Species are distributed across a mosaic of suitable and unsuitable habitat determined by physiological constraints, resource availability, and intrinsic processes, which in turn limit the number and spatial arrangement of individuals within a population. Quantifying the landscape of habitat over which species could occur improves estimates of population density and absolute size, and informs metapopulation-based studies of the probability of local extinction or colonization events. However, surveys that are inclusive of all possible habitats within a species' range are nearly impossible due to the prohibitive financial and time constraints. Websites that feature user-uploaded photographs from around the world are growing rapidly in popularity and allow ecologists to "virtually" visit many more locales than previously feasible. I explored the efficacy of this data-mining technique for a group of conspicuous and photogenic coral reef fish, the anemonefishes (Pomacentridae: Amphiprioninae) using the website Flickr.
Results/Conclusions
Using a suite of general search terms combined with hierarchical location-specific search terms, Over 1,000 unique records were collected from 164 locales spanning the entire geographic range from which these species have been reported. Individuals of all 28 of the widely accepted as valid species of anemonefishes were identified from these records. I then mapped these points over a publicly available dataset of global coral reef distribution, and used GIS to estimate area of habitat available to each species. I report the relationships between our measures of habitat area and geographic range area, ocean basin, and habitat continuity. Additionally, we present two potential applications of these results by mapping habitat availability onto a phylogeny of the group and comparing habitat availability to estimates of population connectivity in several species. This work demonstrates that estimates of species distributions and subsequent applications benefit substantially from the incorporation of such widely available data.