Proposal (1012) to South American Classification Committee

Treat Sphenopsis ochracea and Sphenopsis piurae as separate species from S. melanotis

Note: This is a high-priority issue for WGAC.

Background: Our SACC note on this is as follows:

10. Despite strong differences in some plumage characters, Sphenopsis (formerly Hemispingus) melanotis has been treated a polytypic species based on shared plumage themes (Hellmayr 1936, Zimmer 1947, Storer 1970a, Meyer de Schauensee 1970, Dickinson & Christidis 2014. and others). García-Moreno et al. (2001) and García-Moreno & Fjeldså (2003) found that the distinctive taxon piurae, currently treated as a subspecies of Sphenopsis melanotis (e.g., Meyer de Schauensee 1970), is more distant from the latter than is S. frontalis, and that piurae is basal to frontalis + melanotis; these analyses, however, were based on only ca. 300 base-pairs of mtDNA. Ridgely & Greenfield (2001) treated piurae as a separate species from H. melanotis based on plumage and vocal differences. SACC proposal to recognize piurae as a species did not pass. Hilty (2011) also treated piurae as a separate species. Ridgely & Greenfield (2001) and Hilty (2011) further recognized the subspecies ochracea as a separate species based on plumage differences. Halley (2022) treated piurae and ochracea as species based on distinctiveness of plumage. SACC proposal needed.

We currently treat Sphenopsis (ex-Hemispingus) melanotis (Black-eared Hemispingus) as a highly polytypic species, as did Dickinson & Christidis (2014), who recognized 6 subspecies in 3 groups in the humid Andes:

(1) nominate melanotis (Andes of NW Venezuela to e. Ecuador), berlepschi (e. Peru), and castaneicollis (Eastern Andes of S. Peru and Bolivia)

(2) ochracea ( Western Andes of sw. Colombia and nw. Ecuador)

(3) piurae (s. Ecuador, NW Peru) and macrophrys (Western Andes of c. Peru).

We rejected a proposal in 2007 (SACC 284) based largely on weak genetic data and absence of data on vocalizations.

Here is a crude home-made representation of the distribution of the taxa in question. All photos are from Macaulay Library (melanotis by Ben Jesup, ochracea by Angel Argüello Méndez, piurae and macrophrys by Fernando Angulo, berlepschi by José Martín, castaneicollis by Tini & Jacob Wijpkema). Note the subspecies macrophrys Koepcke 1961 has been generally overlooked (“Near piurae but with broader white superciliaries, a broader and more conspicuous gray band on nape and post-auriculars, and underside of wing more whitish”). See also Halley (2022) for specimen photos.

From the differences in plumage, one can immediately see the problems. It is not immediately obvious why ochracea was ever considered conspecific, although the rationale was likely that it looks like a typical smudgy, obscure Chocó representative of a group of more brightly colored taxa. It also could be argued that nominate melanotis stands apart from all the rest. From the perspective of those most familiar with birds north of the Marañon, such as Ridgely and Hilty (see SACC Note), one might immediately reject piurae as being conspecific with nearby nominate melanotis because of the plumage differences, but for those more familiar with taxa south of the Marañon, such as J. Zimmer and Schulenberg et al. (“Birds of Peru”), piurae looks similar to distant S. m. castaneicollis. I suspect that a formal analysis of plumage similarities could go any number of ways depending on how one scores the plumage characters, e.g. see Halley (2022: 221). Lots of “eye of beholder” reasoning would be involved. It seems we have a classic conundrum concerning taxon rank of distinctive allotaxa, with the only clear case being macrophrys as a subspecies regardless of species limits.

New information: Del Hoyo & Collar (2014) treated ochracea and piurae as a separate species based on the Tobias et al. point scheme as follows (provided by Pam Rasmussen):

“Western: https://birdsoftheworld.org/bow/historic/hbw/bkehem3/1.0/introduction

[ochracea] Usually treated as conspecific with S. melanotis and S. piurae, but each differs markedly in plumage and habitat and hence is tentatively accorded species rank here (although songs of present species and melanotis difficult to distinguish (1) ); review of group warranted. Monotypic.

“Piura: https://birdsoftheworld.org/bow/historic/hbw/bkehem1/1.0/introduction

Usually treated as conspecific with S. melanotis and S. ochracea (see latter); present species exhibits some vocal differences, and limited molecular data suggest that it is at least as distinct from S. melanotis as latter is from S. frontalis; review of group warranted. Two subspecies recognized.”

The habitat difference alluded to likely comes from Ridgely & Greenfield’s statement: “unlike. [melanotis and piurae], the Western Hemispingus does not show any particular predilection for an understory of Chusquea bamboo.” However, photos of ochracea in Macaulay show bamboo in 11 of the small number of photos of the species, so perhaps that needs re-evaluation. Certainly it should not play a role in determining species limits.

