Proposal (1005) to South American Classification Committee

Revise classification of New World nightjar genera

Background: Genetic data (Han et al. 2010, Sigurðsson & Cracraft 2014) has shown that many New World nightjar genera are not monophyletic. Clearly, conservative plumage evolution has confounded phylogenetic classification of this group. We had previously addressed some of the problems piecemeal by resurrecting generic names. However, veterans will remember that we struggled to match phylogeny to existing classification not once but twice: see SACC proposals 465 and 501 for two failed attempts, because of morphological heterogeneity, to expand Hydropsalis to solve problems of non-monophyly. As a consequence, we voted to continue with traditional classifications, although knowing that some of that maintained nonmonophyletic genera, which was highly unsatisfactory to any of us. But we decided to wait for the proper taxonomic revision that we knew would come along.

Our current classification of the family is as follows; taxa shaded in gray are not part of this lineage and remain unchanged:

Chordeiles nacunda Nacunda Nighthawk 3e

Chordeiles pusillus Least Nighthawk 2a

Chordeiles rupestris Sand-colored Nighthawk

Chordeiles acutipennis Lesser Nighthawk

Chordeiles minor Common Nighthawk (NB) 3b

Chordeiles gundlachii Antillean Nighthawk (V) 3c, 3d

Lurocalis semitorquatus Short-tailed Nighthawk 1a

Lurocalis rufiventris Rufous-bellied Nighthawk 2

Nyctiprogne leucopyga Band-tailed Nighthawk 3a

Nyctiprogne vielliardi Bahian Nighthawk 3, 3aa

Nyctipolus nigrescens Blackish Nightjar 14, 14a

Nyctipolus hirundinaceus Pygmy Nightjar 14a

Systellura longirostris Band-winged Nightjar 9a, 9b,

Systellura decussata Tschudi’s Nightjar 9a

Nyctidromus albicollis Common Pauraque 4, 4a, 4b

Nyctidromus anthonyi Scrub Nightjar 12

Eleothreptus candicans White-winged Nightjar 10, 9b

Eleothreptus anomalus Sickle-winged Nightjar 10

Uropsalis segmentata Swallow-tailed Nightjar 16, 17

Uropsalis lyra Lyre-tailed Nightjar

Setopagis heterura Todd's Nightjar 11, 11a

Setopagis parvula Little Nightjar 11, 12, 12a

Setopagis whitelyi Roraiman Nightjar 14, 14a

Setopagis maculosa Cayenne Nightjar 13

Hydropsalis cayennensis White-tailed Nightjar 9b, 9c

Hydropsalis maculicaudus Spot-tailed Nightjar 10a

Hydropsalis climacocerca Ladder-tailed Nightjar 14b

Hydropsalis torquata Scissor-tailed Nightjar 14c, 15

Macropsalis forcipata Long-trained Nightjar 17, 18

Nyctiphrynus rosenbergi Choco Poorwill 5

Nyctiphrynus ocellatus Ocellated Poorwill 5b

Antrostomus sericocaudatus Silky-tailed Nightjar 6b, 8, 9

Antrostomus carolinensis Chuck-will's-widow (NB) 6, 6a, 6b

Antrostomus rufus Rufous Nightjar 6, 6b, 7

Our current Note for the family reads as follows:

1. Although many classifications (e.g., AOU 1998) divide the Caprimulgidae into two subfamilies, Chordeilinae and Caprimulginae, Barrowclough et al. (2006) and Han et al. (2010) found that this creates paraphyletic groupings, with "Chordeilinae" embedded with Caprimulginae. Sibley and Ahlquist (1990), Barrowclough et al. (2006), and Han et al. (2010) found that Australasian Eurostopodus was basal to all other caprimulgid genera sampled. Barrowclough et al. (2006), Larsen et al. (2007), Han et al. (2010), and Sigurdsson & Cracraft (2014) also found that Caprimulgus, as currently constituted, is highly polyphyletic; for example, New World taxa sampled so far indicate that northern “C.” carolinensis and “C.” vociferus fall within a group that includes Phalaenoptilus (extralimital), Nyctiphrynus, and Otophanes, whereas “C.” longirostris, “C.” maculicaudus, and “C.” parvulus fall within a group that includes Hydropsalis, Uropsalis, Nyctidromus, Nyctiprogne, and Lurocalis. Han et al. (2010), with broader taxon and gene sampling, proposed some major revisions to the classification of the Caprimulgidae. SACC proposal for major changes in boundaries of genera did not pass. New SACC proposal passed to revise generic limits in nightjars. Sigurdsson & Cracraft (2014), with extensive taxon-sampling, found that the family consists of seven major clades, three of which occur in South America: a “poorwill” clade (which includes Antrostomus and Nyctiphrynus), a “nighthawk” clade, and a “South American” clade (all other genera); see also White et al. (2016). Proposal needed for new linear sequence.

