Determinants of distribution and abundance of two shrub species, Guiera senegalensis and Piliostigma reticulatum, in Peanut Basin, Senegal

The ability to predict and manage the course of landscape-level ecological change and its longer-term consequences on ecosystem functions (e.g. carbon stabilization and soil degradation mitigation) depends on the ability to understand how a particular ecosystem functions and the mechanisms that control the distribution, configuration and abundance of key species. Guiera senegalensis and Piliostigma reticulatum are two native shrub species that are widely found in Sub-Saharan Africa but unrecognized in their potential role in regulating hydrological and carbon cycles in both natural and agro-ecosystems. Our objective was to conduct a study on the determinants of landscape-level distribution and abundance of these shrub species as a basis for ecological modeling and management of this fragile semiarid environment. Formal Recursive Inference Modeling was used to adduce determinants of species presence while logistic regression and geostatistical approaches were used to estimate shrub abundance within their communities. The results showed that distribution of the shrubs is controlled by four factors: geological substrate, mean annual temperature, mean annual rainfall and landform (profile convexity). Relative abundance within the shrub communities is under the influence of mean annual rainfall, maximum annual temperature and elevation (for G. senegalensis) and mean annual rainfall, mean annual temperature, elevation and landform (profile convexity) (for P. reticulatum). Predictive models for shrub distribution and abundance were generally poor, probably highlighting the weakness of statistical models in analysis and quantification of the spatial structure of ecosystems.