Galloanserae: Anseriformes

Taxa on this Page

1. Anhimidae
2. Anseriformes

Two Approaches to Ducks

We are fortunate in that the basal anseriforms have twice been analyzed using refined cladistic methods. The two papers in fact appeared back to back in a single issue of the Zoological Journal of the Linnean Society in 1997. Ericson (1997); Livezey 1997). The two studies are themselves an interesting study in comparative approaches to the method and, perhaps predictably, come to sometimes very different results.

The good news is that the two studies agree on a good many specifics. In particular, they agree (a) that the anseriforms are monophyletic; (b) that the Anhimidae, Presbyornis, and the Anatidae form successive branches of the anseriform tree; and (c) that the Charadriiformes and Ciconiiformes are closely related but impossible to resolve among themselves. The bad news is that the two studies come to opposite conclusions about the relationship between anseriforms and galliforms. Ericson finds that the galliforms are the sister group to all other neognathous birds (i.e. all modern birds except ratites and their allies). Livezey concludes that galliforms and anseriforms are united in the chicken-duck clade, Galloanserae. It is particularly depressing that the varying results on this issue might be predicted from the academic gharana of the two authors. Livezey has spent many years opposing the Feduccia-Olson school. Ericson is a former student of Storrs Olson.

There is no possible question about the integrity or ability of any of these scholars. The problem is more fundamental. Phylogenetic taxonomy, like all phylogenetic analysis, ultimately rests on ideas about homology. Homology itself is a surprisingly slippery concept. If we say that two structures are homologous, do we mean they are created from the same gene products, the same regulatory sequences, the same embryological cell lines, or something else? What do we mean by "same"? What if a gene sequence is duplicated in the course of evolution? What if (as, for example, in the case of "bone morphogenetic proteins"), a series of gene products and regulatory relationships is conserved, but the ultimate structures are entirely different?

Thus homology itself is, as yet, a difficult and somewhat subjective attribute. When the species are closely related, the calls are relatively easy to make. Thus, the relationships within various avian groups are increasingly agreed by all workers. However, when the relationship is more distant, the element of judgment becomes more important. Worse, the mental pre-conditioning that creates our intuitive ideas about similarity and difference is almost all we have to fall back on in making the detailed decisions about homology which form the data for phylogenetic inference. If, for example, one thinks intuitively that ducks and chickens are similar, it is possibly because one's mind is sensitive to similarities of the kind that, in in the aggregate, unite the two groups. This may be the source of some of the continuing disagreement in this area.

Another interesting difference between these two particular studies is in their style. Ericson clearly loves the look and feel of his specimens, describing each character in some detail. He deliberately chooses not to use all possible characters, but selects those which show reasonable variation within the group of interest and appear to be independent of each other. In fact, he performs two studies: the first to place Presbyornis generally within the Anseriformes, and a second, with somewhat different characters, to find its precise place among the ducks. The computational analysis itself is discussed in a fairly perfunctory manner.

By contrast, Livezey loves the books and concepts. His list of references runs for twenty pages. He describes the choices made in his computer analysis with almost the same affection that Ericson lavishes on the bones. By contrast with Ericson, Livezey simply uses (as he states) as many characters as possible, many of them from the literature rather than personal inspection, and places them all in a rather compact appendix with minimal description.

In part, this difference in approach is determined by the fact that Livezey has been performing cladistic analyses of various portions of the Anseriformes for about 15 years. It would be redundant for him to repeat descriptive work he performed a decade ago. Still, one is left with the impression of two very different personalities: probably equally matched in enormous skill and perceptiveness, but whose judgment calls about homology are likely to be quite different.

Anhimidae

It would really be inaccurate to use the word "consensus" about anything in avian taxonomy. However, it is fair to say that the primitive and early position of a chicken-duck group is widely believed. This galloanseran clade is, in turn, believed to be the cornerstone of a large group of major bird orders, including ratites, ciconiimorphs, chradriiforms, and perhaps others. The anhimids are as close to an intermediate between chickens and ducks as one is likely to find, and they are thus obviously a key taxon.

