Understanding the relative importance of environmental and spatial processes in determining geographic variation in biodiversity remains a fundamental pursuit of macroecology and biogeography. In particular, studying patterns of β diversity is receiving renewed emphasis as a means to understand forces structuring ecological communities. A common method in this context is to use linear (or log-linear) models to relate species turnover to distance and then to compare how rates of turnover (measured as the slope of the regression) vary between regions and/or taxa. Implicit in this approach is an assumption that rates of species turnover remain constant along environmental gradients. We used a non-linear method, Generalized Dissimilarity Modeling, to examine the relative importance of environment and space in determining patterns of beta diversity for groups of plants that differ in seed dispersal mode and we compared these patterns between two regions with contrasting evolutionary and climatic histories: southwest Australia (a global biodiversity hotspot) and northern Europe. We predicted that spatial processes should exert greater influence (1) in Europe where climatic changes since the last ice age has been comparatively extreme and (2) for ostensibly dispersal limited plants, such as those dispersed by passive means or by ants.
Results/Conclusions
Consistent with differences in the climatic histories between the two study regions, we found that the relative importance of space in determining patterns of β diversity was more than three times higher in northern Europe than in southwest Australia. Similarly, temperature was the most important determinant of species turnover in northern Europe, whereas precipitation was the dominant gradient in southwest Australia. In contrast, we found no evidence that differences in dispersal ability (as inferred from seed dispersal mode) produced predictable differences in the extent to which space versus environment explained patterns of β diversity for different groups of plants in either flora. More broadly, we found that relationships between β diversity and both environmental and geographic distance were often highly non-linear and varied as a function of the gradient considered and by dispersal mode. This latter finding calls into question the use of linear models to understand and compare patterns of β diversity between regions and taxa.