Seasonal Effect of Climate Variability on Soil Moisture and Biomass Production: a Case Study of Maasai Mara Rangeland and Naivasha Cropland Ecosystems, Kenya

Abstract

Globally, rangeland and cropland are ecosystems that rely exclusively on soil moisture influenced by dynamic interaction of eco-hydrologic variables caused by climate variability and soil for sustainable biomass/yield production. There are challenges of up to date variable measurements within adequate information of spatio-temporal soil moisture variation and biomass in the selected Kenyan ecosystems. The specific objectives under the study were the determination of spatio-temporal soil moisture storage and retention capacities in the ecosystems; Simulation of the influence of bi-seasonal soil moisture variability on rangeland and cropland biomass yield using coupled Hydrus-1D and Agricultural Production Systems Simulator. Finally, to analyse the impact of bi-seasonal soil moisture variation on land use land cover in rangeland and cropland vegetation. Remote sensing and Geographical Information System derived land use land cover classification maps from Normalized Difference Vegetation Index values for real-time monitoring were obtained and processed via MODIS and Proba-V imagery satellite data. Random undisturbed core soil samples collected from ten (10) sampling points with five varying replication depths of P1 (0-5cm), P2 (5- 10cm), P3 (15-20cm), P4 (35-40cm) and P5 (75-80cm) for ground based (in-situ) installed 5TM-ECH2O probes. Time series variation shows that volumetric water content of spatially distributed probes in wet season ranged between 0.11 and 0.32m3m-3 (0.16m3m-3) and in dry between 0.04 and 0.17m3m-3(0.11m3m-3) across the rangeland respectively. Cropland volumetric water content in wet season ranged between 0.13 to 0.37m3/m3 (0.22m3m-3) and dry between 0.06 to 0.22m3m-3(0.14m3m-3) respectively. Water retention shown that field capacity of soil water content at -3 bars ranged between 0.16cm3H2O/cm3soil and 0.22cm3H2O/cm3soil across the rangeland. APSIM model simulated cropland and rangeland above ground biomass reasonably well, where rangeland model performance gave NSE = 0.988, r = 0.000, RMSE = 0.103tonha-1 and R2 was 0.988. In overall, the rangeland covers approximately 717.203km2(46.75%) with total above ground grass biomass in dry and wet season of 35.094 tonha-1( 2,516,952.208 and 42.123 tonha-1 ( 3,021,074.197) tonnes per season respectively. Land use land cover change indicates gradual encroachment of livestock and commercial wheat farms into the grassland in the last decade (2009-2019). This has decreased (closed, evergreen broadleaved) forest cover while conversion of Naivasha cropland from rain fed to irrigated cropland is also gradually increasing. In conclusion, soil moisture, biomass and change in land use land cover vary seasonally as influenced by climate variability.