Proposal (#361) to South American Classification Committee

Place Cathartidae in their own order

Effect on South American CL: We currently treat the Cathartidae as an Incertae Sedis family. This proposal would elevate the family to the rank of order, placed next to our current Falconiformes.

Background: Our current SACC Note reads as follows:

"Recent genetic data strongly refute a cathartid-stork relationship (Cracraft et al. 2004, Ericson et al. 2006, Gibb et al. 2006, Slack et al. 2007, Hackett et al. 2008). SACC proposal passed to remove from Ciconiiformes. The monophyly of the Cathartidae is strongly supported by multiple data sets (e.g., REFS), and the family is sufficiently distinctive that fossil cathartids can be recognized as far back as the middle Eocene (e.g., Cracraft & Rich 1972). Proposal needed for retention in Incertae Sedis, placement in Falconiformes, or placement in separate, named order, Cathartiformes, as in some paleontological literature."

New information: Hackett et al.'s (2008) comprehensive analysis set new standards for sampling in terms of genes for comparative analysis of the Aves: 19 independent loci and 32 kb of nDNA sequences. Their study targeted resolving the deep branches in the avian tree; resolution among these branches has largely been intractable in previous analyses. Their results confirmed a lack of relationship of the cathartids to the Ciconiiformes (thus supporting our proposal 241) and generally support the traditional relationship of the cathartids to the Accipitriformes (Falconiformes here considered separate - that will be the subject of another proposal). However, the bootstrap support for that relationship is only 61%, below their 70% threshold. Although odds are still good that they will eventually fall out as sisters, technically the current data cannot confirm that the Accipitriformes are monophyletic if Cathartidae is included; in fact, collapsing nodes with less than 70& support the nodes collapse to a polytomy that would include a number of orders, including even the Passeriformes.

Discussion: NACC has already moved Cathartidae back to Falconiformes (sensu lato) and I will do a proposal to follow this if the current one is rejected. I propose that we treat the Cathartidae in their own order, Cathartiformes, as is the case in much paleontological literature. The rationale for this is two-fold: (1) the support for monophyly of the Accipitriformes is weak if Cathartidae is included, and (2) regardless of whether they are shown to be sisters, the Cathartidae deserve the rank of order, in my opinion, because they are diagnosable as a monophyletic group in the fossil record as far back as many or most taxa ranked as orders. If we were to combine orders anytime we found sister relationships, we'd eventually end up with one giant order, so there has to be an independent criterion. Rather than make this something having to do with tradition or some eclectic concept of degree of morphological distinctiveness, I suggest that we use a rough proxy for "time as independent unit", namely how far back in fossil record one can diagnose individuals as belonging to the group. Using the criterion as "mid Eocene or earlier" actually works pretty well for modern nonpasserines. Certainly, the paleo people are comfortable with ranking them as an order. With all appropriate caveats, a Google search today on "Cathartiformes" gets 3400 hits, so we're not using a novel classification.

Recommendation: YES, for elevating Cathartidae to rank of order, to recognize that the group is as old as most orders of modern birds and to acknowledge lingering lack of strong support as the sister group to Accipitriformes. Of minor note is that this proposal would also remove a family from Incertae Sedis status.

Lit Cit: See SACC Biblio

Van Remsen, June 2008

__________________________________________________________________________________________________________________

Comments from Stiles: YES. I agree that a time-based criterion for ordinal rank is at least as instructive as one based on "pure" morphology, especially now that we are aware of a number of morphological "odd bedfellows" (albeit not necessarily in the same order) like flamingos and grebes! Cathartiformes certainly qualifies here, and I do not favor placing the Cathartidae in Falconiformes especially since this weakens support for the latter as a monophyletic group."

Comments from Zimmer: "YES. This approach makes the most sense to me (for reasons already stated by Van and Gary), especially given the date of the split, and the weakened support for monophyly if we put them back in with Falconiformes."

