Phylogeny-based measures of biodiversity when data is scarce: Examples with the vascular flora of Chile and California

The urgency for conserving biodiversity has elicited much research on how to determine its "value" for conservation. The use of phylogenetic diversity (PD) has provided a quantitative framework to assess biodiversity at different taxonomic levels and spatial scales. PD assessments at any spatial resolution need a reasonably well-resolved phylogeny and distribution data for the target taxa. Because of the explosive growth of phylogenetic work, it is becoming easier to get phylogenetic information for a given group. However, distribution data is much harder to obtain. Using examples from world flora, this study explores the correlation between PD-related studies and availability of distribution data. We found that most PD studies in plants have been done in a handful of countries, which correlate with the amount of available distribution information. In order to address the question of whether PD studies should be done in places where information is scarce, we took two recent studies in which PD was calculated - the flora of Chile and California - as examples of poor and good sampling efforts of herbaria data, respectively. We randomly pruned the California database to see if and how spatial patterns of PD change with data depletion. We show that if redundancy (a measure of sampling) is kept at reasonable levels, meaning using larger grid sizes, PD patterns could still be inferred even with 25% of the original data. We argue that these studies are worth doing even with poor data sets, since even coarse PD patterns can point at places where future studies, and conservation efforts, should be focused.