Proposal (173) to South American Classification Committee

Treat Zimmerius chrysops as conspecific with Z. viridiflavus

The Zimmerius viridiflavus complex has been the cause for some taxonomic headaches lately. The group (including named taxa viridiflavus, chrysops, flavidifrons, albigularis, minimus, and cumanensis) is a complex of forms that span much of the northern and central Andes from southern Peru to Venezuela, and it is replaced in south Peru by Z. bolivianus (which, conveniently, continues south and east into the Bolivian Yungas). Historically, the forms within the viridiflavus complex have been considered one species (e.g., Traylor 1979), Two species (e.g., Hellmayr 1927, Zimmer 1941, Ridgely and Tudor 1994, Fitzpatrick 2004), or even three species (Ridgely and Greenfield 2001). Zimmer's (1941) reasoning was based largely on plumage differences between San Martin and Huánuco departments in central Peru, but even he sounded unconvinced, stating:

"I would not be surprised to find this [plumage] difference overcome by specimens from intermediate localities. For the present, however, the two groups may be kept specifically distinct with the realization that they are very closely related."

Ridgely and Tudor (1994) agreed with this decision further stating that a vocal shift accompanies this plumage shift (although no further information is given on the extent to which geographic variation within voice has been studied). Ridgely and Greenfield (2001) suggested that additional study of geographic variation of voice within Ecuador warrants the further split of birds on the west slope of Ecuador and northern Peru into another species-level taxon.

However, in my fieldwork in northern Peru, and through conversations with others (particularly T. Schulenberg, B. Walker, and P. Coopmans), I have come to learn that the vocal-and-plumage shift that is the basis for the split of viridiflavus from chrysops is not real. In fact, the "viridiflavus" voice type persists into "chrysops"-plumaged birds in northern San Martin. It appears that it is really within chrysops that the voice shift occurs, with birds north of the Marañon river, on the east slope of the Andes of northern Peru and eastern Ecuador (and into Colombia?), sounding very different from the "viridiflavus" voice. Bear in mind that even sharp eye of J. T. Zimmer could not find basis for naming additional taxa among the east slope populations of "chrysops" between the north and south banks of the Maranon. In fact, the only real difference between viridiflavus and chrysops is the saturation of yellow in the underparts, which, as demonstrated by Johnson and Brush (1972) is not necessarily reason for splitting biological species (also bear witness to various species that are syntopic with Zimmerius "viridiflavus" (sensu lato), showing differing saturations of yellow pigments among populations, yet still considered biological species: e.g., Basileuterus coronatus, Hemispingus superciliaris).

Furthermore, on the on-line Peruvian birding/ornithology list serve, Sjoerd Mayer (see http://birdingperu.com/picsfiles/photos.asp?idtipopic=1&paginaactual=5, as of 27 April 2005) has astutely pointed out that the "distinctive" voice of Z. flavidifrons, used by Ridgely and Greenfield (2001) as basis for further splitting that form off as a third species within the complex, is simply a reversion to a vocalization remarkably similar to that of the viridiflavus voice-type (simply combining what is two or three notes in the latter into one long note in the former). Conversely, birds farther to the north along the Pacific slope of Ecuador and Colombia have at least one, if not two, additional unique vocal-types (see Krabbe and Nilsson 2003), this geographic variation not yet addressed officially (Ridgely and Greenfield 2001). As an aside, I understand that the "unique" chrysops voice-type (fide P. Coopmans) has been heard on the south bank of the Maranon River in low foothills to the north of the Cordillera Colán. The higher elevations of the Colán appear to have "viridiflavus"-voiced birds (I think. Is that a correct assessment, Tom?).

