Black-headed Nightingale-Thrush Catharus mexicanus (Bonaparte, 1856) is a polytypic species, with several disjunct populations ranging from Mexico to Panama, in need of a taxonomic revision. A recent phylogenetic study of nuclear DNA sequences obtained via ultraconserved elements (UCEs) found 100% support for two reciprocally monophyletic primary clades in the C. [mexicanus] complex (Halley 2021): (1) a southern clade formed by samples from Costa Rica (Alajuela, Cartago), Nicaragua (Matagalpa) and Honduras (Copán), encompassing the type localities of C. fumosus Ridgway, 1888 (Costa Rica), C. m. carrikeri A. R. Phillips, 1991 (Costa Rica) and C. m. yaegeri A. R. Phillips, 1991 (Honduras); and (2) a northern clade that comprised all other C. [mexicanus] samples, ranging from western Guatemala (Huehuetenango, c.210 km west of the type locality of C. m. cantator Griscom, 1930) to the Sierra Madre Oriental in eastern Mexico, c.245 km north-west of Xalapa, Veracruz, the type locality of C. mexicanus (Bonaparte, 1856). The range of the northern clade presumably extends north to Tamaulipas, the type locality of C. m. smithi Nelson, 1909, which was not sampled by Halley (2021); and the southern clade presumably extends to Panama (see Phillips 1991: 111).

In a recent revision of the polytypic Slaty-backed Nightingale-Thrush Catharus [fuscater] Lafresnaye, 1845, complex, Halley et al. (2023) treated populations at species rank when they were genetically diagnosable (monophyletic) and either morphologically or vocally diagnosable. In the C. [mexicanus] complex, southern specimens (i.e., from the range of the southern UCE clade) are said to be ‘Darker gray below than [northern specimens], particularly on the sides [and] Dorsally ruddier (browner, warmer) and darker than [C. m. cantator of Guatemala], particularly on the rump and upper tail-coverts, and darker than [more northern] races’ (Phillips 1991: 111). Vocal studies are not yet available (Collar 2020). The geographic boundary between the clades is apparently located near the Guatemala/Honduras border, but more collecting and genetic sampling across the Río Motagua will be needed to confirm this hypothesis. Although a more formal analysis of phenotypic variation in this complex is unavailable at the present time (e.g., Halley et al. 2017, 2023), irrespective of whether the southern taxon is treated at species or subspecies rank, the nomenclatural conflict needs to be resolved. Here, for simplicity, I recognise two phylogenetic species in the C. [mexicanus] complex, which correspond to the northern and southern UCE clades (see Appendix; Halley 2021).

The oldest available name for the northern clade is C. mexicanus (Bonaparte, 1856), and the oldest available name for the southern clade is C. fumosus Ridgway, 1888 (here referred to by the English name ‘Black-capped Nightingale-Thrush’), which is contested. Phillips (1991: 111) claimed that the C. fumosus holotype (USNM 101765) was a hybrid between ‘C. mexicanus’ and ‘C. fuscater’, the latter taxon referring to ‘Talamanca Nightingale-Thrush’ C. [f.] hellmayri von Berlepsch, 1902, which was elevated to species rank by Halley et al. (2023) and is treated as such here. Phillips (1991) then re-described the southern taxon under the name C. m. carrikeri A. R. Phillips, 1991, setting up nomenclatural conflict. The holotype of C. fumosus is currently identified as ‘Catharus fuscater × mexicanus’ in the USNM online database ( https://collections.nmnh.si.edu/search/birds/, accessed 17 February 2025) and Collar (2020), Gill et al. (2024) and HBW & BirdLife International (HBW 2024) have adopted the name C. m. carrikeri for the southern taxon at subspecies rank. However, Clement (2000), Halley (2021) and Clements et al. (2022, 2024) continue to recognise the priority of C. [m.] fumosus. To resolve this conflict, in the following sections, I review the taxonomic history of C. fumosus and compare the phenotypic characters of its holotype (USNM 101765) to a larger sample of adult male study skins.

