Tuesday, February 24, 2015

Summary of that special issue

(The following is the tenth part of a series of posts on an Annals of the Missouri Botanical Garden special issue on “Evolutionary Systematics and Paraphyly”. All posts in this series are tagged with “that special issue”.)

Okay, time to wrap this overly long series up. This is what the special issue contributes to the discussion around paraphyletic taxa:

Stuessy & Hörandl, Evolutionary Systematics and Paraphyly: An Introduction, pp. 2-5.

Introduces the special issue and provides some background.

Lockhart et al., We are Still Learning About the Nature of Species and Their Evolutionary Relationships, pp. 6-13.

Does not actually try to make the argument that paraphyletic supraspecific taxa should be recognised but merely points out that there is sometimes ongoing hybridisation between very closely related species. The entire contribution appears to be based on the misunderstanding that phylogenetic systematics (AKA cladism) does not accept paraphyletic species, which is further based on the misconception that the term "paraphyletic species" has an actual meaning for sexually reproducing organisms, which, however, it hasn't. Where phylogenetic structure is absent things cannot be "-phyletic", be it mono or para.

Hörandl, Nothing in Taxonomy Makes Sense Except in the Light of Evolution: Examples from the Classification of Ranunculus, pp. 14-31. 

Argues that classifications accepting paraphyletic taxa are more informative than any other classification, and that the approach restricts the options for classification more than any other. I consider both claims to be false: Because the end user cannot know if a taxon in an 'evolutionary' classification is monophyletic or defined by some symplesiomorphy, the information content of such a classification is zero.

Or in other words, the misconception underlying the paper is that "using a lot of information when making the classification" translates into "the end-user can get a lot of information out of it", but that is not a given. In the present case, they cannot deduce the meaning of any individual taxon without backtracking to the original rationale of the taxonomist. The thing is, making that laborious backtracking process unnecessary is precisely the point of having a classification in the first place.

As for the second claim, because there are myriads of ways how a taxon can be circumscribed as paraphyletic and myriads of characters that one could consider 'important' enough to be used as a defining symplesiomorphy, this approach actually has the largest possible number of options for classification.

George, The Case Against the Transfer of Dryandra to Banksia (Proteaceae), pp. 32-49.

The main argument appears to be that Dryandra should not have been sunk into Banksia because the relevant studies had not sampled 100% of the species, some small mistakes were made by the authors, gene trees were incongruent in some irrelevant details, and so on. In other words, nitpicking to distract from the real issue, which is that Dryandra is conclusively known to be nested in Banksia. The author also suspects that Dryandra is polyphyletic, which if true would make the taxon unacceptable to most of his allies.

Stuessy, Paraphyly and Endemic Genera of Oceanic Islands: Implications for Conservation, pp. 50-78.

Argues that the enforcement of monophyly will make currently endemic genera disappear into more widespread genera, and that this would make it harder to conserve the relevant species. It is hard to interpret this contribution as anything but a 29 pages long appeal to base taxonomic decisions and the classification of biological diversity on political convenience. If it isn't I must have missed its point.

Ehrendorfer & Barfuss, Paraphyly and Polyphyly in the Worldwide Tribe Rubieae (Rubiaceae): Challenges for Generic Delimitation, pp. 79-88.

An extremely interesting review of the state of knowledge about phylogenetic relationships in a group that includes such well-known genera as Galium and Asperula. But although the authors want to mentally assign unknown and unavailable ancestors to one of their descendant clades, the classification they propose is nonetheless for all practical purposes an entirely phylogenetic one because those ancestors will not appear in it anyway.

Brummitt, Taxonomy Versus Cladonomy in the Dicot Families, pp. 89-99.

Reiterates an argument that Brummitt made in earlier papers: Linnean ranks and phylogenetic systematics are incompatible because classifying an ancestral species into a genus will make that genus paraphyletic to all the genera its descendants are assigned to except itself. This is true as far as it goes, but of course Brummitt took it for granted that one could ever know that one is faced with an ancestor as opposed to a side lineage, that fossils should be treated as ancestral as opposed to terminals, and that when faced with the incompatibility he argues for one should prefer the pre-Theory-of-Evolution concept of Linnean ranks over phylogenetic systematics. One can certainly challenge all three assumptions, to say the least.

Zander, Support Measures for Caulistic Macroevolutionary Transformations in Evolutionary Trees, pp. 100-107.

Although it also proposes the titular support measures, most of the paper is a mixture of criticism of phylogenetics and explanations of the author's own approach to systematics. The fundamental problem here is that Zander's methodology depends on considering some present-day organisms to be the ancestors of other present-day organisms, but well, they simply aren't. It is like claiming that my brother is my ancestor because he looks more similar to our father than I do. If, however, we sensibly conclude that two contemporary groups have a common ancestor in the past instead of one being the ancestor, the whole argumentation of the paper collapses immediately.

Liu & Viña, Pandas, Plants, and People, pp. 108-125.

I cannot see any connection to the topic of paraphyly. In fact it seems unclear why this paper has been published in that specific journal, and I half suspect that the inclusion of this paper in the special issue is down to some kind of database mix-up.


In summary, not counting the introduction and the Panda article there are seven serious contributions in this special issue advocating the recognition of paraphyletic supraspecific taxa. Of these, two do not actually appear to argue for paraphyletic supraspecific taxa at all and feel as if they are merely based on potentially easily resolved misconceptions. Of the remaining five, one uses a very distinctly non-scientific and political argument, and one argues that you shouldn't make any taxonomic changes the author doesn't like unless you have achieved an unreasonably high level of sampling and of confidence in the results.

Even apart from me disagreeing with the arguments of the final three, this special issue brings to mind an English saying I have heard. Something about barrels and their bottoms. Of course, one good argument would be enough in a case like this, but once more I cannot see it.

As before, the only one who got close is Richard Brummitt, but the problem is that his best argument also contains the seed of destruction for paraphyletic taxa: As our attention is drawn to the problem of classifying ancestors, we start to realise that they break the 'long branches' and significant 'evolutionary divergence' that paraphylists use as cleavage points when circumscribing paraphyletic taxa; evolution is gradual. And then we are soon drawn to the conclusion that a rank-free phylogenetic system is the only solution for a classification across the history of life on this planet. This is not what Brummitt wants, but it is the logical consequence of trying to accommodate ancestors in the classification.

No comments:

Post a Comment