1 Zoologisches Museum, Universitat Zürich (Switzerland), 2 Zoologisches Institut, Universitat Zürich (Switzerland), 3 Division of Biological Sciences, University of Missouri-Columbia (USA), 4 Department of Genetics, University of Washington, Seattle (USA), 5 Research Center for Agricultural and Forest Environment, Polish Academy of Sciences, Poznan (Poland), 6 Department of Ecology, Ethology and Evolution, University of Illinois, Urbana (USA)
Hybrid origin, local abundance, wide range, and inhabitation of harsh or disturbed environments are among the features typically shared by the small fraction of vertebrate taxa that reproduce without genetic recombination. Since the high amount of heterozygosity characterizing interspecies hybrids is thought to buffer against environmental variation and to increase developmental stability, spontaneous heterosis directly caused by hybridity has been hypothesized to explain the ecological success observed in hybrid clonals. We tested this hypothesis in a system of hemiclonal frogs. European water frogs Rana esculenta (RL) are natural hybrids between Rana ridibunda (RR) and Rana lessonae (LL) that reproduce by hybridogenesis: their lessonae genome (L) is excluded in the germ line, haploid gametes with an unrecombined ridibunda genome (R) are produced, and hybridity is restored in the next generation through fertilization by gametes (L) from syntopic Rana lessonae. We compared fitness-related larval life-history traits of natural RL lineages from Poland with those of the two sympatric parental species and with those of newly generated F1 hybrids in an analysis of variance. In a laboratory experiment, 272 individual tadpoles of these 4 genotypes were reared at two food levels. F1 hybrids, compared to either parental species, in the average had higher early growth rates, reached a further developmental stage by day 49 after free-swimming, reached metamorphosis earlier, and had higher survivorship to metamorphosis. Their mean mass at metamorphosis was, however, lower than that of either parental species. Most of these differences are significant in pairwise comparisons (p greater than 0.05, Scheffe's multiple range test). RL hybridogens from a natural population had mean values intermediate between those of F1s and those of the two parental species in these traits. The data match earlier observations on natural RL hybridogens that showed, compared to either parental species, faster larval growth and earlier metamorphosis, as well as higher resistance to hypoxic conditions after metamorphosis. That larval heterosis in early growth rate, time to metamorphosis, and survival to metamorphosis was observed in F1 hybrids, however, at an even higher degree than in hybrid lineages from natural populations, demonstrates that it is "spontaneous" and directly results from hybridity rather than from interclonal selection among natural lineages. Results on clonally reproducing hybrids of other vertebrate groups are equivocal, so the water frog data may not be generalizable to all hybrid clonals. Nevertheless, the data point to the importance of uncoupling the effects of clonality and of hybridity when using hybrid clonal taxa to test ideas about the evolutionary maintenance of sexuality.