Pikas at Baxi Forest, Sichuan

Sid Francis, who organises many Sichuan mammal trips, sent through this piece from Malcolm Peaker on the pikas at Baxi Forest near Ruoergai Town in Sichuan. I imagine some of you will have seen the same animals. I also imagine some of you will now how difficult it is to be confident about pika identification. Yikes.

The animals in this forest have (often) been previously IDd as Gansu Pikas. But Malcom’s phootographs and notes suggest they might be Thomas’s Pika.

But … not so fast. When I sent this to Andre Lissovsky (a pika specialist who’s work is referenced by Malcolm in the blog) he was almost sure they wouldn’t be Thomas’s Pikas given the forest habitat; noted that ID by photograph was all but impossible for some of these species; and thinks Gansu Pikas is still the most likely candidate … (see the comments in the blog).

So I am not sure we are any the wiser but it is an interesting conversation and maybe someone more expert than me (which isn’t hard to be as I know nothing!) has an opinion on it here?





  • Malcolm Peaker

    Thanks for that. The reason for querying the Gansu Pika in the first place was that the published habitat of this species did not fit. The only habitat that did fit was for the Tsing-ling. Then Sid found the Dinets photographs. The stand-out morphological feature of Thomas’s is said to be the long, narrow skull. Clearly, on these grounds and on coloration Vladimir Dinets is inclined (as in the photographs he shows) to have this one as Thomas’s (the long skull shows even better in another of Tim Melling’s photographs). So on habitat the likelihood would be Tsing-ling; on morphology, Thomas’s. While Lissovsky is right in saying it is difficult to identify a pika from photographs, at the end of the day the differences in skull morphology should be reflected in external appearance and what is needed are photographs of live animals in the wild that are also collected as voucher specimens for museums. We don’t even know of course if the habitats stated are unequivocally for the species claimed. At the moment though on current published evidence for morphology and habitat I would put Gansu at the bottom of my list of the three possible species, if indeed there are actually three species in the Baxi area. Much more work on pikas in China is needed.

    • Jon Hall

      Yes, much more work is needed for sure so it great that you are looking into this in such detail. The quest isn’t helped by the mass of wrongly labelled photos online (and even in field guides) for some of these species. If you hear any more please let us know. Thanks Jon

      • Malcolm Peaker

        Absolutely agree. Both Tim and I have found a fair few photographs out there clearly wrongly labelled. The standard field guide was great in its day and the only one that can be carried now. There was some pretty poor professional work published which Lissovsky sorted out in terms of what species there are. The problem is you don’t know even with the latest Mammals of the World volume if some of the information on distribution and habitat has been carried over from some of the poorer studies, although I have to say the authors seem to have done a really good job as a snapshot of what was known shortly before publication. A great area for a bit of citizen science! Best wishes, Malcolm

  • Jon Hall

    Hi Malcolm. Some more on this which you might find interesting. I was in touch with Andrew Smith this week – about his new book “Lagomorphs: Pikas, Rabbits, and Hares of the World” – and asked him to take a look at this conversation to see if he had any comments. He must be one of the world’s most knowledgeable Pika experts. He said I could post his thoughts here ….. in a nutshell its very complicated!

    “True – Ochotona thomasi has the flattest skull of all the pikas (among specimens I have seen). But, it does not occur in Sichuan, nor in the habitat described! And I am the first to agree that identifying Ochotona species by sight (or images) is a fool’s game! But it is even worse than that!

    My most knowledgeable Chinese colleague, and a member of the IUCN/SSC Lagomorph Specialist Group, LIU Shaoying, has told me that even in hand he has falsely identified some pikas as being O. thomasi – when in fact, after doing molecular work – they turned out to be O. cansus.

    And, new collections have even more than muddied the waters of the systematics of pikas in central China. Two papers claim there are a multitude of new species there (whether true or whether LIU and KOJU are splitters has yet to be determined). Liu et al. claim that one of their new pikas (ostensibly O. flatcalvaruyn) – has a very flat skull! Note that neither LIU nor KOJU have made any attempt to compare their specimens with known museum types in western collections – e.g. the British Museum, etc. So I am not about to adjudicate any of these new names until we can come up with a process to clearly determine what is what. Good my book came out when it did!

    I attach the most relevant references – beginning with Lissovsky, and then the up-for-grabs LIU and KOJU manuscripts, with the opaca/pallasii and O. sikimaria papers thrown in for good measure.

    Lissovsky, A. A. 2014. Taxonomic revision of pikas Ochotona (Lagomorpha, Mammalia) at the species level. Mammalia 78:199-216.

