At a larger scale, rocks have also been shown to have an influence on the regional species richness of small mammals, especially in southern Africa, where this phenomenon has been examined in some detail. Coetzee (1969), for example, noted that "The distribution of animal life (especially the small mammals) is "largely influenced by rocky outcrops"; he provided a list of 28 species of mammals that frequent rocky outcrops in Namibia. He later (Coetzee 1983) supported these observations with an analysis of mammalian geographic ranges. In a more detailed ecological study in Namibia, Withers (1979) examined the small mammals of an inselberg (isolated rocky outcrop) community and found that both rock specialists and more-generalist species comprise the community. The community ecology of the mammals was greatly affected by microclimate effects of the inselberg and its associated vegetation, as opposed to the surrounding sparse desert habitat. Although he did not conduct a regional analysis of distributional patterns, Withers showed clearly that the rock-piles were a major factor in small mammal species richness in the Namib Desert. Among the areas used by Mares & Lacher (1987) in their broadly comparative study of rock-dwelling mammals were the Namib Desert, the Caatinga, and the Andean region of Argentina. Mares et al. (1981) and Mares et al. (1985a) had shown that rock-piles in the Caatinga had a significant influence on both species richness and persistence, and Mares (1993) provided information showing that rocky habitats were important components of generic richness for small mammals in Andean deserts.