Proposal (969) to South American Classification Committee

Note from Remsen: This proposal is circulating concurrently at NACC; I have made minor editorial changes in this version.

Revise the taxonomy of Amaurospiza seedeaters:

(a) treat Amaurospiza relicta as a separate species from A. concolor,

and (b) treat A. concolor and A. carrizalensis as conspecific with A. moesta,

or (c) treat A. aequatorialis as a separate species from A. concolor,

or (d) lump the five taxa as subspecies of A. moesta

Description of the problem:

This proposal seeks to revise the taxonomy of Amaurospiza seedeaters and contribute to the efforts of the Working Group on Avian Checklists (WGAC) in reconciling global checklists. The genus Amaurospiza, as currently recognized, is comprised of five taxa of blue seedeaters that show minor differences in plumage coloration and body measurements. Recognition of the five taxa as species or subspecies, including the species to which a subspecies belongs, varies among global avian checklists (Table 1). Howard & Moore and eBird/Clements coincide in the three species they recognize (concolor, carrizalensis, moesta), HBW-BL recognizes two species (relicta, moesta), and IOC recognizes four species (concolor, aequatorialis, carrizalensis, moesta). Classification by the NACC and the SACC agrees with Howard & Moore and eBird/Clements.

Table 1. Current taxonomy of Amaurospiza seedeaters in four global avian checklists. Classification by the NACC and the SACC agrees with Howard & Moore and eBird/Clements.

Taxa	Howard & Moore + eBird/Clements	HBW-BL	IOC
relicta (Griscom, 1934)	A. concolor relicta	A. relicta	A. concolor relicta
concolor Cabanis, 1861	A. concolor concolor	A. moesta concolor	A. concolor concolor
aequatorialis Sharpe, 1888	A. concolor aequatorialis	A. moesta aequatorialis	A. aequatorialis
carrizalensis Lentino & Restall, 2003	A. carrizalensis	A. moesta carrizalensis	A. carrizalensis
moesta (Hartlaub, 1853)	A. moesta	A. moesta moesta	A. moesta

The five taxa in the genus Amaurospiza are allopatric and distributed from central Mexico to northeastern Argentina (Table 2). Two of the three subspecies within A. concolor (relicta and concolor) occur in the area covered by the NACC, from central Mexico to Panama. The third subspecies of A. concolor (aequatorialis), and the species A. carrizalensis and A. moesta, are found in South America, and, therefore, are under the jurisdiction of the SACC.

Table 2. Geographic distribution of Amaurospiza seedeaters.

Taxa	Distribution	NACC	SACC
relicta	Mts. of s Mexico (s Jalisco to Guerrero, Morelos and Oaxaca)	A. concolor Blue Seedeater
concolor	Mts. of s Mexico (Chiapas) to Nicaragua, Costa Rica and Panama	A. concolor Blue Seedeater
aequatorialis	Mountains of SW Colombia (Nariño) to n Peru (Cajamarca)		A. concolor Blue Seedeater
carrizalensis	N Venezuela (lower Río Caroni in Bolívar)		A. carrizalensis Carrizal Seedeater
moesta	Locally from se Paraguay to e Brazil and ne Argentina (Misiones)		A. moesta Blackish-blue Seedeater

Background:

Amaurospiza seedeaters are Neotropical resident species generally associated with bamboo thickets and dense understory (Lopes et al. 2011). They feed on arthropods and bamboo seeds, flowers, petioles, and buds (Areta et al. 2023); in Costa Rica they prefer greener and healthier leaves (Figure 1, Pablo-Castillo 2018).

Figure 1. Male of Amaurospiza concolor feeding on bamboo leaves in Costa Rica (Pablo-Castillo 2018).

The taxa within the genus Amaurospiza have a convoluted history. Here is a brief summary of the five Amaurospiza taxa:

concolor

The genus Amaurospiza was described along with the species A. concolor based on specimens from Costa Rica (Cabanis 1861). Ridgway (1901) measured one Panamanian specimen from the Salvin-Godman collection; he noted that the species was found in both Costa Rica and Panama.

Griscom (1934) recognized three subspecies within concolor: concolor from northwestern Costa Rica (Miravalles and Tenorio); and two additional subspecies that he proposed: grandior from the humid Caribbean forest of eastern Nicaragua, and australis from southwestern Costa Rica to western Panama. Griscom provided measurements of relicta (placed in Amaurospizopsis) and the three taxa within concolor (grandior, concolor, australis). He found overlap among most of the measurements with the exception of the wing of Amaurospizopsis, which was larger than in any of the concolor taxa (Table 3).

