Proposal (946) to South American Classification Committee

Treat Cacicus uropygialis as consisting of three species

Effect on SACC classification: This would split our Cacicus uropygialis (Scarlet-rumped Cacique) into two or three species.

Background: The three taxa involved are:

• microrhynchus: lowlands Honduras to e. Panama (i.e. extralimital to SACC)

• pacificus: lowlands e. Panama south through w. Colombia to w. Ecuador

• uropygialis: Western Andes and Central Andes of Colombia; eastern slope of
Andes from NW Venezuela south patchily, it seems, through Colombia and Ecuador to s. Peru.

The key point to note immediately is that the first two are lowland taxa, whereas uropygialis is strictly montane, and more importantly, that pacificus and uropygialis both occur in w. Colombia; although evidently not precisely parapatric, they come close, e.g. as per Hilty & Brown (1986), pacificus to 1000 m, and uropygialis 1500 m and above (but once to 1000 m).

All three are very similar in plumage, being basically black caciques with red lower backs and rumps, that differ mainly in slight differences in bill size and shape. Here is Peter Burke’s plate from Jaramillo & Burke (1999):

Also marked on the plate are the two subspecies of Cacicus haemorrhous (Red-rumped Cacique), a lowland species of Amazonia and SE Brazil that is not the sister species to this group.

This is a well-known problem in species limits that has been dealt with differently by different authors for at least 120 years. Our current note reads as follows:

Cacicus uropygialis likely includes two, perhaps three, species-level taxa (Hilty & Brown 1986, Ridgely & Tudor 1989); trans-Andean microrhynchus was treated as a separate species by Jaramillo & Burke (1999), Ridgely & Greenfield (2001), and Hilty (2003); Meyer de Schauensee (1966) suspected that the subspecies pacificus of western Colombia, included by Jaramillo and Burke (1999) and others as a subspecies of extralimital C. (u.) microrhynchus, might also deserve species rank. Ridgway (1902) evidently treated microrhynchus as a separate species from uropygialis by omitting mention of the latter. Hellmayr (1938), followed by Wetmore et al. (1984), maintained all as conspecific because of the seemingly intermediate characters of pacificus. SACC proposal to recognize microrhynchus as separate species did not pass because of absence of formal published analysis. Powell et al. (2014) found that pacificus was actually sister to uropygialis, not microrhynchus. SACC proposal needed.

Ridgway (1902) implicitly treated microrhynchus as a separate species by not mentioning extralimital uropygialis or of course then-undescribed pacificus.

Hellmayr (1938) treated all three as conspecific with the following rationale:

“This form [pacificus] combines the general dimensions of C. u. microrhynchus with the powerful bill of C. u. uropygialis, thus occupying in its characters an intermediate position as it does geographically.”

Wetmore (1984) followed Hellmayr and also noted “An occasional adult male of this race [microrhynchus] shows a faint swelling on the outer face of the base of the mandibular rami, an indication of approach to the condition found in C. u. pacificus, but this is not usual. The 2 races are similar in size.” There is no mention of intergradation. Also: “From the somewhat scanty data, there may be a gap between the range of this form [pacificus] and that of C. u. microrhynchus.” And: “The vocalizations of this race include a whistled teeo or keeo, without the burry quality of the corresponding call of microrhynchus in the Canal Zone (Eisenmann, in litt.).”

Hilty & Brown (1986) proposed that pacificus might be a separate species from uropygialis but noted that strict sympatry was not yet known. Ridgely & Tudor (1989) were sure that the lowland taxa (pacificus + extralimital microrhynchus) would be shown to be separate species based on morphological, vocal, and elevational differences; they suggested Subtropical Cacique for uropygialis and retaining Scarlet-rumped for the lowland taxa. Jaramillo & Burke (1999) implemented that split, and their qualitative descriptions of voices are fairly different. Ridgely & Greenfield (& Robbins and Coopmans; 2001) also followed the 2-way split.

In 2003, Jaramillo submitted a SACC proposal to split microrhynchus (including pacificus) from uropygialis was rejected. To summarize the outcome of that proposal. most of the committee thought that two species were involved but did not think the split was adequately supported by published data. Alvaro summarized in detail what was known anecdotally in 2003 concerning differences in voices and jizz – see SACC 73 for all that, which I strongly recommend reading.

