Proposal (469) to South American Classification Committee

Species limits in Paroaria

Effect on South American CL: this proposal would elevate to species rank various forms within Paroaria that are currently treated as subspecies.

Background: Species limits in the genus Paroaria have always been somewhat arbitrary. Members of this genus are all adorned by red somewhere on the face or throat, and variably developed crest from none to well developed. The members differ in distribution or lack of black on the face, bill coloration and upperparts coloration (grayish to blackish). All are whitish below. Here’s what our current footnote states:

66. Paroaria gularis, P. baeri, and P. capitata form a superspecies (Sibley & Monroe 1990); evidence for treating them as separate species is weak (Paynter 1970a); Hellmayr (1938) suspected that P. capitata might best be treated as a subspecies of P. gularis, and Meyer de Schauensee (1966) suspected that baeri might also best be treated as a subspecies of P. gularis. The subspecies P. g. nigrogenis of Venezuela was formerly (e.g., REF) treated as a separate species from Paroaria gularis. Dávalos & Porzecanski (2009) also found evidence that nigrogenis was not most closely to P. gularis; they elevated all diagnosable taxa to species rank based on the phylogenetic species concept.

New information: Dávalos and Porzecanski (2009) published a paper based on molecular as well as a morphological analysis of the genus. They examined over 600 specimens measured and scored them for plumage characters. Molecular data included complete sequences of cytochrome b. Note that they did not have molecular data for xinguensis.

They characterized species based on morphological data, in particular whether characters shared by individuals were of a unique combination and restricted to a geographic region. When these unique characters were fixed in all individuals sampled, the population was considered diagnosably distinct and a hypothesized phylogenetic species. The molecular data was then looked at with at least two individuals per hypothesized phylogenetic species looking for unique fixed mutations for support of these as valid species.

They suggest separating the genus into 8 species based on these data:

coronata - 24 unique substitutions in the cytochrome b gene

dominicana - 33 unique substitutions in the cytochrome b gene

nigrogenis - 20 unique substitutions in the cytochrome b gene

xinguensis – no molecular data.

baeri - 11 unique substitutions in the cytochrome b gene, but since sequences of its sister xinguensis, are lacking, this may be an overestimate.

gularis - two unique substitutions in the cytochrome b gene

cervicalis - two unique substitutions in the cytochrome b gene

capitata - five unique substitutions in the cytochrome b gene

They stated the following: “Lumping of nigrogenis, cervicalis and gularis is not supported by genetic or morphological data.” This is perhaps the most clear-cut result in this study. The taxon nigrogenis does not appear to be closely related to gularis; it is well differentiated based on the molecular data, and well differentiated based on the morphological data. Note also that Restall et al. (2006) separated nigrogenis from gularis and note that the two are sympatric in extreme SW Venezuela with no evidence of hybridization. These data strongly suggest that nigrogenis is a valid biological species.

The relationship of capitata, cervicalis, and gularis is more problematic. All share a similar color pattern with red head, black throat and breast. The species capitata has an entirely yellow bill, unlike cervicalis and gularis. The form cervicalis has no black around the eyes, whereas gularis does have a small dark area around the eyes; this is essentially the main difference between the two and does not seem to be all that impressive to me. Vocal data would be good to have to compare. Furthermore, the type of cervicalis is apparently aberrant, further muddying this situation. The apparent fixed genetic differences are of interest, although only two cervicalis were sampled, a larger sample size would have been better. The range of cervicalis and gularis likely meet somewhere in W Brazil or NE Bolivia and sympatry should be looked for, and any evidence of hybridization. The plot of ML distances showed the gularis-cervicalis comparison to be intermediate between within-species variation and between-species variation, in other words more grey than black or white. At this point it does not seem clear to me that cervicalis warrants separation as a biological species. On the other hand, capitata is geographically isolated from cervicalis-gularis and shows a greater degree of differentiation morphologically and molecularly. Retaining it as separate seems reasonable.

Similarly, the division of xinguensis from baeri is weak. No molecular data were available for xinguensis to begin with. The form is diagnosable in that it has a black throat instead of the red throat of baeri, yet only 5 specimens in total for both of these taxa were studied. It seems very premature to separate xinguensis from baeri at this point.

I think that it is best to separate the issues into two parts:

A. Elevate nigrogenis to species rank. Masked Cardinal (used in Restall et al 2006) is an appropriate English name given that it is the only one with a bold black stripe through the face in this genus of “cardinals.”

B. Also elevate the following to species rank: cervicalis and xinguensis.

Recommendation:

I recommend a YES on A and a NO on B. Dávalos and Porzecanski regard all diagnosable units as species, using PSC, so there is really no evidence for any of the other diagnosable taxa being ranked as species under BSC.

Literature Cited.

