Author(s): D. Tonina; A. Marzadri; A. Bellin; M. M. Dee; J. L. Tank

Linked Author(s): Daniele Tonina

Keywords: Nitrous oxide; River networks; Drought; Denitrification; Land use

Abstract: Climate change is expected to increase extreme events, droughts and floods, in most parts of the world. In this climate warming scenario, expected future droughts may increase riverine emissions of nitrous oxide (N2O), a potent greenhouse gas. Global warming is expected to create changes in extreme weather and climate events: floods and droughts across the globe. In earth's climate history periods of severe droughts are a regular part of nature's cycle; such events are predicted to become more frequent and severe in the coming decades due to climate change. Moreover, these events may interact nonlinearly with other anthropogenic effects, such as the release of known and new contaminants, possibly influencing stream biogeochemical processes. Among solutes, dissolved reactive nitrogen, especially nitrate (NO3), could play an important role because it can fuel an important source of the potent greenhouse gas N2O via the process of microbially-mediated denitrification. Further, N2O is well recognized as one of the most important greenhouse gases responsible for the stratospheric ozone destruction; but estimates of its emissions at the watershed-scale are highly uncertain because of the difficulties to extrapolate local measurements to the river network. With the objective to fill this knowledge gap, we build on a previously proposed 3-equation Damköhler-based scaling law to investigate the effect of droughts on N2O emissions from streams and rivers by using the data recently published by Audet et al. Then, we apply the model to understand the effects of drought on the seasonal and spatial variability of N2O emissions along two Midwestern US rivers with contrasting land-use land-cover: the Manistee River (MI; ~83% forested) and the Tippecanoe River (IN; ~82% agricultural). We combine numerical modeling and geostatistical analysis to upscale network-scale empirical data of N2O emissions collected over four seasonal synoptic sampling campaigns, with the focus of the analysis being emissions during droughts. We investigate the spatial distribution of dissolved inorganic nitrogen and N2O emissions along the two stream networks by comparing the behaviors of streams of different sizes (e.g., headwater stream vs. river mainstem), the role of land use, and contrast surface and subsurface (e.g., benthic-hyporheic zone) stream habitats among different seasons and thermal regimes. The effect of the land use land cover enters in the 3-equation model we propose via the dissolved inorganic nitrogen (DIN) flux (FDIN0) which shows mixed responses along streams and rivers during drought and in its effect on stream and watershed morphology via the Damköhler number. In forested watersheds, DIN concentration is expected to decrease during drought, mainly because of a reduction in catchment inputs and accordingly FDIN0 is expected to decrease with decreasing discharge; as observed along the Font del Regàs stream in Spain which is forested over 90% of its area. In contrast, in watersheds dominated by agricultural land use, DIN concentration is not expected to decrease during droughts, mainly because of agricultural point sources and accordingly FDIN0 is expected to remain nearly constant as water discharge decreases; as observed along the San Joaquin River (California, USA). Results show how the interplay between network topology (i.e., network structure and channel density) and land use land cover strongly control the behavior of the two river networks in contributing to climate change impacts during drought. Along the Tippecanoe River, where network structure reflects its agricultural footprint, with higher NO3 inputs and a high channel density, N2O emissions are high in both the headwaters and in the mainstem, with emissions that are expected to increase during drought. Conversely, along the forested Manistee River, decreased NO3 inputs, combined with a lower channel density, result in a quasi complete transformation of NO3 within the headwaters with undetectable effects of drought on N2O emissions.

DOI: https://doi.org/10.3850/978-981-11-2731-1_119-cd

Year: 2018