Not always young: The first vertebrate ancient origin of true parthenogenesis found in an Amazon leaf litter lizard with evidence of mitochondrial haplotypes surfing on the wave of a range expansion

In vertebrates, true parthenogenesis is found only in squamate reptiles and (mostly) originates via interspecific hybridization after secondary contact. In many cases, parthenogenesis is followed by an increase of ploidy, resulting in triploid lineages. Phylogenetic analyses derived from nuclear and maternally inherited markers can help to clarify the mechanisms of origin and the potential parental species involved. In the Amazon region, parthenogenetic lizards of the Loxopholis percarinatum complex are widely distributed, comprising both diploid and triploid clones. Recently, putative males of L. percarinatum were reported, suggesting the existence of bisexual populations based on morphological data. Here, we used mitochondrial and nuclear data to investigate the origin of parthenogenesis in Loxopholis. Mitochondrial DNA analysis revealed three major lineages: unisexual/2n, unisexual/3n and bisexual, the last of which comprised two sub-lineages placed as the sister taxon to the unisexual/3n lineage. Genetic divergence among the lineages was 10% but was lower between the unisexual/3n and bisexual lineages (-6%). Both mtDNA and nuDNA indicated that individuals from the bisexual lineages might belong to a new species. Nuclear DNA evidence indicates that crossings occasionally occur between unisexual 2n and males from the new bisexual species. Phylogenetic analysis of nuDNA showed L. ferreirai as the closest described bisexual species to the complex. Our results revealed an ancient origin of parthenogenesis in the L. percarinatum complex, in contrast to most young (Pleistocene) parthenogenetic lizards described thus far. Two hybridization events seem to have been involved: the first event occurred in late Miocene, between the ancestral lineage (A) of the new bisexual species (as a maternal species) and the ancestral lineage of L. ferreirai, as a paternal species of L. percarinatum 2n; and the second event occurred in Pliocene-Pleistocene, in a backcross between L. percarinatum 2n and a male from the common ancestor (B) of the new bisexual species giving rise to the lineage of L. percarinatum 3n. With these results, we showed that L. percarinatum complex also includes, at least, one undescribed bisexual species in addition to the two known parthenogenetic lineages (2n and 3n). Finally, we present evidence that diploid individuals of L. percarinatum experienced an event of wide demographic expansion over the past million years under an allele surfing model.