Comment on hess-2021-16

Abstract

Nitrate (NO 3 − ) excess in rivers harms aquatic ecosystems and can induce detrimental algae growths in coastal areas. Riverine NO 3 − uptake is a crucial element of the catchment scale nitrogen balance and can be measured at small spatiotemporal scales while at the scale of entire river networks, uptake measurements are rarely available. Concurrent, low frequency NO 3 − concentration and stream flow (Q) observations at a basin outlet, however, are commonly monitored and can be analyzed in terms of concentration discharge (C-Q) relationships. Previous studies suggest that more positive log(C)-log(Q) slopes under low flow conditions (than under high flows) are linked to biological NO 3 − uptake, creating a bent rather than linear log(C)log(Q) relationship. Here we explore if network scale NO 3 − uptake creates bent log(C)-log(Q) relationships and when in turn uptake can be quantified from observed low frequency C-Q data. To this end we apply a parsimonious mass balance based river network uptake model in 13 mesoscale German catchments (21-1450 km²) and explore the linkages between log(C)log(Q) bending and different model-parameter combinations. The modelling results show that uptake and transport in the river network can create bent log(C)-log(Q) relationships at the basin outlet from log-log linear C-Q relationships describing the NO 3 − land to stream transfer. We find that the bending is mainly shaped by geomorphological parameters that control the channel reactive surface area rather than by the biological uptake velocity itself. Further we show that in this exploratory modelling environment, bending is positively correlated to percentage NO 3 − load removed in the network ( . ) but that network wide flow velocities should be taken into account when interpreting log(C)-log(Q) bending. Classification trees, finally, can successfully predict classes of low (~4 %), intermediate (~32 %) and high (~68 %) . using information on water velocity and log(C)-log(Q) bending. These results can help to identify stream networks that efficiently attenuate NO 3 − loads based on low frequency NO 3 − and Q observations and generally show the importance of the channel geomorphology on the emerging log(C)-log(Q) bending at network scales.and McGlynn, 2018). Increased NO 3 − concentrations in surface waters can induce detrimental algae growths (Beusen et al., 2016;Canfield et al., 2010;Galloway et al., 1995), compromise river ecosystem health and jeopardize drinking water supplies.Since the beginning of the 20 th century, human activities such as agricultural expansion and fossil fuel burning have mobilized additional reactive nitrogen (N), initiating and later exacerbating this problem (Seitzinger et al., 2002). In arable landscapes, which include large parts of Europe, the efficient management of aquatic NO 3 − at network scales is complicated by the spatiotemporal variability of loading patterns and hydrologic regimes as well as the lack of understanding of nutrient pathways, connected transit times and removal processes from in...