Basically, what we have here is a textbook case of "field guide splitting" going astray: splits based on very spotty and poorly documented evidence. Zimmerius viridiflavus (sensu lato) almost certainly contains several species-level taxa (assuming that strong differences in voice among populations of suboscine species also signals breaks in gene flow), but these are not clear-cut cases of voice and plumage changes coinciding nicely. Splitting along plumage lines results in birds with "chrysops" plumage singing "viridiflavus" songs, but splitting along the song break results in no known meaningful morphological characters that coincide (not that this has been an impediment in redefining species limits among the Tyrannidae in the past, consider Empidonax traillii and alnorum, for example. In any case, I imagine that a dedicated study with large series of specimens is likely to turn up something). Regardless, I think that the crux of the problem lies in an exhaustive survey over the complex's entire geographic range (including Colombia and Venezuela), including morphology, voice, and (particularly) genetic studies. Anything less would be a brash assessment of the situation.

Therefore, I suggest that Zimmerius viridiflavus once again absorb all populations (including chrysops, flavidifrons, minimus, albigularis, and cumanensis) formerly considered part of that species (sensu Traylor 1979) until a proper study of the problem can address the real situation. A vote of "yes" supports such a lump.

Literature cited

Fitzpatrick, J. W. 2004. Family Tyrannidae (Tyrant-Flycatchers) in Handbook of Birds of the World. Volume 9. Lynx Edicions, Barcelona, Spain.

Hellmayr, C. E. 1927. Catalogue of birds of the Americas and the adjacent islands in the Field Museum of Natural History. Part V. Field Museum of Natural History Publication 242, Zoological Series Volume XIII.

Johnson, N. K. and A. H. Brush. 1972. Analysis of polymorphism in the Sooty-capped Bush-Tanager. Systematic Zoology 21: 245-262.

Krabbe, N., and J. Nilsson. 2003. Birds of Ecuador, sounds and photographs (DVD-ROM). Bird Songs International, Westernieland, Netherlands.

Ridgely, R. S., and P. J. Greenfield. 2001. The birds of Ecuador. Volume 1. Comstock Publishing Associates, Ithaca, New York.

Ridgely, R. S., and G. Tudor. 1994. The Birds of South America. Volume II. University of Texas Press, Austin, Texas.

Traylor, M. A., Jr. (ed.). 1979. Check-list of birds of the world. Volume VIII. Museum of Comparative Zoology, Cambridge.

Zimmer, J. T. 1941. Studies of Peruvian birds. Number XXXVII. American Museum Novitates 1109: 1-25.

Dan Lane, April 2005

Comments from Remsen: "YES. I think Dan's reasoning for re-lumping these is solid. Until we have a real study of the situation, no taxonomic change from Traylor/Meyer de Schauensee is warranted; further, the current taxon boundaries appear inaccurate, at least according to the largely unpublished data Dan relayed, so the safe course at this point is to return to earlier classification until the details are worked out."

Comments from Stiles: "YES. I agree with both of Dan Lane's proposals regarding lumping unjustified splits. Points up the difficulties and dangers of splitting forms from different areas based on differences, however distinctive, without having experience with the beasties in the intervening regions. The assumption that plumage and vocal differences will sort out together in these regions is not justified in such cases, which may mean that one species or several are really involved - in either case, much more thorough field work and possibly genetic data will be required. (Using genetic data in species-level decisions is risky, but where distributions are continuous such data could well be useful in indicating steps in gene flow, etc.)"

Comments from Zimmer: "YES. I think that more than one biological species is involved (based on clear vocal differences), but clearly, the current arrangement is screwy. Birds in the Abra Patricia region of N Peru look like chrysops but sound like viridiflavus, while birds in the highest, wettest portion of the Tumbes Reserved Zone (presumably flavidifrons) look like chrysops but don't sound like chrysops or viridiflavus. As Dan has pointed out, the morphological and vocal breaks do not correspond. There are some cases where I would advocate a piecemeal approach to revising polytypic species complexes, on the assumption that any improvement in the taxonomy is better than none at all. This is not one of those cases. Better to retain the broad species limits pending a comprehensive analysis of the entire group, than to further confuse the picture by splitting in the wrong places." [but see additional comments from Zimmer below]

