Mantidactylus

Rancilhac, Loïs, Bruy, Teddy, Scherz, Mark D., Pereira, Elvis Almeida, Preick, Michaela, Straube, Nicolas, Lyra, Mariana L., Ohler, Annemarie, Streicher, Jeffrey W., Andreone, Franco, Crottini, Angelica, Hutter, Carl R., Randrianantoandro, J. Christian, Rakotoarison, Andolalao, Glaw, Frank, Hofreiter, Michael & Vences, Miguel, 2020, Target-enriched DNA sequencing from historical type material enables a partial revision of the Madagascar giant stream frogs (genus Mantidactylus), Journal of Natural History (J. Nat. Hist.) 54 (1 - 4), pp. 87-118 : 113-114

The subgenus Mantidactylus shows deep phylogenetic structure

As in other groups of Malagasy frogs, the study of genetic variation within the subgenus Mantidactylus confirmed that the current taxonomy underestimates species richness. Using mitochondrial sequences, we were able to delimit seven well-supported candidate species, all of which are geographically separated, with little to no sympatry. In the present study, we take a first step to revise the taxonomy of this group by determining to which lineages the currently described names should be attributed, and describing a third lineage under a new name. As discussed above, the study of molecular data from museum samples successfully allowed us to attribute the type specimens of M. (M.) guttulatus and M. (M.) grandidieri . Therefore, the former name should be used for the lineage distributed in inland localities of Eastern Madagascar, from Ivohibe in the South to Fierenana in the North. On the other hand, the name M. (M.) grandidieri should be applied to the populations from the northeast of Madagascar, between Mangabe in the South and Marojejy in the North. Unfortunately, the type specimen of Rana pigra , although here treated as junior synonym of M. (M.) guttulatus , could not be identified with full certainty, and we cannot exclude that it may belong to the Mantidactylus sp. Ca66/Ca67 clade at this time.

While the status of the four lineages distributed along the eastern coast is difficult to assess due to our sparse sampling and few loci considered, the northwestern lineage stands out by its high divergence from all the other lineages. The monophyly of this group is fully supported on the 16S phylogenetic tree, and the p-distance matrix shows that it is the most distinct lineage of the group. In addition, it does not share any RAG-1 haplotypes with the other lineages. For all these reasons, as well as its geographic separation from the other lineages, we have described it as a new species: Mantidactylus (M.) radaka . While the molecular analysis of museum samples alone could not rule out that the name Rana pigra refers to that lineage, we are confident that this is very unlikely. Indeed, the type locality for R. pigra (‘forêt d’ Ikongo’) is within the range of M. (M.) guttulatus whereas the range of M. (M.) radaka to our knowledge had not been visited by zoologists at the time of the description of R. pigra (absence of historical amphibian records from this area: see Blommers-Schlösser and Blanc 1991; Glaw and Vences 1994). However, further studies of the R. pigra type specimen will be needed to definitively confirm its synonymy with M. (M.) guttulatus . Despite their molecular distinctiveness, the lineages of Mantidactylus show little morphological divergence. However, the new species tends to have a smaller tympanum, as well as very large and ovoid femoral glands.

The combination of MPS-based molecular study of type specimens with traditional phylogenetic and morphologic analyses proved to be an efficient and effective approach to resolve the taxonomy of the Mantidactylus subgenus. Even if our pipeline remains to be improved in many aspects, it could provide a standardised way to assess taxonomic questions within complexes of cryptic species. Further development of our approach could be the integration of coalescent-based species delimitation and discovery models applied on multi-locus data.