Patterns and drivers of nitrous oxide concentrations and fluxes in headwater streams of the Mara river basin, Kenya

Headwater streams are important inland aquatic ecosystems when it comes to their contributions to the global N2O (a potent greenhouse gas) fluxes. It is estimated that they contribute approximately 10% to the global anthropogenic N2O emission rate. Despite this, flux estimates from them are still poorly constrained and with a lot of uncertainties. Factors, such as stream geomorphology and land use/cover, influence water quality. Variation in nutrient and carbon concentrations have also been studied as controls of these fluxes from the streams. However, processes that regulate their production rates are still not well understood. In Africa, few studies have been done to quantify these fluxes from inland waters and there has been very little focus on headwater streams. This research sought to quantify N2O concentrations and fluxes from the headwater streams of the Mara River in Kenya, access their spatial and temporal variation and determine their relationship with both physical and chemical drivers. A total of 52 sites were sampled in the upper Mara of which 47 were streams of orders 1- 6 across 3 land use types (agricultural, forest and livestock). Water samples collected with the gas samples were analysed for DIN, TDN and DOC. N2O concentrations and fluxes found ranged from 0.33 - 2.6 µg L-1 N2O-N and -11 -323 N2O-N µg m-2 h-1 respectively. Agricultural stream sites had the highest mean fluxes of 26.38 ± 5.37 N2O-N µg m-2h-1 while forest sites exhibited a lower mean of 14.75 ± 42.40 N2ON µg m-2h-1. Analysis was done using linear mixed models to identify relationships between N2O fluxes and concentrations with their physical and chemical drivers. Stream slope and velocity did not significantly explain variation in N2O concentration from the studied sites. However, discharge had a positive relationship to N2O concentration. In stream NO3-N and CO2-C (aq) concentrations were strongly positively correlated with N2O explaining > 50% of their variation. However, DO and C:N ratio showed a negative correlation with N2O concentration. From the results, I hypothesised that nitrification may be the main biogenic process controlling N2O production via autotrophic consumption of dissolved CO2 as an energy source. However, more research is required in order to test this hypothesis with much more accurate measurements of denitrification and nitrification rates