Dynamic seasonal nitrogen cycling in response to anthropogenic N loading in a tropical catchment, Athi–Galana–Sabaki River, Kenya
As part of a broader study on the riverine biogeochemistry in the Athi–Galana–Sabaki (A-G-S) River catchment (Kenya), we present data constraining the sources, transit and transformation of multiple nitrogen (N) species as they flow through the A-G-S catchment (~47 000 km 2). The data set was obtained in August–September 2011, November 2011, and April–May 2012, covering the dry season, short rain season and long rain season respectively. Release of (largely untreated) wastewater from the city of Nairobi had a profound impact on the biogeochemistry of the upper Athi River, leading to low dissolved oxygen (DO) saturation levels (36–67%), high ammonium (NH 4+) concentrations (123–1193 μmol L −1), and high dissolved methane (CH 4) concentrations (3765–6729 nmol L −1). Riverine dissolved inorganic nitrogen (DIN; sum of NH 4+ and nitrate (NO 3−); nitrite was not measured) concentration at the most upstream site on the Athi River was highest during the dry season (1195 μmol L −1), while DIN concentration was an order of magnitude lower during the short and long rain seasons (212 and 193 μmol L −1, respectively). During the rain seasons, low water residence time led to relatively minimal in-stream N cycling prior to discharge to the ocean, whereas during the dry season we speculate that prolonged residence time creates two differences comparative to wet season, where (1) intense N cycling and removal of DIN is possible in the upper to mid-catchment and leads to significantly lower concentrations at the outlet during the dry season, and (2) as a result this leads to the progressive enrichment of 15N in the particulate N (PN) pool, highlighting the dominance of untreated wastewater as the prevailing source of riverine DIN. The rapid removal of NH 4+ in the upper reaches during the dry season was accompanied by a quantitatively similar production of NO 3− and nitrous oxide (N 2O) downstream, pointing towards strong nitrification over this reach during the dry season. Nitrous oxide produced was rapidly degassed downstream, while the elevated NO 3− concentrations steadily decreased to levels observed elsewhere in more pristine African river networks. Low pelagic primary production rates over the same reach suggest that benthic denitrification was the dominant process controlling the removal of NO 3−, although large cyanobacterial blooms further downstream highlight the significant role of DIN assimilation by primary producers also. Consequently, the intense nitrification and uptake of N by algae leads to significant enrichment of 15N in the PN pool during the dry season (mean: +16.5 ± 8.2‰ but reaching as high as +31.5‰) compared to the short (+7.3 ± 2.6‰) and long (+7.6 ± 5.9‰) rain seasons. A strong correlation between the seasonal N stable isotope ratios of PN (δ 15N PN) and oxygen stable isotope ratios of river water (δ 18O H2O; as a proxy of freshwater discharge) presents the possibility of employing a combination of proxies – such as δ 15N PN of sediments, bivalves and near-shore corals – to reconstruct how historical land use changes have influenced nitrogen cycling within the catchment, whilst potentially providing foresight on the impacts of future land management decisions.