Proposal (752) to South American Classification Committee

Split Sclerurus mexicanus into multiple species

The following proposal is a set of hierarchical taxonomic schemes for the Sclerurus mexicanus complex. It expands upon Proposal 603 with newly published vocal analyses (Cooper & Cuervo 2017). This proposal is meant to be considered sequentially (i.e., Part I must be approved to consider Part II).

Part I

Split Sclerurus mexicanus into two species: Sclerurus mexicanus and Sclerurus obscurior.

Effect on SACC: Sclerurus mexicanus would be split into two species that are assumed to be parapatric in the Darién Gap. All known South American populations would become S. obscurior; S. mexicanus would be moved to the hypothetical list.

Background: The genus Sclerurus currently contains six widespread, polytypic species. Sclerurus mexicanus has the broadest distribution of any Sclerurus, with seven subspecies occurring from northern Mexico to southern Brazil (Cooper and Barragan 2017; Fig. 1). Despite the recognized diversity and broad distribution, differences are minimal (Remsen 2003) and subspecies distributions are still incompletely known (Cooper and Barragan 2016; Cooper and Cuervo 2017). d’Horta et al. (2013) found that Sclerurus mexicanus is not monophyletic as currently defined. Rather, three major clades were discovered that form a polytomy in the rufous-throated leaftosser clade: (1) S. mexicanus (subspecies mexicanus and pullus) in North America; (2) S. rufigularis; and (3) S. mexicanus (subspecies obscurior, andinus, macconnelli, and peruvianus) in South America. Similarly, recent vocal analyses only found S. rufigularis to be statistically diagnosable when analyses allowed for multiple species within the S. mexicanus complex (Cooper and Cuervo 2017).

Figure 1. A map of subspecific distributions within Sclerurus mexicanus. Points represent localities from which vocalizations were sampled. Figure from Cooper and Cuervo (2017).

Within the S. mexicanus group, North and South American populations appear wholly allopatric, diagnosably monophyletic (d’Horta et al. 2013), and vocally distinct (Cooper and Cuervo 2017). Unless all rufous-throated leaftossers are considered a single species (including S. rufigularis), the most prudent decision is to split S. mexicanus into two taxa with no known geographic overlap:

1. Sclerurus mexicanus (Sclater 1856). Suggested English name: Central American Leaftosser. Type locality: Cordoba, Veracruz, Mexico. This species contains two subspecies: mexicanus (from northern Nicaragua north to east Mexico) and pullus (from northern Costa Rica south to southern Panama). This includes certus (Chubb 1919, from Guatemala), which is considered a synonym of mexicanus (Hellmayr 1925).

2. Sclerurus obscurior (Hartert 1901). Suggested English name: Dusky Leaftosser. Type locality: Lita, Esmeralda, Ecuador (ca. 600m). This species contains five subspecies: obscurior (the Choco lowlands), andinus (the mid-montane Andes of Colombia, Ecuador, and Venezuela), peruvianus (the eastern foothills of the Andes and western Amazonia in Peru, Ecuador, Colombia, and probably Brazil), macconnelli (eastern Amazonia and the Guianan shield), and bahiae (the Atlantic Forest of Brazil).

Thus, if accepted, all confirmed populations in South America would become subspecies of S. obscurior.

Part IIA

Split Sclerurus mexicanus into Sclerurus mexicanus and Sclerurus pullus.

Effect on SACC: This would only become important if the population pullus is confirmed in Northern Colombia. Sclerurus pullus is known from adjacent eastern Panama. This would have no effect on the main SACC list at the present time; S. mexicanus would be removed from the hypothetical list and replaced with S. pullus.

