Proposal (277) to South American Classification Committee
Merge Pipile into Aburria
Effect on SACC: This would merge Pipile into Aburria.
Background & New information: Here's what we have in the Notes at our SACC website:
"The genus Pipile is merged by some (Delacour & Amadon 1973, Haverschmidt & Mees 1994) into Aburria. Pipile is currently treated as a separate genus in most classifications. Genetic data (Pereira et al. 2002, Pereira & Baker 2004) indicate that Aburria and Pipile are sister taxa, and this is supported by morphological data (Grau et al. 2005). New genetic data (Grau et al. 2005) indicate that Aburria aburri is embedded within Pipile, thus forcing the merger of Pipile into Aburria."
Grau et al. (2005) [let me know if you need pdf] analyzed DNA sequences from two mitochondrial genes and found that P. jacutinga was basal to monotypic Aburria and three other Pipile taxa (cumanensis, pipile, cujubi), with reasonably strong support values (bootstrap values of 72 and 74% for Maximum Parsimony and Maximum Likelihood analyses. A tree that forces Aburria to be basal to all four Pipile had a p value of only 0.32 (Shimodaira-Hasegawa test). On the other hand, Bayesian support for their tree was only 0.76, far below the usual 0.95 level regarded as "strong" for Bayesian support. (The Bayesian support for the node for monophyly of Pipile + Aburria was 1.00). All genetic samples were blood samples taken from aviary birds, with no specimen vouchers.
Grau et al. (2005) also analyzed skeletal characters for all the taxa, and although no formal analyses are presented, they found 5 synapomorphies that support the monophyly of Aburria + Pipile, and no differences between Aburria and Pipile.
Analysis and Recommendation: This is a tough one. On the one hand, this merger has been advocated on subjective grounds by several previous authors, and so even if the branching pattern is not quite right, the monophyly of the group (Aburria + Pipile) is further supported, and so the decision on where to draw the limits of a genus is subjective, not scientific. I see two reasons to vote NO. First, the Bayesian support values for Aburria being nested within Pipile are substandard (and I'm surprised reviewers didn't jump on this, given the emphasis in the title of the paper). Second, they sampled only 1 individual each of the Pipile and 2 Aburria; further, without vouchers, one could make a strong case that this is not science, because it cannot be replicated and the identification of the individuals sampled cannot be corroborated.
Although we do not have a P. jacutinga for comparison, looking at plates in HBW and elsewhere strongly suggests to me that the various Pipile are just geographic representatives of the same monophyletic group, and I find no published evidence that counters this. In fact, some authors have considered them all conspecific. If I am wrong about this, someone please let me know. Biogeographically, one could predict that the other Pipile would be more closely related to each other than any is to jacutinga, an Atlantic forest endemic. But that montane Aburria would be more closely related to the other Pipile than any are to P. jacutinga is so surprising, at least based on superficial phenotypic evidence, that I think a merger based only on the branching pattern of Grau et al. requires additional data. Therefore, I vote NO on this one, at least until someone makes a convincing case otherwise.
GRAU, E. T., S. L. PEREIRA, L. F. SILVEIRA, E. HÖFLING, AND A. WAJNTAL. 2005. Molecular phylogenetics and biogeography of Neotropical piping guans (Aves: Galliformes): Pipile Bonaparte, 1856 is synonym of Aburria Reichenbach, 1853. Molecular Phylogenetics & Evolution 35: 637-645.
Van Remsen, May 2007
Comments from Stiles: "NO. I agree that while Aburria and Pipile definitely form a monophyletic group, the evidence for sinking Pipile into Aburria is not convincing as it stands."
Comment from Thomas Donegan: "Van Remsen's argument above for the maintenance of the genus Pipile pending further investigation has some merit to it. One could also add to those arguments the fact that the bizarre whirring song of Aburria aburri is unique among the Cracids and bears almost no resemblance to Pipile songs (which are, generally, lighter whistles or "pipes"). As generic name use here could be regarded as including consideration of matters of taste, it is of note that the two names "Aburria" and "Pipile" derive from the calls of the species involved and are therefore eminently memorable in the field. A possible other criticism of the lumping of these genera is that a possible "third way" of erecting a new genus for P. jacutinga (presumably, some derivative of "Pipile" or "Jacutinga"; no such genus name exists to my knowledge) also perhaps bears consideration - with molecular, vocal and bare skin pattern characters possibly of use in diagnosis. If Aburria is regarded as a good genus on subjective morphological and vocal grounds, then it is of note that P. jacutinga seems on the basis of published analyses to have diverged at a similar time from other Pipile. Such an approach could maintain the stability of names for the other Pipile species/subspecies. Having said that, the distances between A. aburri, P. jacutinga and other Pipile look rather small (Grau et al. 2005), thus such an approach could itself be subject to criticism.
