Species limits in Ortalis
Proposal (478) to South American Classification Committee
(A) Split Ortalis araucuan from O. guttata
(B) Split Ortalis squamata from O. guttata
Effect of Proposal:
With the passing of Proposal 439, Colombian Chachalaca Ortalis columbiana was split from Speckled Chachalaca O. guttata. Many committee members stated in their comments that they were in favour of taking further steps to split up this group. In particular, many authorities recognise the East Brazilian Chachalaca O. araucuan and the Scaled Chachalaca O. squamata as separate species. This proposal gives committee members an opportunity to consider the rank of these taxa.
History of treatments:
There is no obvious status quo for these birds, with as many leading authorities splitting these taxa as lumping them. Important publications splitting them include Peters (1934), Sick (1993 - the leading ornithological reference work for Brazil for 2 and a half decades) and the recent IOC list (Gills & Donker 2011 online). Vaurie (1965) appears to have been the person first to propose the lumped treatment for the greater guttata group. Other authorities such as Dickinson (2003), Rodriguez et al. (2005), Erize et al. (2006), Brooks (2006) and SACC, among recent treatments, also adopted this approach.
Some vocal data was recently presented in Donegan et al. (2010). That paper did not deal in detail with the status of these taxa (and a much smaller sample was studied of them compared to columbiana) but some sonograms were presented that may be of interest in considering this proposal.
East Brazilian Chachalaca O. araucuan
1. Plumage differences. O. araucuan differs from O. guttata in being white-bellied (not brown/buff-bellied), with ochraceous-cinnamon on the crown and hind neck (not dark brown) and buffier on the undertail coverts. There is a nice illustration in the Erize et al. (2006) non-passerines field guide of this one, alongside O. guttata.
2. Vocal differences. As can be seen from the sonograms in Donegan et al. (2010: bottom row, middle) and as discussed therein, “Recordings of the northern Brazilian populations araucuan examined are all of a three-syllable refrain. Recordings had different relative lengths of the three syllables compared to Amazonian populations (Figure 2).” In the two inspected recordings of songs of this population, the second note of the “Gua-cha-rac” refrain in the song is shorter than in other populations, appearing as a downstroke rather than a blob of noise. That note is followed by a larger gap than that which follows the second note in other populations, which means that the overall rhythm sounds similar and song length is similar to guttata. However, there is a clear difference in phraseology. These differences are not as striking as those between O. columbiana and O. guttata, but available recordings can still be easily told apart on sonograms.
Examples of songs of the guttata group are available on xeno-canto, together with a map showing the distributions of the three groups (with araucuan top right):
Two examples of the araucuan song are available here:
3. Biogeographic issues. The ranges of O. guttata and O. araucuan are not continuous. O. guttata is widespread from East Andean foothills to the Amazonian region. O. araucuan has a highly disjunct distribution (separated by ca. 1300 km from guttata) and occurs in northeasternmost Brazil only. Notably, the range of Buff-browed Chachalaca Ortalis superciliaris occurs in the region between the two. This sort of disjunct distribution pattern (with another related species’ range bisecting them) is rather strange, especially for a chachalaca. The entire genus Ortalis can be regarded as a giant superspecies (Sick 1993), but the lumping of these two looks most odd under any arrangement where it is split up.
Discussion of Scaled Chachalaca O. squamata
1. Plumage differences. O. squamata has a more brownish-cinnamon hue on the cap and hind neck; and is greyer on the underparts than guttata. It has been regarded as intermediate in plumage between O. guttata and O. araucuan (e.g. Erize et al. 2006).
2. Vocal differences. The final note of the Cha-cha-lac refrain in squamata is lower pitched than the other two notes, whilst in both guttata and araucuan each note tends to be of a more similar frequency. We did not ascertain any obvious differences in phraseology between O. squamata and O. guttata in Donegan et al. (2010). Compare the sonogram on the left, last row of Donegan et al. (2010) with the various O. guttata sonograms therein. From what I have seen from studying songs of other groups, like tapaculos and antbirds, I would not consider this sort of a difference (in acoustic frequency of a single note) to be as significant as those in phraseology shown by araucuan, but that is just a hunch, not a reasoned or supported opinion. Either way, the differences seem to be consistent; and there is a greater sample of squamata (compared to araucuan) available on xeno-canto and other published recordings.
3. Biogeographic issues. O. squamata occurs in the Atlantic forest region of Brazil. Its distribution is also disjunct from that of O. guttata by a long distance (c. 1300 km) but no other Ortalis species’ range is interposed between them. The ranges of O. squamata and O. araucuan are separated by some 1100 km.
