Effect of the Focal Taxon on the Selection of Nature Reserves

Among the numerous site—ranking or reserve selection approaches is to select reserves based on the current distribution of a suite of species. This approach only requires distribution data for the species in the "focal taxon." To what extent is the proposed reserve network dependent on the species included in the focal taxon? Can a suite of species act as an umbrella or indicator for other species whose distribution was not directly considered? A simple reserve selection algorithm was applied to species lists from sites where multiple taxa were sampled. The algorithm selected sites so that each species in the focal taxon was represented in at least one reserve. The species lists were complied for birds, mammals, reptiles, and plants on islands in the Gulf of California, and for birds, mammals, and plants in canyons in San Diego County, California. Within both data sets, there were significant rank correlations (P < .01) between the number of species in different taxa per potential reserve. Yet the number of proposed reserves that satisfied the selection algorithm differed strongly between taxa. For example, 2 of 25 canyons were in the San Diego bird reserve network, but 10 of 25 canyons were in the native plant reserve network. The area of the proposed reserves selected by the algorithm ranged from a minimum of 3.2% of the total area for San Diego bird reserves to a maximum of 98.3% of the total area for the Gulf of California plant reserves. The bird reserves included only 45—78% of the species in the other taxa. Conversely, the plant reserve network would contain 88—100% of the vertebrate species. Thus, plants were better indicators than birds for reserve establishment. Plant occurrences were less ordered than bird occurrences, and there were 10 or 20 times more plant species than bird species. The better umbrella taxon was more speciose or had less pattern to species occurrences. Rather than a literal application of plants as reserve indicators, a better approach would be to collect some preliminary distribution data for a number of taxa in the potential reserve network and use these preliminary data to define the focal taxon (or taxa).