We also acknowledge problems with our existing classification in five additional notes that cite Costa et al. (20023) that end with “SACC proposal badly needed.”

New information: Costa et al. (2023) have come to the rescue and have addressed all the problems we had previously with a broadly defined Hydropsalis. The solution required naming two new genera and resurrecting two more. Costa et al.’s paper is outstanding in its thoroughness as a taxonomic revision that integrates the latest genetic data with nomenclature and taxonomy to produce a phylogenetic classification, and I recommend it as a model for analogous revisions. Rather than repeat the information outlined with superior clarity in Costa et al., I refer you to their text for rationale, diagnoses of all genera, photos of specimens, etymology of new genera, etc. Perhaps a classic example of overlooked biodiversity (handy to have in your arsenal) is the transformation of what was treated traditionally and recently as Caprimulgus longirostris decussatus into a monotypic genus: Quechuavis decussata.

Here are the phylogenetic trees on which their classification is based:

Their proposed new classification superimposed on our existing classification would look like this, with the newly recognized genera in red. Sister taxa highlighted in yellow maintain our NW to SE convention for sister species; the sequence was reversed in Costa et al. without explanation that I can find. It can’t be due to subspecies diversity because parvula and maculosa are monotypic.

Nyctiprogne leucopyga Band-tailed Nighthawk 3a

Nyctiprogne vielliardi Bahian Nighthawk 3, 3aa

Lurocalis semitorquatus Short-tailed Nighthawk 1a

Lurocalis rufiventris Rufous-bellied Nighthawk 2

Nyctipolus nigrescens Blackish Nightjar 14, 14a

Nyctipolus hirundinaceus Pygmy Nightjar 14a

Nyctidromus albicollis Common Pauraque 4, 4a, 4b

Nyctidromus anthonyi Scrub Nightjar 12

Tepuiornis whitelyi Roraiman Nightjar 14, 14a

Uropsalis segmentata Swallow-tailed Nightjar 16, 17

Uropsalis lyra Lyre-tailed Nightjar

Quechuavis decussata Tschudi’s Nightjar 9a

Setopagis heterura Todd's Nightjar 11, 11a

Setopagis parvula Little Nightjar 11, 12, 12a

Setopagis maculosa Cayenne Nightjar 13

Antiurus maculicaudus Spot-tailed Nightjar 10a

Macropsalis forcipata Long-trained Nightjar 17, 18

Thermochalcis cayennensis White-tailed Nightjar 9b, 9c

Hydropsalis climacocerca Ladder-tailed Nightjar 14b

Hydropsalis torquata Scissor-tailed Nightjar 14c, 15

Systellura longirostris Band-winged Nightjar 9a, 9b,

Eleothreptus candicans White-winged Nightjar 10, 9b

Eleothreptus anomalus Sickle-winged Nightjar 10

Discussion: I’ve gone through the paper in detail and can find no problems of any kind. I recommend that others check the paper for things I’ve missed. An earlier version of this proposal asked for an all-or-none vote on the proposed changes, but comments on that (see below, gray-shaded) indicated that dividing it into parts would be the best course:

A. Recognize the genus Tepuiornis for Setopagis whitelyi

B. Recognize the genus Quechuavis for Systellura decussata

C. Transfer Hydropsalis maculicaudus to the genus Antiurus

D. Transfer Hydropsalis cayennensis to the genus Thermochalcis

E. Transfer Macropsalis forcipata to Hydropsalis

Recommendation: Based in part on comments on the previous version (gray-shaded, below), I recommend a YES to A and B (NO vote is to retain current classification above, with rationale), and a NO vote to C (YES would be to place maculicaudus in Antiurus).