Molecular studies (e.g. Pereira et al. 2002, Frank-Hoeflich et al. 2007) have not addressed the issue of relations between members of the Ortalis guttata group to date.
O. araucuan is often called “White-bellied Chachalaca” (e.g. Sick 1993, Erize et al. 2006), but that name is pre-occupied by O. leucogastra of Mexico and Central America. The IOC uses “East Brazilian Chachalaca” as a result. If Proposal A passes, I would propose to use the IOC’s English name.
O. squamata is generally known as “Scaled Chachalaca”.
I am basically ambivalent about these two splits (as opposed to that of O. columbiana which is clear-cut). However, I have little field experience with the Atlantic coast populations and would therefore welcome further comments from committee members or others. The three postulated species in this group clearly share many plumage features and have broadly similar voice. Nonetheless, populations are highly disjunct, and there are diagnosable plumage and vocal differences.
Proposal A: O. araucuan. On balance but with no great enthusiasm, I would support the treatment of East Brazilian Chachalaca O. araucuan as a separate species on the basis of its plumage differences, vocal differences and biogeographic considerations (a “YES” vote).
Proposal B: O. squamata. The plumage and vocal differences from guttata are less strong here, and the range of a combined guttata/squamata does not so offend biogeographic ideals. As a result, one cannot help but be even more ambivalent about this one than the first. With even less enthusiasm, I would on balance be inclined to accept this as a species on the basis of plumage and vocal differences and the large range disjunction (which are all symptomatic of long isolation). This approach would be on the basis that accepting both these splits is preferable to adopting some novel treatment; and also on the basis that this split should be re-considered in the future as further data becomes available. (This is on the basis that the SACC should be more open to changing its treatment where there are widely used alternative taxonomies which sense in light of available data.)
Our recent Conservación Colombiana paper did not really speak to the issues addressed in this proposal in much detail, as they were outside scope. This proposal was produced only with a view to helping your committee and presenting data relevant to a widespread alternative treatment that is probably overdue for consideration. I look forward to seeing how this one goes and to hearing any other people’s views. Please don’t regard this as a position paper on splitting these species – an issue the proposal author has no strong views on – and I would appreciate it if committee members would not lay into the author accordingly!
See the previous Ortalis proposal and SACC baseline.
Thomas Donegan, January 2011
Comments from Stiles: “A very tentative NO: I would prefer to have a more thorough analysis of vocalizations and morphometrics and especially genetic data. I suspect that these two will eventually be split because the available data on plumage, vocalizations, and distributions are certainly suggestive.”
Comments from Robbins: “YES. In evaluating this proposal I have attempted to put it in the context of criteria used in defining currently recognized Ortalis species. Virtually all the recognized species are based solely on differences in morphology (plumage pattern and coloration and soft part color and extent), i.e., no vocalizations and genetic data have been available in species status designations. In the current case there are greater differences in plumage between guttata and araucuan (primarily head and neck coloration) than there are between other Ortalis taxa that are recognized as species. The biogeographic pattern also supports recognizing these as separate species. These aspects coupled with the vocal differences that Donegan has identified lead me to support this split. As I mentioned in my review of proposal 439, I would like to see genetic data brought to bear to help clarify relationships within this entire group, but given what we now have available and to be consistent with treatment of other taxa in this genus it seems best to recognize araucuan as a species. Finally, to be consistent with my evaluation of this proposal I now support the recognition of O. columbiana in proposal 439.”
Comments from Nores:
“(A) Split Ortalis araucuan (hírwacan) from O. guttata (characác). YES. Vocal and plumage differences seem to me sufficient reason for splitting these taxa. Biogeographic issues, however, are not evidence for considering them different species. During the moister periods forest belts may have connected Amazonia and the Atlantic forest. Ortalis guttata and other Amazonian birds would have advanced toward the Atlantic forest and become isolated again in the subsequent drier period. Consequently, there are at present both species and subspecies in the Atlantic forest which are different from those of Amazonia.
“(B) Split Ortalis squamata from O. guttata. NO. As I commented in Proposal #439, I consider that the Atlantic Region populations belong to one species (O. araucuan) with two subspecies (O. araucuan araucuan and O. a. squamata).”
Comments from Pacheco: “[YES to both] Influenciado diretamente pela opinião de Sick e conhecendo os três táxons em campo, sinto-me confortável para admitir os táxons em questão como independentes. Neste caso, o meu voto é coerente com aquele exposto na proposta 439.”
Comments from Jaramillo:
“A – YES. Separate araucuan from guttata. The biogeographic pattern, differences in plumage and vocal data available convince me this is legitimate.
