microbial respiration. Streams like this one are common in landscapes with low topographic relief, making it likely that the high contribution of instream metabolism to GHG fluxes that we observed is a widespread yet understudied behavior of many small streams.