Part D. A YES would be for placing cayennensis in the monotypic genus Thermochalcis. A NO would be for retaining it in Hydropsalis.

Part E: Expanding Hydropsalis to also include Macropsalis forcipata is based on the previous set of comments (gray-shaded); see especially those from Nacho and Vitor Piacentini. Normally, I would make this a separate, fully explicated proposal, but no one is happy with our current classification, which was basically only a temporary fix to maintain monophyly; therefore, I think the simplest thing to do is the deal with this issue here.

Literature cited (see SACC Literature Cited for references in Notes):

COSTA, T. V. V., P. VAN ELS, M. J. BRAUN, B. M. WHITNEY, N. CLEERE, S. SIGURÐSSON, L. F. SILVEIRA. 2023. Systematic revision and generic classification of a clade of New World nightjars (Caprimulgidae), with descriptions of new genera from South America. Avian Systematics 2023 1(6): 55–99.

Van Remsen, June 2024

_______________________________________________________________________________________________________________________________________________

Comments solicited from Mike Braun: “Thanks for reaching out. As we said in Costa et al (2023; p.78)—“ Both treatments have pros and cons, and the decision is not straightforward.” Personally, I would generally lean toward larger genera when insuring monophyly, in no small part because they convey some sense of phylogeny that is otherwise lost in binomens and linear classifications. In this case, the behavioral and vocal similarities seem compelling, especially the similar wing sonations of maculicaudus and climacocerca recorded by Dan Lane.  So, I would be quite happy with a 5 member Hydropsalis, to include cayennensis, forcipata and maculicaudus (along with torquata and climacocerca).”

Comments from Areta: “My votes are:

A. Recognize the genus Tepuiornis for Setopagis whitelyi --- YES

B. Recognize the genus Quechuavis for Systellura decussata --- YES

C. Transfer Hydropsalis maculicaudus to the genus Antiurus --- NO

D. Transfer Hydropsalis cayennensis to the genus Thermochalcis --- NO

E. Transfer Macropsalis forcipata to Hydropsalis --- YES”

Comments solicited from Thiago Costa: “Thanks for the opportunity to contribute to this discussion. First of all, I just need to point out that I feel comfortable with any of the two treatments under consideration (the one recommended in Costa et al. vs. a more conservative course of using an expanded Hydropsalis). Establishing genera is often going to require a fair amount of subjectivity, and the final discussions when preparing our paper were centred on the pros and cons of three different treatments: a 5-species Hydropsalis; a 3-species Hydropsalis and two monotypic genera (Antiurus and Macropsalis); and a 2-species Hydropsalis and three monotypic genera (Antiurus, Macropsalis and Thermochalcis).  To deal with this dilemma, as we stressed in the paper, for the whole clade we sought to establish a rationale grounded on two main points: reestablishing generic monophyly and providing diagnoses for the genera. And from the perspective of diagnosability, in my opinion, the more inclusive treatments make Hydropsalis rather heterogeneous, obviously to a greater extent by including maculicaudus and forcipata. This is why we ended up choosing the third treatment listed above.  Morphologically, the five species present remarkable interspecific variation in plumage, including wing and tail morphology and markings, and we have to push it to find a diagnostic plumage feature for the 5-species Hydropsalis (see the details presented in Table I of Costa et al.);  modified tail feathers is one of the arguments in favour of an expanded Hydropsalis, but again, it is the case for forcipata, climacocerca, and torquata only. Regarding behavior, the 5 species are quite similar in many aspects but distinct in others. They are known to perform aerial displays, with males flying, chasing and/or circling females, with some presenting remarkably simple vocalizations, but maculicaudus and cayennensis, on the other hand, have louder, conspicuous loudsongs. Habitat type, from my view, also doesn't contribute too much to the discussion;  not all species inhabit strict open country habitats (it is the case of torquata, cayennensis, and maculicaudus, but forcipata can be found in forest areas in the Atlantic forest; climacocerca in riverine varzea forests). The wing sonations produced by maculicaudus and climacocerca, as showed by Dan Lane, are quite similar indeed! and would be a point in favour of the 5-species treatment. Again, it is not a straightforward decision, and I agree with Mike that larger genera convey a sense of phylogeny that is lost in less inclusive treatments, but the use of the three monotypic genera + 2 Hydropsalis species, with all the cons that we may consider, instead of an expanded, 5-species genus, clearly provides a more objective level of diagnosability for the genera.”