New information (since a previous SACC proposal in 2003):

Fraga (HBW 2011) followed the 2-way split.

Powell et al. (2014), using DNA sequence data, produced the following phylogenetic hypothesis for Cacicus:

Note the result that I think would surprise most: pacificus and uropygialis are actually sisters, not Central American microrhynchus and pacificus. Powell et al. (2014) remarked: “Some authorities (e.g. Jaramillo and Burke, 1999; Fraga, 2011; Gill and Donsker, 2012) recognize Cacicus (uropygialis) microrhynchus as a species and treat C. u. pacificus as a subspecies of C. microrhynchus, but mitochondrial DNA indicates that pacificus is more closely related to C. u. uropygialis.” At face value (as well as eye-balling comparative branch lengths in the broader phylogeny), this would support species rank for all three, in my opinion, given the pronounced vocal differences and near-parapatric distributions of uropygialis and pacificus. The result, however, should be treated with caution because it might be a case of incomplete lineage-sorting. Also, given the surprise, perhaps the results should double-checked --- I wonder if there was a mistake in branch labelling or sample mx-up. On the other hand, the three taxa are so similar that perhaps we should not be surprised that the genetic relationships don’t match our non-genetic assessment. Up until Powell et al. (2014), no one had considered the possibility that the two lowland, parapatric taxa were not sisters.

Boesman (2016) presented sonograms of microrhynchus and pacificus (uropygialis not considered) and stated:

Vocal difference between the two races is quite obvious in all homologous vocalizations:

* single notes: nominate utters irregularly overslurred notes reaching max. frequency of 3.2 - 4kHz, pacificus principally downslurred notes reaching max. frequency of 2.2 - 2.6kHz.

* fast rattling series: A similar difference in max. frequency and nominate often combines two series of different repeated notes.

He presented sonograms of 7 pacificus and 11 microrhynchus (1 Honduras, 5 Costa Rica, 5 Panama). The sonograms look different, but there is a lot of variation within each, as might be expected from the considerable repertoire of most Cacicus. Boesman noted the absence of recordings from eastern Panama and Colombia from near the putative area of contact, but clearly considered the evidence presented as worthy of species rank.

I played around with sample recordings of all three on xeno-canto. With the remarkable variability in vocalizations, it was quickly obvious that casual browsing wouldn’t produce anything but trouble. Those with more patience of course may pull some signal out of all those noises. I will say that uropygialis “sounds different” from the other two, with the notes having a different, querulous quality that somehow reminds me of (don’t laugh) Crotophaga ani. I can certainly see why the genetic results of Powell et al. (2014) should surprise people.

Discussion and Recommendation: I hesitate to make taxonomic changes without solid, published data, but in this case I lean towards a 3-way split for the following reasons. First, in contrast to the previous proposal on Piranga rubra, the original lump of the three was based on a 1938-genre qualitative assessment of size and bill characters that suggested that pacificus was intermediate, and thus is a “bridge” between microrhynchus and uropygialis. This would be insufficient evidence by recent standards. Second, the plumage differences among taxa that are for-sure species in Cacicus are not very large; for example, see the illustration above, in which it is difficult to ascertain differences between C. haemorrhous and the group covered in this proposal, despite them not being at all closely related within the genus. Third, with all appropriate caveats, the voices of the three are evidently different, especially uropygialis vs. the other two, yet genetic data suggest that uropygialis and pacificus are sisters

Fourth, and most important to me, is biogeography. Although perhaps so similar that intergrades would not be detected, microrhynchus and pacificus are nearly parapatric without any signs of gene flow. Although there is no physical barrier between the two, range boundaries that end or begin in Darién, Panama, are numerous --- perhaps one of the most prevalent distribution patterns in Central America. Going way out on a limb …. This implies to me that ecological conditions change fairly abruptly in that region, perhaps caused by differences in rainfall. If that’s the case, then perhaps the microrhynchus and pacificus genomes are incompatible to the extent that interbreeding is prevented or limited. More impressive to me is the near-parapatry of pacificus and uropygialis in western Colombia. Again, there is no physical barrier between the two – the two populations are likely within sight of each other. Yet there is no sign of gene flow. What that tells me is that these two populations have diverged to the point that they have adapted to different ecological conditions, and neither has conquered the conditions in the minor elevational gap (if there really is one). In contrast, if they were the same species, then I would expect free gene flow and a continuous distribution of the two taxa if they were two species, with a zone of intergradation at intermediate elevations. Parapatry without gene flow is prima facie evidence for species rank.