Comments from Robbins: "[NO]. This proposal should be rejected for the following reasons: first, one needs to recognize that plumage is very similar among bona fide species within Zimmerius. In fact, without knowing the provenance of a particular specimen one can misallocate an individual to even the wrong Zimmerius species group. In contrast, there can be relatively major (on a Zimmerius scale) plumage differences within a species with minimal vocal differences, e.g., compare Z. gracilipes specimens from Bolivia and Venezuela. There is a relatively large difference in plumage, but voices between these two populations are virtually indistinguishable to my ear. The point is one should not be surprised that vocal differences do not track plumage variation.

"Second and more importantly, before one can have a meaningful discussion on Zimmerius vocalizations, one must define what vocalization they are talking about and make sure that it is analogous across Zimmerius taxa. The current proposal makes no such distinction, so one has to make some assumptions. My field experience coupled with having listened to many recordings of other field workers indicates that all Zimmerius species have a minimum of two vocal types: 1) a vocalization that can be considered to function as a song that is often given at dawn and less frequently throughout the day. The frequency of when this is delivered during the day varies considerably among species. Often this is a Pachyramphus-like vocalization (a series of musical notes that are very becard-like). This vocalization can be quite similar among species; e.g., west slope Z. chrysops song sounds surprisingly similar to the song of gracilipes (the non-dawn calls notes are diagnostic). In addition, all taxa give one or more single, double or three-noted call notes, and these appear to be quite distinct among taxa.

"Again, given that no vocalization type is defined I'm making the assumption that Dan is referring to non-dawn songs. With regard to flavidifrons: I have recorded flavidifrons from two sites in Loja, Ecuador (MLNS 49433, 57072) and this vocalization is quite distinct from all other Zimmerius. The call of flavidifrons is a unique, upslurred whistle that is quite distinct from west slope chrysops. In fact, if I recall correctly, at the time that when I first recorded flavidifrons (21 August 1989 near Celica, Loja, Ecuador) with Bob Ridgely, we noted how different the vocalization was from west-slope chrysops that I had recorded less than two years earlier. As another aside, note that the distributional map of flavidifrons in Ridgely and Greenfield (2001) does not accurately reflect this taxon's distribution in Ecuador - shading should extend north two provinces to southwestern Chimborazo, the type locality.

"Thus, Dan's implication that flavidifrons' voice is quite similar to viridiflavus in paragraph three of this proposal: "[flavidifrons' voice is] a reversion to a vocalization remarkably similar to that of the viridiflavus voice-type (simply combining what is two or three notes in the latter into one long note in the former)" signifies to me that he does not appreciate the important aspects of Zimmerius vocalizations. The difference between flavidifrons and the bird referred to as viridiflavus in the Sjoerd Mayer web recordings is impressive. If one followed Dan's reasoning that a 2-3 noted primary vocalization is equal to a long, dawn-out, upslurred whistle of flavidifrons then most Zimmerius taxa could be lumped.

"Indeed, Paul Coopman's points out on the Mayer web page (see his posts of 30 Oct & 1 Nov. 2004) that the flavidifrons vocalization is quite distinct from the viridiflavus vocalization. In an email to me on 6 July 2005, Paul indicates that the voice of nominate chrysops is different from flavidifrons, albigularis, and viridiflavus, and birds from Abra Patricia, Peru might even represent an undescribed form.

"In summary, I vote "no" on this proposal because it does not state what vocalizations are being compared among chrysops and viridiflavus (both song and calls need to be addressed before any change is made) and it also oversteps proposing that flavidifrons be lumped into chrysops/viridiflavus. Yes, perhaps one of the Z. chrysops taxa may prove to be viridiflavus, but flavidifrons is not one of those, and I predict that albigularis will prove to be a separate species from east-slope chrysops. Before nominate chrysops is lumped into viridiflavus sound recordings of all vocalizations need to be recorded at the type locality and compared with the full suite of viridiflavus vocalizations. We should keep our current arrangement until someone conducts an in-depth study of this Zimmerius group. "

