Proposal (#12) to South American Classification Committee


Continue to recognize a broad Otus guatemalae (namely to include O. vermiculatus, O. napensis, O roraimae)


RE Otus guatemalae. In this very confusing genus this species complex may have the most complicated taxonomic and nomenclatural problems. Nonetheless, there are a couple of clear decisions that can be proposed. There is an Otus found from at least central Panama (Canal Zone) south to the southern border of Ecuador (west slope) that has a very distinct vocalization (listen to Hardy et al.) from the rest of the guatemalae complex (includes guatemalae, vermiculatus, napensis, roraimae; note that guatemalae has priority). Hekstra's (1982) centralis (holotype from Cerro Mali, Darien, Panama) appears to represent this distinctively vocal form, however, the question that begs to be asked but may not be answered from examination of the holotype, is whether the holotype unequivocally represents the Otus with the distinct vocalization and not guatemalae (includes vermiculatus). From recordings on the Hardy et al. tape, material in LNS, and recordings of Paul Coopmans it appears that guatemalae does not occur in eastern Panama. In fact, every vocalization that I have listened to from Panama is of centralis. Is anyone aware of a guatemalae-type vocalization from anywhere in Panama? I presume Bob and Brett can amplify on what occurs in Panama and Gary perhaps can shed light on whether centralis occurs in Costa Rica. Guatemalae is the screech-owl at La Selva, Costa Rica (recording by P. Coopmans). It would be interesting to know whether the two species are sympatric. So, unless some of the committee members are aware of any guatemalae vocalization from the eastern half of Panama we can be confident that Hekstra's centralis indeed represents the vocally distinct species.


See Marshall et al.'s (1991) frontispiece for depictions of spectrographs and distributions of selected Otus species relevant to the guatemalae complex. Note in the text on page 315 that the Ecuador locality (Paramba) under Otus atricapillus guatemalae is an error and should appear in their O. vermiculatus species account. Finally, the vermiculatus species account *now* refers to O. centralis.


The other issue within this complex concerns the forms in northern South America and along the east slope of the Andes. Individuals that I've recorded on the north slope of Roraima (LNS # 42756-7; 85789) sound very similar to the recording of napensis from prov. Huánuco, Peru on the Hardy et al. tape. Also, my recording of a bird that we collected in the Acari Mtns. (LNS # 42755; KU # 98695) in extreme southern Guyana sounds virtually indistinguishable from the recording from the mountains just south of Caracas, Venezuela (Portachuelo Pass, Aragua) on the Hardy tape. These data (as I conveyed to Terry Taylor in the production of the Hardy tape) were the foundation for considering roraimae and napensis conspecific. Note that König et al.'s range map for vermiculatus is a composite of guatemalae, centralis, and the northern coastal range of roraimae!


So, if we consider roraimae and napensis conspecific (roraimae has priority) the next question that needs to be dealt with is whether roraimae and the Mexican and Central American bird, guatemalae (includes vermiculatus), should be considered conspecific. Although they are widely disjunct (the gap extends from at least central Panama to the Colombia/Venezuela border) and there are morphological differences, they are vocally quite similar. In fact, one could argue that guatemalae, roraimae, and atricapillus (including hoyi) could all be considered conspecific based on vocal similarities (Marshall et al. [1991]did just that, but note what I stated above concerning molecular data). Without any additional information perhaps the conservative thing to do is go with past taxonomic history and treat guatemalae, roraimae, and napensis as conspecific, but for now consider atricapillus as a distinct species.


If we recognize roraimae (with napensis) as a distinct species from guatemalae, then to be consistent, we should recognize colombianus (both have allopatric distributions, some plumage differences, but relatively minor differences in vocalizations).


Mark B. Robbins, December 2001




From Tom Schulenberg:

"Re: Proposal 12A (not yet presented on Van's web site as a formal proposal): To recognize Otus centralis as a species. "No". This vote should not come as any surprise by now, since I am being pedantic throughout, at least with respect to owl taxonomy. The "whole world" seems to agree that there is an Otus with a "different" song in Panama and northwestern South America, and that there is little or no geographic overlap between this entity and other "Vermiculated" screech-owls. But I'd still like to see this better documented than by reference to two recordings on the Hardy et al. tape. Furthermore, someone ought to properly document that Hekstra's name, centralis, does apply to this bird, *and* that no earlier name does. Mark's proposal seems to recognize this. Finally, this case overlaps in jurisdiction with the AOU Check-list Committee for North America. I am not on the committee and I shouldn't presume to speak for how they operate, but I'd be surprised if they would act on this case simply because of the songs presented on the Hardy et al. tape. Not to say that SACC should be the AOUCLC lapdog, but ... So, I think I am going to hold off until someone (some combination of Mark Robbins, Gary Stiles, Joe Marshall, Bob Behrstock???) writes something up for publication that makes the full argument for recognizing this Otus as a species, and that centralis is its name. If this case is as clear-cut as everyone makes it out to be, then "it shouldn't be that complicated to do" (famous last words).


From Alvaro Jaramillo:

"The proposal deals with two questions, one is considering guatemalae to include roraimae and napensis. I know little about this problem, but my understanding from reading the argument and listening to the tapes is that these taxa sound pretty similar and in terms of morphology are different, but only slightly. See my note regarding similarity of song in allopatric owl populations under Proposal 11 for something else to think about. Having said that, the available data does not give much support for retaining all of these taxa as separate so I accept a broad definition of guatemalae until further research is published. The second question is the elevation of centralis, based on different song from guatemalae mainly. On this topic, I have to agree with Tom's comments. There is a need to publish something describing clearly what the song differences are, distribution of song types, and tying the type of centralis to this song type before I would feel comfortable elevating this form to a species."