BIODIVERSITY AND CLIMATE 189Fig. 8. The relationship between North American amphibian and tree species diversity, reflecting the
dependence of both on water availability and temperature.environment. At the coarsest physiological scale
this resolves itself into the distinction between
the ectothermic and endothermic components
of non-avian tetrapod diversity, which is shown
plotted against mean annual temperature
(MAT) in Figure 11. This reveals a strong linear
relationship, with no hemispheric asymmetries
or other regional heterogeneities, and no appar-
ent signal in the residuals (Fig. 12). This is
supported by the high rho values in Table 7.
An additional question is the extent to which
taxonomy masks the physiological diversity
signal. Figure 13 shows the results of a corre-
spondence analysis of modern non-avian tetra-
pod genera in North America and Europe and
these results are presented geographically in
Figure 14. The correlations between each axis
and the environmental variables are assessed
using the Spearman rank test and are listed in
Table 8. These results show how the first axis,
which accounts for 37.2% of the variance, is
dominated by the historical biogeographical
differences between North America and Europe
(this is not seen when Europe and North
Fig. 9. The distribution of mammal species diversity for North America, Europe, southern Africa, Arabia,
tropical South America and Australia.