Proposal (978) to South American Classification Committee

Treat the two subspecies of Parabuteo unicinctus as separate species

Background and Current taxonomy

Parabuteo unicinctus currently consists of two subspecies, P. u. unicinctus (Bay-winged Hawk, Temminck 1824) and P. u. harrisi (Harris’s Hawk, Audubon 1837). From 1837, these two were generally placed in different genera (about 14 different genera) and, uncommonly, in the same genus (Craxirex 1861-1866, Antenor 1875-1882). They were finally stabilized in Parabuteo (1916 to 1922).

These subspecies differ by size, range, and plumage variation. Unicinctus is distinguished from harrisi by having a dark brown ventrum streaked or flecked with white or whitish, a smaller body and a longer tail (Dwyer and Bednarz 2020).

Their ranges are traditionally thought to be:

• P. u. unicinctus is resident from ne. Colombia and w. Venezuela south through e. Bolivia and central and ne. Brazil to s.-central Chile and s. Argentina.

• P. u. harrisi is resident from se. California, east to central Texas and south throughout n. Mexico, then south along the Pacific slope from w.-central Mexico to El Salvador and from Nicaragua to w. Colombia, w. Ecuador, and w. Peru.

New information

Clark and Seipke (2023) presented compelling new morphological data, and they reviewed the behavioral and DNA data, all supporting the elevation of these two subspecies to species status.

Morphological Data

Adult and juvenile plumages differ between the two subspecies. For adult birds, these include throat markings, color and markings on the undersides of the remiges, markings on the belly and breast, markings on the leg feathers, and extent of white at the base and tips of the rectrices.

In juvenile birds, consistent color differences occur in underparts, undertail, upper back, upperwing coverts, and underwing coverts. In addition, P. (u.) unicinctus has delayed plumage maturation with four distinct age classes, and Clark and Seipke (2023) provide the first detailed descriptions of Basic II and Basic III, which are similar in body plumage to juvenile P. (u) harrisi.

See power point by Clark and Seipke.

Ecology

Harrisi individuals hunt and breed cooperatively. Although some examples of co-operative breeding have been noted in unicinctus, there have been no examples noted of co-operative hunting.

DNA

Clark and Seipke (2023) reviewed the limited DNA evidence. In a paper on Buteo phylogeny, Riesing et al. (2003) used two mitochondrial markers (nd6 gene and pseudo-control region (ΨCR) and sequenced two unicinctus specimens and one harrisi. Their ML tree shows the two South American as sister to the harrisi specimen.

Raposo do Amaral et al. (2009) sequenced mtDNA and the nuclear intron Fib5 for three Parabuteo specimens, one from Texas and two from South America. As with the Riesing tree, the two South American clustered together as sister taxa to harrisi.

A quick analysis of Parabuteo cytb sequences (the three from Raposo and Lerner et al. harrisi sequence), using Blast, resulted in two clades – the two harrisi as sister taxa to the two unicinctus (pers. obv.).

Recommendation:

The new morphological data and existing DNA and behavioral data support elevation of the two subspecies and the redefinition of the ranges of the species as follows:

• P. unicinctus - resident in much of lowland tropical and subtropical non-Amazonian South America, including west of the Andes in Chile, Peru, Ecuador.

• P. harrisi - resident from se. California, east to central Texas and south throughout n. Mexico, then south along the Pacific slope from w.-central Mexico to Costa Rica.

Although the current range of harrisi is described as occurring in northwest South America south to northwestern Peru, new information about the Basic II and III plumages of unicinctus suggests that these South American sightings were probably juvenile unicinctus, because observers unaware of these plumages would be unable to distinguish them from juvenile Harris’s Hawk. Clark and Seipke (2023) also noted that researchers in Ecuador have not seen Harris’s there.

Current range maps and eBird sightings for P. unicinctus show geographic separation of South American populations from Central and North American, with a gap in distribution in Panama. This is consistent with an interpretation of the ranges of P. harrisi and P. unicinctus being geographically separated.

Literature Cited

Clark, W. S. and S. H. Seipke. 2023. Taxonomic status of Bay-winged Hawk Parabuteo (unicinctus) unicinctus and Harris’s Hawk P. (u.) harrisi, with documentation of delayed plumage maturation in Bay-winged Hawk. Bull BOC 143(2):142-152.

