Estimating the number of species in a community or ecosystem is a fundamental problem in basic and conservation ecology. Basic researchers use biodiversity estimates to study latitudinal diversity gradients, to determine relationships between local and regional diversity, and as a response variable in manipulative experiments. Conservation ecologists use such estimates to prioritize conservation efforts (Myers et al. 2000) and predict species losses due to fragmentation (Pimm and Askins 1995, Brooks et al. 1997). Much of this work relies on the species-area relationship (SAR) to scale up field measurements of diversity to broader spatial scales. However, these estimates ignore the possibility that species number may be as sensitive to the temporal scale of observation as they are to the spatial scale.