Fernando Po Mannikin Spermestes bicolor poensis (Fraser, 1843) is a subspecies of the widespread African estrildid Black-and-white Mannikin Spermestes bicolor (Fraser, 1843) distributed from southern Cameroon and Bioko south to northern Angola, and east to southern Sudan, south-west Ethiopia and north-west Tanzania (Restall 1996, Payne 2020). It is locally common in tall-grass forest clearings, bushy marshes, forest edge and even cultivation (Kunkel 1965, Restall 1996, Borrow & Demey 2001, Payne 2020). English names follow Restall (1996).

Amadina poensis (type locality ‘Clarence, Fernando Po’) and Amadina bicolor (type locality ‘Cape Palmas’, Liberia) were described on the same page by Fraser (1843), who stated that the latter could be differentiated from the former ‘in the absence, in the adult, of the white markings on the wings, rump, and sides’. A. poensis was described thus (original Latin first, followed by my translation):

Fraser (1843) did not list any previous synonyms; nor did any of the subsequent major works dealing with the taxonomy of the West African avifauna (Sharpe 1890, Shelley & Sclater 1905, Bates 1930, Sclater 1930, Chapin 1954). However, there exist previous obscure references to this species in the scientific literature and at least one scientific binomial (now a nomen oblitum) that have been long overlooked and which are highlighted in this paper.

The ‘Radiated Grosbeak’ of Latham (1783: 156)

John Latham (1740–1837) was one of the earliest British avian taxonomists, publishing the seminal work A general synopsis of birds (Latham 1781–85) in three volumes. With Linnean binomials still not universally accepted at the time of its publication, Latham did not recognise their import and described his birds using only common names. Later, realising his error of judgement, he attempted to provide binomials for the species he had described in the Synopsis in his Index ornithologicus (Latham 1790). Unfortunately for Latham he was beaten to it by Johann Friedrich Gmelin (1748–1804) (Gmelin 1788), who thus appears as the author of many species that Latham described (see below). One of the species described by Latham (1783: 156) was the Radiated Grosbeak, the text for which is reproduced in full below.

Although brief, this text shows obvious similarities to the much later description of Amadina poensis by Fraser (1843), including the clear agreement in body size and the mention of the diagnostic characteristic of the black-and-white markings on the wings and sides. No other bird matches this description.

Margaret Cavendish Bentinck (1715–85), Duchess of Portland, was the richest woman in Great Britain during her lifetime, owning the largest natural history collection in the country and had the stated aim of possessing ‘every unknown species in the three kingdoms of nature described and published to the world’ (Lightfoot 1768, Tobin 2016, Pelling 2019). That Latham based this on a cagebird in her collection means that he was probably unable to provide a type locality for it. This may have been key in this description becoming forgotten.

Loxia lineata of J. F. Gmelin (1789: 858)

The German taxonomist J. F. Gmelin was a student of Linnaeus (Kastner 1977) and edited the 13th edition of the Systema naturae of which the first volume (in two parts) dealt with birds (Gmelin 1788, 1789). He compiled descriptions in non-Linnean works and provided Linnean binomials for them where these did not already exist. Thus Latham's Radiated Grosbeak became Loxia lineata J. F. Gmelin, 1789, with the following derived entirely from Latham (1783):

Loxia lineata of Latham (1790: 395)

The introduction to Index ornithologicus (Latham 1790) expresses the author's displeasure at Gmelin's approach. However, in repeating the scientific name proposed by Gmelin (1789) Latham did respect the Linnean principle of priority. Note, however, how much of the detail included in Latham (1783) has been abbreviated. His text is as follows:

‘No. CXXVII Pico grueso negro y blanco’ of Azara (1802: 458)

