Latitude and latitudinal extent: a global analysis of the Rapoport effect in a tropical mammalian taxon: primates

Aim Analyses of the Rapoport effect (species at high latitudes have broader latitudinal ranges) usually indicate only a weak effect in the tropics. However, studies of tropical taxa are rare. This study used a global analysis of a tropical order; primates, to ask two questions. Is there a Rapoport effect in this tropical, terrestrial mammal, globally or per continent! If so, can the effect be explained by greater adaptability of higher latitude taxa. Location Globe. Methods Geographical data for the majority of primates species are available. Genera, rather than the species, were the main data used because the genus is a more stable taxonomic lever for this order. Regressions and Spearman correlations were used to test for the effect. Generic ranges are the sum of the constituent species' ranges. In addition, phylogenetic effects were controlled for to minimize dependence of data. Results The Rapoport effect is apparent globally when Madagascar and outlier taxa are excluded. It is also present separately in Africa, Asia and Madagascar when outlier genera are excluded, but not in South America. The effect persists globally, and in Africa and Asia when phylogeny is controlled for. Climatic variation (temperature or precipitation) correlates with either mid-latitude or extent of latitude of genera in Africa, Asia and the Americas (insufficient data for Madagascar), but controlling for phylogeny confirms only the African and Asian correlations. The four measures of adaptability of taxa (number of dietary types, number of habitat types, body mass and number of species per genus) correlated variously with latitudinal extent of the taxa, but the correlations did not always match the existence of a Rapoport effect in the respective continents. Main conclusions The intercontinental variation apparent in the Rapoport effect in the tropics should be fruitful material for further study of the causes of the Rapoport effect. The discovery of significant associations of measures of adaptability with latitudinal extent in both the presence and the absence of a Rapoport effect indicates that care is needed in using such associations to explain the Rapoport effect.