Comments from Claramunt:

“A. YES. Recognize the genus Tepuiornis for Setopagis whitelyi. Although not very distinctive, the molecular data show strong support for the uniqueness of this lineage in the tree.

“B. A very reluctant YES. The proposed taxonomy overlooks the fact that there is little statistical support for the position of decussata in the tree. Without additional analyses or data, we cannot rule out several alternative positions for decussata. In particular, given its plumage, I wonder if decussata is sister to heterura+parvula and thus can be placed in Setopagis. In any case, the bird has to be removed from Systellura anyway so a placement in the new genus, even if temporary, may be better.

“C. NO to transfer Hydropsalis maculicaudus to the genus Antiurus.

“D. NO to transfer Hydropsalis cayennensis to the genus Thermochalcis. I think it is completely unnecessary to use two monotypic genera for species that fit well in a diverse Hydropsalis. The new classification has become very fragmented with many small genera. Here we have the opportunity to maintain a monophyletic (strongly supported) and cohesive (even with osteological synapomorphies) genus of 5 species; this would be the most specious genus in the group! I understand the need for naming clades because the experts see differences among these birds, but that can be accommodated by naming subgenera or superspecies. An endless proliferation of names and the atomization of the genus category would not make the classification better.

“E. YES to transfer Macropsalis forcipata to Hydropsalis. This bird fits perfectly well in the genus Hydropsalis.”

Comments from Robbins: “I vote YES for A & B., NO to C & D., and YES for E. See my comments on the earlier version of this proposal.

Comments from Stiles: “The nightjar phylogeny and classification by Costa et al. In the first round, I voted NO because what I considered the best option was not mentioned. However, this was included in the second round, where my votes are: A and B: YES; C and D: NO and E: YES.”

Comments from Mario Cohn-Haft (voting for Del-Rio): “Although I have no strongly principled objection to monotypic genera, I do find myself having a harder and harder time remembering the newly applied ones. So, perhaps more out of tired brain syndrome than anything else, I too am inclined to accept a 5-spp. Hydropsalis. Actually, I genuinely like it, for vocal and morphological and behavioral reasons. I believe that translates as a NO to C and D, and a YES to E. That leaves A and B, which, if we are to respect the annoyingly tyrannical requirement of monophyletic taxa, must get a YES vote, like it or not (and i don't particularly like it), or else we'd have to return to the huge and ugly grab-bag version of Hydropsalis.”

Comments from Zimmer:

“A. Recognize the genus Tepuiornis for Setopagis whitelyi. YES

“B. Recognize the genus Quechuavis for Systellura decussata. YES

“C. Transfer Hydropsalis maculicaudus to the genus Antiurus. NO

“D. Transfer Hydropsalis cayennensis to the genus Thermochalcis. NO

“E. Transfer Macropsalis forcipata to Hydropsalis. A reluctant YES. This one pains me a little bit. I’m ultimately persuaded by the collective comments of others, particularly by some of the things that Dan and Vítor have brought up, although I would note that forcipate doesn’t fit quite so neatly in with the expanded Hydropsalis as some of the others do. Sure, it has the tail modifications that make it a nice fit with torquata and climacocerca, but I don’t see it as primarily an “open country” bird like the others, but as more of a forest-edge bird – it regularly forages along dirt roads through Atlantic Forest, and, in grassy, wet glades within forest or at the forest-edge, and it regularly displays above the forest canopy. Its vocalizations would place it closer to the aforementioned species, but not so much with either maculicaudus or cayennensis (which do remind me of one another). Males, with their long, modified tails, do regularly perch on the ground, with the tail held on the ground, in line with the axis of the body, as in H. torquata, but also perch atop stumps, posts or snags, with the tail hanging downward, as is typical of Uropsalis. Anyway, it’s something of an anomaly to me, but ultimately, I am persuaded that the expanded Hydropsalis is the way to go.”