Dwyer, J. F. and J. C. Bednarz. 2020. Harris's Hawk (Parabuteo unicinctus), version 1.0. In Birds of the World (A. F. Poole, Editor). Cornell Lab of Ornithology, Ithaca, NY, USA. https://doi.org/10.2173/bow.hrshaw.01

Lerner, H. R. L., Kolaver. C. K. & Mindell. D. P. 2008. Molecular phylogenetics of the buteonine bird of prey (Accipitridae). Auk 125: 304–315.

Raposo do Amaral, F. R., Sheldon, F. H., Gamauf, A., Haring, E., Riesing, M., Silveira, L. F. & Wajntal, A. 2009. Patterns and processes of diversification in a widespread and ecologically diverse group, the buteonine hawks (Aves, Accipitridae). Mol. Phylogenetics & Evol. 53: 703‒715.

Carole S. Griffiths, July 2023

Comments from Areta: “YES. The differences in plumages of juvenile and definitive harrisi and unicinctus are striking, and have indeed been long recognized. Although not new to me, it is good to see the evidence for basic II and basic III plumages of unicinctus being published and called to the attention of the ornithological community. It is a pity that the wing molt of the specimens used to substantiate the claims of basic II and III unicinctus is not shown in the paper and not accurately described, and that the specimens examined in museums have not been properly listed with their corresponding molt. The fact that no basic II or III plumages are shown in flight in the paper itself is rather confusing, as unaware readers may be drawn to think that birds can be properly aged by looking at the body aspect: yet as described in the text, proper aging is performed by looking at wing molt (something that Clark and Seipke have championed for a long time, to be sure). I find it surprising that no attempt was made to characterize the vocalizations of the two taxa; someone should explore this. The limited genetic data do not provide strong support for species status of unicinctus and harrisi, and are compatible with different possible scenarios. However, I think that the data collectively tip the scale to consider unicinctus as a separate species from harrisi, and the burden of proof should be inverted. On a personal note, I lament that the authors did not "find" that Pearman & Areta (Field Guide to the Birds of Argentina and Southwest Atlantic, 2020) have also provided accurate illustration of unicinctus (juvenile and adult). To conclude, I vote YES to the split of unicinctus from harrisi, although I think that a much stronger case could have been built to support this treatment.”

Comments from Stiles: “YES, in the sense that although the data supplied by Clark & Seipke leave some notable holes, it definitely shifts the burden of proof onto those who would maintain both groups as subspecies under unicinctus.”

Comments solicited from Therese Catanach: “The fact there is some shallow population structure in a widespread species (especially one that undergoes limited or no seasonal movement in much of its range) is unsurprising. However, level of genetic differentiation between the two currently recognized subspecies of Parabuteo unicinctus is well within the range expected for intraspecific variation. For example, 4 cyt b samples from Parabuteo leucorrhous are also publicly available. Two of these samples, ANSP 180970 (tissue # ANSP 660) and ANSP 186050 (tissue # ANSP 5082) are both from Ecuador and yet exhibit similar levels of genetic differentiation (0.47% divergent) as those observed between North (including a new cytb sequence generated by Catanach et al. 2023) and South American populations of Harris’s Hawk (which range between 0.46 and 0.58% divergence). This level of divergence is low when compared to other hawk species, for example a sampling of Red-tailed Hawks show divergence as high as 1.7% when comparing specimens collected in the United States. When comparing these p-distances to those calculated for sister species, Parabuteo leucorrhous and Parabuteo unicinctus are over 6% divergent. When looking across Buteoninae as a whole (using the data from Catanach et al. 2023), the average cytb divergence is 2.9% divergent between sister species. Only 6 sister species are lower than 1.4% compared to the maximum of 0.58% exhibited within the Harris’s Hawk samples. Of these sister species pairs, the species status of most has fluctuated in the literature (e.g. Haliaeetus sanfordi and H. leucogaster and several Old World Buzzard taxa). In fact, only one, the Buteo galapagoensis and swainsoni pair is only 0.18% divergent and has not been widely questioned. However, these two taxa are morphologically distinct enough that a sister relationship was not suspected until sequencing data recovered this arrangement (Bollmer et al. 2006) making it a very different situation than the two morphologically similar Harris’s Hawk subspecies.”

Bollmer, JL, Kimball, RT, Whiteman, NK, Sarasola, JH, Parker, PG. 2005. Phylogeography of the Galápagos hawk (Buteo galapagoensis): A recent arrival to the Galápagos Islands. Molecular Phylogenetics and Evolution. 39 (1): 237–247.