Response from Dan Lane: "Mark Robbins is correct in criticizing the sloppiness of my wording. What I should have made clear in my original SACC proposal is that I was referring strictly to the "contact call" in all cases. There is far more tape of this vocalization for all populations, and this vocalization is, presumably, on what most of the voice-related taxonomic changes (such as Ridgely and Greenfield) have been based. What I would term "dawn songs" (I assume this is the "Pachyramphus song" to which Mark refers) are much more seldom heard and as a result far less documented by recordings. They may help in assessing the issue, but at the moment there are not enough data available (at least to my knowledge) to say much about the geographic variation among the dawn songs of the various populations in question. So, for the sake of the rest of this response, my mentions of "voice," "vocalizations," etc., will be referring to "calls" (meaning Mark's "contact call") only.

"The problem, as I see it, is in the maintenance of two (or more) species in the viridiflavus complex in the traditional sense, namely chrysops and viridiflavus. These species-level taxa have been assembled by plumage characters, and only recently have voices started to play a role in determining relationships among the populations. Thus, the geographic distribution of the two "species" (for example, in Ridgely and Tudor) shows birds in southern Amazonas and San Martin, Peru, as "chrysops." This is simply continuing the tradition handed down by such taxonomists as J. T. Zimmer. The quote of Zimmer's that I presented in my proposal demonstrates that he was unconvinced (without any knowledge of voice, mind you) that the maintenance of two species in the group was the correct route to take, but followed it since it was "status quo."

"The S Amazonas/San Martin population (hereafter "S chrysops") has, to my eye, and evidently to that of Zimmer, no obvious morphological characters to distinguish it from "nominate chrysops" (hereafter "N chrysops"), and so it traditionally has been placed under that name by all authors. Furthermore, it is as distinct morphologically from viridiflavus as is N chrysops. By call, however, the S chrysops appears to be indistinguishable from true viridiflavus. I believe some field workers such as Paul Coopmans have suggested, therefore, that S chrysops are really a viridiflavus taxon. The placement of this population in "viridiflavus" would be an entirely novel taxonomy, and it is only supported by vocal differences (at least based on current knowledge). By its conservative nature, SACC would not accept this change in taxonomy without a publication supporting it.

"So, what to do about the anomalus S chrysops population? Create a new name for it? Maintain it in "chrysops" or alternatively lump it into "viridiflavus"? Or could it be indicative of a more complex story involving an old (or perhaps even current) hybridization/introgression event between the two parent populations? My best guess is the last, and without a thorough study involving a genetic component, I don't think anyone can deny that it is a very good possibility.

"The story is further complicated by Paul's report of N chrysops-voiced birds at Oracuzar on the south bank of the Marañon river (a relatively low elevation locality, true to the behavior of N chrysops, and apparently unlike S chrysops and viridiflavus. Personally, I have never encountered S chrysops below 1000 m), which lies only about 50 km (straight line) from a series of birds collected by LSU in the Cordillera Colan. This latter population were (as I believe Tom Schulenberg told me at one point) viridiflavus-voiced birds. Thus, S chrysops from the Cordillera Colan, Abra Patricia, the tunnel area NE of Tarapoto, and presumably birds in the "Barbet Peak" area of the Cordillera Azul (no specimens available from the last locality to verify plumage-type, unfortunately, but Barry Walker has told me that they are chrysops based on his sight records there) are all "viridiflavus-voiced" but "chrysops" plumaged.

