Abstract:
The importance of space as an ecological factor is an emerging paradigm in community ecology, particularly as a driving force of biodiversity patterns. We analysed β-diversity linked to spatial structure in four communities (tropical dry forest, savanna, xerophytic vegetation, subdeciduous forest) that occur in a tropical complex landscape of southern Mexico. The landscape was described through an object-oriented classification of a Quickbird satellite image. The classification revealed a highly heterogeneous spatial arrangement of the four communities. Global (landscape-level) β-diversity was 0.12 (mean Sørensen index), a value smaller than those observed for the individual communities (0.20–0.41). By using multivariate classic and partial Mantel tests, and Mantel correlograms based on two distance classes, we analysed β-diversity spatial variation related to landscape configuration. The Mantel statistic values for the four communities combined were negative and very similar both for the classic Mantel (rM = −0.23) and the partial Mantel test (rM = −0.19). Correlograms proved significant spatial autocorrelation across most of the analysed distance classes, except in the case of riparian subdeciduous forest. Tropical dry forest and savanna occupy large, highly connected areas in the landscape. Correlograms for these two communities showed decreasing trends, starting at positive, significant autocorrelation values at short distances. The loss of floristic autocorrelation beyond a 5000 m distance for the tropical dry forest may reflect spatially autocorrelated dispersal, and thus a dispersal-limited community. Xerophytic vegetation displayed a distinctly different correlogram, with a wavy shape showing an alternation of positive and negative values, in agreement with the insular configuration of this community in the landscape. We suggest that the routinely incorporation of a spatial approach in community ecology research will help in furthering our understanding of the elusive issues of species turnover across space.