Catanach, TA, Halley, MR, and Pirro, S. 2023. Enigmas no longer: using Ultraconserved Elements to place several unusual hawk taxa and address the non-monophyly of the genus Accipiter (Accipitriformes: Accipitridae). bioRxiv 2023.07.13.548898; doi: https://doi.org/10.1101/2023.07.13.548898

Comments from Zimmer: “YES, although I don’t consider it a slam-dunk. I’m really not all that impressed by the meager genetic data, but the sum of fairly obvious differences in both adult and juvenile plumages, the differences in molt sequences and number of plumage cycles between the two, and the ecological/behavioral differences, combined with the fact that the two were not only historically treated as separate species, but also were often treated as belonging to separate genera, makes me think that the burden of proof should shift to those favoring maintaining the one-species treatment. I would second Nacho’s call for a vocal analysis to strengthen the case – I suspect that vocalizations might prove diagnostic.”

Comments from Lane: “YES. As others state, I think the overall amount of evidence makes this split quite reasonable. I have long noticed that North American and South American Parabuteo differ. Even those on the Pacific coast of Peru didn't match North American birds, so I never understood how they could be considered harrisi? The present study explains all this satisfactorily to me.”

Comments from Del-Rio: “NO. Although the morphological differences exist, I am afraid the molecular differences are not compelling enough. I will keep my standard criteria here.”

Comments from Robbins: “NO. Clearly, genetic data do not offer support for species recognition. Moreover, like Nacho, I have issues with how the data were (or were not) presented. Likewise as Nacho pointed out, there should have been an attempt to ascertain whether there are differences in vocalizations. Whether differences in juvenile plumage merit species recognition is an open question. I vote NO for species recognition until there is a clearer and more detailed analyses of these two taxa.”

Comments from Brian Sullivan (voting for Remsen): “The authors should be commended for bringing this interesting and vexing taxonomic issue back into the light. In this case, there are clearly two taxa involved—I completely agree! But how and where those taxa come into contact, whether they interbreed, and what their respective ranges are, all require further study. I don’t think the proposed distribution break in Panama is the simple and clean delineation the authors suggest it to be—at least not based on the images I’ve found below, which show Harris’s-like birds all the way south to Ecuador in South America, and more mixed types in Lima down through Santiago. South of there, based on a quick look at photos in ML, it seems like most are Bay-winged types. Everything east of the Andes seems like Bay-winged types. 

“Given that I only spent an hour perusing the Macaulay Library’s collection of these taxa, it seems relatively easy to find exceptions to the hard and fast rules outlined in the paper, such as:

“There are no valid records of Harris’s Hawk in South America.”

https://macaulaylibrary.org/asset/217536651

https://macaulaylibrary.org/asset/427018861

https://macaulaylibrary.org/asset/481506641

https://macaulaylibrary.org/asset/116191271

https://macaulaylibrary.org/asset/328118561

“All adult harrisi have unmarked dark undersides to the remiges”

https://macaulaylibrary.org/asset/121846531

“whereas all adult unicinctus have whitish primaries with narrow dark bands and black tips on the outer ones, but many adult unicinctus have whitish secondaries with narrow dark bands (Fig. 2a), though some have darker secondaries with some narrow white bands”

“See the many examples above that would be unicinctus based on this proposed split that have all-dark remiges like harrisi.

“All adult unicinctus have a broad darker subterminal band on the secondaries”

https://macaulaylibrary.org/asset/559744391

https://macaulaylibrary.org/asset/559744351

“We found no specimens or photographs exhibiting characters of both taxa.”

“See example from Lima above.

“For now, I vote ‘NO’ on this proposed split, as I believe further study is warranted. It could be that two species are involved, but more information about how they interact, especially north and west of the Andes, seems essential.”

Comments from Claramunt: “NO. I don’t see compelling evidence for this split. The new data on juvenile plumages by Clark & Seipke is very interesting and important; they found that the juvenile plumage of birds from South America is different from that of birds in Central and North America. But what is lacking for resolving this problem is a detailed individual-level analysis of the geographic variation of the relevant traits. That analysis would clarify whether there are two species or just a single species that exhibits geographic variation. Framing this problem as an issue of counting differences between two well separated taxa can be misleading.”

