The Publications Warehouse does not have links to digital versions of this publication at this time
Analysis of high resolution diversity data for Mississippian corals in the western interior United States yielded mild latitudinal diversity gradients despite the small geographic area covered by samples and a large influence on diversity patterns by geographic sampling intensity (sample bias). Three competing plate tectonic reconstructions were tested using the diversity patterns. Although none could be forcefully rejected, one reconstruction proved less consistent with diversity patterns than the other two and additional coral diversity data from farther north in Canada would better discriminate the two equivalent reconstructions. Despite the relatively high sampling intensity represented by the analyzed database, diversity patterns were greatly affected by sample abundance and distribution. Hence, some effort at recognizing and accounting for sample bias should be undertaken in any study of latitudinal diversity gradients. Small-scale geographic lumping of sample localities had only small effects on geographic diversity patterns. However, large-scale (e.g., regional) geographic lumping of diversity data may not yield latitudinally sensitive diversity patterns. Temporal changes in coral diversity in this region reflect changes in eustacy, local tectonism, and terrigenous sediment flux, far more than they do shifting latitude. Highest regional diversity occurred during the interval when the studied region occupied the highest latitude. Therefore, diversity data from different regions may not be comparable, in terms of latitudinal inference. Small-scale stratigraphic lumping of the data caused a nearly complete loss of the latitudinal diversity patterns apparent prior to lumping. Hence, the narrowest possible stratigraphic resolution should be maintained in analyzing latitudinal diversity gradients.
Additional Publication Details
Mississippian coral latitudinal diversity gradients (western interior United States): Testing the limits of high resolution diversity data