Félix de Azara (1742–1821) was a Spanish engineer, soldier and naturalist working in Paraguay and the La Plata Basin (Beddall 1983). His three-volume Apuntamientos para la historia natural de los páxaros del Paraguay y Río de la Plata (1802–05) was one of the first systematic attempts to document the regional South American avifauna, but he too neglected to employ Linnean binomials (being greatly influenced by the French school embodied by Buffon). In an analogous situation to that of Latham and Gmelin, many of Azara's descriptions were subsequently published with Linnean names by Louis Jean-Pierre Vieillot (1748–1830) who had commissioned what was a rather unsatisfactory French translation of the Spanish original by Charles Nicolas Sigisbert Sonnini de Manoncourt (1751–1812) and published it as ‘Azara (1809)’. Sonnini, who greatly under-estimated the true extent of avian biodiversity, annotated this work with a series of footnotes (see next section) which attempted (with varying degrees of success) to reconcile Azara’s descriptions with known species. Azara’s life and work is documented by Contreras (2010).

Azara's (1802) description ‘No. CXXVII Pico grueso negro y blanco’ is a remarkable text, for the clarity and detail provided that allows a clear identification as Spermestes bicolor poensis and the fact that it refers to an African cagebird in Argentina in the late 18th century. An important factor to note is that the inches used by Azara are those of the ‘pied du roi de France’ and are equivalent to 27.06 mm (not the standard 25.4 mm used by other authors cited here) and ‘lines’ are equivalent to 2.26 mm.

‘No. CXXVII Gros-bec noir et blanc’ of Sonnini (in Azara 1809: 280)

The translation of Sonnini (in Azara 1809) is, with minor omissions, a faithful translation of Azara (1802) so I will not reproduce it here. However, it is the footnote to this description that is of particular interest:

I have been unable to consult a copy of the Dufart edition of the Oeuvres to which Sonnini refers. Sonnini omits mention of Latham (1783) and may not have been aware of it. His reference to the lack of ‘description’ of the species may concern the absence of a Linnean binomial or perhaps the brevity of the description. However, the suggestion that Azara’s bird and Latham’s Radiated Grosbeak are the same appears well founded.

Discussion

That all of these descriptions refer to the same taxon, Fernando Po Mannikin Spermestes bicolor poensis seems clear. The accuracy of the descriptions by the authors cited above is equal to, and, in at least one case more detailed than, that upon which the scientific name is based (Fraser 1843), and all accounts mention the diagnostic characteristics cited by Fraser (1843) and provide measurements that conform with this taxon.

Pereyra (1945) recognised this to be the case in his review of Azara's ornithological work published in Argentina. Two previous reviewers of Azara's work (Hartlaub 1847, Laubmann 1939) had surprisingly associated his description with White-bellied Seedeater Sporophila leucoptera (Vieillot, 1817), a bird that differs notably from the Pico grueso negro y blanco, most obviously because of clear differences in the colour of the bill and throat, as well as the wing and rump pattern. Furthermore, Azara (1802: 447) described the male Sporophila leucoptera as No. 123 Pico grueso pico trigueño, where the author's careful eye for detail is reaffirmed in the accuracy of that description and its quite different morphometrics. Vieillot (1817) based the scientific name for Sporophila leucoptera solely upon Azara's description.

I suggest that the fact Latham's (1783) work on the Radiated Grosbeak had been ‘watered down’ by the abbreviated description of Gmelin (1789) coupled with his own attempt to apply Linnean binomials to his description (Latham 1790) led to those descriptions being considered unidentifiable, whilst the later association of the same taxon with Azara—an author concerned with the Neotropical avifauna—meant that the possibility the species formed part of the African avifauna was not seriously considered by contemporary authors. Fraser’s (1843) description then provided an apparently stable and consistently applicable name for the common Fernando Po Mannikin, and there was little nomenclatural attention paid to it from that point.

Loxia lineata J. F. Gmelin, 1789, is a senior synonym of Amadina poensis Fraser, 1843; however, it should be considered a nomen oblitum and prevailing usage maintained under Art. 23.9.1 of the Code (ICZN 1999). Loxia lineata J. F. Gmelin, 1789, has not been used as a valid name since 1899 and Amadina poensis Fraser, 1843, has been applied consistently for the taxon that bears the name since the date of its description.