"Notwithstanding the above and Van Remsen's arguments, the Grau et al. study is consistent with the results of another molecular study. Pereira et al. (2002) also held Aburria and Pipile to be the most recently derived Cracidae genera, hypothesised to have diverged in the early Pliocene (one therefore presumes that P. jacutinga and other Pipile diverged around the same time?), with all other Cracidae genera diverging in late Miocene or earlier. Aburria and Pipile species are likely to be as closely related to one another as are species within other single Cracid genera, which could suggest that the maintenance of Pipile is inconsistent with the treatment of the rest of the family. Finally, it is noteworthy that Dan Brooks and his colleagues in the Cracid Specialist Group have recently changed their treatment and recognise the lumping of Aburria and Pipile (e.g. Brooks (ed.) 2006).
"The above is not a recommendation one way or another, but merely some thoughts that may be interest or assistance in coming to a decision.
Brooks, D. (2006). Conserving cracids: the most threatened family of birds in the Americas. Misc. Pub. Houston Mus. Nat. Sci. 6, 169 pp. http://www.cracids.org/AP_Engl_ebook.pdf
Pereira, Sérgio Luiz; Baker, Allan J.& Wajntal, Anita (2002). Combined nuclear and mitochondrial DNA sequences resolve generic relationships within the Cracidae (Galliformes, Aves). Systematic Biology 51(6): 946-958 http://individual.utoronto.ca/sergiolp/pdf/SB2002.pdf."
Comments from Laurent Raty: "I just wanted to make you aware that another recent study exists, that also suggests the merging of Pipile and Aburria. It has up to now only appeared as an OnlineEarly publication on the Synergy website, but I assume this should not be a problem to take it into account. This is: Frank-Hoeflich et al. (2007) http://dx.doi.org/10.1111/j.1439-0469.2007.00396.x."[and in response to Remsen's query about whether the DNA sample used was the same]:
"Yes, they are still unvouchered.
"Also, although the cyt b of many species was re-sequenced for the second study (this I think provides some level of corroboration, if the new sequences behave like the old ones), that of Pipile jacutinga was not – the sequence used in both studies is the same (GenBank AF165476). Hence, irrespective of whether the sequences were vouchered or not, the two studies cannot be viewed as independent regarding the placement of this species, which is critical for the monophyly of Pipile.
"I was a bit intrigued by the very high support of the clade [Aburria + other Pipile] in the molecular analysis of the second paper, as compared to the moderate support this clade received in the first one, so I downloaded the sequences from GenBank and checked them rapidly... Of the three genes in Grau et al.'s data set (cyt b, CR, ND2), the cyt b seems in fact the only one that really has a strong signal for placing jacutinga outside this clade. When this gene is removed, the CR and ND2 place P. jacutinga basal to the other Pipile - although with low/moderate support also – and Aburria basal to this group.
"IOW, the conclusion that Pipile is not monophyletic, in both studies, seems to rest entirely on this single P. jacutinga cyt b sequence. "The second paper has the merit to offer a more explicit analysis of the non-molecular evidence, I think. But, as in Grau et al.'s study, this evidence provides no basis to say that Pipile is not monophyletic. "That said, both papers partly advocate the merger based on other criteria than the paraphyly of Pipile, actually, such as the absence of osteological differences, and clearly established close relationship, with genetic distances lower than e.g. within Penelope."
Comments from Cadena: "NO. Given the overall similarity in external morphology among species currently placed in Pipile and their differences with Aburria, I believe the molecular result (let alone the low Bayesian support mentioned by Van) is an odd one that we should evaluate carefully. I see two problems, one is that owing to the lack of voucher specimens, the identification of the birds employed in the analyses cannot be corroborated, and thus one cannot rule out the possibility that the sample of Aburria used came from a hybrid (or from some sort of backcross), and we know that people that keep cracids in aviaries regularly hybridize them (remember the cases of Crax estudilloi and C. viridirostris). The other problem is a related one, the use only of mtDNA data in phylogenetic analyses. In the absence of unambiguously identified specimens, before making a taxonomic change I would like to see support from a nuclear gene so that we can reject the possibility that one (or more) of the samples used come from hybrid or "mitochondrially introgressed" individuals."
Comments from Robbins: "NO. Until we have a more clear cut data set, I'm for staying with the status quo."
Comments from Sergio Pereira, Anita Wajntal, and Erwin T. Grau: "We would like to take this opportunity to comment on Proposal #277 to the South American Classification Committee: Merge Pipile into Aburria.
