The obscure case of Oscura heterospilla

Oscura heterospila in the aquarium of Lou Rovner, Philadelphia [USA].. 拍摄者是 Juan Miguel Artigas Azas. 鉴定者 Juan Miguel Artigas Azas

In recent years, the historically complex taxonomy of Middle American cichlids has been resolved by applying the new DNA comparison methods available through affordable DNA isolation, coupled with powerful computer systems using refined algorithms. The resulting taxonomic proposal has probably not been what most people were expecting, i.e. a simple classification with a handful of genera to group different lineages in association with four ancient groups, but instead a sound yet complex multi-genera scheme with 34 genera for 112 species. The better part of it is that finally all Middle American cichlid species were covered.

With these proposals, the number of cichlid genera for Middle America far exceeds the current taxonomic organization for the great majority of cichlids, found in Africa and South America. To put this in perspective, we now have for each of the proposed genera for Middle American cichlids an average of 3.3 species, while for Africa, where a currently estimated number of 1,712 species are present, this number is 10.3, and in South America, where a current estimated number of species tallies 522, the number of species per genus is 11.3. What this means is that the Middle American genera group each, on average, fewer than three times the species as for African and South American cichlid genera.

This elevated number is due to the high number of monotypic genera (which comprise only one species) that were described for Middle American cichlids. While monotypic genera are not unusual, they are regularly saved to recognize lineages of species with unique characteristics not found in the most closely related species, and they therefore likely represent the start of a potentially new and different clade. In Middle America, however, just in two recent papers (McMahan et al., 2015; Říčan et al., 2016) seventeen monotypic genera were proposed, fifty percent of the currently recognized genera.

Are all those species recognized in new monotypic genera unique in the way described before? That certainly depends on personal opinion. Some of the genera proposed are given to species that do not, at first sight, seem so striking. The subject of this writing is one of them, the new genus Oscura McMahan & Chakrabarty, 2015, given to Cichlasoma heterospilum Hubbs, 1936.

The new genus was described because using mitochondrial and nuclear DNA data, McMahan et al. (2015:215) produced a phylogenetic tree where Cichlasoma heterospilum appears as a distant relative of species of Vieja (the genus were it was previously classified) and sister with the lotic Rheoheros, which, given the obvious morphological differences between these two groups, results in their proposal of the new genus. Unfortunately, the samples used to compare Cichlasoma heterospilum with the rest of the Middle American cichlids were just two (GenBank HQ424213 and AY843414), and one of them originated from an aquarium specimen (GenBank HQ424213), which adds a degree of uncertainty to the results obtained. It was then, in my opinion, a bold step to propose a new genus based on such little evidence.

McMahan et al. diagnose Oscura having a large, round or oblong dark blotch filling the entirety (or nearly the entirety) of the caudal peduncle, plus the presence of small black spots on scales covering the sides of the body. However, the type species of the genus Vieja Fernández Yépez, 1969: V. maculicauda also has a dark blotch filling the entirety (or nearly the entirety) of the caudal peduncle, with the presence of small black spots on scales covering the sides of the body.

Říčan et al. (Říčan et al., 2016) corroborated McMahan et al. but found a conflict between the results for mitochondrial and nuclear DNA comparisons. In their nuclear comparison using the new ddRAD-technique sequencing, using one juvenile of Cichlasoma heterospilum from the Candelaria River, the species appears as a sister-group of Vieja. Říčan et al. express doubts about the statistical data support given by McMahan et al. for the relationships of Cichlasoma heterospilum and remark the predominance of mitochondrial DNA in such results. In their own mitochondrial DNA comparison, they find differences with the sequence previously obtained in another DNA study, by Concheiro-Pérez et al. (2007, used by McMahan et al.) and in their own mitochondrial DNA sequencing (based on the gene cytochrome b), they place Cichlasoma heterospilum near V. bifasciata and not as the sister-group of Rheoheros like McMahan et al. proposed.

Making things more obscure, Říčan et al. (2016 supplementary material 1) also express doubts about the correct identification of their one examined specimen of Cichlasoma heterospilum, and in their conclusions they decide to act conservatively and wait for more individuals to be examined, however, proposing the new genus Oscura and giving a new diagnosis. In this diagnosis they establish the difference between Oscura and Vieja as having 15 abdominal and 15 caudal vertebrae versus 14 abdominal and 16 caudal vertebrae, and by Oscura having the shortest caudal peduncle in the Theraps-Paraneetroplus clade.

Even if their results were statistically significant (which they cannot be since they mention to have examined only one specimen of Cichlasoma heterospilum) this would hardly stand as a diagnosis that would support a genus. In these conditions, the support for Oscura as a valid genus is very weak. With this in view, If future ddRAD analysis corroborate that Cichlasoma heterospilum is sister to Vieja, there would be no support for Oscura, which would have to be placed in synonymy of Vieja. More speculatively, Oscura could also turn out to be basal (situated at the evolutionary base) in the so-called herichthyines clade, which would mean it has a long evolutionary lineage, and its monotypic status remains.

What is striking, in this case, is that even with very little support, a new monotypic genus was proposed and then validated, and generally accepted without a critical view. As in other aspects of science, every conclusion must be based on copious data, and then subject to scrutiny, discussion and revision.