For that reason, it is a little surprising to find that they are not much studied. There are only three anhimid species, all of which live in fairly inaccessible regions of South America. However, they are not seriously endangered at the moment and not all that difficult to find. Perhaps the problem is that anhimids are clearly rather unique and specialized birds in their own right, and not really anybody's current idea of the basic neornithine stock. Their strange specializations include: lack of uncinate processes, a sharp spur extending from the wrist, enormous feet, and feathers which do not grow in well-defined tracts. Even odder is their extreme pneumatization: essentially every bone is hollow and air-filled. In addition, anhimids have a layer of air-filled tissue between the skin and body wall. Anhimids are strong, soaring flyers with the big wings and substantial feathers necessary for that lifestyle. The net result is a bird the apparent size of a turkey, but which weighs less than half as much.

It is difficult to imagine what combination of circumstances might have led to this combination of features. Soaring flight is typical of predators, but anhimids are generally herbivorous. Lack of feather tracts is found in penguins, for example, which require heavy insulation; but the anhimids are tropical and generally terrestrial. Big, flat feet are usually found in nearly flightless ground birds or in raptors, but anhimids are active flyers, frequent perchers, and never hunt anything larger than a bug. On the basis of vestigial bill lamellae, Feduccia (1999) asserts that anhimids are shorebird filter-feeders who have retooled for a different mode of life. Certainly this is possible, but none of their specializations really demands this result.

One interesting speculation (although the possibility remains that I am simply making the confusion worse) is the anhimids also connect the galloanseran group to the gruimorphs. The large strong feet, downturned bill, soaring habit, and certain behaviors of the anhimids are shared by raptorial gruiforms as well. Shufeldt 1901) reproduces a plate of "Palamedea" (= Anhima) cornuata which, at least to the untrained eye, suggests a gruimorph as much as anything else. Unfortunately the plate cannot be reproduced here. However, a random sample of resigned relatives and bewildered business visitors expressed the opinion that it was in the area of a crane, ibis, cormorant or egret, as well as expressing the opinion that the author had probably lost his marbles, all of which are gruimorphs (except the marbles). Fortunately, taxonomy is not determined by Gallup poll. However, it would be interesting to see the hypothesis tested. Ericson (1997) includes some gruimorphs in his analysis and finds them more closely related to anhimids than are the galliforms. The thrust of his paper is that anseriforms are not closely related at all to galliforms. However, it is possible to reject his conclusion without rejecting his results. The speculation made here in fact goes some distance toward reconciling the results of Ericson with those of Livezey 1997). [The two studies are compared in more detail at Anseriformes.]

So, perhaps the anhimids deserve a bit more attention. There is no question -- that is, there is a true consensus -- that the anhimids are ducks and that they diverged from the main line of anseran evolution very early. There is also little dissension from the proposition that the ducks are a rather early group of birds. See, for example, 75th Abstracts -- Abstract 34. So, however the neornithine radiation began, the anhimids are as close as we can get to an acknowledged primitive group. Where exactly, do they stand with respect to the other elements of that radiation? ATW000729.

Descriptions

Anseriformes: ducks, geese, swans, & screamers.

Anhimidae: Anhima, Chauna (screamers).

Links: Dierentuin.Net Dieren Database Anhima; screamer - article from Britannica.com;  GeoZoo: Screamers; Anhima links to many South American sites, in particular); all about Anseriformes; World Birds Taxonomic List: Genera and species with citations.; AVIFAUNA; 75th Abstracts -- Abstract 34; c/screamer; Anhimidae references, in Spanish); BIRD STUFF AND CLADISTICS; Slider.com Encyclopeadia: screamer; screamer; Anhima cornuta (Spanish); Anhimidae.pdf; Anhimidae (Screamers); Anseriformes; Ducks, Geese, Swans and Screamers; Birds; screamer. The Columbia Encyclopedia, Sixth Edition. 2001.

Image: Chauna torquata from the Barcelona Zoo.