The mid-domain effect cannot explain the diversity gradient of Nearctic birds

A recent explanation for diversity gradients proposes a 'null model' based on how species ranges are constrained by the geometry of bounded domains. We conduct a test of this hypothesis by comparing patterns predicted by two two-dimensional geometric models against observed diversity patterns for terrestrially feeding Nearctic birds. Consistent with previous tests in two-dimensional space, we find empirical support for the hypothesis to be very weak. We also point out a fundamental conceptual flaw in the hypothesis with respect to the key assumption that ranges can exist independently of the environment in which they are embedded that undermines its theoretical basis as well. We conclude that the mid-domain effect has little empirical support and no theoretical support for its existence, and recommend that it be eliminated as a potential explanation for diversity gradients.