Assessing the Influence of Land Use and Catchment Characteristics on Stream Water Chemistry in the South-West Mau, Kenya, Using Spatial Sampling and High Resolution In-situ Monitoring

Monday, 6 June 2016
Suzanne Robin Jacobs1,2, Björn Weeser3,4, Lutz Breuer3,4, Klaus Butterbach-Bahl1,5 and Mariana Rufino2, (1)Karlsruhe Institute of Technology (KIT), Institute of Meteorology and Climate Research, Atmospheric Environmental Research (IMK-IFU), Garmisch-Partenkirchen, Germany, (2)Center for International Forestry Research, Bogor, Indonesia, (3)Centre for International Development and Environmental Research, Justus Liebig University, Giessen, Germany, (4)Institute for Landscape Ecology and Resources Management (ILR), Justus Liebig University, Giessen, Germany, (5)International Livestock Research Institute, Nairobi, Kenya
Abstract:
Forests play an important role in maintaining water quality and evidence suggests that deforestation and land use (LU) change can significantly alter stream water chemistry. However, other catchment characteristics, like geology, topography and climate could also influence the chemical composition. The headwater area of the Sondu basin, located in the South-West Mau, Kenya, has undergone significant change (25% forest loss in the last decades), whereby native tropical montane forest has been converted to smallholder agriculture and commercial tea plantations. Understanding the consequences of such changes for hydrochemistry is essential to manage and maintain water supply and quality in the area.

Three sub-catchments (27–36 km²) with different LU (natural forest, tea/tree plantations and smallholder agriculture) in a 1023 km² headwater catchment have been instrumented with automatic monitoring equipment, recording electrical conductivity (EC), total and dissolved organic carbon (TOC/DOC), nitrate (NO3-N) and turbidity at a 10-min interval. Preliminary wavelet analysis for identification of temporal patterns in NO3-N concentration suggests different behaviour in each of the sub-catchments, with diurnal patterns being more pronounced in the natural forest catchment and timing of daily peaks shifting throughout the year. Comparison of time series and analysis of specific rainfall events also show different responses of C and N to rainfall in the sub-catchments.

High resolution temporal data is complemented by snapshot sampling campaigns of 16 sub-catchments (4.2–103.9 km²), covering a range of LU types, elevations and other physical characteristics. First results suggest that LU influences NO3-N concentrations, while EC and pH are more influenced by geology. Further analysis of the physical parameters will be carried out to identify the importance of the different factors in determining stream water chemistry through the application of Classification and Regression Trees (CART).

<< Previous Abstract | Next Abstract >>