Real and random patterns associated with molluscan spatial and temporal distributions

Abstract

The species-level properties of geographic range and geologic duration are often used as variables in evolutionary studies. However, estimates of species duration are not independent of estimates of geographic range. Before these properties are used in macroevolutionary hypotheses, error associated with these estimates must be quantified. This error may lead to spurious inferences of evolutionary processes. To assess the error associated with estimates of geographic range and geologic duration, we modeled various sampling regimes and calculated the bias associated with these estimates.We present three analyses which document the bias associated with estimates of geographic range and geologic duration. First, we find a positive correlation between local abundance and geographic range for a sample of 180 species of Recent prosobranch gastropods from the northeastern temperate Pacific Ocean. Therefore, geographically short-ranging species are less likely to be represented in the fossil record than geographically long-ranging species because of their local rarity. Second, we demonstrate that the chance of underestimating the geographic range of a species is acute for species with restricted spatial distributions, further compounding the problem of documenting their distribution in space and time. Third, we present a simulation which quantifies the degree of autocorrelation between geographic range and geologic duration for different levels of sampling resolution and spatial distributions of fossil localities.