THE RELATIONSHIP BETWEEN GEOGRAPHIC AREA AND THE LATITUDINAL GRADIENTIN SPECIES RICHNESS IN NEW-WORLD BIRDS

One hypothesis for the latitudinal gradient in species richness observ ed in most animal taxa is that the richness of a region is determined by its geographic area. However, the relationship between geographic a rea and species richness across regions is generally weak. It has been suggested that this is because species from the tropics spill out of this region of high richness, artificially inflating the richness of o ther regions. This generates the interesting prediction that the area and richness of extra-tropical regions should be more strongly correla ted if tropical species are excluded. We test this prediction using th e avifauna of the New World. We find that there is indeed a relationsh ip between the land area and species richness of a region once tropica l species are excluded. This relationship is independent of the latitu de and productivity of regions. Both latitude and productivity can exp lain variance in richness unexplained by land area. There is no relati onship between land area and species richness if tropical species are not excluded from the analysis, suggesting that tropical species do in deed mask the relationship between richness and area. We conclude that our results generally support the geographic area hypothesis, althoug h tests of its other predictions and on other land masses are required .