|Invertebrate habitat specialists||33||0.064||0.010|
|Invertebrate habitat generalists||54||0.149||0.020|
|Vertebrate habitat specialists||82||0.037||0.002|
|Vertebrate habitat generalists||56||0.071||0.001|
|Endemic and relict||96||0.045|
|Note: Data for plants from Hamrick and
Godt 1990: data for animals from Nevo, Beiles, and Ben-Shlomo 1984. Envi-|
ronmental heterogeneity is not measured directly, but is inferred from the extent of the geographic range, successional sta-
tus, or the degree of specialization. N is the number of species in a group: H is the average heterozygosity based on Hardy
Weinberg expectations; and SE is the standard error of the mean heterozygosity of that group.
mately 800 mm per year, but it drops to less than 100 mm in the deserts. As the annual rainfall decreases, the year-to-year variability in rainfall increases dramatically. Thus this transect through Israel describes a gradient of rising stress and environmental uncertainty. Nevo and Beiles ( 1988) summarized allozyme surveys in two plants, two mollusks, two insects, three amphibians, one lizard, and three mammals, testing for correlations with both rainfall and the variation in rainfall across Israel. In the full data set, both heterozygosity and gene diversity increased with the variation in rainfall. Genetic diversity increased toward the deserts in all species studied except the aquatic frog, Rana ridibunda. When the data were analyzed by enzyme locus, rather than by species, variation at 10 of the 13 polymorphic loci increased with the variability of rainfall.
Several population genetic models predict a positive relationship between environmental variability and genetic variability. In laboratory studies with population cages, higher levels of allozyme and additive genetic variation are generally maintained in cages with