"In our 2005 paper (Grau et al, Mol Phylogenet Evol 35: 637-645) we suggested that Pipile and Aburria should be merged not only based on our findings that Aburria aburri was embedded within Pipile according to phylogenetic hypothesis derived from multiple mitochondrial genes, but also as stated in Grau et al 2005: "1) 'The divergence time estimated for Aburria and Pipile based on nuclear and mitochondrial DNA sequence data was around 3.8 MYA (million years ago), well below the range of 8_10 MYA among genera of curassows or 10_18 MYA among other genera of guans (Pereira et al., 2002). Also, this divergence time is within the range estimated for species of curassows within Crax and within Mitu based on mitochondrial DNA sequences (Pereira and Baker, 2004a).', and
"2) 'TN+G distances ranged from 0.7 to 2.9% with a mean of 2.2% for the Pipile-Aburria group. These estimates are in the range for accepted species of guans of the genus Penelope (4.4%; Table 2 [in Grau et al 2005]), and corresponding cyt b and CR fragments of curassows of the genera Crax and Mitu (1.3_3.0% estimated by Grau et al., 2003).' "We believe that "reviewers didn't jump on [the relatively low support for placing Aburria aburri within the former Pipile], given the emphasis in the title of the paper" because they probably agreed with the three reasons given above. "Moreover, a closer association between Andean and Amazonian taxa in exclusion of taxa in Southern South America, Southeastern or Northern Brazil seem to be common in birds (Aleixo 2004, Evolution 58: 1303-1317; Burns et al 2004, Mol. Phylogenet. Evol. 32: 838-854), including other cracids (Pereira and Baker 2004, Auk 121: 682-694). Biogeographically, this outcome seems more plausible than placing Andean taxa as sister lineages to non-Andean South American taxa for groups assumed to have originated in South America.
"Hence, we stand to our view that Pipile should be merged with Aburria based on all the reasons given above.
"In an additional note, we regret that you feel that 'without vouchers, one could make a strong case that this is not science, because it cannot be replicated and the identification of the individuals sampled cannot be corroborated.'
"Although we understand the SACC's point of view regarding lack of vouchers, we would like to ensure you that members of the Laboratory of Genetics and Molecular Evolution of Aves (LGEMA) have carefully curated and stored all blood samples donated to us. Vouchers are not available because most of our blood samples were collected from live animals housed in licensed private breeding facilities and zoos throughout Brazil and some other Latin-American countries. Without these thoughtful contributions, our work on molecular systematics of cracid birds could not have been done as Brazilian museums only recently started tissue collections, and tissue samples for endangered taxa are usually unavailable. Unfortunately, we have no control that the animals which blood samples were taken would be donated to our lab or any museum at the time of their death. Moreover, all blood samples donated from Brazilian institutions were taken in the presence of at least two members of our group. Hence, any specimen misidentification can be ruled out. "Regarding the use of only one specimen per species, our decision was based on:
"1) preliminary studies that Sergio Pereira performed in the early days of his PhD studies back in 1996-1997 at the LGEMA in Brazil under supervision of Anita Wajntal show zero to less than two substitutions in cytochrome b, control region and 12S rDNA for several cracid species, including A. aburri and P. jacutinga; and,
"2) lack of chance to obtain multiple specimens for all species of piping-guans.
"We decided that we should concentrate getting more genes per cracid genus than few genes for many specimens. Funds for fieldwork were not available to us, and museum samples are rarely, if ever, granted to the LGEMA in Brazil because the lab is housed in an institution without tradition in fieldwork and, hence, we cannot offer any reciprocal benefits to international institutions. Moreover, all piping-guans are threatened or endangered, what largely decreased our chances of getting additional samples. "Although the Brazilian government has imposed more restrictive laws to collect, transport, import and export biological samples, anyone interested in studying the samples in Grau et al (2005) are welcome to do so in situ at the LGEMA located at the University of Sčo Paulo, Brazil."
Comments from Zimmer: "NO. As Van points out, the monophyly of Aburria + Pipile does not seem to be in question, and the decision of where to draw the limits of the genus is subjective. The various Pipile species, including jacutinga, are so similar in plumage patterns and vocalizations as to form a cohesive group, the members of which are all more similar to one another than any are to Aburria. This is consistent with my view of what a genus should be. Daniel's cautionary point regarding possible introgression in aviary cracids influencing the molecular results is also well taken."
Comments from Nores: "NO. Aunque sospecho que Grau et al (2005) y otros pueden tener razón que Aburria es congenéricas con Pipile, todavía no hay buenas evidencias genéticas como para demostrar que el plumaje y el canto de las especies de Pipile no son suficientes para separarlas de Aburria. Además, está el hecho seĖalado por Cadena de la posibilidad de híbridos en zoológicos y aviarios. Por otro lado, en dichos trabajos resulta claro que Mitu debe ser incluido en Pauxi.
"Nota: es llamativo que en las propuestas sobre Cracidae, nadie menciona la publicación de F. Vuilleumier.1965. Relationships and evolution within the Cracidae (Aves, Galliformes) (Bulletin of the Museum of Comparative Zoology 134 (1) que para mí fue un gran aporte en ese momento y constituyó el primer trabajo que intentó realizar un análisis biogeográfico y filogenético de las especies."
Comments from Jaramillo: "NO - These are clearly sister groups, however definition of genera are subjective. I don't go for the approach of comparing branch lengths on trees to define genera; at this point sinking Pipile into Aburria would generally decrease the information content of the genus in my opinion. I also worry about the possibility for aviary birds having mixed or odd parentage. We need more data on this one."