Proposal (762) to South American Classification Committee



Treat Cranioleuca baroni and Cranioleuca antisiensis as conspecific


Effect on South American CL: This would lump two taxa that are currently treated as separate species into a single species.


Background: Cranioleuca antisiensis was described by Sclater in (1858) from Cuenca, Azuay, Ecuador and its subspecies palamblae was described by Chapman (1923) from Palambla, Piura, Peru. Cranioleuca baroni was described by Salvin (1895) from central Cajamarca, Peru (Huamachuco and Cajabamba). Later, Zimmer ( 1924) added Cranioleuca baroni capitalis from La Quinua (near Cerro de Pasco), Pasco, Peru, followed by the description of Cranioleuca baroni zaratensis by Koepcke (1961) from Bosque de Zarate, Lima, Peru. Researchers have long recognized C. antisiensis and C. baroni as being a species complex owing to similarities in plumage, vocalizations, and habitat preferences, particularly where their distributional boundaries apparently abut in central Cajamarca (FjeldsĆ & Krabbe 1990). All major taxonomic treatments of the group have maintained them as separate species, but often with the caveat that they may well be conspecific based on further collections and research (Remsen 2003). In 2002, Van Remsen examined the series of C. antisiensis and C. baroni at the Louisiana State University Museum of Natural Science (LSUMNS), which hosts the largest series of the complex, to provide a taxonomic recommendation to Tom Schulenberg for the forthcoming field guide to the Birds of Peru. Two quotes from this correspondence summarizing Van’s main conclusions are worth reiterating, “1. If there is any way/reason to rank [C.] baroni as a separate species, it is not evident from our series -- I'm not sure where to draw the ‘line’. 2.  Almost every locality has its own distinct phenotype…”. Based on this, Schulenberg lumped C. baroni and C. antisiensis in the Birds of Peru (Schulenberg et al. 2007) as C. antisiensis.


New Information and Analysis: Seeholzer and Brumfield (2017) examined morphological, plumage, and genetic variation of C. antisiensis-baroni using a series collected in 2010 and 2011 of 172 individuals from 19 populations spanning the geographic and environmental breadth of the complex (Figure 1). They found that body mass of C. antisiensis-baroni increases clinally almost threefold from north to south (Figure 2a) with individual extremes from 11.5 g to 31.0 g. The cline is remarkably smooth despite some outlier populations on the arid west slope of the Andes in Depts. Ancash and Lima, which were smaller than expected given their transect positions.

Clinal variation in plumage coloration and patterning is also obvious in a visual examination of the series (Figure 3). The southern populations are generally greyer with higher-contrast underparts than the northern populations yet with a smooth transition between the geographic extremes. Within this general trend, however, there is considerable within and among population variation. Seeholzer and Brumfield (2017) quantified plumage coloration and indeed found a strong correlation between north-south transect position and an individual’s plumage score (Figure 2b). However, this relationship was messier than for body mass.

The song of C. antisiensis-baroni, a staccato series of accelerating, descending notes, also varies clinally. GFS examined Macaulay Library and Xeno-Canto songs from across the complex. The songs are variable in length, speed, and acceleration, but these characters do not show any clear geographic pattern. However, the peak frequency (pitch) of the songs decreases from north to south clinally (Figure 4). Although there is no body mass data associated with these vocal data, it is clear that larger birds in the south have lower peak frequencies than smaller birds in the north, as predicted by the scaling of the syrinx with body size (Ryan & Brenowitz 1985). Like body size and plumage, there is no discrete break in this vocal character with which to diagnose C. antisiensis from C. baroni.

Genetic data further reinforce that there is no way to differentiate C. antisiensis from C. baroni. Mitochondrial data suggest the phenotypic cline of C. antisiensis-baroni formed relatively recently, with a maximum sequence divergence between the geographic extremes of 1.1% and a divergence date of ~460,000 years (Derryberry et al. 2011). Although there is variation at the mitochondrial locus ND2 that forms three clades, these clades show little geographic cohesion, with samples from regions normally considered to be C. baroni or C. antisiensis clustering together and samples from the same locality found in different clades (Seeholzer & Brumfield unpublished data). Still, Seeholzer and Brumfield (2017) found a clear signal of clinal geographic population structure among 5,154 single-nucleotide polymorphisms (SNPs) distributed throughout the genome. A principal components analysis of this SNP matrix revealed three spurs of continuous genetic variation consistent with a demographic scenario of isolation-by-distance (Figure 1a). These spurs were geographically structured, and the relative position of the individuals in PC space conformed to the spatial distribution of their respective populations (Figure 1b). Analysis of the SNP data with the clustering algorithm ADMIXTURE (analogous to STRUCTURE) corroborated this pattern of smooth transitions between ancestral populations, as pattern consistent with isolation-by-distance (Figure 1b).


