Ecological impacts of Prosopis invasion in Riverine forests of Kenya

Drylands occupy over 41% of the global land surface, with Africa and Asia accounting for 32% of the global total each. Because of poor resource management, resource overexploitation and periodic droughts, drylands have experienced severe land degradation. Land degradation is manifested in vegetation loss or deterioration, soil erosion and salinization of soil and water. In Kenya, drylands occupy over 87% of the land surface, and support about 30% of the national population, over 70% of national livestock and the bulk of wildlife that support the tourist sector. Following the prolonged sahelian droughts of the 1970’s Kenya’s drylands were seriously degraded through extensive loss of ground vegetation cover; thus threatening the survival of local populations, livestock production and sustenance of tourism sector. Subsequently, exotic trees and shrubs were introduced for land rehabilitation and fodder supply. Trees from from Prosopis genus emerged as the most adapted and were widely planted. Since introduction, Prosopis species have spread from target rehabilitation sites and invaded riverine and wetlands ecosystems but invasion mechanisms and impacts are not yet well understood. In this study we combined geographical information systems techniques; field, greenhouse and laboratory studies, to evaluate riverine habitat invasibility, invasion impacts, invasiveness of Prosopis species and the composition of invasive Prosopis species in Kenya. The following questions were addressed: 1) What abiotic factors make riverine forests vulnerable to Prosopis invasion?, 2) What are the ecological implications of Prosopis invasion in riverine forests?, 3) What mechanisms underlie inhibition of A. tortilis regeneration by Prosopis species invasion?, 4) What are the species composition in Prosopis invaded areas of Kenya, and 5) What are the implications of our results? The present study revealed indiscriminate Prosopis invasion in all land cover and land use types identified through satellite image analysis, field surveys and historical site information provided by local informants. As a result of this trend, we found a contrasting occurrence increase of Prosopis species and decrease of Acacia tortilis between 1998 and 2007. Accordingly, the study has demonstrated that Prosopis species invasion in the Turkwel Riverine forest is invoked more by species invasiveness rather than habitat susceptibility. Consequently, we investigated the invasiveness of Prosopis species by studying invasion impacts and the underpinning mechanisms Our study has shown reduction of herbaceous species ground vegetation cover and herbaceous species diversity, and termination of A. tortilis regeneration by Prosopis invasion. The negative regression coefficients found between herbaceous species ground cover or between herbaceous species diversity and Prosopis canopy dummy, clarifies the partial direct negative effect of Prosopis on herbaceous species. We corroborate this finding by greenhouse studies that show stronger inhibition A. tortilis and Prosopis seed germination by increasing the concentration of fresh Prosopis litter than by increasing the concentration A. tortilis litter in the soil. Indeed, our study demonstrates potential of seed germination termination at 50% fresh Prosopis litter concentration in the soil. After one month of watering of soil-litter mixture, we found no litter effect on seed germination. Since water leaching decreased the concentration of soluble phenols and leached litter had no effect on seed germination, our study has clarified that the inhibition of A. tortilis regeneration by Prosopis canopy was partially the result of allelopathic effect of Prosopis litter on A. tortilis seed germination. There has been great confusion on Prosopis species identity in Prosopis invaded areas of Kenya, because of similar morphology and introduction of several species within sites. Species misidentification may hamper invasion management. In this study we used Random Amplified DNA markers to differentiate species according to sites. Our study shows that only one species or a hybrid is adapted to any one site, despite the number of species that were introduced to any site. We have further clarified that P. juliflora and its hybrid are the most invasive germplasm in Kenya. However, P. juliflora and the hybrid trees tended to have similar tree characteristics in riverine forests and wetlands as we could predict tree volumes in wetlands from equations developed from a distant riverine site. Our study demonstrates potential for perpetual replacement of A. tortilis by Prosopis species in riverine ecosystems. A notable consequence is reduction of both herbaceous species productivity and diversity. Since both A. tortilis and herbaceous species are used for fodder; invasion may have severe consequences on the pastoral economy but this can be reversed by intensified utilization of Prosopis biomass for fuelwood and pods for fodder.