Table 3. From Griscom (1934), morphometric measurements of male Amaurospizopsis and Amaurospiza seedeaters.

Hellmayr (1938) noted that grandior was indistinguishable in color from concolor, and that only one of three specimens had a slightly longer bill, suggesting that grandior was not maintainable as a separate taxon. The australis group has not been reassessed since the original description by Griscom (1934), from which it appeared that the main difference between australis and concolor was the plumage coloration of the immature male (Figure 2). The two taxa grandior and australis are not recognized by any of the four global checklists; they are currently grouped within concolor concolor (Ramos-Ordóñez et al. 2020).

Figure 2. Original description of Amaurospiza concolor australis by Griscom (1934).

moesta

The taxon moesta was described as Sporophila moesta with a type from Brazil (Hartlaub, 1853). Orr and Ray (1945) noted that Hellmayr (1904) found S. moesta to be identical to Amaurospiza axillaris (Sharpe 1888). From then on, axillaris and moesta were synonymized and have been considered part of the genus Amaurospiza. Sharpe, in a key to the Amaurospiza species, distinguished this taxon by its white axillaries, underwing coverts, and quill-lining (Figure 3) [see below, that Hellmayr´s assessment indicates that the immature male type of aequatorialis lacks white underwing coverts as indicated by Sharpe, but that an adult male has white underwing coverts, setting aequatorialis apart from concolor; also see Table 4 in Areta et al. 2023]. However, Sharpe also noted that taxa within Amaurospiza are very closely allied, making it impossible to distinguish them from descriptions alone. The taxon moesta is mainly found in the Atlantic Forest, although records in pre-Amazonian wooded habitats are recently increasing (Rising et al. 2020).

Figure 3. Key to Amaurospiza species from Sharpe (1888).

aequatorialis

Sharpe (1888) described A. aequatorialis, a species from the western foothills of the Andes of Ecuador, as a separate species from concolor and moesta (under the name A. axillaris). This taxon is similar to concolor, but the forehead and eyebrows are slightly paler bluish gray, and the bill is smaller. Hellmayr (1938) treated aequatorialis as a subspecies within concolor after examination of four specimens, two from Ecuador and two from Colombia, and noted that aequatorialis was slightly smaller and paler than concolor (Figure 4).

Figure 4. Note about Amaurospiza concolor aequatorialis by Hellmayr (1938).

AOU/AOS considered aequatorialis as part of the group A. c. concolor (AOU 1983; AOU 1998). Although aequatorialis is not mentioned explicitly in the checklist, the geographic distribution of A. c. concolor encompasses southwestern Colombia and northwestern Ecuador. Recently, the taxon aequatorialis has also been recorded from northwestern Peru (Angulo Pratolongo et al. 2012; Sánchez et al. 2012).

relicta

Griscom (1934) described a new species within a new genus, Amaurospizopsis relictus, based on a specimen from Chilpancingo, Guerrero, Mexico.

Hellmayr (1938) suggested that relictus could be a northern subspecies of concolor, very similar in coloration, but with relictus slightly larger with a deeper, stubbier bill.

Orr and Ray (1945) compared Amaurospizopsis and Amaurospiza, concluded that the differences were not sufficient to warrant separate genera, and proposed that Amaurospizopsis be considered a synonym of Amaurospiza. Griscom, in a letter from 1944, concurred with Orr and Ray, mentioning that his views on avian genera had changed since he proposed the genus Amaurospizopsis.

Orr and Ray (1945) tentatively considered relictus as a separate species, mainly due to the geographic hiatus between relictus and concolor, and the absence of intergradation in the specimens they examined. They noted that in the length of the wing and the tail there is no overlap between A. relictus and A. concolor, although the length of the bill is the same for both species (Table 4). They reported that the color in the adult male of relictus is grayer and duller than the adult males of concolor.

Table 4. From Orr and Ray (1945), morphometric measurements of Amaurospiza seedeaters.

Miller et al. (1957) noted under A. concolor relicta: “Measurements of relicta … essentially bridge the size gap between this form and A. c. concolor of Central America and the color differences appear to be of a magnitude frequent in races.”

