We suggest that pedodiversity, an expression of environmental heterogeneity, should deserve the attention of plant ecologists
interested to study the spatial pattern of plant biodiversity at different scales. Using the FAO pedological and the IUCN-
WCMC biological data bases of the world countries, we show that the prediction on plant biodiversity (number of vascular
plants) improves significantly when the extent of the area is combined with pedodiversity in a multiple polynomial regression.
Partial correlation analysis proves that, by removing the effect of pedodiversity, the correlation “number of species–area”
remains statistically significant for the tropical countries while it loses significance for the countries outside the tropics.
