2005 Salt Lake City Annual Meeting (October 16–19, 2005)

Paper No. 13
Presentation Time: 11:15 AM


SMITH, Krister T., Geology & Geophysics, Yale University, 210 Whitney Ave, New Haven, CT 06511, krister.smith@yale.edu

One major difficulty in discerning true patterns in past biodiversity is the problem of deconvoluting spatial from temporal patterns in the fossil record. But understanding how species richness changes through space—beta diversity, broadly construed—is itself an important desideratum in ecology, and one from which the search for global patterns has distracted us. The fossil record presents a huge repository of diachronic data that could meaningfully be accessed for testing ecological hypotheses, yet we lack many tools that would help us do so.

A new metric for beta diversity is presented. The method is predicated on the observation that the fossil record consists of more or less isolated localities. Numerical methods are used to estimate the path of minimum length that connects all coeval localities of interest (by analogy with species-area curves). Either specimen data or interpolated range-maps may be used for species occurrence data. Cumulative species richness and cumulative length are then measured along the path. Remarkably, for the data sets examined, the relationship between species richness and path length appears to be linear, and I suggest that the slope of this line be taken as beta diversity.

The method is applied to the Australian squamate fauna, which generally shows high alpha and low beta diversity, and to the mammal fauna of the United States, demonstrating that the metric should be meaningful in modern ecological studies. Application to the fossil record, like the use of modern specimen data, requires only that the data be normalized by a simple rarefaction-like procedure. The new metric should contribute to the further unification of paleo- with neoecology and promote the expansion of ecological theory into the past.