AOU/AOS considers relicta as a group within A. concolor (AOU 1983; AOU 1998). The sixth edition (AOU 1983) noted: “The two groups are sometimes recognized as distinct species, A. relicta (Griscom, 1934) [Slate-blue Seedeater] and A. concolor [Blue Seedeater]”. The seventh edition (AOU 1998) mentioned the two groups without referring to the possible recognition of two distinct species.

Some authors treat relicta as a separate species (Eisenmann 1955; Davis 1972; Howell and Webb 1995). The song of relicta is described as similar to concolor but slightly higher and faster (Howell and Webb 1995). Lentino and Restall (2003), considering bill shape, size, color, and song differences, suspected that relicta might represent a separate species from concolor.

HBW-BL split A. relicta from A. concolor based on the following rationale:

“[relicta] commonly treated as conspecific with A. concolor; differs (in this analysis rictal bristles and nostrils accorded equivalence of plumage characters) by its slate-blue vs. dark blue plumage in male (2); longer rictal bristles (Griscom 1934) (allow 1); operculate nostrils (Griscom 1934) (allow 1); shorter, deeper bill (allow 1); longer wing and tail (mean of 3 male tails 57.7 mm vs mean of 5 males 50.8; allow 2); “slightly higher and faster” song (Howell and Webb 1995) (at least 1).”

carrizalensis

The taxon carrizalensis was described by Lentino and Restall (2003) based on specimens collected on the river island Carrizal in eastern Venezuela. The authors measured their specimen series and specimens from the other taxa within the genus (Table 5). They found that carrizalensis has the longest bill and most pointed wing of all the taxa within Amaurospiza. Lentino and Restall diagnosed carrizalensis as “separable from other members of the genus by the density of coloration and black flammulations on the breast, overall size, wing formula, volume and shape of the bill, and general measurements”. Lentino and Restall suggested that carrizalensis should be considered a separate species, which was accepted by the SACC due to the large range disjunction and morphological differences from concolor and moesta (Proposal 74, https://www.museum.lsu.edu/~Remsen/SACCprop74.htm). Subsequently, the English name Carrizal Seedeater was adopted by the SACC (Proposal 92, https://www.museum.lsu.edu/~Remsen/SACCprop92.htm).

Table 5. From Lentino and Restall (2003), morphometric measurements of Amaurospiza seedeaters.

Additional notes that involve more than one of the five taxa

Hellmayr (1938) recognized three species: Amaurospizopsis relictus, Amaurospiza concolor (including grandior, concolor, aequatorialis), and Amaurospiza moesta.

Orr and Ray (1945) proposed that two races of Amaurospiza concolor should be recognized: concolor from Central America and aequatorialis from northern South America. They added that further collecting efforts may show relictus as a large, pale, northern race of concolor.

Monroe (1968) noted that, in Honduras, concolor is a rare resident of the Caribbean lowlands, where it inhabits open rain forests, forest edges, and second growth. Monroe examined a series of relicta and concluded that the Honduran exemplars of concolor are not conspecific with Mexican relicta. Honduran concolor and relicta differ in morphology and habitat, given that relicta inhabits mountain ranges. However, the currently known elevational range of concolor in Central America is 600-2500 m (Howell and Webb 1995), which includes elevations similar to those inhabited by relicta.

Paynter (1970) recognized two species: Amaurospiza concolor (relicta, concolor, aequatorialis) and A. moesta.

Lentino and Restall (2003) suggested that based on wing formula, plumage and morphological differences, and geographic distribution, aequatorialis could be a distinct species from concolor (Figure 5).

Notes from HBW-BL: “plumage and mensural differences are all minor, with the possible exception of the larger bill of carrizalensis; moreover, new records from Brazil as far N as Maranhão suggest that populations of Amaurospiza may generally be more widespread and less disjunct than range maps indicate, as seems often the case with bamboo specialists.“

Figure 5. Comparative wing formulae of Amaurospiza as described by Lentino and Restall (2003).

Howell and Dyer (2022) commented on the similarity in morphology and voice of the taxa within the genus Amaurospiza. However, they noted that relicta is a distinctive taxon, endemic to Mexico, and that it has been considered a separate species. They also considered concolor and aequatorialis to be conspecific.

