Biogeographical characterisation of Egypt based on environmental features and endemic vascular plants distribution

Abstract

Biogeographical studies are considered effective for investigations on macroecology and evolution, as well as for delineating patterns of endemism and identifying the key drivers influencing the historical distribution of species. Clustering techniques, based on environmental datasets and distribution of endemic species, have been largely used in biogeography and for the definition of endemic-rich regions where conservation actions should be implemented. Despite several earlier studies have dealt with the biogeographical territories in Egypt, none of them have provided a replicable method to support such regionalisation. We proposed a two-step procedure for the biogeographical regionalisation of Egypt consisting in (1) the definition of environmental clusters (based on 24 climatic, three topographic, two soil organic carbon content and pH, and four habitat heterogeneity variables) by using the k-nearest neighbours algorithm and (2) the spatial clustering of such clusters according to the distribution of 140 endemic vascular plants. The hierarchical clustering and indicator values analyses, based on the presence-absence matrix of endemic taxa in each cluster were performed to define two cut-off levels of biogeographical sectors and subsectors. A total of six sectors and nine subsectors were identified. Climatic-related variables, elevation and soil organic carbon were the most important determinants for environmental clustering of Egypt. The highest endemic richness was recorded in the Marioutico-Arishian (71 species), Sinaico-Arabian (62 species) and Nilotic (16 species) sectors, and in Sinaic (54 species), Arishian (45 species) and Marioutic (40 species) subsectors. Nonetheless, the most protected sectors were Elbanian and Suezian (62.73 and 29.05%, respectively) while the lowest sectors were Nilotic and Marioutico-Arishian (9.86 and 13.26%, respectively). The already established protected areas in Egypt are not sufficient for conserving the representation of the identified endemic rich plant sectors. This two-step procedure confirms the usefulness of environmental attributes together with the spatial distribution of endemic vascular plants to define the biogeographical units in Egypt. Our regionalisation method could be replicated for other species’ groups, with the ultimate goal of integration all species of interest in a single biogeographical system. Furthermore, the presented regionalisation will help to identify weaknesses in current protection actions and to understand biogeographical processes.