| dc.description.abstract | Tsetse fly control in Lambwe valley has made the region habitable and opened it up for human settlement. The region has, therefore, experienced a steady increase in human population growth. Since most of the households in Lambwe keep livestock as a source of livelihood, the increased human population has consequently resulted in increased livestock numbers and grazing pressure. At the same time, the region is experiencing changes in rainfall patterns which are projected through increased rainfall during wet seasons and prolonged dry periods. Increased livestock grazing and changing rainfall regimes in Lambwe valley Homabay County, Kenya is threatening the existence of plant life forms in Lambwe, but has rarely been studied. The increased grazing and rainfall variability are occurring simultaneously and their implication on the savanna vegetation is not known. Holistic studies that monitor vegetation responses in this valuable ecosystem, are limited. This study analysed the responses of the herbaceous and tree layers to livestock grazing and rainfall variability. The objectives of the study were; i. To determine the interactive effect of livestock grazing and rainfall variability on CO2 exchange, species diversity, composition and palatability of the herbaceous vegetation, ii. To determine the effect of livestock grazing and rainfall variability on CO2 exchange of the tree canopy, iii. To compare productivity responses of tree and herbaceous vegetation under livestock grazing and rainfall variability. Within the herbaceous layer, rain-out shelters were erected above canopy of herbaceous vegetation that was either grazed by livestock or fenced to exclude livestock. The shelters, each measuring 9 m by 6 m were replicated three times on both the grazed and fenced plots. On each of the plots, ambient rainfall was reduced or increased by 50 % using transparent plastic sheets. Within the tree vegetation, three replicates of grazed and fenced plots, each measuring 70 m by 50 m were demarcated for measurements. There was no rainfall manipulation on the tree plots. Measurements included microclimate, soil moisture content, soil bulk density, CO2 fluxes, biomass, species diversity and palatability in the herbaceous layer and sap flow measurements in the trees. Herbaceous Gross Primary Productivity (GPP) was calculated from Net Ecosystem Exchange (NEE) and Ecosystem Respiration (Reco) measured directly with the chambers. Daily GPP for both trees and herbaceous layer was determined using PIXGRO model. Livestock grazing significantly (P ≤ 0.05) reduced herbaceous CO2 exchange but had no influence on tree productivity. The interaction of grazing and rainfall reduction significantly (P ≤ 0.05) reduced GPP, NEE and Reco by 22.5 %, 33% and 39% respectively. The interaction of grazing and rainfall increment significantly increased GPP and NEE by 47 % and 54.8 % respectively, but had no influence on Reco. The interaction of grazing and rainfall reduction significantly (P ≤ 0.05) reduced herbaceous diversity whereas the interaction of grazing and rainfall increment had no significant (P > 0.05) impact on herbaceous diversity. Rainfall manipulation had no significant (P > 0.05) impact on dominance of herbaceous vegetation in either fenced or grazed plots. Herbaceous daily maximum GPP exhibited higher seasonal fluctuations than the tree layer. The overall mean maximum daily ecosystem productivity for the study site was 10.73 ± 1.7 g C m -2 d-1 and 13.75 ± 1.62 g Cm-2 d-1 in the grazed and fenced plots respectively. The herbaceous vegetation, which was the dominant vegetation life form in the valley, contributed about 60 % to the overall mean maximum daily ecosystem productivity. The CO2 fluxes, diversity and composition of the herbaceous layer in the grazed savanna responded disproportionately to the rainfall manipulation, suggesting that grazing and rainfall are not the sole factors driving herbaceous CO2 exchange in this ecosystem. The differential response of trees and herbaceous GPP to livestock grazing was linked to the variations in rooting patterns of the two vegetation types. The significant contribution of the herbaceous layer to the overall ecosystem productivity was connected to higher herbaceous vegetation cover of about 80 %. The results of this study have a significant implication for the theoretical understanding of the tree-herbaceous responses to livestock grazing and rainfall variability. Further, this study provides empirical data that can be used for refining and parameterizing vegetation models that predict future savanna ecosystem responses to livestock grazing and rainfall variability | en_US |