“The main problem is that birds from the Pacific side of Colombia, Ecuador, and Peru, have been considered to be harrisi, not nominate unicinctus. Clark & Seipke try to minimize this issue by implying a “mistake” in the recent literature:

‘Some authorities (e.g., Dickinson & Remsen 2013, Clements et al. 2022) have mistakenly listed Harris’s Hawk for north-west South America.’

“In reality, most previous avian taxonomist have assigned these populations to harrisi based on the examination of specimens that showed uniformly dark underparts like harrisi (e.g. Hellmayr and Conover 1949, also citing Chapman, Blake 1977). Those harrisi-like individuals found from Colombia to Peru can be seen also in the photos posted by Brian Sullivan.

“Therefore, birds from the Pacific side of South America seem to show a mix of harrisi and unicinctus characteristics, exactly what would be expected for a widespread species that has been exchanging traits via gene flow.”

Comments from Jaramillo: “NO. There does seem to be something here that requires further work to adequately describe the situation. Are there two, or are there three populations? The latter has not been evaluated, but it is likely. That third Pacific population, what is it? An intermediate, or a different population altogether, a subspecies of harrisi? What I can say is that this species is a common hawk of Chile’s central zone, yet in the last decade or two it has become more common in the far north, where it is assumed that the Peruvian population moved south. To my eye, those northern birds do not look like the true unicinctus of the south. I have not understood exactly how they look different, but the tendency is to look harrisi like.

“Also, it is worth noting that there are multiple gaps in the distribution of this species, not just the one in Panama. There is an isolated group in Colombia/Venezuela. Then we have the Pacific group that stretches from S Ecuador to (now) northernmost Chile. But it has been and still is largely isolated from the unicinctus of the southern cone. That gap is narrowing. To my untrained eye, the birds in Colombia/Venezuela are harrisi. The map shown in the powerpoint is also not correct: there is no connection of unicinctus to the Colombia/Venezuela population via the area east of the Andes -- there are no Parabuteo there. The fact that we have intermediate looking birds, at least 4 isolated populations, and lots of questions means there is more work needed.”

Additional comments from Lane: “One thing that may be confounding the issue is the presence of escaped uncinatus in coastal Peru thanks to escaped falconers’ birds and efforts to use hawks as pest control in agricultural areas along the coast (this per the attached paper and pers. comm. from Fernando Angulo). This may be changing the dynamics of which taxon is present at least in coastal Peru. So this bit of information may be something that clouded the distributions of the two taxa along the Pacific coast of Peru, and perhaps elsewhere. Whether the two forms interbreed or not would be worthy of study. Perhaps the votes on this proposal should be postponed until such a study is run?”

Additional comments from Stiles: “Given the multiple problems with the data of Clark & Siepke and Brian and Alvaro's comments in particular showing various points of ambiguity, I agree that more detailed information, largely requiring more field studies, would be needed to fully understand this interesting problem, so I will change my vote to NO.”

Comments from Bill Clark and S. Seipke:

“Note that there were three committee members who positively for this split. Two of the three committee members who yoted yes mentioned the differences in immature plumages. However, none of the six committee members who voted no of one mentions this major difference between these taxa: the difference in the number of immature plumages. No other of the 320+ species diurnal raptors differs in the number of immature plumages between or among subspecies. Surely this has a genetic basis and should be a factor in the taxonomic decision.

“Specific comments:

“Sullivan found some photos of Bay-winged Hawks from the west of the Andes that have uniformly dark remiges. What he did not find were any photos of juvenile harrisi from this region. No photos exist in the Macaulay library from Colombia to Peru that are of juvenile harrisi. All in this photo collection and museum specimens from this region are of juvenile and Basic II & III are unicinctus. There may well be a subspecies of unicinctus from there in which the adults have dark remiges. However, the birds west of the Andes are not harrisi, as there are no juveniles of that taxa there, only juveniles and Basic II & III unicinctus.

“Claramunt states that the birds from west of the Andes are harrisi, but there are no photos or specimens of juvenile harrisi from this region. Only adults with dark undersides of their remiges. He mentioned the lack of individual-level analysis of the geographic variation in the paper. He is mistaken in that every juvenile specimen or photo from South America fits the characters of juvenile unicinctus and, likewise, every juvenile specimen and photo from Central America north fits the characters of harrisi, and there is no overlap in these characters.