Background: Sclerurus mexicanus and Sclerurus obscurior have no known regions of overlap, with no confirmed records of Sclerurus mexicanus (sensu Part I) in the SACC region. However, given that S. mexicanus pullus occurs in eastern Panama immediately adjacent to Colombia, it is possible that the taxon will be found in the SACC area in the future. Despite limited vocal data, differences were recovered between S. m. mexicanus and S. m. pullus (Cooper and Cuervo 2017) and populations are reciprocally monophyletic (d’Horta et al. 2013). A split of these two taxa is therefore warranted. This will have no effect at the present time on the SACC list, but will change any future records of S. mexicanus from northern Colombia to S. pullus, as follows:

Sclerurus pullus (Bangs 1902). Suggested English name: Isthmian Leaftosser (thus altering S. mexicanus to Tawny-throated or Mexican Leaftosser). Type locality: Boquete, Panama. Distributed from northern Costa Rica through the Darién in Eastern Panama. This species may occur in the Darién and Urabá regions of Colombia, and may exist parapatrically with S. obscurior. This species includes the synonym anomalus (Bangs and Barbour 1922), which has erroneously been synonymized with andinus in the past (Peters 1951).

Part IIB

Split Sclerurus obscurior into three species: Sclerurus obscurior, Sclerurus andinus, and Sclerurus macconnelli.

Effect on SACC: Sclerurus obscurior would become three allopatric species, with Sclerurus obscurior being retained by lowland Choco populations. Andean populations would become Sclerurus andinus, and Amazonian/Atlantic Forest populations would become Sclerurus macconnelli.

Background: d’Horta et al. (2013) recovered multiple monophyletic clades within South American S. obscurior (sensu Part I). These groups correspond to all subspecies for which genetic data were available (i.e. all subspecies except bahiae), with the shortest branch lengths for the division between S. o. peruvianus and S. o. macconnelli (Fig. 2). Given this short branch length, it is reasonable to consider S. peruvianus and S. macconnelli conspecific (but see Part III below). Cooper and Cuervo (2017) tested multiple different partitioning schemes on Sclerurus songs and found the highest support for a similarly conservative partitioning scheme that splits obscurior and andinus but retains peruvianus, macconnelli, and bahiae as a single polytypic species.

Figure 2. Species partitioning schemes tested for the S. mexicanus complex. Branch lengths are not to scale, but are presented proportionally and the tree depicts monophyletic assemblages recovered by d’Horta et al. (2013). Colored lines, labeled 1-5, represent groupings discussed in Cooper & Cuervo (2017); with respect to this proposal, partition 2 corresponds to Part I, partition 4 to Part II, and partition 5 to Part III. Figure from Cooper and Cuervo (2017).

Genetic data supports the split of andinus from obscurior (d’Horta et al 2013): “Along the continuum of humid forests from the Chocó lowlands to the slopes of the western Andes, two lineages appear segregated elevationally: S. obscurior, and S. andinus found locally from about 1000 m (often up to 2000 m). The two lineages are potentially syntopic at an intermediate point of the elevational and ecological gradient, where no obvious physical barrier is in place… the lowland Chocó (i.e. S. obscurior) and the Andean foothill species (i.e. S. andinus)…[last] shared a common ancestor in the Early Pliocene, between 3.6 and 6.0 Ma”.

Based on these data, Part IIB therefore proposes elevation of S. andinus and S. macconnelli to species level. S. obscurior would then become a monotypic taxon restricted to the Choco (with the suggested English name ‘Dusky Leaftosser’). The descriptions for these new taxa would be as follows:

1. Sclerurus andinus (Chapman 1914). Suggested English name: Andean Leaftosser. Type locality: Buenavista, above Villavicencio, Colombia, on the E slope of the Eastern Andes (ca. 1370 m). This subspecies resides in humid submontane forest in western Venezuela (the main Andes and the Sierra de Perijá), all three Andean ranges of Colombia, and in Western Ecuador. Records from northeastern Ecuador and southwestern Ecuador (El Oro) may refer to this taxon.

2. Sclerurus macconnelli (Chubb 1919). Suggested English name: Long-billed Leaftosser (as opposed to Short-billed Leaftosser S. rufigularis with which this species is sympatric; MacConnell’s, Amazonian, or Guianan Leaftosser [as in Cory & Hellmayr] are also possible English names). Type locality: Ituribisci River, Guyana. This species would consist of three subspecies: peruvianus of the eastern foothills of the Andes and western Amazonia; macconnelli of central and eastern Amazonia and the Guianan shield; and bahiae of the Atlantic Forest of Brazil. Records from Ceará, halfway between the distributions of macconnelli and bahiae, are best left unidentified at this time. While parapatry is assumed between peruvianus and macconnelli, the contact zone is not well known.