I need to invent a term for the distribution pattern in which two taxa may not be in direct physical contact but rather are separated only by habitat that is evidently unsuitable to either population. For reasons outlined above, I consider this as evidence for species rank. If the two populations were separated by a physical barrier, then I would label them as allopatric, regardless of the width of the barrier, because it appears to be physical limits to dispersal ability, i.e. extrinsic factors, that are keeping two populations from contact, in contrast to the intrinsic factors that keep these two pairs of taxa apart. (By the way, I am working on a short paper on this as an operational criterion for species rank, so feedback welcomed. I’m also groping for a term to describe this near-parapatry situation that doesn’t imply a distance criterion, so if anyone has suggestions, fire away. The best I can come up with are unsatisfying: “quasi-parapatry” and “effective parapatry.”

There are actually 4 possible taxonomic treatments of the complex:

A. No change, i.e. one polytypic species.

B. Two species: (1) uropygialis and (2) microrhynchus + pacificus (as in many recent classifications)

C. Two species: (1) uropygialis + pacificus based on the relationships in Powell et al., and (2) microrhynchus

D. Three species: (1) uropygialis, (2) microrhynchus, (3) pacificus

So, for voting purposes, a YES means 3 species, i.e. option D, and a NO means one of the other options, which will then be voted on in a subsequent proposal if D (three species, the title of the proposal) is rejected.

English names: I favor a separate proposal on English names if this one passes. The simplest solution would be to go with the flow as in recent literature and use Pacific Cacique for pacificus, Subtropical Cacique for uropygialis, and retain Scarlet-rumped Cacique for microrhynchus s.s. Those names don’t have too much traction, however, so reasons for a second look are as follows. Many, including me, don’t like using “Pacific” for non-marine or non-insular species. There is precedent for it, yes, but that doesn’t mean it’s good. Second, retaining Scarlet-rumped for microrhynchus bumps up against our guidelines for English names in parent-daughter splits, and all three have identical scarlet rumps, so the name is not useful and causes perpetual confusion because it has been applied to THREE separate taxonomic concepts: broadly defined uropygialis, microrhynchus + pacificus, and just microrhynchus. On the other hand, Scarlet-rumped works well within the context of Central America, and although its range is small, certainly it is the most frequently seen and studied scarlet-rumped cacique because of its presence in heavily visited Costa Rica and Panama. Ridgway (1902), by the way, used Small-billed Cacique for microrhynchus, and that indeed is one of the only differences between it and pacificus. This one is really a NACC issue, so they should have say over that, however. Also, some might favor a hyphenated group-name approach, e.g. Subtropical Scarlet-rumped, etc.

Van Remsen, June 2022

_______________________________________________________________________________

Comments from Areta:

“I vote NO to options A and B, as both seem untenable on morphological and (to some degree) phylogenetic grounds. Now, deciding whether one goes for C or D is more difficult. In principle, I am happy to adopt option C, and may extend onto D if enough convincing data is provided.

“YES to the split of microrhynchus as a separate species. The deep split from uropygialis and pacificus coupled to their geographic distributions satisfies me to grant species status to microrhynchus. I examined a few hundred of eBird photographs. As the name implies, the bill of microrhynchus does look more slender (i.e., relatively shallow and long bill, note also the possibly straighter culmen) in comparison to the thicker, relatively broad and proportionally shorter bills with presumably more curved culmens of uropygialis and pacificus. Very importantly yet possibly omitted from all descriptions, the bill of microrhynchus is also noticeably paler than the bills of both uropygialis and pacificus (i.e., the bill of microrhynchus is of an ill-defined greenish-yellowish tone, with a somewhat brighter base, whereas the bills of uropygialis and pacificus have a striking yellow base; possibly more extensive in uropygialis?).