“Jaramillo offers little in taxonomic insight other than there may be a separate taxon west of the Andes.

“Catanach and Del-Rio think that the level of genetic differences is too low for species, but the genetic analyses done to date are few and shallow. Again, why did they not discuss how two subspecies can have a differing number of immature plumages?

“Robbins questions the differences in juvenile plumage of the two taxa, when it has been clearly shown in Clark & Seipke (2023) that they differ greatly. The reason we did not use vocalizations is discussed below under Areta.

“Areta (a positive vote) Mentions that Clark & Seipke (2023) do not show Basic II and Basic II unicinctus in flight; they do however, refer to URLs in the paper pointing to flying immature unicinctus that show this primary molt of both ages of immatures.

“Areta suggests that vocalizations should be compared. We did listen to some and felt that they were too similar to use for differences. Sister species can have the same call. For Example:

• Amur Falcon

https://xeno-canto.org/species/Falco-amurensis

• Red-footed Falcon

https://xeno-canto.org/species/Falco-vespertinus

“Their calls sound the same. See also the attached sonograms below.

“In summary, we feel that the consideration of this proposed split by some committee members was shallow and did not take into account the great differences in annual plumage sequences between these taxa, as well as the great differences in plumages, especially juveniles.

“David Mindell, noted raptor taxonomist, supports our arguments above.”

Comments from Fernando Angulo:

“I have a couple of comments regarding the proposal and the PowerPoint presentation.

“1. Lima’s Parabuteo unicinctus is a population formed from escaped birds that were used by falconers, including me. We got our birds at that time, from local trade. I am referring to years 1990 approximately, until more or less 2000 (This stopped when there were available birds from captive breeding). I myself have contributed with at least one escaped bird. Birds that were acquired on those days were from unknown origin (sellers told us “From the north”, referring to Piura or Tumbes, but we have no certainty of that). So, the Lima population is formed from unknown subspecies of Parabuteo unicinctus, basically, from birds of unknown origin.

“2. Distribution in coastal Peru, as far as I know, has to be taken carefully. The Lima population is from escaped birds, which are expanding north, west, and east from the Lima city. Also, there is a native population (you can check Koepcke’s Aves de Lima book, where she mentions “rarely seen on the coast and Andean slopes”). Other coastal populations can be safely regarded as wild, such as in the dry forest of Tumbes, Piura, and Lambayeque, for example. Here, part of the problem is that with growing agriculture activities on what was desert on the Peruvian coast, “bird control” using trained Parabuteo unicinctus has increased a lot, especially in Lima, Ica, La Libertad, Lambayeque, and lower Piura. This means that any Parabuteo unicinctus found on any of the coastal cities has to be carefully evaluated, because there is a great chance of being escaped from trained birds.

“3. Trained birds currently come from breeding centers that are using Parabuteo unicinctus captured or obtained from the Lima population, or not. The Lima Parabuteo unicinctus population originated from birds of unknown origin, and the captive breeding centers uses birds from unknown origin (it can be from injured birds from Lima population x a wild bird confiscated from illegal trade, or any of the imaginable possibilities), so the Parabuteo unicinctus used for bird control, i.e. the ones likely to be found on coastal cities, can’t be regarded as typical looking wild birds.

“4. A problem I see, is that several pics from Peru on the Clark & Seipke power-point are taken by Lee Jones. Lee was a veterinarian working at El Huayco some 5 or more years ago. We cannot eliminate the possibility that the pics he took belong to Lima Parabuteo unicinctus (not wild birds) that are simply, from unknown/uncertain origin or identity. So, those pics can’t be used to separate these two spp into sp. (slides 8 and 21). On slide 21, you can notice infrastructure of El Huayco in the background.

“5. On the other hand, I can see plenty of birds from Peru that look more closely to what they call harrisi, and are supposed to be as south as Panama, according to their ppt. https://ebird.org/checklist/S93955615 https://ebird.org/checklist/S79099153, from Lima, Piura, and I guess a thorough search will find more.”

Additional comments by Claramunt: [in response to Clark’s points]: "My point is that there seems to be a discrepancy between adult plumages and juvenile plumages in SA birds E of the Andes: the adult plumage matches harrisi, the juvenile plumage matches unicinctus. Which one is correct? We don't know. We can't just assume that the juvenile plumage is correct in indicating affinities here. Maybe both are "correct" and these birds share traits and genes with both North American and eastern S American birds."