"If Paul's report represents a population of N chrysops on the S bank of the Marañon (rather than a straggler), presumably such a population is able to contact the Colan birds, then it would be interesting to see if there is active gene flow between them. Remember, we have no idea if gene determining "call type" in this group is governed by a single 1/0 type locus-change or not! "Intermediate voices" suggesting hybridization may not exist simply because of the genetic properties of the voice. In that vein, my own experience with the birds from San Martin suggests that they can give a two to five-noted phrase as "diurnal contact calls" (I am not referring to the "dawn" or "Pachyramphus songs" here), and thus demonstrate that there is within-individual, much less within-population variation in this call. Thus, if the voices of flavifrons (I assume Sjoerd is correct in labeling his sonograms this name) differ only in number of notes making up this "tuuEE?" call, I don't see this as being as huge a difference as Mark makes it out to be, but this point is irrelevant to the current issue. Basically, you have Loja birds giving a single "tuuEE?" versus S chrysops' "tuu-huuEE?" or even "tu-hu-hu-huEE?"

"I wish to make clear one point where I think Mark has misinterpreted my original proposal: the species status of flavifrons is not the issue here. Granting flavidifrons full Biological Species status would be an entirely novel species assignment, and thus would require, by SACC protocol, a paper defending the change. Other than the (rather poorly documented, but not necessarily incorrect) summary in Ridgely and Greenfield, no such publication exists.

"I hope that Mark understands that my proposal is not meant to say that I "believe" that there is only one species involved (in fact, I explicitly state in my proposal that I believe this *not* to be the case), but rather that taking the conservative route is the AOU/SACC modus operandi under such circumstances, as far as I can see. Thus, to follow that logic, temporarily lumping everything into an expanded viridiflavus (which is not a novel taxonomy, by the way, see Traylor 1979) is the only option until a better understanding of the entire complex can be gained (meaning: is there active gene-flow across different plumages and breaks at voice changes, or is there even gene flow across the "voice barrier"? If the latter, then the flavidifrons case would have to be reconsidered as a "no brainer"... as would rather a few strictly voice-based taxonomic changes among Tyrannids). Now, it would be intuitive to suspect that voice trumps plumage as a taxonomic key in all Tyrannidae, but it's an untested assumption, and may not tell the real story. To maintain the two-species concept as it stands now is not satisfactory.

"Furthermore, I believe that a Cotinga paper with sonograms (as Paul Coopmans has suggested to me privately) will not address the concerns I have laid out above. It simply will give Ridgely's taxonomy a "half-assed" paper in a journal of fairly low repute (in taxonomic circles) to support it. Certainly this will not suffice for a SACC-endorsed change?

"In conclusion, I'm not sure that Mark's points really disagree with my own in their logic. He simply has misconstrued my inclusion flavidifrons among the taxa to be lumped with viridiflavus as meaning that I do not believe that it merits species status on its own. My view is that the jury is still out (no publication exists, and the form was never given species status by SACC to start with). Basically, I am proposing taking the conservative route, one that is *not* a novel taxonomy, in assessing the relationships among the taxa of this very interesting complex. I think that this move is entirely in the spirit of SACC's taxonomic changes."

Comments from Stotz: "NO. This and Prop. 174 are similar cases, and I have a similar response. In both cases vocal evidence suggests at least two species are involved. It looks to me in both cases that the names chrysops/viridiflavus and gujanensis/albilora apply to the distinct vocal types. In both cases Dan believes, and presents evidence to support, that the division between the two vocal types has been drawn in the wrong place. Dan's suggestion is that we return to a classification that is even more wrong (i.e. one species) because of the fact that the separation was not done correctly.

"I have to say I think that is an error. I think we should maintain our current treatment, acknowledging that there are problems with how certain populations are allocated. To go to one species, because there is a problem with how certain populations should be assigned is a step backward I think. To me, this would be like deciding King and Clapper Rail should be treated as one species because we don't really know whether the western populations belong with King Rail or Clapper Rail."

Response from Dan Lane: "Why is lumping all taxa (in both the Zimmerius and Synallaxis cases) involved into one species "wrong" in your view? As the conservative move, it seems like the only logical one, especially since it conforms to the majority of historic taxonomic publications, particularly those explicitly used by SACC as the source of the original baseline list.