“Maybe yes to the split of pacificus from uropygialis. The overall bill proportions, shape and color (assessed through eBird photographs) are similar, in nice agreement with phylogenetic data in Powell et al. (2011). The differences in size reported in Jaramillo & Burke (1999) are striking and, if constant throughout their ranges, are possibly the best additional bit of information in favor of species status for pacificus, which is the "lowland" form though possibly reaching altitudes equivalent to those of the montane uropygialis. May these sizes be adaptive? I have not studied the vocalizations in detail and although my cursory examination suggests that the calls are different (simpler and bolder in pacificus, more complex and softer in uropygialis; some birds do not fit the bill, perhaps misidentified to ssp.?) I would like to see more rigorous bioacoustic comparisons (these are caciques!), and I have not assessed whether the proportionally longer tail of uropygialis reported in the literature passes a photographic test (though looking at the measurements it should be obvious). I would also like to see a stringent morphological study across space, in order to test whether the size differences hold or there is clinal variation and an assessment on whether the populations to the east and west of the Andes spill to one or the other side, meet somewhere, etc. In sum, although a split seems likely, I would like to see a more detailed case.

“I also note that the reduced number of samples in Powell et al. (2011) will be unable to uncover more complex gene flow or contact areas scenarios, and is largely driven by mitochondrial DNA. The two microrhynchus samples are from Panama, the single sample of pacificus is from Ecuador (Esmeraldas) and the single sample of uropygialis is from Ecuador (Morona-Santiago).”

Comments from Stiles: [YES]

“Gathering what data I can from my field notes and collections, I can make some contributions regarding proposals 946 through 948. In this, I have put together data on weights with those for wing lengths and tail lengths. Taking the cube roots of the weights of each taxon and dividing the linear values for wings and tails, I obtain values of these adjusted to their absolute lengths: relative values independent of absolute size. I have found such values very useful for hummingbirds, and they here provide useful insights for the Cacicus as well. I haven’t gone into this for bills because I suspect that the most useful measurement would be the height of the bill at the nostril, for which I have no data for microrhynchus. I didn’t take these in Costa Rica (except for hummingbirds, and then only in visits following most of my work there). Because these caciques are typically canopy birds, they don’t often get caught in mist nets, so sample sizes are small, and there are holes in the data (I lack data for females of uropygialis, tails for microrhynchus, and weights of females for pacificus, so here I give only data for males.

C. microrhynchus: mean weight: 68.7g; mean wing length: 136.5mm; relative wing length: 33.32.

C. pacificus: mean weight 80.1g; mean wing length 136.25mm; mean tail length 90.15mm; relative wing length 31.60; relative tail length 20.11.

C. uropygialis: mean weight 87.6g; ,mean wing length 143mm; mean tail length 103.7mm; relative wing length 32.24; relative tail length 23.35.

“These data tend to confirm the suggestions that microrhynchus does have considerably relatively long wings, and that the tail of uropygialis is indeed relatively longer than that of pacificus. The latter, appreciably heavier than microrhynchus but with wings very similar in length, would seem to be a more robust bird, probably with higher wing loading and differing flight characteristics (but wing areas would be needed to confirm this). All this would add evidence favoring recognition of pacificus as a separate species. Hence, I will vote YES to adopting the three-species option D in proposal 946.”

Comments from Fernando Machado-Stredel: “I think it is important to note that the plumage similarity between haemorrhous and the other caciques with red rumps may not be rare, with some recurrent plumage patterns being common between non-sister taxa within Cacicinae. For instance, one could think on the pairs Cacicus koepckeae - Cacicus chrysonotus chrysonotus (both with yellow rumps), Cacicus sclateri - Cacicus solitarius (all black), or Cacicus cela - Cacicus chrysonotus leucoramphus (both with yellow rumps and a yellow wing patch, although in cela the yellow extends to the vent). In these cases, one taxon is larger than the other.

“The measurement data from Jaramillo & Burke (1999) shows that in the "red" caciques, haemorrhous is the largest, also differing in the size of the rump patch (as Jaramillo mentioned in proposal 73), followed by nominate uropygialis, which has longer mean wing and tail, as well as no overlapping ranges with the other two smaller taxa in these measurements. Cacicus u. microrhynchus and C. u. pacificus do overlap in these traits. Here are the data for males:

C. h. haemorrhous (n=28): wing 175.8 (168-187.5); tail 105.9 (100-115)

C. uropygialis: wing (n=8) 156.5 (145-165); tail (n=9) 129.6 (107-142)

C. m. pacificus (n=10): wing 132.2 (122.2-137.2); tail 91.1 (87.8-94.9)

C. m. microrhynchus (n=10): wing 129.0 (122.0-136.5); tail 89.7 (83.3-96.5)

“Lastly, I wonder if the pacificus sample (ANSP 182884) from Ecuador used in Powell et al. (2014) could be compromised, in which case, contamination may play a role in the observed phylogenetic relationships.”

