Proposal (599) to South American Classification Committee

Revise classification of the Psittaciformes

Effect on SACC: This would recognize families and subfamilies within the Psittaciformes.

Background: Our current classification treats the Psittaciformes as containing a single family, Psittacidae, although we do not state this explicitly. The addition to our list of introduced Psittacula krameri and including it in the Psittacidae implies that we recognize only one family, so it is time to be explicit on this. We also do not recognize subfamilies within the Psittacidae.

I will not attempt a complete history of parrot classification, but it is safe to say that treatments have varied strongly in whether to recognize 1, 2, or more families, and which genera go in which families. Sibley & Monroe (1990) recognized only 1 family, for example. The Handbook of Birds of the World and Forshaw (1996, parrot identification guide; need to see what he did in his 2010 book) recognized only two, Cacatuidae and Psittacidae. AOU (1998) recognized only one family, but three subfamilies: Platycercinae (for Australian parrot group) and Psittacinae (for African and Asian parrots), and Arinae (for New World parrots).

New information: Joseph et al. (2012) synthesized all recent evidence, including molecular, paleontological, and morphological, to produce a consensus classification with formalized nomenclature. Greatly improved taxon sampling over the last 5 years or so has led to a much better knowledge of the overall relationships of parrots at the genus level: very few genera have not been sampled. Joseph et al. (2012) divided the order into 3 superfamilies, further divided into 6 families: Strigopidae (for the Kakapo, Strigops, of New Zealand), Nestoridae (for the Kea and Kaka, Nestor, of New Zealand), Cacatuidae (cockatoos), Psittrichasidae (for Psittrichas and Coracopsis), Psittacidae (for African Psittacus and Poicephalus plus all New World parrots), and Psittaculidae (for all remaining Old World parrots, which thus includes our introduced Psittacula):

Cracraft (2013) reduced this to four families: Strigopidae, Cacatuidae, Psittacidae, and Psittaculidae. Within the Psittacidae, Joseph et al. (2012) recognized two subfamilies, Psittacinae for the African genera, Arinae for the New World genera, and this was followed by Cracraft (2013) and implemented, therefore, by Dickinson & Remsen (2013). As for Psittacula, both classifications placed it in a separate family Psittaculidae, and in the subfamily Psittaculinae.

Discussion and Recommendation: The vast majority of the “problem” is extralimital to SACC. However, we are approaching outlier status in terms of implicit treatment of all parrots in same family.

As you all know, there are not formal, objective criteria for ranking groups at higher levels, so this is all largely subjective. Psittaciformes have been around since the Eocene, thus as long as most groups ranked as orders (with the first fossil that everyone seems to agree is a real parrot being Psittascopes). Schweizer et al. (2011) found evidence for multiple transoceanic colonization events in Psittaciformes, and criticized the previously used 85 MYA calibration point, and dated the age of the lineage at ca. 58 MYA (Paleocene), which would make it one of the oldest orders of modern birds.

Although in my simplistic viewpoint, parrots are remarkably homogeneous in terms of core morphology and fundamental bill shape, for them to have such a wide distribution, it seems likely that they are a fairly old group. If, as the genetic data indicate, the African and New World groups are sisters, then that implies a fairly ancient connection, either a Europe to North America hard connection on the Atlantic side or a Beringia connection on the Pacific side, but the latter would also require a much less parsimonious explanation of presence of the African lineages in Asia followed by total extinction there. A European connection seems much more likely – lots of African vertebrate taxa were once known from Europe. Looking at continental drift maps (http://www.odsn.de/cgi-bin/make_map.pl), it looks like a crossing would get difficult starting at 40 MYA unless warm climates extended to at least 75 degrees N, but the hard connection between the two was gone by 15 MYA. So, that’s a huge range of potential separation dates between the two lineages, but the point is that the two have likely been separated for a “long time.”