Taxonomic history

José Castulo Zeledón (1846–1923) collected USNM 101765 at an unknown locality in Costa Rica, on 20 October 1884, then deposited the specimen at the National Museum of Natural History, Smithsonian Institution, Washington (USNM), in January 1885, where Robert Ridgway (1850–1929) catalogued it under the name ‘Catharus mexicanus’ (Fig. 1). Later that year, Zeledón (1885: 104) listed ‘Catharus mexicanus (Bonap.)’ in his catalogue of the Costa Rican avifauna, with an asterisk drawing attention to the USNM specimen. It was the first evidence of the species in Central America, and the first C. mexicanus specimen in the USNM collection (note the geographic range given by Baird 1874: 11). Ridgway and Zeledón apparently identified USNM 101765 by consulting published descriptions because no other specimens were immediately available for comparison (see Ridgway 1888). Bonaparte’s (1856) type had been collected by Auguste Sallé (1820–96) in ‘Jalappa’ (= Xalapa, Veracruz, Mexico) but its whereabouts are uncertain.¹

Figure 1.

Three views of USNM 101765, the holotype of Black-capped Nightingale-Thrush Catharus fumosus Ridgway, 1888, and its two labels (courtesy of USNM Division of Birds, Smithsonian Institution)

Three years later, Ridgway (1888: 505) described USNM 101765 as the holotype of a new species, Catharus fumosus Ridgway, 1888, after comparing it to a study skin from Guatemala in the private collection of George N. Lawrence (1806–95): ‘Although I have only one specimen of each for comparison, there cannot, I think, be any question of the propriety of separating this bird from true C. mexicanus … [because the Guatemalan specimen] agrees very closely with descriptions of Mexican specimens, I have no doubt it belongs to the same form.’ Lawrence’s skin was probably AMNH 39095 (apparently the only Guatemalan skin of C. mexicanus in his large collection), which was catalogued at the American Museum of Natural History, New York, on 23 May 1889, and which Ridgway (1907: 23) later listed in his C. m. mexicanus sample.

However, after two decades of collecting efforts, the size of Ridgway's (1907: 24) sample had barely improved. Now, in addition to USNM 101765, he had five or six adults from Costa Rica,² but this series was more similar in colour to two northern specimens (one each from Mexico and Guatemala) than to the aberrant USNM 101765. Ridgway (1907: 24) wrote:

However, unwilling to concede that the dark ventral plumage of USNM 101765 was merely an extreme example of individual variation in the Costa Rican sample, Ridgway (1907: 24) recognised both C. m. mexicanus (Mexico to Costa Rica) and C. m. fumosus (Costa Rica). The latter taxon, he proposed, was endemic to the ‘highlands of Costa Rica (precise locality unknown)’ and only one specimen was known.³ However, Carriker (1910: 749), who collected that specimen, dissented:

Griscom (1930) eventually restricted C. m. mexicanus to Mexico, and C. m. fumosus to Costa Rica and western Panama, while describing Guatemalan specimens from the mountains north-west of the Río Motagua under the novel name C. m. cantator Griscom, 1930. Hellmayr (1934: 464) expressed doubts about C. m. cantator but nonetheless recognised all three subspecies and extended the range of C. m. fumosus to north-western Nicaragua. Ripley (1952: 40) inexplicably synonymised C. mexicanus with C. dryas (Gould, 1855), but later followed Hellmayr (1934) in applying C. m. fumosus to populations ranging from Nicaragua to Panama (Ripley 1964: 169). This was the status quo until the late 20th century, when Phillips (1991) re-examined the C. fumosus holotype and proposed that it was a hybrid.