Part III

Split Sclerurus macconnelli into two species: Sclerurus macconnelli and Sclerurus peruvianus.

Effect on SACC: Amazonian populations of S. macconnelli would be split into two species, with peruvianus occupying the western basin and Andean foothills and macconnelli occupying the lower basin, the Guianan shield, and the Atlantic Forest.

Background: While defined vocal differences between S. macconnelli subspecies (sensu Part IIB) were not recovered (Cooper & Cuervo 2017), d’Horta et al. (2013) recovered two monophyletic groups aligning to the described populations of peruvianus and macconnelli. Per d’Horta et al. (2013): “... S. macconnelli and S. peruvianus … are in close geographical proximity in southern Peru and Bolivia but seem to occupy different elevations along the cis-Andean foothills”. The magnitude of the differentiation between macconnelli and peruvianus is less than any other branching event within S. mexicanus sensu lato, including the relationship between S. mexicanus and S. pullus.

Regrettably, no genetic data is presently available for bahiae, and its relationship to the rest of the S. macconnelli complex is uncertain. Thus, Part III opts for the elevation of S. peruvianus to species level while retaining bahiae as a subspecies of the geographically proximate S. macconnelli.

S. peruvianus (Chubb 1919). Suggested English name: Peruvian Leaftosser (as in Cory & Hellmayr). Type locality: Yurimaguas, Loreto, Peru. This species replaces macconnelli in northwestern Amazonia and in the higher elevations of the Andean foothills in southern Peru and Bolivia. Exact range limits are unknown, but it is known to occur on both sides of the Napo/Amazon Rivers in Ecuador and Colombia.

Recommendation

Given the amount of genetic and vocal data available, we recommend the acceptance of Part I and Part II (both A and B). Five species are repeatedly delimited using both vocal and genetic data, and best represent the diversity within S. mexicanus sensu lato. These decisions would remove S. mexicanus from the SACC list and replace it with three species: S. obscurior, S. andinus, and S. macconnelli. One additional species should be considered hypothetical within the SACC area (S. pullus) pending further surveys of the Darién Gap.

Given the present data, we recommend holding off on Part III. The relationship of bahiae to peruvianus and macconnelli is not resolved, and there is no reliable way to identify these taxa in the field at the present time short of collecting vouchers or genetic samples.

Literature cited

Cooper, J. C. & A. M. Cuervo. 2017. Vocal variation and species limits in the Sclerurus mexicanus complex. The Wilson Journal of Ornithology 129:13-24.

Cooper, J. C. & Barragán, D. 2017. Tawny-throated Leaftosser Sclerurus mexicanus. Neotropical Birds Online <http://neotropical.birds.cornell.edu/>.

d’Horta, F. M., A. M. Cuervo, C. C. Ribas, R. T. Brumfield & C. Y. Miyaki. 2013. Phylogeny and comparative phylogeography of Sclerurus (Aves: Furnariidae) reveal constant and cryptic diversification in an old radiation of rain forest understory specialists. Journal of Biogeography 40:37-49.

Other references in SACC bibliography.

Jacob C. Cooper & Andres M. Cuervo, May 2017

========================================================

Comments from Jaramillo: “Part I – YES, data look solid to me, and include multiple independent sets of data. Note that Central American Leaftosser does not work for me.

“Part IIA – YES, assuming it eventually is found in Colombia.

“Part IIB – YES, I find it powerful that obscurior and andinus are syntopic, and replace each other elevationally. It is unfortunate that bahiae was not sampled molecularly, as it may shift the tree perhaps?

Part III – NO, unresolved as noted in the proposal. Data needed for bahiae, as well as no reliable way to identify taxa in field!”