The point (?) of my long-winded outburst of speculation is that the two groups likely diverged somewhere 15-40 MYA, and this is in the range of taxa typically ranked at the subfamily or family level. Obviously, the age of the deeper branches has to be even greater, and so these various lineages of parrots are old, as old as many groups ranked at the family level. Schweizer et al. (2011) dating of the divergence of parrot groups is pasted in below:

Their estimate of the separation between African and American lineages is at ca. 30-35 MYA, within the range of the rough calculation above. Therefore, the New World parrots are an old lineage, probably older than some nonpasserine groups ranked at the family level, and thus in my opinion worthy of formal taxonomic rank.

In my view, the only hope for having objective criteria for ranking higher-level taxa is to have some “age” guidelines, i.e., for a group to be ranked as a family, the estimate of the age of the lineage would have to fall within certain broad limits, say at least early Miocene or Oligocene for “Family”. Whatever pitfalls there might be in such a scheme, at least it provides objective criteria and testable hypotheses, in contrast to the current scheme, which is basically a set of historical conventions. The good news is that almost all nonpasserine groups we traditionally label as orders or families have fossil records extending to Paleogene, or are sister groups to other groups with such fossils. Therefore, the turbulence caused by such a criterion would be minimal. As molecular dating gets more sophisticated and as the fossil record improves, the hope for adoption of such a scheme improves (and I’m contemplating whether to publish a little paper on this). But for now, I think that broad use of comparative lineage ages should at least be taken into account. Therefore, I am in favor of adding families and subfamilies to the Psittaciformes.

For the limited involvement of SACC with parrot classification, this proposal contains two parts:

A. Divide Psittaciformes into two families, namely Psittaculidae for Psittacula and Psittacidae for all New World parrots. I recommend a YES on this one. This follows Joseph et al. (2012) and Cracraft (2013).

B. Treat the New World parrots as a subfamily, Arinae, and treat our introduced Psittacula in another subfamily, Psittaculinae, as in Joseph et al. (2012) and Cracraft (2013). Note that this is actually independent of “A” above – one could have only one family of parrots, Psittacidae, as we do now, but divided up into multiple subfamilies. I recommend a YES on this one.

Literature Cited:

CRACRAFT, J. 2013. Avian higher-level relationships and classification: nonpasseriforms. Pp. xxi-xliii in The Howard and Moore Complete Checklist of the Birds of the World, 4^th Edition, Vol. 1. Non-passerines (E. C. Dickinson & J. V. Remsen, Jr., eds.). Aves Press, Eastbourne, U.K.

DICKINSON, E. C., AND J. V. REMSEN, JR. (eds.). 2013. The Howard and Moore complete checklist of the birds of the World. Vol. 1. Non-passerines. Aves Press, Eastbourne, U.K., 461 pp.

JOSEPH, L., A. TOON, E. E. SCHIRTZINGER, T. F. WRIGHT, AND R. SCHODDE. 2012. A revised nomenclature and classification for family-group taxa of parrots (Psittaciformes). Zootaxa 3205: 26-40.

SCHWEIZER, M., O. SEEHAUSEN, AND S. T. HERTWIG. 2011. Macroevolutionary patterns in the diversification of parrots: effects of climate change, geological events and key innovations. Journal of Biogeography 38: 2176-2194.

Van Remsen, October 2013

______________________________________________________________________________________________

Comments from Stiles: “As I understand it, part A would split the parrots we treat into two families, and part B would place each in a subfamily: Ariinae in the Psittacidae for the New World parrots, and Psittaculinae in the Psittaculidae. I see no problem with this, and vote YES if this is what is intended – however, as stated part B could be interpreted to mean that Psittaculinae should be a subfamily of Psittacidae, which would be in conflict with the recommendation in part A.”

Comments from Zimmer: “YES on both parts, assuming that the intent of Part B is to place the proposed subfamilies in two separate families as per Part A.”

Comments from Pacheco: “[YES to both] A adoção das subfamílias, implicadas nesta proposta, é consistente com as duas mais amplas e recentes classificações do grupo.”

Comments from Cadena: “599A. YES. 599B. NO. See my comments on proposal 591C; I think it would be best not to recognize subfamilies at all in our classification. The other alternative, recognizing subfamilies in all families, would be a rather difficult task, which I do not think we should embark on.”

Comments from Jaramillo: “A – YES. These groups have been separate for a long time; it seems consistent (compared to other groups) to rank them as families. B – Yes.”