Boesman (2016k) analyzed an unspecified number of recordings from unspecified locations. Apparently, vocalizations of berlepschi and hanieli were not analyzed. Here are his main conclusions:

“From the above, it is clear that duet song of all three species is structurally similar. There seems to be however a closer resemblance between duets of H. ochraceus and H. melanotis.”

“All in all, we can conclude that voice of H. ochraceus is about identical to H. melanotis.”

“Voice of piurae at the other hand is quite distinctive and can be safely told apart.”

“Race castaneicollis of H. melanotis also differs from other races of this species”

Price-Waldman (2019), an unpublished dissertation from Kevin Burns’ lab, used UCE data to construct a phylogeny of the tanagers. As cited by Halley (2022), the topology for this group was (((ochracea, piurae), frontalis), melanotis), which would thus require separation, minimally, of ochracea + piurae as a separate species from melanotis.

Expanding frontalis to include all of the melanotis group is clearly untenable because frontalis is broadly sympatric with melanotis representatives. Tangentially, note that the subspecies of frontalis of the Coastal Range of Venezuela, S. m. hanieli, superficially looks as much like a melanotis type as a representative of frontalis, and Halley (2022) emphasized the distinctiveness of this taxon and its tentative placement within S. frontalis. Here’s a photo from Macaulay by Margaret Wieser from Aragua:

So, species limits need to be changed, but what do we do with berlepschi, castaneicollis, and even hanieli? Certainly by extrapolation, the differences in plumage between nominate melanotis and berlepschi + castaneicollis are of the same general magnitude as those between nominate melanotis and ochracea, again depending on how one weights characters and determines homoplasy.

Halley (2022) made two important contributions. First, he pointed out that the Garcia-Moreno et al. data-set actually lacked a sample of ochracea despite its claims otherwise, and second, he assembled the few existing specimens of ochracea to show that it had been inaccurately illustrated in all published works, with respect to coloration of the underparts, which are indeed ochraceous, and in some cases with respect to having a mask, which is minimal. Halley did some solid, baseline alpha-taxonomy that will be useful for all future analyses.

Discussion and Recommendation: Just to understand the history of all this, I have recorded here my wanderings down several rabbit holes (to use a worn-out metaphor). I was tempted just to delete everything and just present the Price-Waldman tree, which requires at least a 2-way split. But that makes me uncomfortable. The thesis is unpublished. Given the surprising finding, were vouchers double-checked? And what about unsampled berlepschi-castaneicollis, for which Boesman has evidence for vocal distinctiveness (at least for castaneicollis) and which, as a group, are phenotypically distinctive? Do we keep them with melanotis despite plumage and vocal differences? And what about the largely neglected hanieli issue? My gut instinct is not to meddle with our current classification, despite the possibility of a paraphyletic melanotis, until all these other issues are sorted out. In contrast to the philosophy of many colleagues, I actually like conflict among world classifications because it signals honestly and appropriately that there is considerable uncertainty, which would otherwise be masked. Nevertheless, I don’t have a strong recommendation and will wait to see what others say.

English names: “Western Hemispingus” and “Piura Hemispingus” have more than two decades of traction in the literature. I suspect “Western” is not a name anyone really likes. Hellmayr (1936) called it “Ochraceous-bellied Hemispingus”. The reason why Ridgely & Greenfield (2001) didn’t go with this is because they described the underparts as “drab buffy olivaceous”, although Greenfield’s illustration of the head shows the throat at least to be ochraceus. The ochraceous belly (vs. all of the underparts) highlights what seems to be a unique feature (and it also helps remember the scientific name ochracea). I predict if Bob Ridgely knew what we do now about ochracea, he would gave stuck with Ochraceous-bellied. I am willing to write a short proposal on this if anyone else is favors overturning 20+ years of stability. If it were 50+ years of stability or if all classifications treated it as a sperate species “Western Hemispingus”, then I would object to changing the name, but if we are ever going to do it, now is the time (if the proposal passes). “Ochraceous-bellied Hemispingus” would have the additional advantage of pointing out the problems with illustrations of ochracea, as noted by Halley (2022). In contrast, “Piura Hemispingus” was used by Hellmayr (1936) and thus has been in the literature for at least ca. 88 years.

This dodges the problem that none of these species are in the genus Hemispingus any longer, but that’s a separate problem that would have to consider the broader diaspora of former Hemispingus. For now, I’m personally content with “Hemispingus” as a vague morphotype. i.e. roughly as a “half finch”, as in the derivation of the genus name, in the interests of stability, but if anyone wants to tackle this problem, feel free.

References: (see SACC Bibliography for standard references)

HALLEY, M. R. 2022. Taxonomic status of the Western Hemispingus Sphenopsis ochracea (Thraupidae) and a review of species limits in the genus Sphenopsis P. L. Sclater, 1861. Bulletin British Ornithologists’ Club 142: 209-223.

