WHAT DETERMINES THE PROBABILITY OF DISCOVERING A SPECIES - A STUDY OFSOUTH-AMERICAN OSCINE PASSERINE BIRDS

The spread of dates of description for species within taxa is normally broad, clearly indicating that there is considerable variation in the probability of species being discovered. The precise reason for this variation is unknown, although newly described species have been chara cterised as small, cryptic and locally distributed. Here, we explore t he hypotheses that the range (both geographic and altitudinal), abunda nce and body size of species determine their probability of discovery, using South American oscine passerine birds. A species geographic ran ge is the best predictor of its date of description: the two are stron gly negatively correlated both across all species and within taxa, and when controlling for other variables. Altitudinal range and body mass are negatively correlated with date of description across species, an d abundance and altitudinal range are negatively correlated with descr iption date within taxa. However, the amounts of variation explained b y these variables are low, and none of the variables remain significan tly correlated with date of description if geographic range is control led for using multiple regression. Of those variables tested, species body mass is the poorest predictor of species date of description, at least in South American oscine passerine birds. Whether the visual con spicuousness (e.g. plumage and behaviour) of a species affects its pro bability of description remains untested.