A biogeographic–ecological approach to disentangle reticulate evolution in the Triatoma phyllosoma species group (Heteroptera: Triatominae), vectors of Chagas disease

Introgression and incomplete lineage sorting (ILS) are two of the main sources of gene-tree incongruence; both can confound the assessment of phylogenetic relationships among closely related species. The Triatoma phyllosoma species group is a clade of partially co-distributed and cross-fertile Chagas disease vectors. Despite previous efforts, the phylogeny of this group remains unresolved, largely because of substantial gene-tree incongruence. Here, we sequentially address introgression and ILS to provide a robust phylogenetic hypothesis for the T. phyllosoma species group. To identify likely instances of introgression prior to molecular scrutiny, we assessed biogeographic data and information on fertility of inter-specific crosses. We first derived a few explicit hybridization hypotheses by considering the degree of spatial overlap within each species pair. Then, we assessed the plausibility of these hypotheses in the light of each species pair's cross-fertility. Using this contextual information, we evaluated mito-nuclear (cyt b, ITS-2) gene-tree incongruence and found evidence suggesting introgression within two species pairs. Finally, we modeled ILS using a Bayesian multispecies coalescent approach and either (a) a “complete” dataset with all the specimens in our sample, or (b) a “filtered” dataset without putatively introgressed specimens. The “filtered tree” had higher posterior-probability support, as well as more plausible topology and divergence times, than the “complete tree.” Detecting and filtering out introgression and modeling ILS allowed us to derive an improved phylogenetic hypothesis for the T. phyllosoma species group. Our results illustrate how biogeographic and ecological-reproductive contextual information can help clarify the systematics and evolution of recently diverged taxa prone to introgression and ILS.