The hybrid hypothesis

The origins of the ‘hybrid hypothesis’ are found in draft copies of Phillips’s manuscript, now at the Delaware Museum of Nature & Science (DMNH, formerly Delaware Museum of Natural History), Greenville. In the earliest draft, Phillips stated that USNM 101765 ‘appears to me to be a hybrid’, then edited the sentence to read, ‘[it] is evidently a hybrid’ (Fig. 2). Finally, he removed all semblance of doubt in his final (published) draft, stating that ‘[it] is a hybrid’ (Phillips 1991: 111):

Figure 2.

Early draft of Phillips's (1991) comments on the type of Black-capped Nightingale-Thrush Catharus fumosus, written in pencil: ‘Hybrid: The unique type of Catharus fumosus Ridgway, which has long puzzled ornithologists … appears to me to be is evidently a hybrid × C. fuscater; its dark colors, rather long wing (88.5 mm), rather heavy bill, and intermediate tail-graduation (4.5 mm), and perhaps [a] somewhat shorter “crest”, all point in that direction, though it resembles mexicanus in most ways (except underparts).’ Reproduced courtesy of the Delaware Museum of Nature & Science (DMNH).

Art. 23 of the Code, then in its third edition, stipulated that ‘A species-group name established for an animal later found to be a hybrid [Art. 17(1)] must not be used as the valid name of either of the parental species’ (ICZN 1985: 23). This rule was essentially unchanged in the fourth edition, despite some light editing (ICZN 1999, Art. 23.8). Therefore, after proposing the hybrid hypothesis, Phillips (1991) declared that Ridgway’s (1888) description was invalid and re-described the Costa Rican taxon under the name C. m. carrikeri A. R. Phillips, 1991, based on CM P13557, the Costa Rican female examined by Ridgway (1907). At that time, there had been no credible reports of hybridisation between any Catharus species (McCarthy 2006); and even now, 36 years later, the only reports of early generation hybrids come from three closely related migratory species (FitzGerald et al. 2017, Martinsen et al. 2017, Termignoni-Garcia et al. 2022). No reports of hybridisation have been confirmed in any of the Neotropical resident Catharus lineages, to my knowledge, despite genetic evidence of prehistoric introgression (Everson et al. 2019).

Phillips (1991) offered scant evidence in support of the hybrid hypothesis. With respect to the dark ventral plumage, he merely implied, as Ridgway (1907) had done, that all C. [mexicanus] specimens from Costa Rica are uniformly bright in their ventral plumage, and therefore the ‘dark’ underparts of USNM 101765 needed an explanation. He contended that the ‘long wing’ and ‘rather heavy bill’ were evidence of hybrid origin but presented no evidence that his measurements were statistical outliers, or that they were intermediate between the putative parental species. Here, to resolve nomenclature, I compared the phenotypic characters (plumage colour and quantitative morphometrics) of USNM 101765 to a large and geographically representative sample of adult male study skins.

Methods

I examined and measured adult male study skins of C. mexicanus (n = 56) and C. fumosus (n = 72, including USNM 101765), identified by their collecting localities north-west and southeast of the Río Motagua, respectively (i.e., corresponding to the UCE clades; see Appendix), and C. hellmayri (n = 41), in the collections of the Academy of Natural Sciences of Drexel University, Philadelphia (ANSP), AMNH, CM, DMNH, Field Museum of Natural History, Chicago (FMNH), Museum of Comparative Zoology, Cambridge, MA (MCZ), and USNM. I used an iPhone to capture images of specimens in the USNM collection, to illustrate relevant patterns of colour variation. I used a metric wing ruler to measure the flattened length of the right wing of each specimen, from the carpal joint to the tip of the longest primary remex (WG); and the length of the tail from the insertion point of the two central rectrices to the tip of the longest rectrix (TL); and digital callipers to measure the length of the right tarsometatarsus (or left if the right was broken), from the intertarsal joint to the end of the final leg scale (TR); the length of the bill, from the anterior edge of the right naris to the tip (BL); and the width of the bill at the anterior edge of the naris (BW). I excluded ‘tail graduation’ from the analysis, although it was given by Phillips (1991) as an intermediate character, because of uncertainty about his measurement technique. All statistics were performed with R-Studio (R Core Team 2020). To examine potential clinal variation in the C. [mexicanus] complex, I estimated the collecting latitude of each specimen from label data or with Google Earth Pro. I performed linear regression analyses to model the relationships between latitude and the five morphometric variables, and interpreted models with p-values less than 0.05 as statistically significant. I also used boxplots to compare the measurements of USNM 101765 to the C. mexicanus, C. fumosus and C. hellmayri samples.

