Proposal (1003x) to South American Classification Committee

Species limits in the Myioborus melanocephalus complex, revisited.

As originally presented, this proposal (see below) did not pass, but it was suggested that I present a revised version with more details on specific parts of the proposal, progressing from north to south.

A. Continue to recognize M. albifrons as a species distinct from the rest of the taxa related to the hybrid zone. The alternative would be effectively lumping all of the complex into a single species ornatus, which has priority (a possible option suggested by two members of SACC). However, no evidence of hybrids between albifrons and ornatus exists; moreover, recent eBird records of albifrons from the Tamá area on the border between Colombia and Venezuela effectively establish parapatry between albifrons and ornatus which would render untenable lumping them. Hence, I recommend NO for this option.

B. Recognize M. chrysops as a distinct species from M. ornatus. The original proposal failed to pass principally because of doubts regarding this split. However, it was handicapped by the rather poor illustration of chrysops and the lack of a good illustration of M. o. ornatus. Here are two photos that I think represent these taxa more clearly:

M. o. ornatus M. chrysops

Both taxa show a generally similar pattern of brightly colored faces that present strong contrasts with the dark irides, and with black napes, set off from dark gray backs and wings. In both, note a fine white line in the black of the side of the neck, approximately coinciding with the lower posterior border of the auriculars. These patterns occur throughout the respective distributions of each taxon. The principal difference between them is in coloration of the face: glossy white in ornatus vs. glossy orange-yellow in chrysops. The color of the underparts of ornatus is bright yellow, as are the forehead and crown; chrysops is a more orange-yellow below and has a notably different distribution of colors on the head, with the orange forehead much more prominent and extending back over most of the orange-yellow crown. The current distributions of the two do not overlap: chrysops occurs widely in the Central Andes, wherever the elevation extends well above the 2000 m ridgeline – essentially an archipelago of high Andean islands – and more locally in the Western Andes, where such islands are much fewer and more isolated. Note that the isolated population of chrysops where the Eastern Andes unite with the Central Andes is isolated by ca. 100 km from the southernmost limit of ornatus and is ca. 50 km east of the hybrid zone between chrysops and the melanocephalus group. With the current global warming, the geographic separation of chrysops and ornatus can only increase, s.s.as the white postauricular line) do not represent current or recent hybridization but more likely, the retention from a considerably earlier common ancestor. To summarize:

The hybrid zone in the melanocephalus complex is strictly with chrysops: ornatus is not directly involved. It thus more clearly expresses the established precedent that the hybridization is occurring between two full species. The distributions of ornatus s.s. and chrysops are separated completely by at least 100 km of unsuitable habitat and no hybrids between them are known. Their plumages differ strikingly in the head region, of a magnitude similar to those distinguishing many other species of Myioborus; plumage similarities between them are not evidence of current or recent hybridization but retention of more ancestral characters. Perhaps pertinent here is that the phenotypic differences in color and pattern between ornatus s.s. and chrysops are if anything greater than the differences between ornatus s.s. and albifrons, which appear to be parapatric species.

Points in favor of a NO: The shallow divergence in mitochondrial genetics between both species, although this could be invoked for the complex generally, including M. albiceps, which is a close sister to the remaining members, but which is apparently parapatric with M. o. ornatus. The lack of direct evidence that the latter could interbreed with chrysops were they to enter into contact is a piece of the puzzle missing, given their disjunct distributions of some antiquity and is thus impossible to resolve. A NO vote here would imply that the entire complex represents a single species and consequently, the hybridization event occurs between subspecies of a phenotypically extraordinarily diverse species.

My personal opinion here favors a YES.

C. Suppress the name ruficoronatus due to the hybrid nature of its type, thus rendering this name inapplicable to any described taxon. This should be an undisputed YES given its clear genetic identity and distribution.

D. Select bairdi as the second parental species involved in the hybridization, reflecting its adjacent distribution, its stable phenotype through central and southwestern Ecuador and the compatibility of its phenotype for integration with the hybrid zone. This should be an easy YES: no other taxon unites all of its qualifications.

E. Recognize griseonuchus as a separate species from bairdi. At its southern limit around the Ecuador-Peru boundary, bairdi meets griseonuchus, a poorly known taxon that is phenotypically most similar to bairdi but the two are reciprocally diagnosable. The current treatment appears to favor treatment of griseonuchus as a subspecies of bairdi, but apparently there is no evidence of hybridization between them, but only further collections and observations can fully resolve its status as a subspecies or separate species.

