Drivers of water quality in Afromontane-savanna rivers

Although several studies have investigated the relationships between water quality in rivers and the types of land use within their catchments, many aspects of these relationships remain unclear in Afromontane-savanna rivers, especially the interactions between catchment land use, seasonality and stream size. Afromontane-savanna catchments present a unique situation where headwater regions and lowlands have experienced more dramatic land cover change, but mid-elevation regions remained rather natural. We examined the influence of seasonality, catchment land use and stream size, including their interactions, on water physico-chemistry, nutrients and major ions in the Afromontane-savanna Mara River in Kenya, using data collected from 2010 to 2018 at >150 sampling sites in the Kenyan part of the river. We developed generalized linear mixed models (GLMMs) to explore the influence of seasonality (dry and wet seasons), land use (forest, mixed, agriculture and grasslands), stream size (stream orders 1–7), and their interactions on river water quality. Water quality variables included physico-chemical measures (pH, dissolved oxygen [DO] concentration, temperature, electrical conductivity, total dissolved solids [TDS], turbidity, total suspended solids [TSS] and particulate organic matter [POM]), nutrients ( NH4+ , NO3− , total dissolved nitrogen [TDN], total nitrogen [TN], soluble reactive phosphorus [SRP], total phosphorus [TP] and dissolved organic carbon), and major ions (Cl−, F−, SO42− , Na+, K+, Ca2+, Mg2+, Fe2+, HCO3− and Si). There were clear differences in average values of most water quality variables among land uses with sites in savanna grasslands having high levels of major ions, ammonium and P, while agricultural sites had higher dissolved fraction of N (except ammonium). Stream order was a poor predictor of water quality, and most parameters did not display any relationship (either linear or non-linear) with stream size. Our results can be used to efficiently enhance water quality by developing strategies for stream restoration and management based on the predomination type of land use in the catchments.