Plumage colour variation

The dark ventral colour of USNM 101765 was emphasised by Ridgway (1888, 1907) in the original and secondary descriptions of C. fumosus, and later by Phillips (1991: 111), who concluded that it was a hybrid character. However, there are other C. fumosus study skins that are nearly as dark (especially those collected in Panama, as acknowledged by Phillips 1991: 111, e.g., MCZ 137397 and 137398), and the ventral plumage of USNM 101765 is actually darker than that of most C. hellmayri males, not lighter as one would expect from a bird with an intermediate (hybrid) phenotype (Fig. 3). It appears that the ‘dark’ ventral plumage of USNM 101765 merely reflects one end of a continuum of intrapopulation (individual) variation within C. fumosus. For example, USNM 198803, an adult male from Bonilla, Limón, Costa Rica, collected by A. Alfaro with Ridgway present (see Halley 2024: 124) on 30 March 1905, is nearly as dark as USNM 101765 (Fig. 3), and digital photos of wild birds in the Macaulay Library (ML), taken between Volcán Tenorio and San José, show a considerable range of individual variation. The underparts of some individuals are pale grey (e.g., ML 557487321, 393335501), whereas others have a bright (whitish) throat that contrasts with a dusky grey belt across the lower breast (e.g., ML 525014371, 496253471), and others still have a dark grey throat like USNM 101765 (e.g., ML 550113391, 417916291, 352993681). The evolutionary history of this variation and its adaptive function (if any) are unknown, but subtle patterns of polychromatism (sometimes called ‘colour phases’ or gradients) have been demonstrated in the closely related C. [fuscater] complex (Halley et al. 2023) and in two migratory species, Grey-cheeked Thrush C. minimus (Lafresnaye, 1848) and Bicknell’s Thrush C. bicknelli (Ridgway, 1882), with the variation being evident among adults in the same breeding population (e.g., Wallace 1939, Burleigh & Peters 1948), as Phillips (1991: 95) acknowledged. Therefore, there is no reason to assume that the ventral gradient in C. fumosus, or the ‘dark’ colour of USNM 101765 (i.e., one endpoint of that gradient), are the consequence of hybridisation with C. hellmayri or any other species (contra Phillips 1991: 111).

Figure 3.

Ventral and dorsal views of (A) a ‘dark’ adult male Black-capped Nightingale-Thrush Catharus fumosus (USNM 198803) collected in 1905 at Bonilla, Limón, Costa Rica; (B) the adult male holotype of C. fumosus Ridgway, 1888 (USNM 101765) collected at an unknown locality in Costa Rica; and (C) an adult male of Talamanca Nightingale-Thrush C. hellmayri von Berlepsch, 1902, collected near Dota, San José, Costa Rica (USNM 209929). The difference between these photos, and those of USNM 101765 in Fig. 1, provide a valuable demonstration of how variable light conditions may deceive taxonomists working with Catharus specimens; specimens should only be compared within the same photo, not between photos taken under different conditions (Matthew R. Halley; reproduced courtesy of the USNM Division of Birds, Smithsonian Institution)