I lean toward a YES here, but tentatively, pending more data from the potential zone of contact between them.

F. Split the remaining southern members of the complex as a separate species, M. melanocephalus. This reflects the fact that the three included species have black crowns, and zoogeographically, all occur east of the río Marañón valley whereas bairdi and griseonuchus occur to the west; this low, relatively dry valley is well recognized as a major barrier for the distributions of taxa of the wet highland forests of opposite sides of the valley. This decision should logically be YES.

G. Continue to recognize the currently known subspecies within M. melanocephalus. Its three subspecies are distributed sequentially from northwest to southeast, collectively ranging from northern Peru south to central Bolivia: malaris, melanocephalus and bolivianus. Cuervo and Céspedes presented brief descriptions, but emphasize that their respective distributional limits are poorly documented, and the genetic characterizations are relatively incomplete; clearly this group merits further study, but for the present they considered it best to continue recognizing all three as described. I recommend a YES here.

Note from Remsen: At Dan Lane’s suggestion, here are the published trees for Myioborus in case they are helpful. Both are based on mtDNA sequence data:

Lovette et al. (2010; MPE):

Pérez-Emán (2005; MPE):

Gary Stiles, September 2024

Comments from Zimmer: “

A. I’m a little confused by how this one is stated in the Proposal. It begins, by saying “Continue to recognize M. albifrons as a species distinct from the rest of the taxa related to the hybrid zone.” I would vote YES to that for the reasons stated by Gary in the sentences that follow (e.g. the alternative would be lumping all of the complex into a single species ornatus… but there is no evidence of hybrids between albifrons and ornatus…and eBird records effectively establish parapatry between albifrons and ornatus, which “would render untenable lumping them”), but then, he says “Hence, I recommend NO for this option.” It seems to me that Gary is recommending NO to the option of lumping all taxa in the complex into a single species, ornatus, to which I would also vote NO, but the initial question was whether to continue to recognize albifrons as a distinct species, to which I would reiterate my YES vote. Or, am I missing something here?

B. Recognize M. chrysops as a distinct species from M. ornatus. I’m persuaded by the clarification presented in the revised proposal, so a YES for me on this one.

C. Suppress the name ruficoronatus due to the hybrid nature of its type. This would seem to demand an obvious YES.

D. Select bairdi as the second parental species involved in the hybridization. Another obvious YES.

E. Recognize griseonuchus as a separate species from bairdi. I’m uncommitted on this one. I think I would need more information to pull the trigger on splitting these, so I lean toward a NO for now vote. The same logic that calls for treating the three black-crowned taxa that occur east of the rio Marañón valley as a single polytypic species (see Part F), would seem to argue for treating rufous-crowned bairdi and griseonuchus from west of the Marañón as a single polytypic species, rather than as two distinct species that are similar but diagnosable.

F. Split the remaining southern members of the complex as a separate species, M. melanocephalus. YES.

G. Continue to recognize the currently known subspecies within M. melanocephalus. YES.

“What I don’t see here, is a sub-part of the Proposal that explicitly addresses the question of bairdi versus chrysops, given the broad hybrid zone, with these two as the sole recognized parental types. If we are treating them as a single species, and, if griseonuchus is treated as a subspecies of bairdi, then, by my count, my votes would support recognition of 4 species: albifrons, ornatus, chrysops (including bairdi and griseinuchus), and melanocephalus (including 3 sspp), which, I believe, is what both Mark and Dan suggested as an alternative in the original Proposal. That would be my leaning too, for 4 species.”

__________________________________________________________________________

Proposal (1003) to South American Classification Committee

Species limits the Myioborus melanocephalus complex: Taxonomic options for resolving the classification of the species forming an extensive hybrid zone between southern Colombia and northern Ecuador, and related taxa

Antecedentes: Céspedes-Arias et al. (2021) described in detail an extensive hybrid zone between a northern taxon, Myioborus ornatus chrysops, and a southern taxon (M. melanocephalus ruficoronatus (the northernmost subspecies of M. melanocephalus). This hybrid zone is exceptionally long, covering ca. 200 km between southwestern Colombia and northwestern Ecuador and is apparently stable (no “pure” parental phenotypes occurring within the central hybrid zone). They found that the mitochondrial genetic differences along this zone were extremely small but the phenotypic differences, particularly in plumages of the head, were visually striking. Treating the entire melanocephalus complex as a cline, apparent gene flow along the occurs with only a distinct a break at the dry Marañon valley of Peru (a well-known barrier separating highland wet-forest taxa on either side).