From the Home Page of the SACC website: "The primary goal is to provide references for all changes from Meyer de Schauensee's (1966, 1970) foundational classification, as well as the "Peters Checklist" series and the "Cory-Hellmayr" series, so that the user can determine how and why (if known) changes were made..."

"Hellmayr, Peters, and both Meyer de Schauensee (1966, 1970) considered albilora part of Synallaxis gujanensis. In fact, of all these publications, only Hellmayr considers S. maranonica a separate species. So, how did albilora even get onto the baseline SACC list as a separate species from the start? The only publications that predated SACC that consider it a separate species are Vaurie, Ridgely and Tudor, and Sibley and Monroe, and none of these are acknowledged as sources of the SACC baseline list. So, by this logic, my proposal should never have been necessary from the start! It's only because of the erroneous voice information in Ridgely and Tudor that their taxonomy has persisted through recent publications. They made the same case (although lacking a prior publication by another author, largely due to the late description of the form) for separating S. chinchipensis from S. stictothorax, yet in rejected Proposal 37, you have no problem saying that the revision was unwarranted until hard data are published (I'd argue that the case for maintaining albilora is in fact weaker in the current literature!). Also, the same logic as I am using here was used in accepted Proposal 117 lumping Momota aequatorialis with M. momota (although I note that you voted "no" on that particular case, as well).

"In the case of the Zimmerius, of the sources for the SACC baseline list, only Hellmayr regarded chrysops a separate species from viridiflavus, and a footnote there also suggests that it is "probably conspecific." Thus, again, the maintenance of two species here actually should have been the original SACC stance, since it was the overwhelming preferred taxonomy in the source publications.

"Finally, using the King and Clapper rails as a "similar" case to the two above does not really seem similar to me at all. In both of my cases, the population in question is intermediate in geographic distribution, as well as physical or vocal characters. Thus, to make the rails similar, there would have to be a population between the eastern North American salt and fresh water marshes that exhibited the appearance of one, but the voice of the other [as an aside, I am not convinced that intermediate rail populations *don't* exist here on the Gulf Coast. Maintenance of this species pair has always struck me as very poorly supported by real data. We have specimens here at LSU that are unidentifiable. Vocally, there is little support for a consistent species-level distinction. But I suspect a genetic/vocal project truly is needed to see if my gut feeling is borne out or not, especially since the AOU NACC does not operate by outsider proposal as does SACC.].

"In the case of S. gujanensis, about 1/2 of the population currently considered by SACC to be "gujanensis" sings the song Ridgely claims is the decisive character for "albilora." This, to me, suggests that the current taxonomy is very unsatisfactory. How can lumping the two voice types be any "more wrong" than the current situation?

"I do not mean to put you on the defensive about the topic, but I am simply very interested in knowing your thinking on it as I greatly respect your knowledge and opinions on taxonomy. If I am overlooking something crucial in my argument, I would greatly appreciate the opportunity to further refine my own methods in approaching the topic."

Response from Stotz: "First off, I will say that I knew when I used the King/Clapper Rail as the analogy that there were weaknesses with that as an analogy. Doesn't really matter, although I would say that the fact that the populations that may be misassigned are not geographically intermediate is immaterial. In the two cases in question you aren't claiming that the geographically intermediate populations show signs of intergradation, they just happen to be intermediately positioned geographically [Note in looking back at your comments I see that you do talk about the possibility that the chrysops-plumaged birds with viridiflavus voice are introgressed, so this is not entirely true; I think that is an unlikely explanation for that pattern, but not impossible]. In the Zimmerius case one of the problematic taxa flavidifrons is not geographically intermediate.

