CORRECTING FOR THE SPECIES-AREA EFFECT WHEN ESTIMATING PALEODIVERSITY

The impact of the species-area effect, the positive relationship between species richness and the area of geographic sampling, has been well-documented among extant organisms. However, the effects of this relationship on estimates of paleodiversity have never been fully explored, complicating our understanding of evolution, biogeography, and the response of ecosystems to climate change. We used the MIOMAP database of western North American mammals to assess how the species-area effect influences estimates of paleodiversity. Species lists were compiled for the entire USA and by geographic region for each of the fifteen subdivisions of the North American Land Mammal Ages from the Arikareean through the Hemphillian. All resulting species richness numbers were rarefied to account for sampling biases, such as uneven sampling, preservation biases, and the varying length of time bins. Using the online MIOMAP mapping program, the geographic area encompassed by each set of samples was calculated by plotting the points and drawing the area of the smallest polygon that would surround all of the samples per time bin. Plots of rarefied species versus geographic area revealed a clear species-area bias. We show how one can use these plots to correct traditional species diversity curves. Accounting for the species-area bias shows that only a few of the peaks and valleys in the mammalian paleodiversity curve remain. We investigate whether these peaks likely reflect biotic events or result from additional sampling biases in the mammalian fossil record, and then interpret how the remaining biodiversity peaks and valleys might relate to major tectonic and climatic events during the late Oligocene and Miocene.