Proposal (11) to South American Classification Committee


Treat Otus atricapillus, Otus hoyi, and Otus sanctaecatarinae as a single species, O. atricapillus


RE atricapillus/hoyi/sanctaecatarinae. I'd be interested in hearing what others think of this proposed split. Atricapillus and hoyi sound very similar, but sanctaecatarinae sounds quite distinct (listen to Hardy et al.). I wouldn't hesitate adopting Marshall and Behrstock's suggestion of treating atricapillus and hoyi as conspecific (in the liner notes of Hardy et al.), but König et al.'s molecular data (page 46) has atricapillus sister to usta and equally surprising, hoyi sister to petersoni and widely separated from both atricapillus and sanctaecatarinae. The problem with their data set is that they lack many taxa and the bootstrap values are quite low for several of the nodes. Nonetheless, I believe the data are suggestive enough to treat atricapillus, hoyi, and sanctaecatarinae all as species.


Mark B. Robbins, December 2001




Comments from Schulenberg: "NO. What I mean is that I am voting to continue to recognize all three taxa as species, which I *think* is the status quo. As arranged on Van's proposal tracking web page, the title of the proposal ("Lump Otus atricapillus, Otus hoyi, and Otus sanctaecatarinae into a single species, O. atricapillus") seems at odds with what Mark is suggesting ("I believe the data are suggestive enough to treat atricapillus, hoyi, and sanctaecatarinae all as species"). What I wish to endorse is Mark's suggestion to treat atricapillus, hoyi and sanctaecatarinae all as species.  I am not convinced that this is correct (recognizing all three as species). However, I am reluctant to endorse any changes based on listening to a handful of recordings, or after looking at a handful or fewer of sonograms.  I don't think that we are going to get anywhere in our understanding of the Otus until someone gets serious about this and does a true (quantitative, reasonably exhaustive) analysis of vocalizations. This approach is all the more important when songs are similar, as seems to be the case here (*how* similar is "similar"?, when is "similar" the same as "identical"?, etc.).  (I have recordings of what I took at the time to be hoyi (Chuquisaca, Bolivia) and of (what I think is) atricapillus from Espírito Santo, Brazil: available for study to anyone who wants them.)  Clearly a genetic perspective on this would help as well. That said, I am not very impressed with the work by Wink and Heidrich as presented in König et al. It is not clear whether some samples are blood only or are tied to museum specimens (a bit more detail, but not much, is presented in an earlier paper on Otus molecular systematics by Heidrich et al., 1995, Z. Naturforsch. 50c: 294-302), and if there are specimens, where these are deposited. Therefore, confirming any of their identifications will be difficult if not impossible, which does not seem like a good place to start. So although their genetic data so far support the recognition of these three taxa as separate species (as Mark pointed out), I don't find this genetic evidence particularly compelling either. In other words, still waiting for someone to do it right.  Just as an example, Wink and Heidrich (in König et al.) have three samples of Otus "usta" from "South America". I assume that these are the same three samples of Otus "usta" referred to by Heidrich et al. 1995., where we learn that all three samples are from the LSUMZ (although they don't specify LSUMZ numbers for these samples), and that the these come from La Paz (their sample 1), the Rio Napo region (2) and Pando (3). Their sample 2 of usta, therefore, is from a region where some authors (Chapman, HBW) tell us watsonii would be expected. Heidrich et al. do not tell us how they identified their sample 2 as usta: by examining study skins (and if so, by what characters?)? by the vocal type prevalent at the site where the sample came from? or simply because its genetic profile matched that of birds from southwestern Amazonia (presumed "certified" usta), and so therefore it has to be? THEY MAY BE CORRECT, but the lack of attention paid to documenting the sources and identifications of their samples greatly undermines my confidence in anything else that they have to say. "


Comments from Nores:  "NO. El Handbook dice de hoyi "vocally distinct, and treatment as separated species supported by DNA evidence", y de sanctaecatarinae "Formerly considered conspecific with O. atricapillus, but vocally and morphologically distinct, and treatment as separate species further supported by DNA evidence." Además, sanctaecatarinae y atricapillus tienen una distribución muy similar. Los ojos oscuros de atricapillus parece ser una característica importante de la especie que las separa de hoyi y sanctaecatarinae, y también de guatemalae.


From Alvaro Jaramillo:  "No. In another words, do not lump these three taxa together. There are several reservations I have regarding a lump. One of which is that I recall that both sanctaecatarinae and atricapillus are recorded from Iguazu National Park in Argentina (and presumably on the Brazil side too). If the two are sympatric, well it needs to be addressed before a lump. I do not know the validity of reports, if specimens are involved or any details but do recall that both are apparently there. Hopefully I am not wrong on this. I think there has been some confusion in the identification of specimens and recordings of atricapillus and sanctaecatarinae, which also needs to be resolved before we tackle a lump. Recordings of sanctaecatarinae I have heard sound quite unlike atricapillus, being much harsher, guttural, in tone. Having said that atricapillus does indeed sound like hoyi, if a lump was to be made I feel better about these two taxa being together and leaving sanctaecatarinae separate. One topic that does bother me is if one can conclude anything if songs of two or more taxa of owl are similar when they are allopatric. I don't know that this tells us much, if anything about what to do in terms of lumping or retaining taxa as separate species. Presumably if two populations of owl separate, in many situations time alone will cause a divergence in song type, particularly if the two populations end up in radically different habitats. However, there may be strong mate choice or other factors that may keep the song from diverging after separation, even after a great deal of time has elapsed and speciation has occurred. There may also be a limited palette of sounds that New World Otus can make, so some song types may emerge independently even in non-sister taxa. I think this is clearly what goes on in oriole plumage patterns (based on work of Omland and Lanyon's work) and presumably could happen in groups of birds where song has a high genetic component. The similarly recurring patterns in Scytalopus tapaculo vocalizations may be a similar situation, although I don't know if similar songs are always shared by closely related taxa in that case. "