Biogeographical regions and phytogeography of the eucalypts

Aim: To map spatial patterns of species richness, species endemism and species turnover of the eucalypts; to propose a biogeographical regionalization of eucalypts based on species turnover; and to identify the environmental correlates of these patterns.\ud \ud Location: Australia and Malesia.\ud \ud Methods: We analysed 798 eucalypt species (Angophora, Corymbia and Eucalyptus) with distributions across Australia and Malesia using square cells with a resolution of 100 × 100 km. Species richness, endemism and species turnover were calculated. Phytogeographical regions were identified using an agglomerative cluster analysis derived from a matrix of pairwise Simpson's beta (βsim) dissimilarity values. Eleven environmental variables were used to analyse the environmental correlates of species turnover. Non-metric multidimensional scaling (NMDS) of the βsim, Getis-Ord Gi* hotspot spatial statistics and an ordination of the βsim -NMDS were used to investigate the environmental drivers at the continental level and for each of the phytogeographical regions.\ud \ud Results: We identified three centres of species richness and fourteen of endemism, of which several are newly identified. The main centres of species richness agree with previous studies. Six major eucalypt phytogeographical regions are proposed based on the species turnover: monsoon, tropical/subtropical, south-east, south-west, Eremaean north and Eremaean south. These findings are supported by significant environmental differences of the NMDS vectors and the Gi* statistics. The environmental drivers of species turnover are broadly consistent with the continental patterns of summer and winter rainfall below and above the Tropic of Capricorn.\ud \ud Main conclusions: The proposed phytogeographical regions are similar to the Australian biomes. Climate is the main driver of the phytogeographical regions, varying from region to region. Comprehensive bioregionalization frameworks and phytogeography updates, as proposed here, are fundamental for enhancing our understanding of the spatial distribution of biodiversity and therefore benefit global biogeography and help planners to identify regions of high conservation relevance