Genomic heterogeneity of historical gene flow between two species of newts inferred from transcriptome data

Abstract

The role of gene flow in species formation is a major unresolved issue in speciation biology. Progress in this area requires information on the long‐term patterns of gene flow between diverging species. Here, we used thousands of single‐nucleotide polymorphisms derived from transcriptome resequencing and a method modeling the joint frequency spectrum of these polymorphisms to reconstruct patterns of historical gene flow between two Lissotriton newts: L. vulgaris (Lv) and L. montandoni (Lm). We tested several models of divergence including complete isolation and various scenarios of historical gene flow. The model of secondary contact received the highest support. According to this model, the species split from their common ancestor ca. 5.5 million years (MY) ago, evolved in isolation for ca. 2 MY, and have been exchanging genes for the last 3.5 MY Demographic changes have been inferred in both species, with the current effective population size of ca. 0.7 million in Lv and 0.2 million in Lm. The postdivergence gene flow resulted in two‐directional introgression which affected the genomes of both species, but was more pronounced from Lv to Lm. Interestingly, we found evidence for genomic heterogeneity of interspecific gene flow. This study demonstrates the complexity of long‐term gene flow between distinct but incompletely reproductively isolated taxa which divergence was initiated millions of years ago.

Divergence history of the Carpathian and smooth newts modelled in space and time

“…). A model of secondary contact but with a much longer period of genetic exchange spanning the entire Pleistocene received the highest support in the transcriptome resequencing study of Stuglik & Babik (in press). The most complex scenarios assuming temporal changes in demography and gene flow were analysed in two‐population models.…”

“…This possibility could be evaluated by using information contained in haplotype spectra to estimate the time of secondary contact more precisely (Harris & Nielsen ). Genomic heterogeneity in the pattern of gene flow between Lm and Lv inferred from the site frequency spectrum (Stuglik & Babik in press) argues against the invasion‐related neutral explanation. This is because asymmetric gene flow was found only in the portion of the genome characterized by a low migration rate, where selection may oppose introgression.…”

Differential introgression across newt hybrid zones: Evidence from replicated transects

“…Twelve single‐copy BS genes were identified previously using two composite tests looking for an excess of high‐frequency nonsynonymous polymorphisms and an excess of polymorphism over divergence (Fijarczyk et al, ). Forty‐one F ST genes and 33 d xy genes were identified from transcriptome resequencing of six Lm and six Lv individuals sampled to encompass the genetic diversity of the species (Stuglik & Babik, ). HDM genes identified in an F 2 hybrid family clustered in two blocks on linkage groups (LG) 4 and 10 (Niedzicka et al, ).…”

Isolation and Gene Flow in a Speciation Continuum in Newts

“…Estimates of long-term, coalescent effective population sizes (N e ) for the Lissotriton lineages are quite large, ranging roughly between 80,000 and 800,000 (Table S9). Independent assessments place the divergence time of L. montandoni and L. vulgaris between 3.7 to 6.3 mya (Stuglik and Babik, 2016), while mtDNA lineages within L. vulgaris split in close succession during the late Pliocene and early Pleistocene (Babik et al, 2005; Pabijan et al, 2015). This timeframe suggests that most Lissotriton lineages have not attained the 4N e generations when roughly half of the loci in the genome achieve reciprocal monophyly (Hudson and Coyne, 2002; Degnan and Rosenberg, 2009), thus widespread incomplete lineage sorting (Table 2) and the presence of identical haplotypes amongst all lineage pairs are expected (Fig.…”