New information:

Genetics

Bryson et al. (2014) studied the diversification of the “blue cardinals” across the New World. They generated multilocus sequence data from one mitochondrial gene (ND2) and three nuclear introns (ACO1, MYC, FGB-I5) and estimated time-calibrated species trees. The authors included four Amaurospiza taxa, all except for relicta. The mtDNA phylogeny recovered two main clades: a first clade consisting of concolor (concolor) from southern Mexico and Central America, and a second clade formed by the South American subspecies of concolor (aequatorialis), moesta, and carrizalensis (Figure 6). Therefore, Amaurospiza concolor was not recovered as a monophyletic taxon, although only one sample from concolor aequatorialis was included, and as noted before, concolor relicta was not included. Support for the node uniting moesta and carrizalensis was a middling PP ≥0.70, whereas support for all others was ≥0.95.

Figure 6. Relevant part of Figure 2 of Bryson et al. (2014), mitochondrial ND2 Bayesian phylogeny.

The multilocus phylogeny from Bryson et al. (2014) included a smaller sample size, one individual per Amaurospiza taxa: concolor A (concolor), concolor B (aequatorialis), moesta, and carrizalensis. However, the authors noted that it was not possible to obtain any nuclear data for concolor B (aequatorialis), and this individual was represented only by mtDNA. The divergence between Central American concolor and South American aequatorialis, moesta, and carrizalensis was supported by a PP ≥0.95 (Figure 7). The nodes within the South American clade had lower support (PP ≥0.70). Branch lengths in Amaurospiza were comparable to intraspecific divergence in Cyanocompsa parellina and C. cyanoides, although branch length between P. ciris and P. versicolor was shorter, and the branch length between Cyanoloxia and “Cyanocompsa” brissonii was comparable to Amaurospiza.

Genetic evidence, based solely on mitochondrial DNA, suggests that concolor is more distantly related to aequatorialis than the latter is to carrizalensis and moesta; therefore, the authors suggested that the geographically and genetically distinctive aequatorialis be elevated to species status (Bryson et al. 2014).

Figure 7. Multilocus *BEAST phylogeny from Bryson et al. (2014).

The IOC list split aequatorialis based on Bryson et al. (2014)

The SACC assessed the split of aequatorialis in 2016, analyzing the new phylogenetic information from Bryson et al. The split was rejected mainly because vocal data were not published (Proposal 728, https://www.museum.lsu.edu/~Remsen/SACCprop728.htm).

Areta et al. (2023) developed the first phylogenetic analysis that included multiple samples from each of the five taxa within the genus Amaurospiza. The mitochondrial gene ND2 was sequenced for all 19 ingroup samples, and three nuclear introns (ACO1, FGB5, MB) for a subset of samples (one sample per taxon, with the exception of relicta). ND2 and multilocus phylogenetic analyses confirmed the monophyly of the genus Amaurospiza, recovered A. moesta and A. carrizalensis as sister species, and supported the relationship of aequatorialis as sister to the moesta-carrizalensis clade, thus confirming the paraphyly of A. concolor (Figure 8). ND2 haplotypes of relicta were recovered as monophyletic, either within a polytomy of concolor haplotypes in the ND2 gene tree or as sister to concolor in the BEAST tree. The relationship of concolor + relicta (ND2) or concolor (multilocus) was recovered as sister to all the other taxa.

Additionally, Areta et al. (2023) estimated mean ND2 pairwise distances, showing that the distances between concolor and aequatorialis were greater (8.3%) than those between moesta and carrizalensis (5.7%). The two relicta samples diverged on average by 1.0% from nominate concolor. Importantly, the authors uncovered low levels of intraspecific genetic differentiation between geographically distant populations, which contrasts with the deep divergences between allopatric species. Divergence times estimated from ND2 suggest that the Central and South American groups diverged 6.1 Ma, that populations of relicta diverged from concolor about 1 million years ago, and that the differentiation of South American lineages started about 3.4 Ma (Figure 9).

Figure 8. Phylogenetic hypothesis of relationships within the genus Amaurospiza from Areta et al. (2023). (a) mtDNA and (b) multilocus datasets. Numbers on nodes represent maximum likelihood bootstrap (* 100%) / Bayesian posterior probabilities (* 1.0).

Figure 9. Bayesian phylogenetic reconstruction of Amaurospiza based on ND2 data from Areta et al. (2023).

Vocalizations

Boesman (2016), using songs available in Xeno Canto (XC), analyzed and compared the voices of concolor (including concolor and aequatorialis), moesta, and carrizalensis. The taxon relicta was not included; there are no songs available in XC or the Macaulay Library (only calls in XC). Boesman concluded that the “song of all three species is very similar, given the range of variation within each species”. He added:

“All basic sound parameters have a largely overlapping range (min. frequency, max. frequency, number of notes, note length, phrase length,...). Note shapes are also quite similar, with many about identical between species.