ZIMMER, J. T. 1947. Studies of Peruvian birds, No. 52. The genera Sericossypha, Chlorospingus, Cnemoscopus, Hemispingus, Conothraupis, Chlorornis, Lamprospiza, Cissopis, and Schistochlamys. American Museum Novitates 1367: 1-26.

Price-Waldman, R. M. 2019. Phylogenomics, trait evolution, and diversification of the tanagers (Aves:Thraupidae). M.Sc. thesis. San Diego State Univ.

Van Remsen, June 2024

Comments from Robbins: “NO. Another messy situation for all the reasons that are pointed out in the proposal. Given the number of issues that need to be addressed, as summarized by Van, for now I vote NO.”

Comments from Jaramillo: “YES – You only need to look at the map, compare to other similar Andean chain “superspecies,” look at the photos and you have the informed suspicion that more than one species is involved in melanotis. The unpublished (thus far) DNA work, the vocal information that is available adds more pieces to the puzzle, and they also lean in the way of separating this chain of taxa into multiple species. I am comfortable in voting yes, and comfortable in understanding that in the future there may be future changes to this complex.”

Comments from Areta: “A conflicted NO. I think that there are at least 3 species in S. melanotis (and probably more), but I feel quite uncomfortable with the lack of data on berlepschi and melanotis (taxa sampled by Price-Waldman 2019: Sphenopsis melanotis castaneicollis FMNH 430079 / Sphenopsis melanotis ochracea ANSP 149722 / Sphenopsis melanotis piurae FMNH 480966) and I don´t like to guess. berlepschi is clearly more similar to melanotis, so it might not be much of a problem (perhaps...), and I could accept that it is conspecific with melanotis. But castaneicollis, with its broad black mask and chin, contrasting white supercilium and rufous-chestnut large pectoral band is much more like the widely allopatric piurae, while also being quite distinct from melanotis/berlepschi. Ochracea is ridiculously different from the rest, yet based on Price-Waldman it is sister (how deeply diverged? we don´t know) to the distinctive piurae, both from the W slope of the Andes. Therefore, the drabbest and the most colourful forms are sister, suggesting that concluding on the phylogenetic placement of other taxa based on plumage is risky. Should we go with geography then? Risky again. Then there are also Van´s worries about macrophrys and S. frontalis.

“This is one of the situations in which there is (barely) enough information to split some of the taxa, but in which information on key taxa is missing to properly establish species limits and in which we must guess in order to decide. Shall we stick to the old taxonomy or move to a newer one that may have other problems? Also, the Price-Waldman phylogeny has not been adequately published, and all we seem to have at hand is the topology of a tree. Halley (2022) provides an excellent overview of the taxa involved and clarifies some problems while proposing an alternative taxonomy. Although it seems clear that the single-species treatment will fall, I am not convinced about how many species we should recognize and how each of them should be confirmed. Therefore, I vote NO to any split for the time being.”

Comments from Stiles: “NO. The whole situation is simply at the stage where too little is known about the genetics and vocalizations of too many taxa, and their geographic distributions appear to present a more complicated picture. A more comprehensive study of Sphenopsis is badly needed - splitting these two now might be jumping the gun.”

Comments from Lane: “NO. Another situation where I think we are voting without all the necessary information in hand to make an informed decision. What harm is there in waiting for someone to compile those data?”

Comments from Niels Krabbe (voting for Remsen): “NO. There is no indication that Price-Waldman (2019) double-checked the identity of his frontalis (MSB 31856 (Peru: dpto. Amazonas; prov. Utcubamba; dist. Lonya Grande; ca. 4.5 km N Tullanya). Halley (2022) certainly did not. However, as both Price-Waldman and García-Moreno & Fjeldså (2012), the latter using a double-checked frontalis from Imbabura, W Ecuador, found frontalis to be embedded in the melanotis complex, there is no reason to doubt the correct identification of the Amazonas specimen.

“A recent, as yet unpublished find by Jonas Nilsson (pers. comm.) of a population in the Chilla Mts of southern Ecuador (a ridge connecting the east slope with the west slope) that appears intermediate in plumage between the vocally similar ochracea and melanotis further underscores the need for a detailed study that includes all the Sphenopsis forms before changing the current classification.”

Comments from Claramunt: “YES. The only problem here is our regressive tradition in ornithology of stuffing taxa into polytypic species based on pure speculations about degree of differentiation. The actual information we have is: 1) subspecies now included in melanotis are not each other’s close relatives because of the position of S. frontalis (Garcia-Moreno et al, Price-Waldman). 2) Examination of study skins point to the existence of three well delineated taxa (100% diagnosable, Halley 2022): melanotis, ochracea, and piurae. The solution is simple: split the polytypic entity into its fundamental units. And the result is totally coherent from a biogeographical point of view. I fail to see problems with the proposed taxonomy. In contrast, I don’t see any evidence supporting the traditional treatment. Where is the evidence that ochracea or piurae are reproductively compatible with melanotis?”