The taxonomy underlying the aforementioned study was effectively relegated to the dustbin by Cuervo & Céspedes-Arias (2023) when they examined the photographs, original illustrations and description of the type specimen of ruficoronatus and found that its plumage was identical to those of hybrid specimens from near the midpoint of the hybrid zone, with its probable provenance from the area of Pasto, Nariño, Colombia. Here is he plate from Cuervo and Céspedes-Arias:

This effectively excludes ruficoronatus as the name of any valid taxon, and prompted a more far-reaching examination of other taxa, including their type descriptions, photographs of same, and distributions, initially to find the “pure” species-level taxon that would represent the southern parent of the hybrid zone. For brevity here, I omit details of type descriptions and photographs, for which see Cuervo & Céspedes-Arias (2023). The appropriate southern parental taxon found was bairdi once the confusion regarding its distribution and erroneous synonymization under ruficoronatus had been cleared up. This taxon is found through central and southern Ecuador. They then proceeded to examine the distributions, type specimens and distributions of the other taxa of the clade that includes the hybrid zone, and including M. albifrons of Venezuela, the sister and near outlier to this clade. North of the hybrid zone, three taxa occur: M. o. ornatus of the Eastern Andes and M. o. chrysops of the Central Andes of Colombia, as well as albifrons. Proceeding south on the eastern slope of the Andes, the taxa include griseonucha (currently considered a subspecies of bairdi), malaris, melanocephalus and bolivianus (the last three considered subspecies of melanocephalus). This chain of taxa is bisected by the dry Marañón valley of eastern Peru. Of the aforementioned taxa, bairdi and griseonucha occur to the north and west of this barrier, with the three taxa of melanocephalus to the south and east.

Here is the distribution map from Cuervo & Céspedes-Arias (2023):

The authors found that genetic variation over the entire complex permitted recognition of four genetic clusters, largely correlated with geography and partly with phenotypic patterns: 1) M. albifrons, isolated from the remining taxa in the Venezuelan Andes; 2) M. melanocephalus (including malaris and bolivianus), east and south of the Marañón; 3) M. bairdi and griseonucha, north and west of the Marañón, and 4) M. o. ornatus and M. o. chrysops, north of the hybrid zone. Hybrid genotypes and phenotypes clustered with groups 3 and 4. They then proposed three taxonomic hypotheses for the species represented among these taxa:

A. Three species: M. albifrons, M. ornatus (all of the taxa north of the Marañon and including the hybrid zone) and M. melanocephalus, south of the Marañón. Thus, ornatus includes a variable mixture of yellow- and white-faced and partly to entirely black to rufous-crowned taxa; albifrons has a white eyering and forehead and black crown, and melanocephalus includes all black-crowned taxa, and is excluded from participation in the hybrid zone by the intervening, rufous-crowned bairdi and griseonuchus. The species status of albifrons is not affected; in fact, its distribution approaches rather closely to nominate ornatus from the north, but no hybrids between the two are known.

B. Five species: M. albifrons, ornatus, chrysops, bairdi and melanocephalus. This option differs in splitting chrysops and bairdi from ornatus. This is advantageous in recognizing the hybrid zone as strictly between chrysops and bairdi. Both melanocephalus s.s. and ornatus s.s. are isolated from this zone: the former by bairdi (and griseonuchus) on the opposite side of the Marañón valley, and ornatus from chrysops by the wide Magdalena River valley. No phenotypic evidence of recent hybridization between the latter two exists, and their similarity in mitochondrial genetics possibly best indicates incomplete lineage sorting during their separation. Moreover, the temperature increase due to ongoing climate change further reduces probability of future genetic exchange between ornatus and chrysops, and some checklists have in fact accorded separate species status to each. Cuervo & Céspedes-Arias (2023) favored this option.

C. Six species, by the additional separation of griseonuchus from bairdi reflecting their phenotypic diagnosability and the apparent absence of hybrid phenotypes between them. The problem here is that very few specimens are available from the area of possible contact across the border between Ecuador and Peru; further collecting and field observations are needed to evaluate this possible split.

For the SACC, I recommend voting YES for one of the above three options: if any one receives a majority of votes, it passes; if no option passes, the voting could be repeated after eliminating the option least voted.

A. Three species

B. Five species

C. Six species

Literature cited:

Céspedes-Arias, L. et al. 2021. Extensive hybridization between two Andean warbler species with shallow genetic divergence. Ornithology 138:1-28.