“Other features that may allow differentiation such as e.g. at start or end of a song phrase could not be found.

“It is probably impossible to assign any recording with a reasonable level of confidence to any species. A multivariate statistical analysis may allow to separate song of the different taxa (once more recordings become available), but in any case differences will be small, and will not lead to scores higher than e.g. 1 + 1 applying Tobias criteria.”

Sample of the sonograms included by Boesman (2016):

Notes from HBW-BL: “while relicta is here separated as a full species, the other taxa appear to be very weakly differentiated: available acoustic evidence reveals identical songs (Boesman 2016).”

Areta et al. (2023) performed a quantitative vocal analysis that included the five taxa within the genus Amaurospiza. They showed that vocalizations are quite conserved in the group, but that they also provide taxonomically useful information. The authors found consistent differences between the Central and the South American clades: the number of inflections/second exhibited a stepped pattern, with concolor and relicta on the lower end and carrizalensis, aequatorialis, and moesta on the upper end; the South American taxa averaged more inflections per note than concolor and relicta (Table 6). A linear discriminant analysis using nine acoustic variables correctly assigned all 62 songs to the correct taxon (but note that there were single recordings for relicta [most similar to nominate concolor] and carrizalensis [most similar to moesta]. The first linear discriminant consisted mainly of maximum frequency, peak frequency average of all notes per song, and song duration on the first three notes; this first linear discriminant separated the South American taxa from the Central American taxa (Figure 10).

Table 6. From Areta et al. (2023), quantitative characterization of the inflections in the songs of Amaurospiza seedeaters.

Figure 10. Linear discriminant analysis of songs of Amaurospiza seedeaters from Areta et al. (2023). Note the distinctive cluster of aequatorialis, and the placement of the single recordings of relicta (close to nominate concolor) and carrizalensis (close to moesta).

Recommendation:

Species limits in Amaurospiza seedeaters are a complex issue mainly due to their morphological similarity and allopatric distributions. Each of the five Amaurospiza taxa is considered a subspecies in at least one of the four global avian checklists, and each has also been considered a separate species at some point in history. Total evidence should be considered to reconcile the taxonomy of these seedeaters. They all are allopatric with no evidence of intergradation, show morphological differences, have similar but distinctive songs, and are phylogenetically closely related. The recent integrative study by Areta et al. (2023), which analyzed phylogenetic data, vocalizations, morphology, and plumage, suggested that four species should be recognized within the genus Amaurospiza: A. concolor (relicta + concolor), A. aequatorialis, A. carrizalensis, and A. moesta.

We present four separate subproposals to revise the taxonomy of Amaurospiza seedeaters:

(a) Split Amaurospiza relicta from Blue Seedeater A. concolor.

(b) Lump two subspecies of A. concolor (concolor + aequatorialis) and A. carrizalensis with A. moesta.

(c) Split A. aequatorialis from A. concolor.

(d) Lump the five taxa (relicta, concolor, aequatorialis, carrizalensis, moesta) as subspecies of Amaurospiza moesta.

Approval of subproposals (a) and (b) would reconcile NACC (and SACC) with HBW-BL. Approval of subproposal (c) would reconcile NACC (and SACC) with IOC and follows the recommendation by Areta et al. (2023). Approval of (a) and (c) would result in five species; conversely, approval of subproposal (d) would lump the five taxa in a single species, A. moesta.

We recommend the following votes:

(a) NO, different lines of evidence (genetics, plumage, morphology, and vocalization) suggest that relicta should not be given species status but considered a subspecies of A. concolor. However, Areta et al. 2023 recommend more rigorous studies on the taxonomic status of relicta.

(b) NO, neither Bryson et al. (2014) nor Areta et al. (2023) provided phylogenetic support for this lump. HBW-BL considers concolor, aequatorialis, carrizalensis, and moesta as subspecies within A. moesta, leaving A. relicta as a separate species. However, the two large clades (relicta, concolor / aequatorialis, carrizalensis, moesta) supported by phylogenetic data do not correspond with that classification.

(c) YES, all evidence support aequatorialis as a separate species from A. concolor, aequatorialis is more closely related to carrizalensis and moesta than to concolor, and also differs from the latter in having white underwing coverts (at least in adult males) and in song. Considering aequatorialis as a separate species requires a change in the geographic distribution of A. concolor to include only the area from Mexico to Panama (eliminating Colombia and Ecuador).

