Rangeland management practices and their influence on soil properties, vegetation metrics and wildlife in Lewa Wildlife Conservancy, Kenya

Abstract

Rangeland landscapes, comprising forbs, grasses, and shrubs, are primarily managed for wildlife and livestock grazing. These ecosystems, encompassing various vegetation types like shrublands, grasslands, and savannahs, cover a significant portion of the Earth's landmass. With low and unpredictable annual rainfall, rangelands contribute over 30% to terrestrial net primary productivity, playing a crucial role in natural ecosystems. Given their sensitivity to human activities, effective management interventions are essential for sustaining forage quality and quantity for wildlife. The uneven wildlife utilisation of Lewa Wildlife Conservancy (LWC) rangelands prompted the application of recognised improvement methods, yet their effects remain understudied. This research therefore investigated the impacts of mowing of grasslands and carrying away (MO), prescribed grazing (PG), and unprescribed grazing (UG) on select soil and vegetation chemical elements, above-ground biomass, basal gaps, diversity, and wildlife abundance across 62,000 acres of LWC rangeland in Meru County, Kenya. Data collection was undertaken 18 months after treatment for MO and PG, while UG was continuous. Treated blocks were selected in a random systematic way, where adjacent untreated plots that had the same physical soil and site characteristics as determined using the Land-Potential Knowledge System (LandPKS) application acted as controls. Blocks were divided into 100 m × 100 m grid cells using ArcGIS 10.8.1, where sampling plots were drawn. Data analysis, both descriptive and inferential statistics were done in the R version 4.2.2 environment, with a significance level set at 0.05 (α = 0.05). A two-sample t-test was used for above-ground biomass, basal gaps, and soil and vegetation chemical elements data to discern variations between treatments and their controls. Additionally, a one-way analysis of variance (ANOVA) was utilized to assess the extent of change (treatment minus control) among treatments. Wilcoxon rank-sum test (WRST) was used on diversity data to compare differences between treatments and their controls while the Kruskal Wallis H test was used to compare the magnitude of change between treatments. Duncan's multiple-range test and Conover's all-pairs test were used as post hoc tests for one-way ANOVA and H test respectively. The vegetation P concentration was significantly higher in MO (t = -2.5164, p = 0.0455) but significantly lower in UG (t = 2.6222, p = 0.0399) compared to their controls. Vegetation K was significantly higher in PG compared to its control (t = -3.6222, p = 0.0225). The mean above-ground biomass was significantly lower in MO (t = 4.8861, p = 0.0029) and UG (t = 5.4866, p = 0.0068) compared to their controls while no significant difference was observed between PG and its control (t = 1.1916, p = 0.2867). The mean length of basal gaps of MO (t = 7.0687, p = 0.0001) and UG (t = -4.0531, p = 0.0001) was significantly lower and higher respectively compared to their controls. MO decreased mean basal gaps by a larger magnitude compared to UG where mean basal gaps increased (p = 0.0008). A significantly higher wildlife density was observed in MO compared to its control (t = -4.6696, p = 0.0034) as well as other treatments (F (2, 9) = 5.216, p = 0.0313). In conclusion, this study establishes that various management practices exert distinct effects on rangelands. The significant rise in wildlife densities observed in MO, coupled with its positive impact on several metrics examined, positions it as the most favourable practice. Furthermore, the study recommends time series data be collected to understand changes in these metrics at time intervals and the time in which effects are neutralised.