Cuervo, A. M & L. Céspedes-Arias. 2023. The type of Setophaga ruficoronata is a hybrid: implications for the taxonomy of Myioborus warblers. Zootaxa 5383(4): 476-490.

Gary Stiles, June 2024

_______________________________________________________________________________________________________________________________________________

Comments from Del-Rio: “YES for B, although I believe whole genome + habitat studies would benefit the understanding of the strength of potential barriers to reproduction between chrysops and bairdi. In the meantime the five taxa should be treated as species.”

Comments from Areta: “YES for A. The species-limits in this complex are not easy to sort out given the lack of samples from some key areas and the seeming ability of taxa to interbreed wherever they meet. What a fascinating system! Genetic divergences are shallow, and support for many of the mtDNA relationships is poor, although they make a lot of sense in geographic and plumage terms. Pérez-Emán (2005) uncovered a close relationship between "ruficoronatus", chrysops and ornatus. Vocalizations are structurally very similar across the geographic range of the ornatus-melanocephalus complex, and this applies to albifrons too: all taxa give a rapid series of tonal notes that includes some shorter and some longer whistles ascending or descending in frequency that rise and fall in pitch and speed up and increase in amplitude as the song progresses. All can also duet.

Given that the quite different-looking bairdi and chrysops hybridize freely and massively across 200 km, it seems that there are no barriers to interbreeding and that these two can be considered as part of the same species. The white-faced ornatus is very similar to chrysops, and if one can use the bairdi-chrysops case as a predictor of what would happen if chrysops meets ornatus, the answer would seem to be massive interbreeding. Do the eastern population of chrysops meet ornatus somewhere? It seems, based on the examination of a few photographs, that ornatus at Sumapaz at the southern end of the range have considerably less white on the face-sides and on the upper-throat (below the bill) [e.g., https://macaulaylibrary.org/asset/151865601; https://macaulaylibrary.org/asset/192970091] than birds in the northern end [e.g., https://macaulaylibrary.org/asset/205397261]. This suggests that the unsampled area between ornatus and eastern chrysops may well show a broad area of phenotypic transition.

“The situation between bairdi and griseonuchus seems relatively straightforward: the plumage variation reported by Cuervo & Céspedes-Arias (2023), and the genetic data are consistent with considering them as part of the same species. Therefore, in my view, also conspecific with chrysops and ornatus.

“The most complicated part to me is the trans-Marañón divergence between griseonuchus and malaris. Céspedes-Arias et al. (2021: 12) put it this way: "Genetic structure was shallow, and only clearly associated with 1 topographic discontinuity of the cloud forest belt. We found evidence of mtDNA differentiation across the Marañón River Valley, a dry area that dissects the cloud forest distribution of M. melanocephalus and coincides with the transition between rufous-crowned (M. m. griseonuchus) and black-crowned (M. m. malaris) forms (Zimmer 1949, Curson et al. 1994). This arid valley is important for differentiation in many other cloud forest birds (Bates and Zink 1994, Chaves et al. 2011, Gutiérrez- Pinto et al. 2012, Winger and Bates 2015), and likely acts as a strong barrier for Myioborus taxa restricted to humid high-elevation forests and scrub (Curson et al. 1994). The exception are 3 specimens from south of the Marañón (Amazonas, Peru), corresponding to the black-crowned subspecies M. m. malaris, which clustered with individuals of M. m. griseonuchus and M. m. ruficoronatus occurring north of the Marañón Valley. This pattern might reflect trans-Marañón introgression (Winger 2017) or incomplete lineage sorting (Maddison and Knowles 2006)." Thus, 3 out of 8 samples of the black-crowned malaris from immediately south of the Marañón valley clustered with the rufous-crowned griseonuchus, while 5 out of 8 samples of malaris from the same area clustered with the black-crowned melanocephalus-bolivianus (see Figures 1A and 2A in Céspedes-Arias et al. 2021). This is only mtDNA data, but the fact that there is trans-Marañón mtDNA sharing does not instill confidence on the degree of isolation between the similar (yet differently crowned) griseonuchus and malaris: if there is leakage across this barrier, these two sets of taxa would very likely interbreed freely if given a chance to do so. Here is when having genomic data would provide us with much needed insights to inform our taxonomic decisions.

“I think that the most conservative course of action in the light of the great works by C-A et al. 2021 and C & C-A 2023 is to vote for their "three-species alternative" (Option A). Maybe a genomic dataset can make my understanding change and result in more splits, but at present, I think that even the three-species treatment might be on the generous side of splitting.”