(d) NO, phenotypic and genotypic data do not support the lump of the five taxa within a single species.

English names:

Similarly to taxonomic treatment for Amaurospiza seedeaters, there is no consensus in English names among global avian checklists (Table 7).

Table 7. English names currently used for Amaurospiza seedeaters in four global avian checklists.

	Howard & Moore	eBird/Clements	HBW-BL	IOC
relicta	Blue Seedeater	Blue Seedeater (Slate-blue)	Slate-blue Seedeater	Cabanis's Seedeater
concolor	Blue Seedeater	Blue Seedeater (Blue)	Blue Seedeater	Cabanis's Seedeater
aequatorialis	Blue Seedeater	Blue Seedeater (Equatorial)	Blue Seedeater	Ecuadorian Seedeater
carrizalensis	Carrizal Seedeater	Carrizal Seedeater	Blue Seedeater	Carrizal Seedeater
moesta	Blackish-blue Seedeater	Blackish-blue Seedeater	Blue Seedeater	Blackish-blue Seedeater

The NACC proposal reads as follows: “Therefore, according to passing subproposals, please consider the following:

“• If (a) passes and relicta is separated, Slate-blue Seedeater could be used.

“• If (b) passes and concolor (concolor + aequatorialis), carrizalensis, and moesta are merged, the name Blue Seedeater could continue to be used or a new name could be proposed.

“• If (c) passes and aequatorialis is separated, Areta et al. (2023) suggested the English name Ecuadorian Seedeater because most of its range occurs in Ecuador, whereas the previously proposed name, Equatorial Seedeater, could suggest a lowland distribution rather than the montane range that the species occupies. If you vote YES on (c), please vote either for Ecuadorian Seedeater or Equatorial Seedeater.

“• If (d) passes and the five taxa become subspecies within A. moesta, a new English name should be proposed.

“• If (a) and/or (c) pass, the name Blue Seedeater could continue to be used for concolor or a new name could be proposed.

[Note from Remsen to SACC voters: no need to vote now, but preliminary opinions welcomed, A separate proposal that includes all the members of our English names subgroup is a better way to go once we have voted on species limits.]

Literature cited:

Angulo Pratolongo, F., J. N. M. Flanagan, W. P. Vellinga, and N. Durand. (2012). Notes on the birds of Laquipampa Wildlife Refuge, Lambayeque, Peru. Bulletin of the British Ornithologists’ Club 132(3):162–174.

American Ornithologists' Union. (1983). Check-list of North American birds. 6th edition.

American Ornithologists' Union. (1998). Check-list of North American birds. 7th edition.

Areta, J. I., Benítez Saldívar, M. J., Lentino, M., Miranda, J., Ferreira, M., Klicka, J., & Pérez‐Emán, J. (2023?) Phylogenetic relationships and systematics of the bamboo‐specialist Amaurospiza blue‐seedeaters. Ibis (in press).

Boesman, P. (2016). Notes on the vocalizations of Blackish-blue Seedeater (Amaurospiza moesta). HBW Alive Ornithological Note 388. In: Handbook of the Birds of the World Alive. Lynx Edicions, Barcelona. https://doi.org/10.2173/bow-on.100388

Bryson Jr, R. W., Chaves, J., Smith, B. T., Miller, M. J., Winker, K., Pérez‐Emán, J. L., & Klicka, J. (2014). Diversification across the New World within the ‘blue’ cardinalids (Aves: Cardinalidae). Journal of Biogeography, 41(3), 587-599.

Cabanis, J. 1861. Uebersicht der im Berliner Museum befindlichen Vögel von Costa Rica, Vom Herausgeber. Journal für Ornithologie 9: 1-11.

Davis, L. I. (1972). A Field Guide to the Birds of Mexico and Central America. University of Texas Press, Austin.

Eisenmann, E. (1955). The Species of Middle American Birds. Transactions of the Linnaean Society, Volume VII, New York.

Griscom, L. (1934). The ornithology of Guerrero, Mexico. Bull. Mus. Comp. Zool. 75: 367–422.

Hellmayr, C.E. 1938. Catalogue of birds of the Americas. Part XI. Field Museum of Natural History Zoological Series, volume 13, part 11.

Howell, S. N. G., and S. Webb (1995). A Guide to the Birds of Mexico and Northern Central America. Oxford University Press, New York, NY, USA.

Lentino, M. and Restall, R. (2003). A new species of Amaurospiza blue seedeater from Venezuela. Auk 12115(3): 6115115–61156.

