When homoplasy mimics hybridization: a case study of Cape hakes (Merluccius capensis and M. paradoxus)

In the marine environment, an increasing number of studies have documented introgression and hybridization using genetic markers. Hybridization appears to occur preferentially between sister-species, with the probability of introgression decreasing with an increase in evolutionary divergence. Exceptions to this pattern were reported for the Cape hakes (Merluccius capensis and M. paradoxus), two distantly related Merluciidae species that diverged 3-4.2 million years ago. Yet, it is expected that contemporary hybridization between such divergent species would result in reduced hybrid fitness. We analysed 1,137 hake individuals using nine microsatellite markers and control region mtDNA data to assess the validity of the described hybridization event. To distinguish between interbreeding, ancestral polymorphism and homplasy we sequenced the flanking region of the most divergent microsatellite marker. Simulation and empirical analyses showed that hybrid identification significantly varied with the number of markers, model and approach used. Phylogenetic analyses based on the sequences of the flanking region of Mmerhk-3b, combined with the absence of mito-nuclear discordance, suggest that previously reported hybridization between M. paradoxus and M. capensis cannot be substantiated. Our findings highlight the need to conduct a priori simulation studies to establish the suitability of a particular set of microsatellite loci for detecting multiple hybridization events. In our example, the identification of hybrids was severely influenced by the number of loci and their variability, as well as the different models employed. More importantly, we provide quantifiable evidence showing that homoplasy mimics the effects of heterospecific crossings which can lead to the incorrect identification of hybridization.