    Lissovsky, A. A., S. P. Yatsentyuk, and D. Ge. 2016. Phylogeny and taxonomic reassessment of pikas Ochotona pallasii and O. argentata (Mammalia, Lagomorpha). Zoologica Scripta 45:583-594.

    Liu, S., W. Jin, R. Liau, et al. 2017. Phylogenetic study of Ochotona based on mitochondrial Cyt b and morphology with a description of one new subgenus and five new species. Acta Theriologica Sinica 37:1-43.

    Koju, N.P., K. He, M. K. Chalise, C. Ray, Z. Chen, B. Zhang, T. Wan, S. Chen, and X. Jiang. 2017. Multilocus approaches reveal underestimated species diversity and inter-specific gene flow in pikas (Ochotona) from southwestern China. Molecular Phylogenetics and Evolution 107: 239-245.

    Dahal, N., A. A. Lissovsky, Z. Lin, K. Solari, E. A. Hadly, X. Zhan, and U. Ramakrishnan. 2017. Genetics, morphology and ecology reveal a cryptic pika lineage in the Sikkim Himalaya. Molecular Phylogenetics and Evolution 106:55-60.”

    • Malcolm Peaker

      Thanks Jon.

      I had a look at the new papers from China when I spotted one on ResearchGate a couple of weeks ago. While the molecular work with a single mitochondrial gene can be dismissed as inadequate (including trying to check identities by that method), the paper by Koju et al is informative.

      Andrew Smith’s comments on O. thomasi distribution and habitat are interesting because that’s what I thought originally. When I saw the stuff gathered more recently, the distribution had been extended south to north-west Sichuan (see the IUCN distribution for O. thomasi, for example) and the habitat had changed! In that respect, in the Koju et al paper is a map showing where their specimens were collected. It is difficult to make out the symbols but it looks like one of the spots for thomasi is on the Sichuan border. There were no collections as far as I could see from the Baxi area. So the map I showed gathered from the distributions shown in Mammals of the World may still be about right. The problem of course is if the animals recognised as thomasi in the Baxi area were actually cansus, then that error might simply being perpetuated.

      The introgression, shown by the difference between mitochondrial and nuclear gene trees in the Koju paper, between O. curzoniae and O. cansus, illustrates the dangers of using only mitochondrial genes as has been done in the past and is still being done. I cannot comment on the modelling of the trees – but some may argue which method is more appropriate. Koju et al’s unknown species (op 1, op 2 and op 3) which nestle with O. thibetana or O. syrinx also come into play in Sichuan but whether they actually constitute good biological species is open to question.

      This is a problem that is going to run and run and until we know the number of species (always open to doubt and interpretation by different species concepts) and have unequivocal information on distribution and habitat, then spotters in the field will be left scratching their heads. My guess is that it should become possible in the future to match cranial morphology to photographic identification. Given stereo photographs or many shots from different angles some clever image analysis expert may just come up with an estimate of length:breadth for example. Until then and until we can get a rapid faecal DNA profile in the field (not impossible to imagine) we just have to have fun discussing the fast-changing current view.

      For those who have not seen the paper by Koju et al., their last paragraph in the discussion explains the fascinating biological problem of untangling the evolution and speciation of pikas (references omitted):

      Pika inhabit and are adapted to cold, high-elevation areas. Low elevations, deep river valleys and high mountains that become impassable during glacial periods all restrict their dispersal. Such barriers to dispersal lead to population isolation, triggering speciation and the increase of genetic diversity. The mountains of southwest China and adjacent areas harbour high biodiversity. Formation of the mountains of the Qinghai-Tibetan Plateau in southwest China created a complex montane system, with the glaciation and complex topography here affecting genetic diversity in populations adjacent to the edge of the Tibetan Plateau. This phenomenon may have led to the divergence and current distribution of the major clades/subgenera. Effective cryptic divergence was explored in different taxa in this region. Moreover, this area has passed through multiple geological changes as Tibetan plateau uplift and temperature fluctuation in complex terrain have forced species to diversify in response to geological events and abiotic factors that typically have strong effects on evolution.

      For Baxi, the old admission of failure, Ochotona sp. inc. may have to suffice for the present.

  • Jon Hall

    You clearly know a LOT more than I do about all this! I also live for the day when there is some pocket DNA reader available… certainly not beyond the realms of possibiiity I imagine. I wonder if Sid works with any academics on his trips… maybe he can combine a tour with some research to collect one of those Pikas and have the museum look at it.

Leave a Reply