Lopes, L. E., De Pinho, J. B., & Benfica, C. E. R. (2011). Seasonal distribution and range of the Blackish-blue Seedeater (Amaurospiza moesta): a bamboo-associated bird. The Wilson Journal of Ornithology, 123(4), 797-802.

Miller, A.H., H. Friedmann, L. Griscom, and R.T. Moore. 1957. Distributional check-list of the birds of Mexico. Part 2. Pacific Coast Avifauna number 33.

Monroe, B. L., Jr. (1968). A Distributional Survey of the Birds of Honduras. Ornithological Monographs 7. American Ornithologists’ Union, Washington, DC, USA. 458 pp.

Orr, R.T., and M.S. Ray. 1945. Critical comments on seed-eaters of the genus Amaurospiza. Condor 47: 225-228.

Pablo-Castillo, J. (2018). Nuevos aportes a la ecología e historia natural del semillero azulado (Amaurospiza concolor, Cardinalidae) en Costa Rica. Comité editorial.

Paynter, R. A., Jr., Ed. (1970). Checklist of Birds of the World, vol. 13. Museum of Comparative Zoology, Harvard University Cambridge, Massachusetts.

Ramos-Ordoñez, M. F., C. I. Rodríguez-Flores, C. A. Soberanes-González, M. d. C. Arizmendi, A. Jaramillo, and T. S. Schulenberg (2020). Blue Seedeater (Amaurospiza concolor), version 1.0. In Birds of the World (T. S. Schulenberg, Editor). Cornell Lab of Ornithology, Ithaca, NY, USA. https://doi.org/10.2173/bow.blusee1.01

Ridgway, R. 1901. The birds of North and Middle America. Part I. United States National Museum Bulletin 50.

Rising, J. D., A. Jaramillo, G. M. Kirwan, and E. de Juana (2020). Blackish-blue Seedeater (Amaurospiza moesta), version 1.0. In Birds of the World (S. M. Billerman, B. K. Keeney, P. G. Rodewald, and T. S. Schulenberg, Editors). Cornell Lab of Ornithology, Ithaca, NY, USA. https://doi.org/10.2173/bow.blbsee3.01

Sánchez, C., J. R. Saucier, P. M. Benham, D. F. Lane, R. E. Gibbons, T. Valqui, S. A. Figueroa, C. J. Schmitt, C. Sánchez, B. K. Schmidt, C. M. Milensky, A. García Bravo, and D. García Olaechea. (2012). New and noteworthy records from northwestern Peru, Department of Tumbes. Boletín de la Unión de Ornitólogos del Perú 7(2): 18–36.

Sharpe, R.B. 1888. Catalogue of the birds in the British Museum. Volume XII. British Museum (Natural History), London.

Rosa Alicia Jiménez, Terry Chesser and Juan I. Areta, February 2023

Comments from Remsen: “Yes to the proposal’s recommendations, i.e. YES to “(c)” and NO to a-b-d. But only because this seems to be the best overall taxonomy given the collection of weak data we have and the lack of stability in treatment of these taxa in various classifications. The proposal does a terrific job of assembling every bit of published information on the group, but even combining all this information, I see big problems. The voices may be separable in DFA space, are those differences really significant? DFA on the geographic dialects of a related species, Passerina cyanea, would likely be able to discriminate them as well. I lean more towards Boesman’s qualitative impression that the songs are very similar, remarkably so in my opinion given the great distances involved from one end of this group to the other and the gaps in distribution. Playback trials would seem to be a necessary step The genetic data are of interest, of course, but are weak in terms of sampling of genes (mostly ND2), taxa, N, and geographic sampling with populations. That’s not the authors’ fault – these are mostly hard-to-find, rare-to-uncommon birds – the studies did the best they could with the material available. The minor differences in bill size, body size, wing formulas, and coloration are all matched by intraspecific variation in these characters in broadly distributed polytypic species of other passerines and in themselves seem only to confirm that the taxa are phenotypically diagnosable and thus worthy of at least subspecies rank under the BSC. Nonetheless, some of these differences are roughly comparable to phenotypic differences in related taxa we rank as species, specifically Cyanoloxia cyanoides/C. rothschildii, for which we have better data.”

