Proposal (8) to South American Classification Committee


Change linear sequence of orders to place Galliformes and Anseriformes after Tinamiformes


Synopsis: An accumulating body of independent evidence indicates that the (1) the Galliformes and Anseriformes are sister taxa, and (2) they represent the earliest branch in the living class Aves after the palaeognaths. These two orders have traditionally been separated by the Falconiformes and placed after a group of orders (Gaviiformes to Ciconiiformes) that almost certainly appeared later in avian evolution. To maintain these two orders in their traditional places in linear sequences obscures patterns of avian evolution.


Background: See Sibley-Ahlquist (1990) tome for an unsynthesized compilation of pre-1990 literature. I noted that although Beddard's 1898 review concluded that both (1) and (2) above were correct, this was not generally recognized, and it seems that the current linear sequence has been maintained in most classifications for most of the 20th Century. A study of cranial morphology led Simonetta (1963) to suggest a sister relation between Anseriformes and Galliformes. But I think it was Sibley and Ahlquist's results (1990) that provided the first prominently recognized genetic data for this relationship and the basal origin of the two orders. We did not make this change in AOU (1998) because we require multiple independent data sets to revise at the family/order level, especially one that affects linear sequence so radically.


New data: Now we have multiple independent data-sets that support the original Sibley-Ahlquist finding. Here's a quote from Cracraft (2001; Proc. Royal Soc. London 268: 459-469):


"There is compelling evidence from immunological distances (Ho et al. 1976), amino-acid sequences from conservative alpha crystallin genes (Caspers et al. 1997), DNA hybridization (Sibley & Ahlquist 1990), whole and partial mitochondrial gene sequences (Mindell et al. 1997; Van Tuinen et al. 2000), nuclear gene sequences (Groth & Barrowclough 1999) and morphological characters (Cracraft 1988; Dzerhinsky 1995, Livezey 1997; Cracraft & Clark 2001) that the Galliformes and Anseriformes are each other's closest relative (united in a group called Galloanserae) and are the basal sister group to all remaining neognaths, the Neoaves (see Cracraft & Clarke [2001] for a review)."


Two parts of this proposal require separate evaluation:


1. Sister relationship of Galliformes and Anseriformes: I am unable to find any recent data set that refutes this. In addition to the papers cited by Cracraft, Harshman's (1994) reanalyses of the Sibley-Ahlquist DNA hybridization data supported their original finding. Waddell et al.'s (1999) analysis of mtDNA sequences supported the sister relationship of duck + chicken (compared to rhea, ostrich, falcon, and passerine). Zusi and Livezey (2000)'s analysis of cranial morphology supported this relationship (and explained problems in previous analyses that did not find this relationship). Johnson's (2001) analysis of cytochrome b supported the Galliformes-Anseriformes clade to the exclusion of 703 species of birds from an exceptional range of families and orders. Müller and Weber's (1998) analysis of tongue musculature also supported the sister relationship of Galliformes and Anseriformes. Ericson's (1997) analysis of osteological characters was ambiguous: he was unable to corroborate their sister relationship but also was unable to refute it.


2. Basal position in Neognaths: Support for this is not as solid. Mindell et al. (1997), Härlid et al. (1998), Waddell et al. (1999), and Johnson et al. (2001) found that Passeriformes were basal to all Neognaths or even all living birds. These studies can be faulted, as the authors themselves often pointed out, for combinations of limited taxon sampling, rooting the tree with distantly related alligator sequence, or assuming equal rates of mtDNA evolution within lineages. The analysis of van Tuinen et al. (2000) showed very strong support for the Galloanserae as the sister to all other Neognathae. They used complete mitochondrial gene sequences from 3 genes from 41 taxa spanning a broad taxonomic range, and also sequence data from a nuclear gene from representatives of 32 taxa (including all avian orders); thus their analysis stands apart from the others in depth and breadth of sampling. They also explained why the paraphyly of the neognaths found by the Mindell lab was an artifact of poor taxon sampling. Also, Zusi and Livezey (2000)'s analysis of cranial morphology supported a basal position of the Galloanserae. These studies, in conjunction with the studies cited by Cracraft, in my opinion provide sufficient evidence for the basal position of the Galloanserae.


Recommendation: I am by no means an expert on higher-level phylogeny. With varying degrees of misunderstanding, I recognize that all the studies above can be faulted to varying degrees for incomplete taxon sampling, tree-rooting problems, failure to provide bootstrap support, use of only one analytical paradigm (parsimony vs. likelihood), problems in character-coding, etc. Yet, standing on the sidelines, my view is that it is statistically impossible that so many independent and heterogeneous analyses would arrive at the same conclusion (Galloanserae clade) by chance or artifact alone. (Besides, I like the best-tasting birds all bunched together, early in the sequence; if you've tasted ostrich or tinamou, you know what I mean.) I am less comfortable with the basal position of the Galloanserae, but because retaining the Galloanserae in one of the two linear positions currently occupied by Anseriformes or Galliformes is not supported by any data other than tradition, I think that moving them to follow the Tinamiformes is actually a conservative action, given that it has received major support.


V. Remsen, June 2001


Literature Cited (in part)

Dzerhinsky, R. Y. 1995. Evidence for common ancestry of Galliformes and Anseriformes. Courier Forsch. Senckenberg 181: 325-336.

Ericson, P. G. P. 1997. Systematic relationships of the Palaeogene family Presbyornithidae (Aves: Anseriformes). Zool. J. Linn. Soc. 121: 429-483.

Groth, J. G. & G. F. Barrowclough. 1999. Basal divergences in birds and the phylogenetic utility of the nuclear RAG-1 gene. Mol. Phylogenetics Evolution 12: 115-123.