Dávalos, L.M., Porzecanski, A.L. (2009) Accounting for molecular stochasticity in systematic revisions: species limits and phylogeny of Paroaria. Molecular Phylogenetics and Evolution 53: 234-248.

RESTALL, R., C. RODNER, AND M. LENTINO. 2006. Birds of northern South America. An identification guide. Christopher Helm, London.

Alvaro Jaramillo, October 2010

=========================================================

Comments from Stiles: A. YES – both gularis and nigrogenis also occur in NE Colombia at no great distance from one another; although I am not aware of any locality where both have been taken here, I can also detect no indication of intermediacy in any of the specimens here. B. NO – I agree with Alvaro that evidence for splitting these two is at present inconclusive.”

Comments from Nores: “A. YES. Aunque por distribución parece corresponder a una subespecie, los datos moleculares y el análisis hecho en la propuesta por Álvaro indican que se trata de una especie. B. NO. Coincido con Álvaro que no hay evidencias concluyentes como para elevar estas subespecies al rango de especies.”

Comments from Robbins: “A. YES. Both molecular and morphological data support this conclusion. B. NO, until additional data are presented supporting elevating these to species status.”

Comments from Remsen: A. YES. But only because of Restall et al.’s report of sympatry, of which I was not aware, and Gary’s report of near parapatry. Otherwise, I see no evidence in this paper, either morphological or genetic, that supports species rank for nigrogenis. That nigrogularis does not group with gularis might be a simple gene tree vs. species tree problem, given that only one mtDNA was sequenced. B. NO. The authors of the paper have documented that there are 8 diagnosable taxa based on plumage. Whether those diagnosable units are ranked as valid subspecies or PSC species is a matter of taste - - the results are still the same. Any use of comparative genetic distances for a single locus for assigning taxon rank is naïve. That there are fixed genetic differences in cyt b among taxa known to also have diagnosable, presumably genetically based plumage differences is of interest but how that relates to taxon rank is a mystery to me. If there were no fixed differences in cyt b, but fixed plumage differences, would the authors have arrived at a separate taxonomic decision?”

Comments from Zimmer: “A. YES, based on reports of sympatry and parapatry with gularis, coupled with diagnosable plumage differences that are on a scale with some of the differences between currently recognized species in the genus. B. NO. I see nothing in the paper by Dávalos and Porzecanski to indicate that cervicalis and xinguensis are anything more than diagnosable subspecies/phylogenetic species, which is how we currently treat them.”

Comments from Pacheco: “A. Yes. A julgar pela aparente ausência de espécimes intermediários entre os táxons nigrogenis e gularis, no sw. Venezuela e, possivelmente, no nw Colômbia – sou favorável ao status de espécie para nigrogenis sob o BSC. B. No. Em vista das informações disponíveis é prematuro considerar cervicalis e, sobretudo, xinguensis como espécies plenas sob o BSC.”

Comments from Perez: “Paroaria gularis and nigrogenis are largely divergent in both molecular and morphological characters; however, the key data supporting species rank was the following information:

“Note also that Restall et al. (2006) separated nigrogenis from gularis and note that the two are sympatric in extreme SW Venezuela with no evidence of hybridization.  These data strongly suggest that nigrogenis is a valid biological species.”

“Although Gary commented on another potential zone of potential parapatry without records of intermediate specimens, the SW Venezuelan zone of sympatry was the key issue.

“Following this decision, I was interested in doing a more in-depth research in the area to describe this zone of sympatry or find a potential area for a contact zone study between both gularis and nigrogenis. Together with Miguel Lentino, from the Colección Ornitológica Phelps (COP), we produced a map (see below) including all records of these species present in three of the most important ornithological collections in Venezuela (with specimens from SW Venezuela). To our surprise, we did not find any confirmation of sympatry; in fact, the closest records between both taxa were about 200 km from each other. We also requested information for many birdwatchers, who routinely travel around the area, and found no evidence for such claimed sympatry. We did find, however, two misidentified COP specimens collected at El Dorado, at SW Venezuela. Those specimens were labeled as nigrogenis whereas the rest of specimens in the same locality were gularis, potentially misleading description of biogeographic patterns. The question now is if we will keep our decision to elevate nigrogenis to species rank, based only on the potential parapatry in Colombia, and the large morphological (even in juveniles) and molecular differentiation between gularis and nigrogenis. At least, I think we need to discuss if such differences are enough to claim reproductive isolation under the biological species concept we use here (which I think will be difficult); if not, we might need to reconsider our decision.”

New Comment from Remsen: “Jorge’s comments and research leads me to change my vote to NO. The primary reason I voted for this was the alleged sympatry (and given what I perceive as limited meaningful differences among these parapatric taxa, I was always suspicious).”