Integrating Ecosystem Contributions to Stream Corridor Carbon Dioxide and Methane Fluxes

Abstract:

The heterogeneity of CO₂ and CH₄ sources within and across watersheds presents a challenge to understanding the contributions of different ecosystem types to stream corridor carbon cycling. Stream carbon fluxes integrate biogeochemical processes from their contributing valleys and upstream corridors. Changing hydrology and diverse landscape patches (e.g., surface, subsurface, and riparian) can have dynamic influences on stream corridor greenhouse gas emissions. To identify patterns and sources of carbon emissions across stream corridors, we measured gas concentrations and fluxes over 2 summers at Coweeta Hydrologic Laboratory, NC. We sampled CO₂ and CH₄ along four streams (including flowing and dry channels), adjacent wetlands, and riparian hillslopes. Stream CO₂ concentrations varied as much over space as they did time (550-2500 μatm), and all streams were sources of CO₂ to the atmosphere (median from all stream reaches = 93.9 mmol m^-2d^-1). Streams were sources or sinks of CH₄ depending on sampling location (-0.0001 to 0.158 mmol m^-2d^-1). Hillslopes were sources of CO₂ (median 259 mmol m^-2d^-1) and sinks of CH₄ (-0.086 mmol m^-2d^-1); stream dry beds were sources of both gases (median 62 mmol CO₂ m^-2d^-1 and 0.003 mmol CH₄ m^-2d^-1). Wetlands were consistently sources of CO₂ (median 211 mmol CO₂ m^-2d^-1); however, wetland CH₄ emissions were highly heterogeneous (range -1.1 - 2713 mmol m^-2d^-1). Ongoing work seeks to integrate stream discharge with high-frequency dissolved CO₂ sensor data with within-reach spatial CO₂ data to identify spatiotemporal patterns of variation. Future expected hydrologic and climatic extremes will change carbon cycling through watersheds. A better understanding of carbon fluxes from diverse habitat patches within and between stream corridors will improve our quantifications of freshwater contributions to landscape and regional carbon emissions as ecosystems respond to global change.