Interactive influence of rainfall manipulation and livestock grazing on species diversity of the herbaceous layer community in a humid savannah in Kenya

Abstract

Changes in rainfall regime and grazing pressure affect vegetation composition and diversity with ecological implications for savannahs. The savannah in East Africa has experienced increased livestock grazing and rainfall variability but the impacts associated with those changes on the herbaceous layer have rarely been documented. We investigated the effect of livestock grazing, rainfall manipulation and their interaction on the composition and diversity of the herbaceous community in the savannah for two years in Lambwe, Kenya. Rainfall manipulation plots were set up for vegetation sampling; these plots received either 50% more or 50% less rainfall than control plots. Simpson's diversity and Berger-Parker indices were used to determine diversity changes and dominance respectively. The frequency of species was used to compute their abundance and their life forms as determined from the literature. Grazing significantly increased species diversity through suppression of dominant species. Rainfall manipulation had no significant impact on plant diversity in fenced plots, but rainfall reduction significantly reduced diversity in grazed plots. In contrast, rainfall manipulation had no impact on dominance in either fenced or grazed plots. The interaction of grazing and rainfall manipulation is complex and will require additional survey campaigns to create a complete picture of the implications for savannah structure and composition.

Keywords: Grazing; Moist savannah; Plant diversity; Rainfall.

Fig. 1

(A) The layout of rainfall manipulation plots. Parallel arrows show transparent rainout gutters used to simulate rainfall reduction and direction to which excluded rainfall flows (rainfall increment plot/150% rainfall). (B) Photograph of rainfall manipulation plot with a slanting roof partially covered by transparent rainout gutters to exclude 50% of the rainfall and redirect it to the plot designated to receive more rainfall (i.e., 150% rainfall).

Fig. 2

Mean monthly rainfall amount (mm) and soil water content (%) within 0–30 cm soil profile recorded in 2014 (A) and 2015 (B) when measurements were conducted.

Fig. 3

(A and B) Herbaceous plant diversity within grazed and fenced plots in 2014 and 2015 respectively. Interaction of grazing and rainfall manipulation on herbaceous plant diversity (C) and dominance (D) for the entire study period. Bars are means (±SD). Bars within plots not sharing the same letters are significantly different from each other (Tukey-LSD, p < 0.05).