Comments from Robbins: "Before accepting the "mid-Eocene or earlier" criterion, I'd appreciate the opinions from avian paleo experts, such as Dave Steadman. In addition, given the understandably contentious nature surrounding avian fossil identifications, I had questions about the claim that cathartids have been identified from the Eocene. I consulted with Univ. of Kansas vertebrate paleontologist Larry Martin about the Eocene cathartid skeleton and he stated that if it is based on the Eocathartes described by Wetmore, that specimen has been reidentified as a non-cathartid. Thus, it behooves us to confirm that there is indeed a cathartid from the Eocene before voting on this proposal."

Comments from Schulenberg: "NO. We probably should have a fuller discussion of the issue of "support for monophyly of the Accipitriformes is weak if Cathartidae is included." The study with the most data, Hackett et al., seem to lean towards a sister-clade relationship between Cathartidae and accipitriform raptors, although there are scant details in their paper. (I don't know when or if that team will get around to publishing on this question in more detail, but we can always hope. In the meantime, Hackett et al. have unanalyzed data on 10,000 more base pairs, and 30 more taxa - this work is ongoing.)

"Beyond that, I have reservations about relying so heavily on "paleontologists." There aren't that many paleontologists who work on ancient birds, and I'm not sure if I trust them. It wasn't that long ago that major figures in the field (Feduccia, Olson) were claiming some strange phylogenetic relationships, such as that half or more of all birds could be traced back to Charadriiformes, or that ducks, flamingos, and one (!) genus of avocets were related. Perhaps there's a difference between accurately identifying the fossilized tip of a tibiotarsus to an order, and inferring phylogenetic relationships based on a small mound of fossilized tips of tibiotarsi, but I need to be convinced that the difference is great enough to overcome my distrust of the entire field.

"All this is not to say that I know a bird order when I see one, only that I'm not sure that we want to put this question in the hands of "paleontologists."

Additional comments from Remsen: "Cathartid presence in Eocene is based on Diatropornis Milne-Edwards, 1892 (fide Paleobiology online data-base + confirming comment on that diagnosis from Emslie at same site, and a Cracraft paper in Condor on evolution of cathartids in Old World). You can all Google your way to the official citations, etc."

Comments from Jaramillo: "YES - Higher level taxon changes are always problematic, partially because it is quite a subjective matter. One person's order is another person's family etc. To me Cathartidae are rather unique, and at most they appear to be only distantly related to the Accipitriformes. The fact that they have bounced around between Ciconiiformes and Falconiformes is telling, this group does not fit neatly anywhere. My reading of the data is that they will come out to be sisters with the Accipitriformes; nevertheless they are an old lineage. I think it is a good resolution to give them their own order, and if data ever comes out to challenge this, then we can change then."

Comments from Cadena: "YES, albeit hesitantly. The Hackett et al. (2008) analyses show that Cathartidae and Accipitridae are sister taxa. Although support for this result is not great, it is consistent with the long-term status quo in which these two lineages are included in the same order, Falconiformes. Thus, in retrospect, perhaps the wisest thing would have been to leave Cathartidae in the Falconiformes and not place it incertae sedis. Now that this move has been made (and I realize I voted yes on that proposal) and the Cathartidae are unplaced, I guess recognizing Cathartiformes is OK, particularly considering that this rank has been applied to these birds in the past. Leaving Tom's valid comments on paleontology aside, I sympathize with Van's argument about the age of the vulture lineage. However, I don't think lineage age is a sufficient reason to make taxonomic changes; adopting this as a standard would require changing ranks for a whole bunch of clades, and is not strictly necessary to maintain a classification that is consistent with phylogeny. But well, that's a more general discussion that we might want to have at some other time. I would personally favor not changing higher-level classification unless they are obvious problems of non-monophyly."

Comments from Pacheco: "YES. Considero o critério paleológico uma boa ferramenta para arbitrar e calibrar o ranking em categorias taxonômicas. Um tratamento em nível de Ordem para o "New World Vultures" é reivindicado desde Henry Seebohm (1890) "Classification of birds; an attempt to diagnose the subclasses, orders, suborders, and families of existing birds. London: K.H. Porter. 54 pp."