1. Species diversity is most simply measured by counting species. More complicated measures, which take into account the relative abundance of the species, have been derived from information theory or from parameters of statistical distributions fitted to the census data. The information theory formulae can also be used to measure habitat diversity and differences between communities or habitats. In this way, changes in the pattern of species diversity can be compared with changes in the environment.

2. Small or remote islands and islands with uniform topography have fewer species than large or complex islands or islands nearer the source of colonization. For birds and some orders of insects it appears that the rate of colonization of new species is virtually balanced by the rate of extinction, so that the number of species has reached equilibrium. For other organisms, such as mammals, and for all organisms on the most remote islands, this equilibrium has probably not been reached and further increases in the fauna may be expected. The comparison of impoverished island faunas with the mainland faunas whence they were derived shows the effect of relaxed competition.

3. Local variations in the species diversity of small uniform habitats can usually be predicted in terms of the structure and productivity of the habitat. Habitats of similar structure on islands and mainland often have similar species diversities; the impoverishment of the island is reflected in the fact that different habitats on the island have nearly the same species, while different habitats on the mainland have more different species. This is interpreted as evidence that uniform habitats are nearly saturated with species and that new species usually colonize by occupying different habitats from present species.

4. The theory of competition and the facts of character displacement indicate that there is a limiting similarity to species which co-exist within a habitat. Species more similar than this limiting value must occupy different habitats. According to the theory, this limiting value should be less where productivity is high, where family size is low and where the seasons are relatively uniform. It should also be less for pursuing hunters than for species which search for stationary prey.

5. Total species diversities, from areas composed of many types of habitat, are usually, but not always, much greater in the tropics than in temperate regions. This is accomplished by a finer subdivision of habitats (habitat selection) more than by a marked increase in diversity within habitats. This total diversity may still be increasing and may have not reached saturation.