Proposal (963) to South American Classification Committee

Split Myiophobus fasciatus into three species

Our SACC notes read:

"82. The subspecies rufescens of arid western Peru and northern Chile was formerly (e.g., Cory & Hellmayr 1927) considered a separate species from Myiophobus fasciatus, but Zimmer (1939c) and Koepcke (1961) reported specimens that showed signs of intergradation between rufescens and M. f. crypterythrus (cf. Ridgely & Tudor 1994); thus, Meyer de Schauensee (1966) considered them conspecific, and this has been followed by subsequent authors. Jaramillo (2003), however, suggested that rufescens should be considered a separate species."

Vocally, the three groups (fasciatus, crypterythrus, and rufescens) are quite distinct, and these distinctions match the marked plumage differences described in the literature. It is remarkable how perfectly similar the calls and songs of birds in the fasciatus group are, even when they cover southern Central America and most South America. It is not clear how many of the northern individuals would represent southern migrants (populations in Argentina are complete migrants), but even then, birds giving the full dawn song in southern Central and northern South America sound like those at the southernmost portions of the range. The evidence for the continuity of this vocal time is overwhelming, and contrasts with the clear change to vocalizations of crypterythrus and rufescens.

For an overview of vocalizations see here:

fasciatus: https://birdsoftheworld.org/bow/species/brcfly1/cur/sounds

crypterythrus: https://birdsoftheworld.org/bow/species/brcfly4/cur/sounds

rufescens: https://birdsoftheworld.org/bow/species/brcfly3/cur/sounds

You will immediately notice that it is easy to track homologous dawn song and diurnal call (or song) between fasciatus and crypterythrus while at the same time recognizing the obvious differences. The vocalizations of rufescens are totally different, and have some reminiscence to some Ochthoeca voices and duets.

There are two literature reports of intergradation to be dealt with.

Zimmer (1939: page 6) reported a specimen from Pacasmayo, northern La Libertad, Peru that he considered to be "nicely intermediate between crypterythrus and rufescens, being too pale beneath for typical rufescens and too buffy for crypterythrus, with pectoral streaking also intermediate. It furnishes additional evidence that the two forms are conspecific". The specimen in question is this one (photos by Paul Sweet):

The specimen is in terrible shape, making it difficult to evaluate in terms of its characters. Beyond that, one specimen, even if an intergrade, does not tell us much. Zimmer, of course a man of his times, did not think in terms of hybrid zones or the incidence of hybridization, but we definitely would take that into account. As for the incidence of introgression, Zimmer also reports specimens of apparently typical rufescens in AMNH from Trembladera [= Tembladera], Cajamarca, which is upstream from Pacasmayo; and as Zimmer was aware (he cites these records, without comment), rufescens has been collected at two other sites, Paucal and Guadalupe, both in Cajamarca and both a short distance northeast of Cajamarca. So his single, reported intergrade is south of the 'front line' of wherever rufescens and crypterythrus would be expected to come into contact, and even if an intergrade (which is not clear at all by looking at this specimen), there is no other specimen like it from the same region. Furthermore, there's no indication that Zimmer noticed anything odd or unusual about the geographic distribution of these specimens.

Koepcke (1961: pages 17-18) reported two apparent intergrades from Yantán, Ancash. This is a site that is almost 300 km (!) south of Zimmer's reported intergrade, and of course there are many specimens of accepted rufescens from throughout that intervening area. These two specimens are in Lima, but we have not seen them. One has to wonder if there is some plumage of rufescens (juvenile or formative plumage?) that Zimmer, Koepcke, or both were unfamiliar with, and that simply confused them. Certainly it makes no great sense to have uncovered instances of 'introgression' between resident populations at sites that are distant (Zimmer) to very, very distant (Koepcke) from any contact zone.

So where would we expect crypterythrus and rufescens to meet? The southernmost specimens of crypterythrus from west of the Andes that we aware of are from near Taulis, Cajamarca (specimens in Lima, not seen); and the northernmost specimen of rufescens we know of is from near Chiclayo, Lambayeque (MVZ, Berkeley; again, not seen). These two localities are at about the same latitude, but Taulis is further inland. Anyway, the turnover between these two taxa must be pretty abrupt, in view of the absence of specimens or sight records of intergrades from this area.

Genetic data add little to the case. Harvey et al. (2020) included a single sample of fasciatus that was sister to cryptoxanthus (no samples of crypterythrus or rufescens included). Note that Myiophobus is paraphyletic (type species fasciatus):

Recommendation: We recommend the recognition of 3 species, as follows:

1. Myiophobus fasciatus (including Central American furfurosus and all South American subspecies east of Andes)

2. Myiophobus crypterythrus (monotypic; SW Colombia in Nariño south to NW Peru in Piura and Marañon Valley)

3. Myiophobus rufescens (monotypic: NW Peru in Lambayeque south to N Chile)

Note from Remsen on voting: A NO vote is for retaining broad M. fasciatus or for splitting just rufescens (leaving crypterythrus with M. fasciatus until additional data are published). Please indicate your preference between these two options just in case a 2/3 majority prefers splitting just rufescens.

