1 Zoologisches Museum, Universität Zürich (Switzerland), 2 Department of Genetics, University of Washington, Seattle (USA), 3 Division of Biological Sciences, University of Missouri-Columbia (USA), 4 Department of Ecology, Ethology and Evolution, University of Illinois, Urbana (USA)
Genotypic diversity in spatially heterogeneous environments is thought to confer a short-term fitness advantage to organisms by permitting niche subdivision and thus a more efficient exploitation of natural resources than in genotypically uniform populations. This idea has been put forward for progeny of a sexually recombining individual as well as for coexisting clones of an organism that reprocuces without sexual recombination. Vrijenhoek's "Frozen Niche Variation" model assumes that an assemblage of coexisting, closely-related, clonally-reproducing organisms consists of an array of clones, each with a unique, relatively narrow niche utilization, representing a particular combination of fitness-related life history traits. Such genotypes are "frozen", not broken up in each generation by recombination. The model predicts that local abundance of clonals relative to related sexuals is only possible with genetic variation among clones: frequency of clonals is positively correlated with the amount of clonal diversity. We tested this prediction in populations of hemiclonally reproducing European water frog hybrids and their sexual host species. The European water frog Rana esculenta (RL), a natural hybrid between Rana ridibunda (RR) and Rana lessonae (LL), reproduces without meiotic recombination. The L parental genome is excluded in the germ line and haploid gametes with an intact R genome are produced (hybridogenesis). Hybridity is maintained via fertilization of such gametes by L gametes of Rana lessonae, with which the hybrids coexist (the L-E system). We compared amount of clonal diversity (different R haplotypes in coexisting hybrid lineages, determined by protein electrophoresis) with local abundance of hybrids relative to their sexual host species. We used 13 L-E systems in Switzerland and Italy, outside the native range of Rana ridibunda, where no hemiclonal hybrid lineages can presently be founded by primary hybridization. Clonal diversity in L-E systems is markedly higher in regions where both parental species are sympatric; this is explained by recurrent formation of new hemiclonal lineages. Outside the range of Rana ridibunda, we observed no positive correlation between clonal diversity and hybrid frequency. We found both populations with relatively low hybrid frequency but relatively high clonal diversity (such as 24% hybrids, 4 coexisting hemiclones), and populations with high hybrid frequency and no clonal diversity (100% hybrids in sample, 1 hemiclone). Overall, there was an insignificant negative correlation between relative hybrid frequency and clonal diversity (r = -0.28, p>0.3). This contrasts distinctly with data on hemiclonally reproducing Mexican fish hybrids of the genus Poeciliopsis, for which a significant positive correlation has been reported (r = +0.93, p<0.01). Our results suggest that the amount of local clonal diversity observed in Rana hybridogens may be shaped by historical factors; for example, it may be limited by the absence of different hemiclones in neighboring areas. We hypothesize that such historical effects can be strong enough to conceal the signal from local operation of the Frozen Niche Variation model. This is testable in extended surveys of natural L-E systems.