The new set of 14 polymorphic microsatellite markers is a valuable tool for population studies on B. spinosus. The species, formerly abundant throughout its Iberian range, is now undergoing a slow but sustained decline (Ortiz Santaliestra, 2014), and molecular tools can provide relevant information about patterns of genetic diversity and structure to guide conservation efforts. Furthermore, the high cross-amplification rates and observed levels of polymorphism in B. bufo make these markers valuable for fine-scale studies of the contact zone with B. spinosus as well as to complement previously developed markers in local or regional studies on B. bufo. Additional populations of the latter species should be assayed with the new markers to better characterize actual polymorphism, which was certainly underestimated in this study (for instance, markers Bspi3.02, 3.11 and 4.30 are monomorphic in the two French populations of B. bufo analyzed here, but additional polymorphism was found in a reduced sample of eastern European samples, data not shown).
Based on the new molecular data, the morphologically intermediate population of Moyaux is shown to represent a hybrid population, where mtDNA haplotypes of the two species can be found in syntopy. Co-occurrence of mtDNA haplotypes of the two species in a single site is rare, but had been previously reported (Recuero et al., 2012; Arntzen et al., 2013a). However, microsatellites also reveal a clearly admixed nuclear DNA genetic pool at Moyaux, suggestive of interspecific gene flow. This is thus the first solid evidence of hybridization between the two species (see Arntzen et al., 2013a for a discussion about alternative interpretations of allozyme data on the taxonomic status of B. bufo and B. spinosus). While more detailed transects need to be studied to better characterize the contact zone in terms of its relative width or patterns of symmetry/asymmetry of gene flow across species, our preliminary results show a consistent signal of hybridization across independent sets of markers, including species-diagnostic morphological characters, mtDNA haplotypes and microsatellite loci. Thus, at least some of the previously observed instances of nuclear DNA haplotype sharing in slowly evolving nuclear markers (POMC, RAG1) may also result from interspecific gene flow rather than from incomplete lineage sorting alone (Arntzen et al., 2013a, 2014).
The lack of F1 hybrids and the high frequency of F2 and backcrossed individuals in the hybrid population at Moyaux suggest no restrictions to gene flow (random mating), which is in accordance with the observation of amplectant mates with mtDNA from the two different species (in both directions). This contrasts with the high incidence of deviations from linkage disequilibrium in the hybrid population, which may be indicative of assortative mating or selection rather than result from random demographic effects, because hybridization tends to break down species-specific physical associations between markers in hybrid zones and create new ones (Jiggins and Mallet, 2000). In any case, the high frequency of admixed individuals and relative scarcity of representatives of the parental taxa indicates this would be a relatively old contact zone, with many generations of hybrids gradually increasing their frequency in the population and becoming the dominant parental class, although more detailed, replicated transects are required to learn more about the potential role of selection in shaping the contact zone.
Whereas the distinction of B. spinosus as a separate species from B. bufo has long been subject to confusion, close inspection of the accumulating morphological and molecular evidence for species differentiation indicates that the two species have diagnostic morphological features and are highly distinct in their mitochondrial and nuclear DNA, except, as shown here, in parts of the contact zone where some hybridization occurs (including some of the reference B. spinosus populations, which according to NewHybrids results may also include some hybrid individuals, see Fig. 2). More detailed analyses of different sections of this contact zone are required to further delineate it and investigate whether observed patterns are generalizable. Additionally, comparative data about demographic and life history traits across species may be relevant to understand the evolutionary processes maintaining species boundaries in the face of gene flow. For instance, the skeletochronological work of Hemelaar (1988) showed remarkable differences in age at maturity and growth rates in Bufo populations from different latitudes. Of the five populations compared, one from southern France represents B. spinosus as inferred from recent molecular studies, whereas the other four would correspond to B. bufo on geographic grounds. Several characteristics of the French population stand out in comparison with populations of B. bufo, for instance much faster growth rates, larger maximum body size and earlier age at maturity, suggesting genetically, rather than environmentally induced differences (Hemelaar, 1988). Also, B. spinosus appears to start breeding a few weeks earlier than B. bufo and the breeding season is also more prolonged than in B. bufo, which is usually considered an explosive breeder. These differences, in concert with potential mechanisms of species recognition, which do not seem to play a role in preventing interspecific mating in Moyaux but may be apparent elsewhere, can shape patterns of interspecific gene flow in the contact zone, ultimately determining its evolutionary fate.