J. I. Areta & T. S. Schulenberg, February 2023

Note from Remsen on English names: if this passes, a separate proposal would be needed, with the starting point presumably the names already adopted in the BLI/BOW accounts, e.g. retaining Bran-colored Flycatcher for widespread M. fasciatus, and adopting Mouse-gray Flycatcher for M. crypterythrus and Rufescent Flycatcher for M. rufescens.

Comments from Lane: “YES to recognizing 3 species (M. fasciatus, M. crypterythrus, and M. rufescens). Fernando Angulo tells me that M. crypterythrus and M. rufescens overlap broadly between La Libertad and Lambayeque depts in Peru, with the former occurring in the deciduous thorn scrub of the foothills and the latter fairly strictly in the river valleys that pass through (generally more coastal than the former). That there is no obvious interbreeding given this broad overlap, it seems clear that Zimmer misinterpreted the specimen that has stymied this taxonomic issue for so long. However, I will note that I was surprised to hear M. crypterythrus giving vocalizations very similar to typical vocalizations of M. rufescens (https://macaulaylibrary.org/asset/495506591), suggesting that such vocalizations are within the repertoire of the latter, but seem not to be used regularly. By contrast, I am unaware of the "standard trill" of most Myiophobus appearing in the repertoire of M. rufescens.”

Comments from Stiles: “A- split rufescens from the rest- this at least seems clear so YES; B- recognize crypterythrus as separate- YES, another example of a split of a Pacific form of extreme SW Colombia and Ecuador; C- further splits? NO at this point, though if there are various sspp. of fasciatus in E South America, their relationships to the Mesoamerican furfurosus should be looked into.”

Comments from Zimmer: “YES for treating rufescens and crypterythrus as separate species from C American and cis-Andean fasciatus. As stated in the proposal, the dawn songs, diurnal songs, and rattle calls of crypterythrus are similar enough to the homologous vocalizations of cis-Andean and C American fasciatus that it is easy to recognize each vocalization as a homolog, but are distinct enough to suggest separate species status, whereas the rattles, single-notes, and duet vocalizations of rufescens are qualitatively very different to my ears. Plumage characters, particularly of rufescens, are also quite different, and I am unimpressed by the alleged “nicely intermediate” specimen from Pacasmayo – it is in such poor condition that it is difficult to make any concrete assertions regarding intermediacy. The clincher is the information passed by Fernando Angulo to Dan regarding the broad overlap in range, coupled with habitat segregation, of crypterythrus and rufescens in La Libertad and Lambayeque depts., Peru, without evidence of interbreeding or intergradation.”

Comments from Claramunt: “YES. It would have been nice to see some modern and more detailed analysis of geographic variation, but the plumage and song differences seem well demarcated. This complex is clearly another victim of “the big lumpings” of the mid-XX century.”

Comments from Niels Krabbe (voting for Remsen): “YES. I clearly must vote yes to ranking rufescens as a species for the reasons given by Dan (with Fernando) and Kevin. However, the vocal differences between crypterythrus and the fasciatus group are much less convincing. As described in Brds Wrld, crypterythrus is much higher pitched and has longer notes, but those are only two differences. The wider frequency span (bandwidth) is a natural consequence of the higher pitch and the slower pace is correlated with the longer notes. However, to my ear the dawn song of the fasciatus group is also louder and of a different, almost oscine-like quality that I am not sure how to quantify, so I will go along with three species.”

Comments from Robbins: “YES. Although like proposal 961 (Tolmomyias viridiceps split) there is no published evaluation of the Myiophobus fasciatus complex, I believe information provided in the proposal supports the recognition of three species. The primary song of fasciatus and crypterythrus appears quite similar, with the latter’s higher in frequency. At first listen, I thought the vocalizations of rufescens were quite distinct from the other two, until I listened to Dan’s recording of crypterythrus that sounds quite similar to rufescens. However, given what Dan relates from Fernando Angulo that crypterythrus and rufescens apparently overlap without interbreeding over a relatively broad area, I’m swayed to vote Yes for the recognition of three species. Clearly, an in-depth study is needed in the area that Angulo has identified, but until then, I believe it is better to recognize three species.”

Comments from Bonaccorso: “YES. Plumage and song differences (at least those available) support the split. I am a little uncomfortable about the lack of good genetic sampling, but I guess the evidence available is sufficient.”