Over much of its range, Black-crested Antshrike Sakesphorus canadensis is locally common in semi-open habitats such as lighter woodland, arid scrub, savannas and even mangroves. It is largely restricted to northern South America, north of the Amazon River, across the Guiana Shield, with populations in north-east and central-east Peru penetrating south of the Amazon in its headwaters (Ridgely & Tudor 1994, Sick 1997, Zimmer & Isler 2003, Zimmer et al. 2020). Here I draw attention to a specimen of this species at the Museu Nacional, Universidade Federal do Rio de Janeiro ornithological collection (MN), allegedly from the Serra do Cachimbo, southern Pará, which has not been previously mentioned in the ornithological literature. My aim here is to clarify the distribution of the species.
Other specimens housed at MN, and the field diaries belonging to the Cachimbo specimen's most likely collectors (Helmut Sick and José Hidasi) at MN and in Dr L. P. Gonzaga's personal library were examined.
Meyer de Schauensee & Eisenmann (1966: 269) were the first authors to attest to the Serra do Cachimbo being part of the species' known range: ‘...and south of the Amazon in northern Mato Grosso (Serra do Cachimbo, fide Sick).’ Subsequent general works followed this statement, by including the Serra do Cachimbo, Pará state, Brazil, within the range of S. canadensis (Meyer de Schauensee 1970, Pinto 19781, Ridgely & Tudor 1994, Sick 1997), but neither Zimmer & Isler (2003) nor Zimmer et al. (2020) mentioned it south of the Amazon in Brazil, nor do recent world checklists (Dickinson & Christidis 2014, del Hoyo & Collar 2016) and an updated work on South American passerines (Ridgely & Tudor 2009).
However, a specimen of the species (Fig. 1), labelled as being from the Serra do Cachimbo, Pará, and collected in June 1956, was found within the series of S. canadensis at MN, along with six others from north of the Amazon: two from northern Amazonas (MN 29635–636) and four from Roraima (MN 30205, 30207, 50019, 50020). Unlike these other specimens, the specimen purported to have been collected in Cachimbo lacks an accession number, despite being the oldest of the seven. The label data provide the locality, date and species identification alone.
Until recently, at least half of the specimens at MN from Fundação Brasil Central's Roncador-Xingú-Tapajós expedition (mainly Passeriformes), on which Helmut Sick was the naturalist, have lacked accession numbers, and are only now being studied, registered and catalogued. All of the ornithological specimens from this expedition are easily identified, bearing labels stamped ‘Fundação Brasil Central’, with the data in Helmut Sick's distinctive handwriting. The great majority of the estimated >3,400 specimens (Gonzaga 1991, Sick 1997) possess the same general data: locality, date, sex, collector, field number and identification on the reverse (Fig. 2). The S. canadensis specimen is now registered as MN 52533.
It is worth stressing the importance of the field numbers to these specimens. Helmut Sick and his assistants were punctilious, and practically every specimen has much more complete data in the original field diaries than were noted on the labels (Fig. 3).
Because the Cachimbo specimen may represent the only record from south of the Amazon in Brazil, and, because other expeditions to the Serra do Cachimbo, including one fairly recent MN-sponsored survey that collected birds in September–October 2010 did not record the species (Pinto & Camargo 1957, Santos et al. 2011), nor are there any recent south-bank records documented in citizen science platforms such as Wikiaves and eBird, it is worth examining some possibilities for the specimen's provenance.
1. Sick personally collected the bird in Serra do Cachimbo.—This is the most parsimonious hypothesis. Sick visited the Serra do Cachimbo in June 1956. There is a mounted Tiny Hawk Accipiter superciliosus collected by him on 15 June 1956 in Rio (MN 22379), but unlike the Black-crested Antshrike, it has a field number and detailed entry in the diary (A.2767). The absence of a field number, a diary entry and the poor condition of the S. canadensis (lacking almost all of the right wing, right leg and damage to the right flank) raises the question of how Sick obtained the specimen. All of the expedition diary entries from 1956 have been reviewed, and no information about the species has been found (L. P. Gonzaga pers. comm.).
2. Sick received the specimen from a third party.—Given the bird's condition, Sick might have received this poorly prepared specimen from a third party, and, considering the importance of the record, incorporated it into the collection with the available data. Why he elected not to publish the record is unknown, but the fact that the species' presence in the Serra do Cachimbo was clearly reported to Meyer de Schauensee by Sick and that a specimen in such poor condition was not discarded provide clear indication that Sick believed the provenance to be correct. The absence of a field number belonging to Sick or one of his assistants, and the specimen's poor preparation compared to the expedition's other material (all in very good condition) mean this assumption is also plausible.
3. José Hidasi collected the specimen.—José Hidasi is one of the most prolific of recent Brazilian collectors and for several years was Sick's assistant on the Roncador-Xingu expedition, undertaking some surveys alone, sending the material either to Sick or to the Museu Paraense Emílio Goeldi, Belém (Santos et al. 2011). However, Hidasi's diaries reveal that he visited the Serra do Cachimbo only in late 1955 (December) returning to Aragarças (the only available route to central Brazil at the time) in January 1956, making collection of the S. canadensis by him improbable. Hidasi visited Roraima (where S. canadensis is fairly common) only in 1962, making the possibility that MN 52533 is mislabelled also unlikely (Novaes 1965, Santos & Silva 2007).
The Cachimbo specimen is indistinguishable from the remaining northern Brazilian specimens of this species at MN (all tentatively assigned to S. c. loretoyacuensis given the grey flanks and browner back, but no comparable material from the nominate race is available).
While it is plausible that S. canadensis occurs in the Serra do Cachimbo, the lack of a clear trail of evidence must leave some doubts as to the specimen's provenance. Several other common birds of savanna and campina / campinarana habitats on white-sand soils, which were previously believed to be restricted to the Guianan Shield and north bank of the Solimões / Amazon River, are now known to be present in savanna enclaves on white-sand soils throughout the Amazon Basin (Sanaiotti & Cintra 2001, Vasconcelos et al. 2011, Borges et al. 2016). As S. canadensis is a more generalist species of scrubby, second-growth habitats in both semi-arid and humid areas, as well as riverine and nearby forests in parts of its range, to some extent competition with Glossy Antshrike S. luctuosus in this region of Amazonian Brazil might explain the rarity or even displacement of S. canadensis to white-sand habitats in Serra do Cachimbo. Efforts to re-encounter Black-crested Antshrike south of the Amazon in Brazil, in sandy soil campinarana woodland and arid scrub, as well riverine forests, might further elucidate the true range of this species.
I thank Luis Pedreira Gonzaga for information, access to and discussions concerning Helmut Sick's diaries in his possession, José Fernando Pacheco for early discussions that led to this manuscript and help with literature, Guy Kirwan and Gustavo Bravo for their input, and Tomas Capdeville for curatorial assistance. Fernando Pacheco and Kevin Zimmer reviewed the submitted version, and proffered useful and constructive criticism. I am grateful to the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) for the Programa de Apoio ao Pós-doutorado (PAPD) scholarship.
 1 In the final version of his catalogue, Pinto (1978), the first author after Meyer de Shauensee's general works (1966, 1970) to review the species' range in Brazil, considered the Serra do Cachimbo population to involve S. c. loretoyacuensis, without indicating the material concerned, nor providing any explanatory footnote. Given that Pinto (1937, 1938) did not examine any specimens of S. canadensis, the update appears to have been exclusively based on Meyer de Schauensee & Eisenmann (1966).