Nashon K. R. Musimba

Competition indices of three perennial grasses used to rehabilitate degraded semi-arid rangelands in Kenya

Rangeland degradation is one of the most serious environmental challenges in arid and semi-arid environments in Africa. Grass reseeding technology, using indigenous perennial grass species has a great potential for restoring ecosystem functionality in these degraded lands. This study was aimed at investigating how the establishment of indigenous grasses used to rehabilitate degraded semi-arid rangelands in Kenya as monocultures or binary mixtures affect dry matter biomass yields and plant competition indices at different phenological stages of development: seedling, elongation and reproduction. Using plant competition indices, we demonstrated which among these grass species showed a complementary vis-a-vis competitive interaction in binary mixtures across their phenological stages. Monocultures and binary mixtures of three perennial grass species: Cenchrus ciliaris L. (Buffel grass), Enteropogon macrostachyus Munro ex Benth. (Bush ryegrass) and Eragrostis superba Peyr. (Wilman lovegrass) native to these semi-arid rangelands were established in a field experiment. Pure stands of E. macrostachyus yielded higher biomass than C. ciliaris and E. superba only at the seedling and elongation phases but not the reproduction phase, where the yields were lowest among pure stands. Negative plant interactions in binary mixtures were observed at seedling and reproduction stages, compared with positive interactions observed at the elongation phase. Binary grass mixtures had a significant advantage for exploiting available environmental resources only at the elongation phase but not at the seedling and reproduction phases. Based on the estimated competition indices, E. superba best complements C. ciliaris across the phenological stages. We therefore conclude that this binary mixture is the one most likely to ensure successful rehabilitation of degraded semi-arid rangelands in this region.

Range Use and Trophic Interactions by Agropastoral Herds in Southeastern Kenya

Abstract Habitat utilization patterns and feeding interaction of free ranging agropastoral herds were investigated in two cycles of four consecutive grazing periods, in a semiarid environment, southeastern Kenya. The bites count and herd locations per area methods were used. During the dry season, areas of concentrated drainage; river valleys, bottomlands and ephemeral drainage ways absorbed a greater feeding load, taking 57.1 to 60% of the grazing time by the animals. In contrast, areas of limited moisture concentration, the open sandy/clay plains, were mainly exploited in the wet season and accounted for 52.6 to 55.6% of the grazing time. The trophic interaction patterns indicated that goats and cattle had a seasonal mean diet overlap index of less than 0.5 for all forage classes. Sheep and cattle, and sheep and goats had a seasonal mean diet overlap index of greater than 0.5 on grass and forbs, and browse and forbs, respectively. This indicated that during periods of resource scarcity, sheep and cattle or sheep and goats could become competitive feeders for same feed resources. Grazing management strategies aimed at even distribution of grazing pressure and enhancing complementary trophic interactions could be central to sustainable livestock production in such environments.

Determinants and rates of land degradation: Application of stationary time-series model to data from a semi-arid environment in Kenya

The causes of land degradation in the African drylands have been shown to vary. Some researchers consider climate to be the major contributor to degradation, with anthropogenic factors playing a minor role. Others reverse the significance of these two factors. A third group attributes land degradation to climate and anthropogenic factors equally. This study was undertaken to establish the factors influencing land degradation in a semi-arid environment in southeastern Kenya and the rate of change in vegetation types for a period of 35 years (1973–2007). The reduction in grassland cover was used as an indicator of land degradation. Causes of land degradation were determined by a multiple regression analysis. A log-linear regression analysis was used to establish the rate of vegetation change. The multiple and log-linear regression analyses showed: (1) woody vegetation, livestock population and cultivated area to be the main contributors of reduction in grassland cover in the area, and (2) an increase in undesirable woody species, livestock population and cultivated area had a significant (P<0.05) negative effect on grassland vegetation. Increased human population, low amounts of rainfall and drought showed no significant negative effect on grassland vegetation cover. In conclusion, human and livestock population growth and increased agricultural land have contributed to intensive crop cultivation and overgrazing in the semi-arid lands. This overuse of the semi-arid rangelands has worsened the deterioration of the natural grassland vegetation.

Land Use and Spatial Distribution of Two Gum And Incense Producing Tree Species In The Blue-nile Valley of Wogidi District, Ethiopia

Abstract The study was conducted in the Blue-Nile valley of Wogidi district in Ethiopia, where woody vegetation degradation associated with intensive land use activities has become a national concern. It was undertaken with an overall objective of documenting the impacts of land use and topography on the spatial distribution of two gum and incense producing tree species, Boswellia papyrifera (Del.) Hochst and Commiphora africana (A. Rich.) Engl. Five land use units, each with two slope categories (⩽ 25% and > 25%) were subjected to a 5x2 factorial arrangement and a total of 50 sample points were employed to measure density, crown cover, and age ratio of trees. Analysis of vegetation attributes revealed that, except for density of C. africana, all the vegetation attributes did show lower status in the heavily populated land use units than the control site reflecting the differential impact of intensive land use pressure against protection. In the heavily populated land use unit (Luu-3) for example, density, cover and age ratio of B. papyrifera trees were as low as 8.3 trees/ha, 7.7% and 0.21 respectively. In contrast to this, i.e. in the site of low population pressure (the control) the same attributes appeared to be measured as 77 trees/ha, 7.7% and 0.51 respectively. However, attached to the proximity ecological position, density and crown cover of B. papyrifera showed no significant difference (P. > 0.05) between Luu-1 and Luu-2. Density of C. africana, most likely due to its affinity to a specific microhabitat, it did not differ among all land use units (P > 0.05). Between the slope aspects of the area, B. papyrifera and C. africana were found to dominate the steep (>25%) and gentle (⩽ 25%) slopes respectively. Regarding density of B. papyrifera, while it was 196 trees/ha in the steep slope, 146 trees/ha were documented from the lower slope aspect. Inversely related to this, C. africana had 68 and 102 trees/ha on >25% and ⩽25% slopes respectively. Age ratio analysis of B. papyrifera trees, however, did not show significant difference (P > 0.05) between these two slope categories. In the study area therefore, land use and slope were identified as determinant factors to the spatial distribution of woody vegetation and policy frameworks to account these factors are suggested to be substantial in the management of such resources.