Phylogenetic Relationships in Earthworm Megascolex Species (Oligochaeta: Megascolecidae) with Addition of Two New Species

Megascolex (Oligochaeta: Megascolecidae) are endemic species to India and Sri Lanka, however, to date their molecular taxonomy and phylogenetic relationships have not been reported. We applied the first integrative approach using morpho-anatomical features and a COI dataset to unveil species delimitation (SD), molecular taxonomy, and phylogenetic relationships in Megascolex species. Our morpho-anatomical results revealed nine Megascolex species, namely, M. auriculata, M. cochinensis cochinensis, M. filiciseta, M. ratus, M. travancorensis travancorensis, M. triangularis, M. konkanensis konkanensis, M. polytheca polytheca, and M. polytheca zonatus. We also reported the occurrence of two new species, namely, M. papparensis sp. nov, and M. vazhichlensis sp. nov. Such findings were also supported by the analysed COI dataset, in which these new species appeared distinct on the phylogenetic trees with strong support. The studied Megascolex species appeared paraphyletic and formed three subclades on Bayesian inference (BI) and Maximum Likelihood (ML) phylogenetic trees. The first clade consisted of six species: M. cochinensis cochinensis, M. polytheca polytheca, M. polytheca zonatus, M. konkanensis konkanensis, M. filiciseta, and M. auriculata with strong posterior probability support. The second clade consisted of M. travancorensis travancorensis, M. papparensis sp. nov, and M. vazhichlensis sp. nov with strong support. The third clade consisted of M. ratus and M. triangularis with good support. In addition, the validation of species was confirmed by SD methods, in which the congruence among OTUs was observed with the clear barcode gap of 12-14% suggested by ABGD analysis. However, the species M. ratus and M. travancorensis travancorensis show deep intraspecific divergence and, therefore, require more sampling data. Such findings are essential to study the phylogenetics and evolution of the genus and, nonetheless, demand larger COI datasets to make concrete conclusions.