“B – YES. Although this is less clear cut, for consistency (compared to other recognized Ortalis species), this separation appears to be necessary. Because vocal and genetic data have not been used to define species in this genus, at least consistent use of differences in morphology as a way to separate species is appropriate.”
Comments from Cadena: “A. YES, B. YES. The data available are not ideal, as no quantitative analysis of vocal variation with appropriate geographic coverage and sample sizes has been conducted. However, I agree with Mark in that vocal differences do appear suggestive and in the context of how species are currently recognized in the genus, these taxa likely merit species rank. Others have mentioned the need for genetic data, but in cases like this, where taxa are all allopatric, I think genetic data have actually little to offer. Some genetic differentiation in neutral markers will undoubtedly exist, but such data would say very little regarding the status of these populations as reproductively isolated entities, which is what we are trying to judge here. So, based on how other species in the group are defined, I am inclined to split them, although I am by no means directly familiar with the taxa involved.”
Comments from Zimmer: “A) “YES, and (B) “YES”. Splitting out both araucuan and squamata makes the most sense to me given the prevailing concept of species limits within the rest of Ortalis. As Sick (1993) and Donegan (in the Proposal) point out, the entire genus could be regarded as a superspecies. We are talking about a large number of geographically isolated replacement species, which have traditionally been recognized on the basis of morphological differences and biogeography. Morphology within this group appears to be evolutionarily conservative, with most species representing minor variations on common themes. Given all of this, it does strike me as odd that disjunct araucuan and squamata, each occupying distinct biogeographic regions with high degrees of endemism, and each morphologically distinct from other Ortalis, would continue to be treated as subspecies. This is especially true for araucuan, given that the range of another Ortalis species (superciliaris) separates its range from that of its presumed congener (guttata). I think that vocal differences between Ortalis species are more significant than realized, but as with parrots, the distinctions are often difficult to assess because in many (most?) recordings of these birds it is difficult to impossible to know just how many birds are vocalizing at once (one, two, a group, or even multiple groups). One chachalaca of any given species will sound different from two, which will sound different from several. In order to compare apples to apples, you have to have adequate sample sizes of recordings of a single individual, and even then, there may be sexual differences that can’t be accounted for. So, it’s tough to use vocalizations in this group, but I do think that once sound archives have accumulated enough recordings of known numbers of individuals, we will see vocal analyses that reveal significant differences between these various populations.”
Comments from Remsen: “A) YES, and (B) NO. Although I think the case is defensible for araucuan (including squamata) as a separate species from O. guttata, the same cannot be said for squamata as is evident in the proposal, and in agreement with Manuel, I think it is best to treat it as a subspecies of O. araucuan pending more data.”
Comments from Pérez-Emán: “A: YES. B: YES. Morphological differences among these taxa seem to be small to elevate both taxa to species. Also, vocal differences are difficult to evaluate (few recordings and different vocalization scenarios as Kevin pointed out). However, if current Ortalis taxonomy is based on small plumage differences, then it makes sense to elevate these taxa to species considering congruent (but small) morphological and vocal characters. Biogeographically it also makes sense. I assume we will include subaffinis as a subspecies of guttata.”
Additional comments from T. Donegan: “As set out in the proposal, the splitting of araucuan only (and not squamata) is a reasonable approach. The SACC does not deal in subspecies. However, as set out in the proposal, Part A, the better approach under a two-species approach contra some comments above would involve splitting only araucuan but retaining squamata with guttata. Whilst plumage differences are pretty similarly differentiated among all three, the latter two taxa are vocally most similar (as set out in the proposal and illustrated briefly in sonograms in Donegan et al. 2010) whilst araucuan has most different note shapes in its songs, particularly in the second note. I would be more inclined to take vocal cues as an indication of species limits than seek to lump disjunct Atlantic populations on biogeographic, plumage or other grounds. Lumping squamata with guttata but splitting araucuan produces an Atlantic plus Amazon guttata group with two subspecies and without any other species' distribution bisecting their ranges. This is a sensible outcome and guttata's range would reflect a pattern found in many other species (see e.g. the outcome of the recent Schiffornis turdina proposal).
“Separately, we did carry out an analysis using sonograms of the nature suggested by Kevin Zimmer in connection with the aforementioned paper. Vocalisations with two birds calling can be analysed when one looks carefully at sonograms and picks out the calls of different individuals, which are typically at different frequencies: when two birds call together this seems likely to be males and females which use different acoustic frequencies. Where more than two birds call, it's certainly very difficult to pick out what is happening, and one must scroll further through sonograms to find instances where fewer birds vocalise in order to analyse song structures."