Proposal (281) to South American Classification Committee

Classification of the Falconidae

Effect on SACC: This would add subfamily structure to and rearrange linear sequence of Falconidae.

Background & New information: Our current classification has no subfamily structure, and the linear sequence basically goes from caracaras to Micrastur to Falco. Griffiths (1999) produced a phylogenetic hypothesis for the family based on mtDNA sequences and syringeal morphological characters that proposed two subfamilies, Falconinae (caracaras and Falco, each designated as a tribe) and Herpetotherinae (Micrastur + Herpetotheres); some of the basal nodes, however, did not have strong support.

Griffiths et al. (2004) analyzed nuclear DNA sequences (RAG-1). This analysis provided strong support for the basal nodes that divide the Falconidae into two subfamilies, with the same composition as in Griffiths (1999), although Griffiths et al. (2004) did not use subfamily names directly in their RAG-1 tree (Fig. 2).

Analysis and Recommendation: Our classification has the opportunity to increase its information content by using these phylogenetic data to recognize this deep division in the Falconidae. I propose that we formally recognize the two subfamilies, Herpetotherinae and Falconinae. The linear sequence that incorporates this structure, along with the branching pattern found by Griffiths et al. (1999), with the least amount of change to the current sequence, is as follows:

Falconidae

Herpetotherinae

Herpetotheres

Micrastur

Falconinae

Spiziapteryx

Caracara

Ibycter

Phalcoboenus

Daptrius

Milvago

Falco

The only weakly supported node (66% Maximum Likelihood bootstrap) is the one that links Daptrius and Milvago as sister genera (with respect to Phalcoboenus. The cyt-b tree (Fig. 3B), however, strongly supports this node (93% Maximum Parsimony bootstrap).

We could also recognize a major split within Falconinae with tribe designations (Falconini and Caracarini), but that should be a separate proposal. For now, I recommend a YES on the subfamilies and linear sequence.

References:

GRIFFITHS, C. S. 1999. Phylogeny of the Falconidae inferred from molecular and morphological data. Auk 116: 116-130.

GRIFFITHS, C. S., G. F. BARROWCLOUGH, J. G. GROTH, AND L. MERTZ. 2004. Phylogeny of the Falconidae (Aves): a comparison of the efficacy of morphological, mitochondrial, and nuclear data. Molecular Phylogenetics & Evolution 32: 101-109.

Van Remsen, June 2007

______________________________________________________________________________________________________

Comments solicited from Carole Griffiths: "The proposed classification of the Falconidae for the SACC looks good. I agree with you that classification based on phylogenies increased the information content."

Comments from Stiles: "YES, seems reasonable given the data. A question: If so, does this imply that we will be including subfamilies as a general policy in the Checklist? This could generate a whole flock of proposals, not least in the hummingbirds."

Comments from Robbins: "YES. I concur with Gary's concern about using subfamilies."

Comments from Zimmer: "YES, for reasons summarized by Van. This does make our classification more informative."

Comments from Nores: "SI, al ordenamiento (aunque no entiendo cómo fue hecho porque no coincide con ninguno de los árboles que aparecen en Griffiths 1999 y Griffiths et al. 2004). NO, a la inclusión de las subfamilias, porque eso significaría empezar de nuevo con un aspecto que no se ha tenido en cuenta hasta el momento. Incluso sacaría las sufamilias de las pocas familias que se les ha arbitrariamente puesto. Por ejemplo: Hydrobatidae, Anatidae, Laridae, Cuculidae, etc."

Comments from Stotz: "YES. I accept the linear sequence as a solid improvement over our previous arrangement.  The recognition of subfamilies makes sense to me, but does open the issue of whether we want to recognize subfamilies more generally.  It may be that we want them for situations like this with well-defined groups and deep nodes, but this could get messy it some families."

Comments from Jaramillo: "YES - In general, I agree, and this does add informational content. It is a good change. However, I would find it hard to believe that Milvago is not sister to Phalcoboenus. If there is an equally valid organization that would put Milvago and Phalcoboenus nest to each other in the linear sequence, well, I would give that the thumbs up."

Comments from Pacheco: "YES. Os resultados disponíveis e bem sumarizados aqui dão um suporte robusto a tais proposições."