Härlid, A., A. Janke, and U. Arnason. 1998. The complete mitochondrial genome of Rhea americana and early avian divergences. J. Mol. Evol. 46: 669-679.

Harshman, J. 1994. Reweaving the tapestry: what can we learn from Sibley and Ahlquist (1990)? Auk 111: 377-388.

Johnson, K. P. 2001. Taxon sampling and the phylogenetic position of Passeriformes: evidence from 916 avian cytochrome b sequences. Syst. Biol. 50: 128-136.

Livezey, B. C. 1997. A phylogenetic analysis of basal Anseriformes, the fossil Presbyornis, and the interordinal relationships of waterfowl. Zool. J. Linn. Soc. 121: 361-428.

Mindell, D. P. et al. 1997. Phylogenetic relationships among and within select avian orders based on mitochondrial DNA. Pp. 213-247 in Avian molecular evolution and systematics (D. P. Mindell, ed.). Academic Press, San Diego.

Mindell, D. P. et al. 1999. Interordinal relationships of birds and other reptiles based on whole mitochondrial genomes. Syst. Biol. 48: 138-152.

Müller, W. and E. Weber. 1998. Re-discovery of a supposedly lost muscle in palaeognathous birds and its phylogenetic implications. Mitt. Mus. Nat.kd. Berlin, Zool. Reihe 74: 11-18.

Van Tuinen, M., C. G. Sibley, and S. B. Hedges. 2000. The early history of modern birds inferred from DNA sequences of nuclear and mitochondrial ribosomal genes. Mol. Biol. Evol 17: 451-457.

Waddell, P.J. et al. 1999. Assessing Cretaceous superordinal divergence times within birds and placental mammals by using whole mitochondrial protein sequences and an extended statistical framework. Syst. Biol. 119-137.

Zusi, R.L. and B. C. Livezey. 2000. Homology and phylogenetic implications of some enigmatic cranial features in galliform and anseriform birds. Annals Carnegie Museum 69: 157-193.




Comments (from Carla Cicero to AOU CLC on same proposal): " Date: Mon, 11 Feb 2002 15:36:45 -0800 - I read the Cracraft and Clarke paper on "The basal clades of modern birds" over the weekend. The paper primarily analyzes - or re-analyzes - a suite of 44 morphologic characters, and discusses those as well as molecular data to address 4 major questions. Here's a summary, for what it's worth: (1) Monophyly of modern birds, using as outgroups Ichthyornis, Hesperornis, and outgroups outside of those fossil clades (they don't specify) - Strongly supported (99%) by 16 derived morphologic characters, 8 of which were unambiguously optimized on tree. They also discuss other morphologic characters that are now found to be synapomorphic at other levels. Three amino acid replacements in alpha-crystallin A are consistent with neornithine monophyly. (2) Monophyly of palaeognaths - Strongly supported (100%) by 5 derived characters, all unambiguously optimized on tree. They also discuss other morphologic characters (previously proposed for monophyly) that are found to be synapomorphic at other levels, or where polarity unclear. Four molecular data sets support monophyly of palaeognaths: immunological distances, 2 derived amino acid replacements in alpha-crystallin A, DNA hybridization, RAG-1 sequences (3) Monophyly of neognaths - Strongly supported (100%) by 11 derived characters, 6 of which are unambiguously optimized on tree. Two molecular data sets consistent with monophyly of neognaths include 2 amino acid replacements in alpha-crystallin A, and RAG-1 sequences. (4) Monophyly of Galloanseres - Strongly supported (100%) by 12 derived characters, 11 of which are unambiguously optimized on tree. They also review synapomorphies postulated by Livezey (1997) to support monophyly of Galloanseres (some, not all, incorporated by Cracraft and Clarke; others not included pending further comparative analysis within neognaths and outgroups). Molecular data that support monophyly include: immunological distances, DNA hybridization, DNA sequences from alpha-A and alpha-B-crystallin genes, RAG-1 sequences, unpublished mtDNA and RAG-2 sequences. Furthermore, 3 amino-acid replacements in alpha-crystallin A are consistent with a neognath clade that excludes Galloanseres, and a 5-codon deletion in RAG-1 also diagnoses this clade. They state: "Whereas a sister group relationship between palaeognaths and neognaths is strongly supported by a variety of data, the basal divergences of the neognaths has remained controversial. Recently, however, interpretations of both molecular and morphological data have begun to converge on the hypothesis that Galliformes and Anseriformes are sister taxa and are together the sister group of all other neognaths...Although no viable alternative hypothesis...has been put forth, a monophyletic Galloanseres has been doubted by some (e.g., Feduccia 1996). In particular, Ericson (1996) questioned the validity of many of the morphological characters listed above, and in a later study (1997) on the systematic position of Presbyornis placed Galliformes as the sister group of other neognaths and imbedded the Anseriformes deep within the neognaths in a clade also containing ciconiiforms. However, as already noted, Livezey (1997) examined many of the same taxa during a parallel study of Presbyornis, used a much larger morphological data set than Ericson (1997), especially for the skull, and concluded that the morphological evidence supported a monophyletic Galloanseres...Thus, a broad suite of data, from many independent sources, all point to a monophyletic Galloanseres. These data now place a substantial homoplasy burden on all other hypotheses..." They go on to talk about problems with rooting and taxon sampling in molecular studies. Basically, their conclusion is that we can confidently place Galloanseres as sister to a clade of other Neognathes, but relationships of the latter are still poorly understood. This is nothing new, but their paper is a nice summary of the data. Though Galliformes and Anseriformes may be sister taxa in a cladistic sense (i.e., supported by numerous derived morphologic as well as molecular characters), which makes them closest living relatives, they are not closely related. Nonetheless, in view of the fact that they have been evolving independently for a very long time, the number of characters shared by these groups is surprising - more so than the differences between them.