Comments from Robbins: “Of the options given, I vote for A. Given the data at hand, I believe Nacho’s interpretation of how best to treat these taxa is the best option. Another option that hasn’t been given is a four species treatment, i.e., albifrons, ornatus, chrysops, and melanocephalus. By doing the latter, one does not have to speculate on what might happen if ornatus and eastern chrysops should be found in contact. Nonetheless, I find Nacho’s observation that there may be a cline in white in ornatus or may even indicate hybridization with eastern chrysops to be plausible. So, if four species treatment isn’t an option, I vote for option A. Clearly, more study is needed on what is going on in the Marañón region with regard to griseonuchus and malaris.”

Comments from Jaramillo: “YES on B – I can see both the 3 and 5 species solutions as valid given what we know now. When the genomic analysis happens, if ever, we will know more. But to me the 5 species solution seems cleaner to me actually. Although super on the fence here.

Comments from Stiles: “B – 5 species. Here, I comment on Nacho’s conclusion that M. ornatus s.s. and M. chrysops would show massive interbreeding should they come into contact, because this has occurred between chrysops and bairdi, given the shallow genetic divergence in both cases. However, the use of the latter as a yardstick for predicting the former ignores the complex Andean topography. The long hybrid zone between chrysops and bairdi occurs along a continuous range of upper temperate-zone forest habitat, north of which chrysops occurs alone along the Central Andes at similar elevations for several hundred km more. The distribution of ornatus s.s. lies between the northern end of the Cordillera Oriental south to Sumapaz massif. South of this massif, this cordillera narrows abruptly to a long ridge separated from the Cordillera Central by the upper Magdalena Valley with tropical elevations of ca. 350-400m as far south as the latitude of San Agustín. Southward, this cordillera connects with the Cordillera Central via an eastward extension of temperate-zone habitat in southeastern Cauca – western Putumayo. The important point here is that the long ridge south of the Sumapaz massif lowers to subtropical elevations over much of its length, where the only Myioborus present is miniatus. At the temperate-zone connection to the south, only chrysops is found along the highest ridge of this zone – which adjoins the hybrid zone with bairdi. To the north, the two main cordilleras are separated by the lower and middle Magdalena Valley which here supports a zone of hot, dry to moist tropical forest at elevations of 100-200m, over a 50-100+ km wide zone. Hence, there is NO point of direct contact between chrysops and ornatus s.s. anywhere in their respective distributions – they are isolated at least since the latest glacial maximum and given the warming trend of climatic change, this degree of isolation can only increase in the coming years.”

Comments from Bonaccorso: “YES for A. Based on the available evidence, three species (Myioborus albifrons, M. ornatus, and M. melanocephalus) make sense (five also, but I think the move would be too bold). Céspedes et al. are working on a genomic treatment of the whole group, and knowing Laura (Céspedes), it will be superb. So, we will have more than enough genetic evidence to consider further splits.”

Comments from Lane: “I am not entirely clear of the accepted phylogeny within this clade, but the evidence provided makes me think that there are a couple of other taxonomic options that are not offered here (namely, 2 species [M. albifrons and M. melanocephalus, including all remaining taxa of the complex] and 4 species. The one I would favor, based on what I am seeing here is: 4 species.

1. M. albifrons (monotypic)

2. M. ornatus (monotypic)

3. M. chrysops (assuming that it is decided to be the name with priority over M. bairdi, and also including M. griseonuchus. The broad hybrid swarm zone seems to make this a necessity)

4. M. melanocephalus (including malaris and bolivianus).

“Did I miss some reasoning why this taxonomy was not considered an option? So, NO for any of the three options offered in the proposal.”

Additional comments from Areta: “Thanks Gary for clarifying that the gap between ornatus and chrysops is real. In no way have I ignored the complex Andean topography in my reasoning. However, the meager plumage, vocal, and genetic differences that we know at present, leave ornata as a very weak species in my perspective, and I still stand by the yardstick approach that it is better considered a subspecies of chrysops. It is great to hear that Laura is working on a genomic perspective on these Myioborus. I am eager to see what she finds, and to revise my vote accordingly.”

Comments from Claramunt: “YES to A. Three species. The new information clearly shows gene flow and introgression between chrysops and bairdi, demonstrating that they are not separate species. Not much we can say about the rest of the complex, but I agree in maintaining albifrons and melanocephalus as separate species for now.”