Proposal (148) to South American Classification Committee

Change linear sequence of families in Charadriiformes

Effect on South American CL: This would make minor changes in the placement of five families to make our sequence reflect recent genetic data.

Background: Our current sequence of families in the Charadriiformes is a conventional one, with "kind-of-rail-like" families first (Jacanidae + Rostratulidae), followed by the "shorebird" families, and ending with the "gull-like" families. Oddballs like Thinocoridae and Chionidae are placed between "shorebirds" and "gulls." Relationships among these families have been controversial -- see references at our SACC site.

Our current sequence =

Jacanidae
Rostratulidae
Haematopodidae
Recurvirostridae
Burhinidae
Charadriidae
Pluvianellidae
Scolopacidae
Thinocoridae
Chionidae
Stercorariidae
Laridae
Rynchopidae

New information: Three recent data sets (Ericson et al. 2003, Paton et al. 2003, Fain & Houde 2004) using DNA sequence data have produced largely concordant results, which in turn are reasonably consistent with the DNA hybridization data of Sibley & Ahlquist. The first two data sets both use the RAG-1 gene, so are not really independent (but it's comforting to see similar results); Ericson et al. (I have pdf if interested) also used an intron of the myoglobin gene. The third (Fain & Houde 2004; I have pdf if interested) used an intron of the beta-fibrinogen gene.

The results with decent bootstrap support that are consistent among the three studies and relevant to our sequence are as follows:

(1) the Charadriiformes consists of three major groups: (1) the Scolopaci [Scolopacidae, Thinocoridae, Pedionomidae, Rostratulidae, and Jacanidae]; (2) the Charadrii [Charadriidae, Recurvirostridae, Haematopodidae, Burhinidae, and Chionidae]; and (3) the Lari [Laridae, Rynchopidae, Stercorariidae, Alcidae, and Glareolidae]. This is also consistent with Sibley & Ahlquist (1990).

(2) Jacanidae and Rostratulidae are sister families (consistent with our current sequence as well as Sibley & Ahlquist).

(3) Thinocoridae is sister to Jacanidae + Rostratulidae (Sibley & Ahlquist have it as sister to other Scolopaci but still with Scolopaci)

(4) The Scolopacidae are sister to the other Scolopaci.

(5) Haematopodidae and Recurvirostridae are sister families (consistent with our current sequence as well as Sibley & Ahlquist).

(6) Charadriidae is the sister to Haematopodidae + Recurvirostridae (consistent with Sibley & Ahlquist).

(7) Stercorariidae must be ranked as a family if Rynchopidae and Alcidae are also ranked as families; Stercorariidae is not the sister to Laridae + Rynchopidae (in contrast to Sibley & Ahlquist).

The following result is consistent between Ericson et al. and Paton et al. but is unresolved in Fain & Houde:

(8) Chionidae and Burhinidae are more closely related to each other than to any other charadriiform family (except Pluvianellidae: see #9 below). Sibley & Ahlquist "almost" found that relationship -- they found: Chionidae + (Burhinidae + [other Scolopaci]). Ericson et al. also showed that this relationship was not just due to RAG-1 but also was obtained from myoglobin alone, so in this case we can treat this as evidence from two independent analyses.

The following result was obtained only by Paton et al. because the others did not have Pluvianellus:

(9) Pluvianellidae and Chionidae are sisters. This result received 100% bootstrap support, as did the node uniting these two with four species of Burhinus.

[There are some additional minor points that I have incorporated directly into our Notes.]

Of these 9 points, 5 (#1, #3, #6, #8, #9) are not reflected in our linear sequence.

Analysis and Proposal: I am always impressed by congruence among data sets. Even the conflicts with Sibley & Ahlquist are relatively minor, involving single node shifts in each case. To make our sequence reflect their combined phylogenetic hypotheses and to also cause minimum disturbance, I propose the following (with the changes in red):

Charadriidae (moved to front to reflect basal position; has added trivial benefit of having nominate family first)

Haematopodidae
Recurvirostridae
Burhinidae
Chionidae
Pluvianellidae
Scolopacidae
Thinocoridae
Jacanidae
Rostratulidae
Stercorariidae
Laridae
Rynchopidae

Recommendation: Because our linear sequence and classification should reflect phylogenetic data, and because the data appear solid, I will vote YES on this. Whatever problems there might be with this sequence, it is grounded in phylogenetic hypotheses and data and is certainly closer to the true phylogeny of the order than any other sequence currently in use. If this passes, I'll submit another proposal to recognize formally as suborders the three major groups.

References:

ERICSON, P.G.P., I. ENVALL, M. IRESTEDT, AND J. A. NORMAN. 2003. Inter- familial relationships of the shorebirds (Aves: Charadriiformes) based on nuclear DNA sequence data. BMC Evolutionary Biology 3 (14 pp).

FAIN, M. G., & P. HOUDE. 2004. Parallel radiations in the primary clades of birds. Evolution 58: 2558-2573.

PATON, T. A., A. J. BAKER, J. G. GROTH, AND G. F. BARROWCLOUGH. 2003. RAG-1 sequences resolve phylogenetic relationships within charadriiform birds. Molecular Phylogenetics Evolution 29: 268-278.

Van Remsen, December 2004

_____________________________________________________________________________________

Comments from Robbins: "YES. The data presented in the three studies support the changes proposed by Van, so I vote "yes" for the new linear family arrangement within the Charadriiformes."

Comments from Jaramillo: "YES - impressed by the congruence of data sets, and it makes a great deal of sense to me. I like it!"

Comments from Stiles: "YES. Genetic data seem well-substantiated, and the linear sequence should reflect phylogenetic patterns insofar as possible."

Comments from Silva: "YES. The congruence among the datasets is quite impressive."