SEARCH

SEARCH BY CITATION

Keywords:

  • Artiodactyla;
  • bioclimatology;
  • biogeography;
  • biome;
  • Carnivora;
  • ecological pattern;
  • ecological specialization;
  • macroecology;
  • Mammalia;
  • Primates

Abstract

Aim  One of the mechanisms proposed to explain the tendency for geographical range size to increase from the equator to the poles, known as the Rapoport effect, is the climatic variability hypothesis. It states that, towards higher latitudes, greater seasonal climatic variability is the most important pressure that selectively promotes greater general climatic tolerance of species, and therefore also more extensive species ranges. In order to test this hypothesis, we explore the influence of climate, area and biome diversity on the latitudinal gradient of climatic specialization.

Location  The study used the large mammal assemblage from Africa.

Methods  The degree of climatic specialization of African large mammals (Primates, Carnivora, Proboscidea, Perissodactyla, Hyracoidea, Tubulidentata, Artiodactyla and Pholidota) is investigated using the biomic specialization index (BSI) for each mammal species, based on the number of biomes it inhabits. We studied the influence of 11 climatic and biogeographical predictors in the latitudinal pattern of biomic specialization. Stepwise multiple regressions were used to identify the strongest predictors of biomic specialization in Africa and, separately, in both continental hemispheres. We also studied differences among taxonomical groups (primates, carnivores and artiodactyls). We used correlograms generated using Moran's I coefficients to control for spatial autocorrelation in all these analyses.

Results  Average BSI values for successive 1°-latitude bands generally decline towards the equator and temperature variability emerged as the most predictive factor in the regression model for the whole continent, thus supporting the climatic variability hypothesis. Nevertheless, there are differences between hemispheres and among taxa. While temperature variability is the most important predictor of latitudinal variability in biomic specialization in most of the regression models for the northern hemisphere, continental area for each latitudinal band is the best predictor in all the regression models in the southern hemisphere.

Main conclusions  It appears that similar patterns in latitudinal variation in average BSI may be caused by different factors in the two hemispheres. We suggest that the strong north–south geographical asymmetry of Africa, which influences its biogeographical structure, and the presence of land connections with Eurasia in the northern hemisphere are responsible for the observed patterns. Our data illustrate the influence of continental biogeographical structure and history on macroecological patterns.