Phenotypic and life-history diversification in Amazonian frogs despite past introgressions.

ABSTRACT

The advent of genomics in phylogenetics and population genetics strengthened the perception that conflicts among gene trees are frequent and often due to introgression. However, hybridization occurs mostly among species that exhibit little phenotypic differentiation. A recent study delineating species in Anomaloglossus, a frog genus endemic to the Guiana Shield, identified an intriguing pattern in the A. baeobatrachus species complex. This complex occurs in French Guiana and Amapá (Brazil) and comprises two sympatric phenotypes contrasting not only in body size, habitat, and advertisement call, but also in larval development mode (endotrophic vs exotrophic tadpoles). However, molecular and phenotypic divergences are, in some cases, incongruent, i.e specimens sharing mtDNA haplotypes are phenotypically distinct, suggesting a complex evolutionary history. Therefore, we genotyped 106 Anomaloglossus individuals using ddRADseq to test whether this phenotype/genotype incongruence was a product of phenotypic plasticity, incomplete lineage sorting, multiple speciation events, or admixture. Based on more than 16,000 SNPs, phylogenetic and population genetic approaches demonstrated that exotrophic populations are paraphyletic. Species tree and admixture analyses revealed a strikingly reticulate pattern, suggesting multiple historical introgression events. The evolutionary history of one exotrophic population in northern French Guiana is particularly compelling given that it received genetic material from exotrophic ancestors but shows very strong genetic affinity with the nearby endotrophic populations. This suggests strong selection on larval development and mating call after secondary contact and hybridization. The case of A. baeobatrachus represents a striking example of introgression among lineages that are phenotypically distinct, even in their larval development mode, and highlights how high-resolution genomic data can unravel unexpectedly complex evolutionary scenarios.