Revising the Rhombophryne serratopalpebrosa species group

Today (6 June 2017), my colleagues and I have published a new paper in the journal Zootaxa, entitled ‘A review of the taxonomy and osteology of the Rhombophryne serratopalpebrosa species group (Anura: Microhylidae) from Madagascar, with comments on the value of volume rendering of micro-CT data to taxonomists.’ This is the latest installation in our revision of the genus Rhomobphryne (see other recent contributions here and here) and in particular of the Rhombophryne serratopalpebrosa species group. Over the last four years, we have described R. vaventy from Marojejy, and R. ornata and R. tany from Tsaratanana, but we also re-described the type species of the group, R. serratopalpebrosa, which is also from Marojejy. In this paper, we re-describe another species, R. guentherpetersi, from high altitude on Tsaratanana, and describe two additional new species, R. diadema from the forests of Sorata (the same place R. longicrus was described from), and R. regalis from the forests west and southwest of Marojejy, namely Ambolokopatrika and Anjanaharibe-Sud.

The full Rhombophryne serratopalpebrosa species group. Photos are by Miguel Vences, Frank Glaw, and Franco Andreone.

Like all members of the genus Rhombophryne, the new species are quite cryptic, and difficult to discern from other species. In fact, R. diadema is almost identical morphologically to R. coronata, a species found several hundred kilometres to the south, in the vicinity of Andasibe—but the two species are genetically not closely related. On the other hand R. guentherpetersi, which was described over 40 years ago by Jean Guibé, is highly distinctive, having swollen glands on its tibias. Interestingly, the relationship of R. guentherpetersi had been slowly established based on genetic grounds in recent papers, but here we showed that the type series also has small superciliary spines, making this species a perfect member of the species group.

The bulk of the new paper focuses not so much on the taxonomy but rather the osteology of this species group. Since we started working on the group with our first paper in 2014, osteology has been highly important in its revision. This is in part because of the wealth of informative taxonomic characters to be found in the skeletons of microhylid frogs, but mostly because the external morphology of the holotype of R. serratopalpebrosa is difficult to assess due to the condition of the specimen. We produced micro-CT scans from all members of the species group, including six specimens of the new species R. regalis for a little perspective on the intraspecific variation to be expected.

An illustration of the skull of the new species Rhombophryne diadema

The hip of Rhombophryne serratopalpebrosa. Note how dense the pubis appears on the left (surface rendering) compared to the version on the right (volume rendering).

Whilst performing this survey of the osteology of the R. serratopalpebrosa species group, we happened upon inconsistencies in our dataset compared to ones published previously. Bones that we had previously thought were ossified were in fact cartilaginous. We traced this error back to the method of representing the data. In the new study, we based our analysis of the micro-CT dataset on volume rendering, which takes the various grey values of the rectangular prism created by computed tomography, and assigns transparency and colour based on the grey value of each three-dimensional pixel (voxel). By contrast, our previous work, especially our first treatment of the group, had been mostly founded on surface rendering, which takes thresholds of grey values (and other methods of selection such as brushes) to establish limits at which to draw on/off boundaries, resulting in much smaller files that can be easily manipulated. While both methods have their advantages, failure to understand the limitations of either method can easily result in a misinterpretation of the data. This is indeed what happened in our case: we misinterpreted a pubis that had been represented as ‘on’ in the surface data of R. serratopalpebrosa as being ossified, when in fact this was a decision made during the surface construction; volume rendering shows clearly that the pubis of the specimen in question has a much lower density than other bones, and it is thus clearly unossified.

This latest contribution to the genus Rhombophryne is going to be an important stepping stone in the revision of the genus as a whole. This is our next and ongoing project, and one that is likely to see the number of Rhomobophryne species, today extended to 21, rise by over a third. Osteology is clearly going to play an important role in this work, but we are also going to draw on bioacoustics, external morphology, and possibly some more advanced morphometrics to delimit and distinguish other species groups and new taxa within Rhomobphryne.

The new paper is available from me upon request!

Leave a comment

Your email address will not be published. Required fields are marked *