Comments on the earlier version of this proposal, which was an all-or-none vote on adopting the classification of Costa et al.:

Comments from Areta: “NO. This is a most welcome paper for someone like me who has been battling against massive lumping of genera in nightjars. I agree with most of the recommendations, but from my perspective, the behavioral and vocal similarities argue for a 4-species Hydropsalis, including forcipata, cayennensis, torquata, and climacocerca (all tend to have modified tail feathers, some longer than others, but all give high-pitched vocalizations and make similar displays during chases and also likely while displaying and copulating, the latter at least in torquata and forcipata), while maculicaudus seems to deserve its own genus based on its unique display and different vocal repertoire, and this is how I vote. Thus, I don´t think that recognizing Thermochalcis or Macropsalis is really well justified, and instead they expand a very neatly coherent Hydropsalis.”

Comments from Robbins: “NO. This is a very comprehensive, well researched paper and I agree with most of their treatment, but like Nacho, I too would have included forcipata, cayennensis, torquata and climacocerca in Hydropsalis not only based on overall morphology but also the molecular data. It could also be argued that anthonyi, albicollis, nigrescens and hirundinaceus all be placed in Nyctidromus. Finally, interesting what they suggest might be the case with maculosa, that it might be an aberrant Nyctidromus albicollis…..no wonder we couldn’t find it during our work in Guyana!”

Comments from Jaramillo: “YES – I am actually fine with the more restricted genera, rather than a more inclusive Hydropsalis as Nacho suggests. It is subjective, these are genera, and I am at this point OK with the suggestions made by the authors.”

Comments from Stiles: “NO. Here, I tend to agree with Nacho and Mark – adding two species to Hydropsalis produces a very coherent genus group, agrees as well with the genetic data and makes two monospecific genera unnecessary.”

Comments from Lane: “YES to the revised order within the family. YES to the adoption of Tepuiornis and Quechuavis. NO to the generic names Antiurus and Thermochalcis. Unlike Nacho, I am actually inclined to retain these latter two species within Hydropsalis thanks to similar vocalizations (both between the two species involved and over the larger Hydropsalis group), open country habitats of all members of the group, and, most importantly, because the wing sonations that H. climacocerca and [H.] maculicauda produce in flight displays are nearly identical (compare: min 1:35 of https://macaulaylibrary.org/asset/228688 with 2:18 of https://macaulaylibrary.org/asset/371384901). By the way, ever since seeing the photos of the holotype of “Caprimulgus maculosus” in Cleere (2010) I have been convinced that it is simply a juvenile (so, not fully-grown) Nyctidromus albicollis, so I am glad that Costa et al. seem to have come to the same conclusion.”

Additional comments from Areta: “I am glad that my expanded Hydropsalis is meeting some general agreement, and I am fine with expanding it to include maculicauda on account of its high-pitched and simple song and the presence of mechanical sounds as in the other Hydropsalis that I advocated. I just was not aware of this sound in maculicauda, which is great to see and hear. So YES to a 5-species Hydropsalis, Quechuavis, and Tepuiornis.”

Additional comment from Piacentini: “Regarding the Caprimulgidae, perhaps it is worth mentioning that Thiago himself found osteological synapomorphies that can be taken as diagnosis for the expanded Hydropsalis (which group ecologically makes all sense to me):

“Character 25: Maxilla; em aspecto lateral, abaulamento dorsoventral da sua região caudal (IC = 1) [ausente/presente]

“Character 86: Coracoide; processo lateral (Processus lateralis) (IC =  0.5) [largo/estreito]

“Comments about char. 86: "A largura do processo lateral do coracóide apresenta-se estreito na maioria das espécies, enquanto que no gênero Hydropsalis é mais largo"

“Details (including photos illustrating the states of each character) can be seen in his thesis, available at:

https://teses.usp.br/teses/disponiveis/41/41133/tde-09032015-080107/publico/Thiago_Costa.pdf.”