4 August 2020

New species described

A couple new species descriptions:
1. A new Sorex shrew described from Honduras. The description is based on principal component analysis of just two available specimens; my understanding is that it’s very bad methodology. Personally I think descriptions of new species done without comparing nuclear DNA shouldn’t be accepted for publication anymore, unless it’s something strikingly different. I also think that all species in S. veraepacis species group (veraepacis, chiapensis, ibarrai, madrensis, mutabilis and the new cruzi) are in fact conspecific, but wouldn’t mind being proven wrong.
2. A new mouse lemur, Microcebus jonahi, described. It is the larger of the two species occurring around Mananara, and the only one near the coast there, so if you’ve seen a mouse lemur on your trip to “Aye Aye Island” that’s what it was. The authors compared it only with other species in that part of Madagascar, so it might be conspecific with some species from other parts of the island. Notably, the authors found M. mittermeyeri to be conspecific with M. lehilahystara. I expect some other recently described lemurs to be eventually proven conspecific as well, as many of those descriptions were extremely lousy.

Post Views: 650

5 Comments

  • 6 August 2020
    Theo Linders

    It is a shame so much of mammal taxonomy is dominated by heavy splitters, which completely swamps the good work done by a less vocal group…. It doesn’t help that some of the more influential books (Lynx Edicions….) don’t seem to take taxonomy serious at all. I am looking forward to anyone taking up a checklist and evidence-based approach like is done with birds….

  • 8 August 2020
    Eric Renman

    The more splitting it is the less value it will be of each species. In money-based economy it is called inflation. Generally, I find a species is not a species until it is proved a real species (I am well aware that it can be terrific difficult to prove that a taxon is a valid (biological) species). 11 species of Klipspringers in the Lynx handbook is one of the most ridiculous examples of splitting. Wonder what our other fellow mammal species would say of Homo sapiens, should at least be 5-6 species of Homo today…
    /Eric Renman (Vallentuna, Sweden)

    • 9 August 2020
      Vladimir Dinets

      Homo sapiens taxonomy is a very politically charged subject, as you can imagine. Linnaeus described 4 subspecies; in reality there are over 100, plus, of course, lots of hybrid swarms. But there’s not enough genetic difference to justify splitting into species, unless you use really lousy approaches popular among primatologists. Also, all humans except Khoisan peoples are technically hybrids (with Neanderthals, Denisovians, etc.), so it’s all complicated. One consequence of all this is that you can’t have H. sapiens on your life list unless you’ve been to some Khoisan village in Africa 🙂

  • 10 August 2020
    Theo Linders

    On using PCA and other such analyses in taxonomy, there are three factors that make it’s use in taxonomy far from ideal:

    – Colinearity, these models assume that the different predictor variables put in the model (e.g. skull measurements), are independent from each other, which in most taxonomic cases they are off-course not.

    – Uneven sample size of the different groups, the higher sample sizes and more evenly distribution, the better the chance that the model actually works

    – Small sample sizes, it is advised that the sample size of each group must be larger than the number of predictor variables, otherwise the risk of a false positive increases.

    Finally, many papers use purely qualitative measures and just interpret graphs. They don’t do any follow-up analyses to test whether found differences are actually significant or not…

    There is a pretty good paper on this here:
    https://www.sciencedirect.com/science/article/pii/S0305440311002263?casa_token=Y9EVmrSzMTAAAAAA:ig9muHXeiqfkLq6lPHrvra1DitYhe6drOh-OcHgbMO05lDfeffVLvQhpxNMQe_rPvPtMQToXZac

Leave a Reply