Comments from Claramunt: “NO. The situation is far from clear. Previous evidence was interpreted as suggesting that pacificus should be separated but the new genetic evidence contradicts that. We clearly don’t have much evidence for splitting (and how) this species. I always like to see measurements, but just looking at taxon averages, without an assessment of variation or potential clines, does not inform much regarding potential species limits.”

Comments from Robbins: “YES. Taking the Powell et al. (2014) genetic data as a correct representation of relationships, this is still a difficult proposal to evaluate. I would prefer, like the recent Trogon rufus proposal, to have data from that key area of potential contact in northwestern Colombia. If uropygialis (sensu stricto) and microrhynchus do not hybridize or is limited then clearly at least two species should be recognized. That coupled with pacificus being more distantly related (a surprise, because of the biogeography) to the other two, would argue for it also being recognized as a species. Finally, if Boseman (2016) is correct in stating that microrhynchus and pacificus are so different in vocalizations, then that supports species recognition for those. Although I’m still on the fence on this, the preponderance of information at this point makes me think we should go ahead and recognize three species until there are data to indicate otherwise.”

Comments from Pacheco: “YES to D. Considering the distinctions available for morphology, genetics and vocal repertoire. Added to this is the suggestive information gathered by Gary and, finally, to realize that the historical lumping of the three taxa was merely arbitrary.”

Additional comments from Stiles: “Upon consulting Ridgway, vol. 4 (19xx), I found that he also presented data on length of culmen from base (CfB) and Bill height at base (BHb) for both sexes of C. microrhynchus (as well as wing and tail lengths) for both sexes. To compare these with our specimens of pacificus, I measured the same bill dimensions of these: The results-Valle del Cauca

“For bills: the ratio of CfB/BHb) for males: microrhynchus = 2.702; ; for pacificus = 2.221; therefore, the bills at base are taller relative to their lengths in pacificus, thus supporting the statement that their bills are “more powerful” than those of microrhynchus.

“Considering distributions, Hilty (1986, 2021) showed that the ranges of pacificus and uropygialis in Colombia are mostly separated by the dry valley of the río Cauca, and only intersect in two areas: the upper río San Juan region in Valle del Cauca (the only extension of uropygialis onto the Pacific slope – and also the same area as the W part of the hybrid zone in Ramphocelus studied by Sibley and Morales, but Hilty explicitly noted that here the two Cacicus separate by elevation), and at the extreme northern end of the Central Andes, where uropygialis is mapped in the mountains and pacificus in the adjacent Sinú lowlands (but I know of no observations or specimens that explicitly document such a separation here).

Comments from Bonaccorso: “NO (no change for now). Plumage differences are weak at best. The DNA data is for one sample of each subspecies (yes, the branch lengths are long, but the sampling is too small!) and only mitochondrial genes. Even if the tree showed the actual relationships among these taxa, there is no way of detecting admixture between C. u. pacificus and C. u. uropygialis with such a sparse sampling. Also, if they intergraded, it would be almost impossible to notice it by plumage. It seems that the vocal information from Boesman and other sources would be a good start but need to be analyzed in a quantitative way.”

Additional comments from Stiles: “A minor disagreement with Elisa, that the plumages of pacificus and uropygialis are indistinguishable. Strictly on plumage, yes, but given the great difference in measurements between these two, any intermediate specimen would be easy to detect by measurements (and so far, none has been detected though, admittedly, no one so far has specifically tried to collect at intermediate elevations). Regarding the possible contact zone in N Antioquia, Andrés Cuervo has definitely recorded uropygialis at > 1500m, but the rest of the transect remains unstudied (probably unsafe to try at present).