Recommendation: Although Seeholzer and Brumfield (2017) did not address taxonomy in their study, the taxonomic implications were clear and corroborated Remsen’s assessment. The variation in morphology, plumage, song, and genetics exhibit clinal variation that bridge the phenotypic extremes that were described as C. antisiensis and C. baroni. With no discontinuities in any characters with which to distinguish C. antisiensis from C. baroni, their current taxonomic status as distinct species is not justified under any species concept. We recommend that these species be lumped under C. antisiensis, the name with priority. It makes sense to retain the English name, Line-cheeked Spinetail, for stability and because all populations exhibit the eponymous auricular streaking. However, we would like the committee to consider changing the English name to the scientifically evocative Clinal Spinetail, highlighting the striking pattern of clinal geographic variation exhibited by C. antisiensis.


Literature Cited:

Chapman FM (1923) Descriptions of proposed new Formicariidae and Dendrocolaptidae. American Museum Novitates, 86, 1–20.

Derryberry EP, Claramunt S, Derryberry G et al. (2011) Lineage diversification and morphological evolution in a large-scale continental radiation: the Neotropical Ovenbirds and Woodcreepers (Aves: Furnariidae). Evolution, 65, 2973–2986.

FjeldsĆ J, Krabbe N (1990) Birds of the High Andes. Zoological Museum, University of Copenhagen and Apollo Books, Svendborg, Denmark.

Koepcke M (1961) Las razas geográficas de Cranioleuca antisiensis (Furnariidae, Aves), con la descripción de una nueva subespecie. Publ. Mus. Hist. Nat. Javier Prado (Ser. A. Zool.), 20, 1–17.

Remsen J V (2003) Family Furnariidae (ovenbirds). In: Handbook of the birds of the world (eds Hoyo J, Elliot A, Christie DA), pp. 162–201. Lynx Editions, Barcelona.

Salvin O (1895) On birds collected in Peru by Mr. O. T. Baron. Novitates Zoologicae, 2, 1–22.

Schulenberg TS, Stotz DF, Lane DF, O’neill JP, Parker, Theodore A I (2007) Birds of Peru. Princton University Press, Princeton, NJ.

Sclater PL (1858) List of Birds Collected by Mr. Louis Fraser at Cuenca, Gualaquiza, and Zamora in the Republic of Ecuador. Proc. Zool. Soc. London, 449–461.

Seeholzer GF, Brumfield RT (2017) Isolation-by-distance, not incipient ecological speciation, explains genetic differentiation in an Andean songbird (Aves: Furnariidae: Cranioleuca antisiensis, Line-cheeked Spinetail) despite near three-fold body size change across an environmental gradie. Molecular Ecology, 1–18.

Zimmer JT (1924) New birds from central Peru. Field Museum of Natural History (Zoological Series), 12, 51–67.





Figure 1. Genetic variation and geographic distribution of C. antisiensis. (a) The first two principal components of the matrix of 5,154 SNPs across 172 individuals of C. antisiensis explained 20.3% of the genetic variance. Proportion of genetic variance explained by each axis in parentheses. Pie charts represent the individual ADMIXTURE assignment probabilities to ancestral populations at K = 7. (b) Species distribution model of C. antisiensis. Pie charts are population averages of ADMIXTURE assignment probabilities. Red dots with white centers are localities used in the MAXENT model that did not have genetic data. Locality labels correspond to Table 1. Red dashed line represents transect from orthogonal regression. The Central Andean Wet Puna ecoregion (brown shading) forms a biogeographic barrier to dispersal between the central Andean populations (10-15) and the SW slope populations (16-19).




Figure 2. Clinal variation in (a) body mass and (d) plumage color reproduced from Seeholzer and Brumfield (2017) Figure 2a and 2d. Points represent individuals and are overlain on the population means (larger grey circle) and standard deviations (vertical grey bars). Individuals from the SW slope are represented as hollow circles. The relationships for body mass were stronger when the SW slope populations are excluded (solid line) than among all populations (dashed line).





Figure 3. (a) Representatives of the phenotypic cline of Cranioleuca antisiensis-baroni taken from points along the north (left) to south (right) transect. b) Side-by-side comparison of the phenotypic extremes. Upper specimen from Cerros de Amotape, Tumbes (pop. 1). Lower specimen from La Quinua, Pasco (pop. 15).



Figure 4. Clinal variation in the frequency (pitch) of C. antisiensis-baroni song (Seeholzer unpublished data).


Glenn F. Seeholzer and Robb T. Brumfield, 19 December 2017





Comments from Stiles: "YES. A thorough analysis of a curious situation, and a good example of a study where genetic variation clearly trumps morphological evidence for NOT splitting a species!"


Comments from Areta: "YES. A great quantitative example showing clinality in another Andean furnariid."


Comments from Zimmer: “YES.  The thorough examination of multiple data sets from the length and breadth of the geographic distribution of the complex clearly establishes the clinal nature of the rather radical morphological variation.  Nicely done!”


Comments from Remsen: "YES.  I was on Glenn’s committee and know these data and this paper well.  The conclusion that a single species is involved is inescapable.  To me, this is THE most spectacular example of geographic variation in birds.”


Comments from Cadena: “YES. A beautiful example of geographic variation and of the perils of using specimens from extremes of the geographic distribution of taxa to make taxonomic inferences.”