Comments from Areta: “YES to C, given the closer relationship of aequatorialis to South American taxa evidenced in both mitochondrial and nuclear markers, the plumage differences, and the vocal distinctions. NO to A, B and D. The songs of Amaurospiza are rather conserved, but not identical among taxa (note that there is just one recording for carrizalensis and one for relicta, that in our analyses were more similar to their closest relatives). However, note that relicta is apparently very recently diverged from nominate concolor (mtDNA only), suggesting that for the time being and until there is more data, the current evidence supports retaining it as a ssp. of concolor. I don´t see any evidence supporting the merger of any other taxon with moesta.”

Comments from Robbins: “Kudos to the authors of this proposal for pulling all the information together in a digestible format. With some hesitation, I vote YES to subproposal C, for recognizing aequatorialis as a species, and NO to the other subproposals.

“As Van pointed out, there are issues with saying anything meaningful about vocalizations beyond that they are very similar. Similarly, plumage and morphometrics don’t offer much insight into this taxonomically difficult complex. As we increasingly see, phylogenetics based solely on a single mitochondrial gene can be misleading, but given the rather large mean pairwise distances between aequatorialis and concolor, in comparison to related taxa, leads me to put more weight on that factor than any other. Hence, my vote to recognize aequatorialis as a species.”

Comments from Stiles: “The genetic data favor at least the split of Mesoamerican concolor from South American aequatorialis, carrizalensis and moesta and the vocal data clearly separate aequatorialis from moesta; the separation of carrizalensis from moesta is less clear, but given that SACC recognized it, I’ll tentatively continue to do so as well. So, (b) NO, (c) YES, (d) NO.”

Comments from Lane: A) NO, B) NO, C) YES, D) NO. This result seems best to reflect the molecular and voice data presented in the proposal. These taxa are all quite rare and difficult to document in my limited experience, so it will be a challenge to try to augment the dataset here to see if it will change the results. I do think “Ecuadorian Seedeater” would be an appropriate name for A. aequatorialis. Luckily, dealing with relicta is not our burden to bear.”

Comments from Bonaccorso:

“969a. NO. The data from Areta et al (2023) do not show much of a genetic difference between A. c. concolor and A. c. relicta. Given that “The song of relicta is described as similar to concolor but slightly higher and faster (Howell and Webb 1995),” I don’t see much of a strong argument to split them. Minor bill shape, size, and color differences may be just as those expected among subspecies.

“969b. NO. I don´t see how lumping A. moesta and A. carrizalensis makes sense; they show almost the same amount of genetic differentiation seen between A. moesta + A. carrizalensis and A. aequatorialis. Because morphological differences are overlapping and song data for carrizalensis in the LD1 is so scant, we need to go with the phylogenetic data.

“969c. YES. Clearly, A. aequatorialis and A. concolor are not monophyletic. So, splitting A. aequatorialis from A. concolor is the sensible option.

“969d. NO. For the reasons stated in the proposal (phenotypic and genotypic data do not support the lump of the five taxa).”

Comments from Claramunt: “YES to (c) treat A. aequatorialis as a separate species from A. concolor. This is the important piece of evidence provided by the molecular phylogeny, as pointed out in the previous proposal 728: A. aequatorialis is not sister to concolor but to the other South American clade, and is genetically and phenotypically (maybe in subtle ways) differentiated from all.”

“NO to a,b,d. No evidence supports these changes.”

Comments from Zimmer: “Revise the taxonomy of Amaurospiza seedeaters. This is, as several others have already pointed out, a tough one, given that sample sizes for genetic material and vocal material are very limited, and, in general, this entire group is pretty poorly known. But here goes…

A) “NO. Range disjunctions, non-overlapping morphometric characters, and possible subtle distinctions in vocalizations all point to continued recognition of relicta as a distinct subspecies of concolor, but lacking genetic data corroborating greater divergence and a better sampled, quantitative vocal analysis, I don’t think there’s enough here to support a split.

B) “NO to lumping concolor + aequatorialis and carrizalensis into moesta. This runs counter to the genetic data, my impression of vocal differences, and ecological and biogeographic considerations.

C) “YES to splitting aequatorialis from concolor, based on genetic data (demonstrating paraphyly of concolor if aequatorialis is included, and, the sister relationship of aequatorialis to the moesta-carrizalensis clade), differences in underwing covert color in adult males (seemingly an important signaling mechanism, based on how conspicuous the white underwing “flash” is in the field for those “seedeaters” that have it), apparent differences (no matter how subtle) in vocalizations, and biogeographic and ecological considerations.

D) “NO. Morphological/vocal data, genetic data, and ecological distinctions fail to provide support for lumping the five taxa, in my opinion.”