Comments from Lane: NO. Whereas I suspect that this complex will require splitting at some point (much as C. cela, and probably C. chrysonotus will as well), I don’t feel that Powell’s phylogeny is sufficient for us to assess the true phylogenetic relationships among these taxa. A far better-sampled study would go a long way to illuminate these relationships satisfactorily. Furthermore, I don’t hear super-distinctive vocalizations from the three taxa for me to say that they can’t be simply subspecific variation. Therefore, NO on splitting this complex based on the present information.”

Comments from Jaramillo: “YES – Treat the three as different species. Icterids are troublesome is what I have to say! The issue is that you would expect that birds which are colorful, or have a single distinct color region on a largely black body, would sort out easily based on plumage. That is to say, they are not tyrants or furnariids in which species can look exactly the same, or we discount plumage as a feature to look at, because we have little to look at. Icterids are often boldly colored, so our expectation is that different taxa should differ significantly in plumage, but ironically, they do not do this. Look at Tricolored and Red-winged blackbirds; Western, Eastern and Chihuahuan meadowlarks, or recurring patterns that are extremely similar from birds in separate genera, or subgenera (Yellow-winged vs Yellow-shouldered blackbirds; various orioles…. Etc.). Then we have the big mess of yellow rumped, black rumped, and red-rumped caciques. In short, I think that we are expecting that good species in this group should show a distinct feature, differences in the rump color perhaps? But they do not. The differences tend to be in their ecology, voice (although difficult to sort out given how complex the voices are), or size and structure. This is the problem here.

“Basically, we almost don’t believe the molecular data because our expectation is that they should look more different than they do. Yet if we had the same phylogeny and they were Zimmerius tyrannulets, that would not concern us. It is not a fair comparison, as the tyrannulets would likely have quite obvious simple differences in voice. Although icterids are complicated in voice, so much so that even when there are clear differences as in eastern vs western Great-tailed Grackles, the multiple voices given in different contexts confuses the listener so much that most experienced birders have not noticed the differences when they visit the range of the western vs eastern clade. I think something similar is going on here. The voices are different, but it takes some analysis to get at that. Fortunately, a bit of that has been completed, as noted in the proposal.

“There are other differences, as Gary notes the measurements, the bill structure and relative tail lengths etc. Notably if you riffle through photos of microrhynchus and pacificus and just concentrate on the bills, they are quite different. The larger billed pacificus has a similar length bill to microrhynchus, but the base is thicker, specifically it looks thicker on the maxilla. This creates almost a casqued look on some individuals. The form microrhynchus has a thinner bill, but also a more noticeable downward curve nearer the tip. One would surmise the different bill sizes on a similar size bird is telling us something about their differing ecology? As well, there are subtle coloration differences. Bill colors differ quite a bit in the cacique – oropendola group; it is unclear if these have any role in mate selection. If you look carefully at a series of microrhynchus bills, they are yellow but show a dull greenish wash to the bills. Overall, they are duller; whereas on pacificus, they are yellowish but have a distinctly brighter base, some nearly orange at the base, and becoming more greenish yellow towards the tip. Adding in the much larger uropygialis, it is more extreme in the bill thickness, and has a much more noticeable bright colored base to the bill.

“It would be great to have more information on coloniality in these birds. It seems that some populations are colonial, and others are solitary. Yet these details are not well known, and unclear how they map on to the three taxa. The three are divided geographically, but as you would expect also ecologically. This has been one of the key issues that had led many to separate “Subtropical Cacique” uropygialis from the others given its different elevational range. The idea Van focused on, regarding the break in the Darien is important. The two smaller subspecies were formerly lumped due to similarity in size, and plumage. But the fact that they are nearly touching, no evidence of intermediates, and differences in bill shape suggesting differences in what they feed on are important. Voice is different as well for those two.

“I think these are three different species and feel comfortable making the change. The molecular phylogeny is vital here, but take the similarities in plumage with a grain of salt. This is not unusual in icterids, and surely, we will have more splits in the future in this family where plumage is not as informative as we would think it should be in brightly colored species.”

Comments from Zimmer: “YES to D, for reasons stated by Mark, Gary, and Alvaro. As Alvaro says, “Icterids are troublesome”. Plumage patterns are clearly highly conserved in this group, and vocalizations are complex (including advanced mimicry abilities among some species), making traditional vocal analyses difficult. In this case, the morphometric and genetic differences, combined with lack of evidence of intermediacy from potential contact zones, and apparent qualitative differences in at least some vocalizations, are enough to sway me.”