Genetic diversity and climatic determinants of tree frogs in Israel

Allozymic variation in proteins encoded by 27 loci was analyzed electrophoretically in 218 adult specimens, mostly males, representing 8 populations, 5 central and 3 marginal, of lemon yellow treefrogs, Hyla arborea savignyi, in Israel along two transects of increasing aridity: (a) north to south and (b) west to east. The results indicate that (a) Of the 27 loci examined, 5 are monomorphic in all 8 populations; 9 are locally and weakly polymorphic; 6 are regionally and weakly polymorphic; and 7 are regionally and strongly polymorphic; (b) In the populations studied, no alternative fixations were found in any of the 27 loci, except in the Wasit population. The commonest allele predominates across all populations, except Wasit, central as well as marginal; (c) Clinal patterns associated with increasing aridity southwards and eastwards occur in polymorphism, P, heterozygosity, H, and in allele frequencies of Ldh-1, Fum, Got-1, and Sdh; (d) For habitat generalists, treefrogs have above average number of alleles per locus, A, and polymorphism, P, while the heterozygosity, H, is average. All three estimates A, P, and H exhibit wide geographic variation decreasing progressively southwards and eastwards; (e) Central populations harbor more genic variation then marginal populations; (f) Genic similarity between populations is high; (g) Significant gametic phase disequilibria were found in several tests in 2 populations; (h) P, H, and allozymic variation in several gene loci are significantly correlated and predictable by environmental variables, primarily those related to water; (i) Morphological and allozymic variations are uncorrelated. The spatial patterns and environmental correlates and predictors of genic variation in Hyla arborea savignyi in Israel suggest that (i) protein polymorphisms are largely adaptive and are molded primarily by climatic selection rather than by stochastic processes or neutrality, and (ii) the environmental variation model seems to be the best predictor of genic variation in treefrogs.