The variation in 10 morphometric characters and the degree of sexual dimorphism in size and shape was studied in pure and hybrid populations of smooth Triturus vulgaris and Montandon's T. montandoni newts. Multivariate analyses showed pronounced interspecific differences in both sexes. Females differed mainly in traits related to general size, whereas males differed mainly in body proportions. Triturus vulgaris males possessed a relatively longer trunk and tail. Sexual size dimorphism was much stronger in T. montandoni, the females being the larger sex. In hybrid zone populations a general correlation between morphometric traits and nuclear genotypes was found. This does not, however, fully explain the variation in size and shape. Whereas the majority of genetically identified hybrids fell into the range of morphologically pure species, many genetically pure individuals were classified as morphometric intermediates. The linear relationship between the morphometric and genetic variation suggested no strong differential selection on the molecular markers and loci responsible for morphometric differences between species. The differences in the level of sexual size and shape dimorphism, however, suggest that females of both species show dissimilar mate preferences, and thus hybrid males may experience reduced fitness resulting from lower mating success.

Identification of hybrids is important for the study of introgression between related species, but can be difficult when morphological differences between parental species are small. In the case of the newts Triturus vulgaris and T. helveticus, males are readily identifiable during the breeding season when secondary sexual characters are highly developed, but female dimorphism is always much less. Male hybrids with intermediate characters are therefore more likely to be detected than female ones. RAPD (random amplified polymorphic DNA) analysis of three putative newt hybrids and examples of the two parent species T. vulgaris and T. helveticus confirmed hybrid status in all three cases. The maternal parents of each hybrid were identified by amplification of mitochondrial cytochrome b and ATPase genes, including partial sequencing of the latter. One hybrid had a T. vulgaris mother while the other two had T. helveticus mothers. The PCR-based methods therefore provided a quick and simple approach for identifying both hybrid status and parentage in crosses between these two urodele species. The RAPD technique should also prove useful for the identification of larvae, which are morphologically indistinguishable in T. vulgaris and T. helveticus.