"So back to these two cases. First your legalistic argument that we say "The primary goal is to provide references for all changes from Meyer de Schauensee's (1966,1970) foundational classification, as well as the "Peters Checklist" series and the "Cory-Hellmayr" series, so that the user can determine how and why (if known) changes were made..." and that based on that we should never have split these two cases anyway. However, the SACC list does provide the references to publications that make these splits and provide rationales for the splits. There are dozens of such cases, where we adopt treatments different from Meyer de Schauensee, Peters and/or Hellmayr. Would you argue that we need to have proposals for those changes? So, we should go back to Cymbilaimus sanctaemariae as a subspecies of lineatus until somebody does a proposal to change it? In fact our baseline is actually Howard & Moore, which Van put together. He attempted there, and has expanded here on the attempt to indicate the basis for not following Meyer de Schauensee or Peters, but I don't think that Meyer de Schauensee should be the default taxonomy.

"As I stated in my comments, based on voice there do appear to be two species (or more) in each complex and I think you probably agree. Although we may not have proof that vocal data trump other data in suboscines, we sure act like we believe it (think of how many splits we've put into place due to a certain population having a distinct voice, not because of the details of the plumage, although typically in these cases there are plumage characters that line up with the voice. So, to me the problem is that the chrysops south of the Maranon are misassigned based on the vocal information. I fail to see how reverting back to a single species is better than maintaining two species, given that I think we have based on voice at least two species in the complex. It would seem to me that the solution is to maintain the two species we currently have and work on the note that accompanies it to make clear the nature of the problem you raise here, that populations from south of the Maranon that are morphologically like chrysops are vocally like viridiflavus. Resolution of their status awaits further work.

"On albilora, as I stated the situation seems pretty similar to me. We have two vocal types that seem to be pretty clearly separate species. There are a series of taxa south of the Amazon that were placed by Ridgely and Tudor in gujanensis based on plumage, but vocally are related to albilora. Again it seems to me that improving the note in the list to represent the problem better, rather than lumping is the best way of handling this problem.

"A couple of other points are: 1) in both cases the names that we recognize as distinct species are the correct names for vocal types, no matter what the outcome of the placement of the taxa you indicate are misplaced vocally. From the point of view of the list, these issues don't affect the taxa recognized at distinct species nomenclaturally; 2) I am not sure I agree that changing our baseline is a more conservative approach. It recognizes fewer species, but it means a change, and that is inherently not conservative. That is why the committee rules require that changes to our baseline require a supermajority.

"With respect to the case for Synallaxis chinchipensis, my problem there was not the lack of published hard data, but the fact that I have not seen any data indicating vocal differences. Additionally, our baseline list treats chinchipensis as conspecific, and I typically am looking for more data to make a change than to maintain our status quo. I should note that based on comments I have made earlier if Ridgely and Tudor and convincingly described vocal differences, I would have been willing to split chinchipensis. I am not as opposed to "field guide taxonomy" as others on the committee, because there are hundreds of cases of these unrecognized species with distinct voices. I really feel like we will be waiting forever if we require the definitive work on each of these cases. In part, it is also the recognition that for these allopatric forms even the definitive work is likely going to consist of basically publication of sonograms of a couple of songs of each type, along with some arm-waving about plumage differences.

"I did vote to keep M. aequatorialis separate for M. momota. I also consider this a similar case, in that the majority of the committee appears to believe that aequatorialis is a distinct species, but because the re were potentially other species hidden away in momotus, they lumped aequatorialis back into momotus. I think that was a step backwards."

Additional comments from Zimmer: "Based on the exchange between Dan and Doug (regarding proposals 173 & 174), I do have some comments that I would like to add to the discussion, and I would also like to change my votes on both proposals to a "NO". Following are my comments to the committee:

"I've followed the exchange between Dan and Doug with great interest. Both have made cogent arguments for their respective cases, and I can see the merit in both lines of reasoning. However, I find Doug's argument more compelling. I really do think that one purpose of the committee is to move the taxonomy of South American forward, and we're not doing that by reverting to species-limits that we know are wrong just because we don't have enough information to draw all of the limits in the complex correctly. Although I certainly recognize the merit in being conservative as regards change, I also see a danger in this approach. By failing to recognize any change until all information is in, we stifle any progress at all, because few people have the resources, opportunity or desire to tackle broad problems in whole. The reason that some of these species groups such as Sittasomus or Tolmomyias sulphurescens and assimilis have remained static in spite of the fact that everyone knows there are multiple species involved, is because the prospect of tackling such diverse, complex and geographically widespread species-groups in whole is so daunting. If we never accept piecemeal advances, we almost guarantee a status quo that reflects the same level of knowledge that was available to Hellmayr and Zimmer, even though our understanding may have moved far beyond that.

