Hydro-geomorphic controls of greenhouse gas fluxes in riparian buffers of the White River watershed, IN (USA)

Abstract

Riparian ecosystems are defined by the nature and regularity of the interactions between a given river system and its floodplains, and past studies have often presented vegetation cover as the exclusive expression of these interactions. There has been to our knowledge, no systematic attempt at linking greenhouse gases (GHG) fluxes and types of riparian buffers. The present study was conducted to investigate the intensity and seasonality of carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O) fluxes in riparian buffers in three common hydro-geomorphic settings (HGM) across the White River watershed (Indiana, USA). These classes included riparian sites located: (i) in till plain depressions near 1st order streams (HGM-1), (ii) in incised narrow valleys with thin alluvium layers above glacial till (HGM-2), and (iii) along 3rd–4th order streams in broad floodplains with thick alluvial and glacial outwash deposits (HGM-3). For each class, 3 sites were selected and GHG fluxes were measured during the wet (May) and dry seasons (August). Strong relationships were found between GHG fluxes, soil properties and environmental factors, but these relationships varied with season and gas species, making it challenging to rely on these relationships for GHG fluxes upscaling. Analysis of variance and discriminant analysis showed that the HGM-defined riparian buffers were distinct in terms of GHG flux intensity. Regardless of season, the HGM-1 sites emitted CO2 at rates 1.6 times higher than at the other sites, likely due to difference in soil C quality. During the wet season, N2O emission was significantly higher at the HGM-3 than at the other sites (0.88 vs 0.27 mg N m− 2 d− 1), and was negatively related with the gradient of the adjacent channel (r2: 0.69). The riparian buffers acted as CH4 sinks, with the HGM-2 sites exhibiting CH4 uptake rates significantly greater than the other riparian types (− 0.80 vs − 0.34 mg CH4-C m− 2 d− 1). The consistency of these results underscores the potential of an HGM-based monitoring approach to derive watershed-scale GHG budgets for riparian buffers.