Genetic Divergence for the Amphibian Pleurodeles waltl in Southwest Portugal: Dispersal Barriers Shaping Geographic Patterns
Dispersal and vicariance are the principal mechanisms responsible for the formation of biogeographic patterns, driven or maintained by the role of past and current barriers to dispersal. Southwest Portugal harbors a rich endemic fauna and flora where strongly differentiated taxa can be observed, suggesting ancient isolating mechanisms acted in this region. In this study, we used information from two partial mitochondrial sequences, combined with information from microsatellite loci and amplification success, to investigate biogeographic patterns of genetic divergence in populations of Pleurodeles waltl in Portugal. Our results demonstrate genetic differentiation at different time frames creating distinct populations in southwest Portugal, which might have arisen following isolation by a large river estuary and mountainous barriers. Habitat destruction associated with agricultural intensification raises conservation concerns over the unique biodiversity in this region.Abstract

Distribution of individuals of Pleurodeles waltl in the Iberian Peninsula sampled for mitochondrial analyses. Clades in Spain are redrawn from Carranza and Arnold (2004) where black arrows illustrate the presence of clade one in an expanded region (according to our results) from central Spain covering a large part of Portugal and clade two covering East Spain. Locations of published haplotype samples (AY-, DQ numbers, GenBank) are indicated. An important inferred barrier is illustrated by the red line coinciding with the Rio Mira estuary. The presence of this barrier was supported by a clear increase in amplification failure (dashed arrow). Significant FST genetic distance across this barrier is illustrated for both mitochondrial fragments (P < 0.001). Triangles indicate mountainous areas (Serra de Monchique and Serra do Caldeirão). Corresponding regional locations are described in supplementary Table S1.

Haplotype networks constructed using statistical parsimony (95% statistical confidence limit). (A) cytochrome b; (B) 16S rRNA. The relative sizes of the circles reflect the frequencies of the different haplotypes. Squares indicate the ancestral haplotype. Number of sequenced individuals (between brackets), for each location, are indicated. Lines between circles represent single mutational steps; black dots represent hypothetical haplotype which connect observed haplotypes. Haplogroups found for regions south or north-northeast of the dominant barrier “Rio Mira estuary” are indicated as dark and light grey, respectively.

Phylogenetic tree of Pleurodeles waltl for individuals from the Iberian Peninsula inferred from mitochondrial cytochrome b sequences. Numbers in the major nodes correspond to NJ bootstrap proportions (above branches) and below branches BI (Bayesian inference) posterior probabilities and maximum likelihood, respectively. AY-accession numbers (GenBank) for published sequences and haplogroup numbers (1–10) corresponding to accession numbers JF803794–JF803803 (GenBank) are indicated, including ID-locations and within brackets number of sequenced individuals.

Factorial Correspondence Analysis (FCA) showing differentiation of individuals across the Rio Mira estuary. Analyses were based on either 14 loci (A) or the six loci showing neither amplification failure nor null alleles (B). Different symbols for individuals indicate the three inferred clusters based on STRUCTURE results using 14 loci (supplementary Fig. S3). Alm: two clusters (open circles, 72 individuals//six ponds; solid circles, 28 individuals//1 pond). VNM: one cluster (squares, 41 individuals//eight sites), including individuals with the northern mithocondrial haplotype (open squares), and individuals with the southern mithocondrial haplotype and unique alleles (solid squares) (supplementary Fig. S2).
Contributor Notes