"It seems to me that part of the reason for assembling a committee of people with great personal experience with the avifauna involved is to apply that collective experience to the difficult problems and try to get the taxonomy to reflect what is currently known. If we rely solely on a dogmatic adherence to previously published compilations, even when we know their taxonomy is wrong, then it seems we could just as easily be replaced by anyone capable of doing a literature search. Mort Isler and I faced a similar dilemma in compiling species accounts for the antbird chapter for HBW. Had we relied solely on published sources, many (perhaps even a majority) antbird species accounts would have contained entries that simply read "nothing known". It seemed criminal to me to write "nothing known" as regards foraging behavior for a species for which I had masses of foraging data, simply because none of that data was already published. To me, the logic behind getting authors that really knew the families they were writing about was to bring that personal experience to the table. Otherwise, again, all we're talking about is a literature search.

"As Doug points out, there are scores, if not hundreds of cases where taxa currently treated as subspecies of geographically widespread, polytypic species, are known (based on distinct vocal differences that also coincide with more subtle plumage differences) to represent good biological species. It will take forever, or at least an exponential increase in the current number of taxonomists working in the Neotropics, to sort this all out in peer-reviewed venues if we adhere strictly to a philosophy of not making any change until the entire picture is clear. I'm not advocating that we cast aside all standards and accept wholesale taxonomic revisions simply because they've been advocated in a field guide-type publication. And there are clearly cases where tinkering with the taxonomy of a complex group based on limited information about one population will create more problems than it solves. But I do think that when there is compelling published information (regardless of whether or not it is peer-reviewed) that coincides with the personal experience and knowledge of the SACC committee members, that we are doing a disservice not to recognize these advances in understanding. By adding a brick here and there, we elevate the entire wall by making it that much easier for others to build upon our changes. Doug's comments about making the Notes section reflect the unresolved problems and uncertainties of any given species group seem right-on to me. I think it is at least as important to point out what isn't known as what is.

"For the above reasons, for reasons stated by Doug and Mark, and because my own field experience clearly indicates that the "single species" approach as regards both the Zimmerius chrysops complex and the Synallaxis gujanensis complex is incorrect, I am changing my votes on Proposals 173 and 174 to "NO".

Comments from Silva: "NO. Doug's arguments are strong enough to get my vote."

Comments from Jaramillo: "NO - this was a very interesting series of interchanges to read through. The details of my thoughts are in the notes from Stotz and Zimmer. However, if I had to boil it down it is that I do not think that lumping these taxa is the conservative stance on the matter, as we all seem to agree that more than one species is involved here. Lumping these taxa is historically conservative; it takes us back to another plateau of knowledge and understanding in this group but the move itself causes instability at the present time and that is not a conservative move from the viewpoint of decreasing nomenclatural volatility. This is particularly important when we are pretty sure that the lump will not stand the test of time, as is the case with this issue."

Comments from Pacheco: "YES. Não conheço diretamente os táxons envolvidos nesta questão; logo, o meu voto baseia-se especialmente nos argumentos e fontes apresentados pelos demais. O meu voto deriva da concordância com a interpretação do caso apresentada por Stotz. Há sim diversos casos na taxonomia das aves sul-americanas nos quais o consenso suplantou a existência de uma análise formal publicada."

Comments from Nores: "NO. Después de leer los múltiples comentarios, pienso que las razones para juntar las especies no son mayores que para separarlas. Yo pienso que hasta tanto no haya estudios genéticos o comportamentales que demuestren claramente que son la misma especie, es mejor no innovar."