Morphometric variation

In the C. [mexicanus] complex, I detected significant relationships with latitude in four of five morphometric variables (Table 1, Fig. 4). WG showed a decreasing trend from north to south, and the ‘long wing’ of USNM 101765 (Phillips 1991: 111) fell comfortably within the C. fumosus distribution (Fig. 5). TR also decreased from north to south, but the effect was weak (ß = 0.051, t = 2.948, Table 1, Fig. 4). USNM 101765 had a shorter than average TR for the C. fumosus sample—not longer, as one would expect of a hybrid with a C. hellmayri parent (Fig. 5). TL also decreased from north to south, with a medium effect size (ß = 0.384, t = 8.175, Table 1) and the tail of USNM 101765 (58 mm) was below average for C. fumosus, not above average as predicted by the hybrid hypothesis (n = 35, mean TL = 58.2 mm, Fig. 5). BL increased weakly with latitude (ß = 0.019, t = 2.686) and this relationship was not significant (Table 1, Fig. 4). The BW of USNM 101765 was slightly below average for the C. fumosus sample (Fig. 5), not above average as predicted by the hybrid hypothesis.

Figure 4.

Relationships between latitude (predictor) and five morphometric variables in a sample of adult male study skins of Black-headed Nightingale-Thrush Catharus mexicanus (yellow dots: Mexico to Guatemala) and Black-capped Nightingale-Thrush C. fumosus (blue dots: Honduras to Panama): wing length (WG), tail length (TL), tarsometatarsus length (TR), bill length (BL) and bill width (BW). Latitude is shown on the y-axis for visualisation purposes. All variables have millimetres (mm) as units. Linear regression trendlines for the combined data are shown in black. Red vertical lines denote the measurements of the C. fumosus holotype (USNM 101765).

TABLE 1 Results of linear regression models of relationships between latitude (predictor) and five morphometric variables (WG, TL, TR, BL, BW) in a range-wide sample of adult male Catharus [mexicanus] study skins (n = 128).

Figure 5.

Geographic variation in five morphometric variables in a sample of adult male study skins of Black-headed Nightingale-Thrush Catharus mexicanus (Mexico to Guatemala), Black-capped Nightingale-Thrush C. fumosus (Honduras to Panama) and Talamanca Nightingale-Thrush C. hellmayri (Costa Rica to Panama): wing length (WG), tail length (TL), tarsometatarsus length (TR), bill length (BL) and bill width (BW). All variables have millimetres (mm) as units. Red horizontal lines denote the measurements of the C. fumosus holotype (USNM 101765).

Taxonomy and nomenclature

The supposedly hybrid characters of USNM 101765, the adult male holotype of C. fumosus Ridgway, 1888, fall within the normal range of phenotypic variation of the southern clade of the C. [mexicanus] complex, for which the name C. fumosus has priority (Halley 2021). With respect to its ‘darker’ ventral plumage, Phillips (1991) was evidently misled (like Ridgway 1888, 1907) by individual (intrapopulation) variation, reminiscent of polychromatic patterns in the C. [fuscater] complex (Halley et al. 2023). Specimens and photographed individuals of C. fumosus exhibit a range of brightness in the ventral plumage, perhaps but not necessarily related to age or sex. Therefore, the most parsimonious explanation is that USNM 101765 merely represents one extreme of a plumage brightness gradient. The ‘long wing’ and ‘rather heavy bill’ of USNM 101765, which Phillips (1991) emphasised as potentially hybrid characters, also fall within the normal range of variation in C. fumosus. Analysis of three other morphometric characters (TL, TR, BL), which were not mentioned by Phillips (1991), similarly failed to produce support for the hybrid hypothesis. Therefore, I recommend that the name C. fumosus Ridgway, 1888 be retained for the southern clade of the C. [mexicanus] complex, irrespective of its taxonomic rank. The taxon C. m. carrikeri A. R. Phillips, 1991, is its junior synonym (see Appendix). If additional research finds support for the hybrid hypothesis, in the interest of nomenclatural stability, I recommend that a neotype be designated to preserve the long-standing nomenclature (ICZN 1999, Art. 75).