Fourteen loci consistently amplified and were polymorphic in the two populations of B. spinosus from Spain (San Martín de la Vega and El Pardo). The original sequences were deposited in GenBank under accession numbers KX271337-KX271350.
The number of observed alleles per locus across the 14 loci optimized in B. spinosus ranged from 2 to 14 (mean NA =8.1) in San Martín de la Vega and from 2 to 12 (mean NA =6.4) in El Pardo. Observed and expected heterozygosities ranged from 0.20 to 0.91 (mean HO =0.66) and 0.31 to 0.86 (mean HE =0.71) respectively in San Martín de la Vega; and from 0.13 to 0.97 (mean HO =0.66) and 0.12 to 0.89 (mean HE =0.70) in El Pardo. Micro-Checker inferred the likely presence of null alleles in loci Bspi3.19, Bspi3.20 and Bspi4.25, associated with deviations from Hardy Weinberg equilibrium due to heterozygote deficits in both populations (Bspi3.20 and 4.25) and San Martín de la Vega (Bspi3.19); therefore, caution should be taken when using these markers for population studies in B. spinosus. Four locus x locus combinations significantly deviated from linkage disequilibrium (after accounting for the Bonferroni correction, P<0.001) in Iberian populations (two in each population, with no consistent pattern across populations).
Locus Bspi4.07 only amplified in B. spinosus and locus Bspi3.20 did not consistently amplify in B. bufo. Therefore, we calculated the number of alleles (NA ), observed (HO ) and expected (HE ) heterozygosity in the remaining populations using twelve microsatellites: Bspi3.11, Bspi4.24, Bspi4.25, Bspi4.16, Bspi4.30, Bspi4.27, Bspi4.28, Bspi4.29, Bspi3.02, Bspi3.19, Bspi3.26 and Bspi4.14. Results are shown in Table 2.
Loci Bspi4.16, 3.02, 3.11 and 4.30 were monomorphic in the two B. bufo populations analyzed. Significant deviations from Hardy-Weinberg equilibrium (H1 : heterozygote deficit) were detected for loci Bspi3.19 in Durtal, Bspi4.28 in Perré and Bspi4.24 in Moyaux and Erloy. Significant linkage disequilibrium were only observed in Moyaux (four locus x locus combinations: Bspi4.24 × Bspi4.27, Bspi4.24 × Bspi4.29, Bspi4.28 × Bspi4.29 and Bspi4.28 × Bspi4.14).
All individuals analyzed from Perré (n=8), Durtal (n=5) and Saumur (n=10) had the characteristic mtDNA profile of B. spinosus, whereas all individuals from Audresselles (n=22) and Erloy (n=28) had B. bufo mtDNA. In Moyaux, six individuals had B. spinosus mtDNA and 16 individuals had B. bufo mtDNA haplotypes. We found two interspecific amplexed pairs in our sample, a male B. spinosus mating with a female B. bufo and a male B. bufo with a female B. spinosus.
The first principal component of the microsatellite genotype data explained 25.8% of variation; when individual scores were plotted against geography, a pattern emerged of three groups corresponding to 1) the populations of Saumur, Durtal and Perré; 2) Moyaux and 3) Audresselles and Erloy (Fig. 2).
Structure analyses grouped individuals in two clusters (Fig. 2), with high assignment probabilities. A first cluster, corresponding to B. spinosus, included the populations Saumur (average probability of assignment, PA : 0.99), Perré (PA : 0.97), Durtal (PA : 0.99), whereas the second cluster (B. bufo) included the populations Erloy (PA : 0.99) and Audresselles (PA : 0.99). Individuals from Moyaux were generally admixed, with a higher probability of assignment to the B. bufo cluster (PA : 0.85, range 0.54-0.98, Fig. 2).
According to the results of NewHybrids analyses (Fig. 2) most individuals in the populations of Saumur, Perré and Durtal correspond to B. spinosus (population average proportions of 1.00, 0.63 and 0.88 respectively), while the remainder of the individuals from these populations (three individuals in Perré and one individual in Durtal) was assigned to the F2 class. In the populations Audresselles and Erloy all individuals were classified as B. bufo with high probability, whereas in Moyaux, most individuals (64%) were classified as backcrosses with B. bufo, 27% were classified as F2 individuals and the remaining 9% individuals were classified as B. bufo (Fig. 2).
The morphometric characteristics of the population Moyaux were intermediate between those of B. bufo and B. spinosus, as shown by the analysis of the diagnostic characters paratoid angle / paratoid divergence and metatarsus tubercle shape and size (Fig. 3).
Fig. 3. Morphometric characteristics of adult toads from the population Moyaux (solid round symbols, 20 males, two females) relative to reference individuals from northern France, with ellipses showing one standard deviation around the mean, for B. bufo (shaded) and B. spinosus (unshaded). Reference data are taken from Arntzen et al. (2013b).