Wednesday, May 20, 2015

'Monophyletic species' once more

Originally I had decided not to write anything on the "open letter to scientific community" [sic] from about two weeks ago.

Because the author not so much refutes but rather rejects the logic of the arguments of those he criticises, a response could at best be a rephrasing in different words of what I wrote in the paper that incensed him so. It consequently seemed pointless to write a reply, and indeed I would just refer again to that paper anybody who is interested in arguing about the information content of 'evolutionary' classifications, the feasibility of delimiting taxa based on long branches, and the relevance of a distinction between paraphyletic and polyphyletic taxa.

However, a few days later I picked the manuscript attached to the letter up another time and gave more attention than before to the second half, the one where the author's ire is directed towards the publications of Frank Zachos. There I found a section that motivated me to write something after all; not much, but something, simply because the section in question is so depressingly typical of much of the opposition to phylogenetic systematics:
Another subterfuge is to exempt species from the ban on paraphyly ("the concept of paraphyly does not apply to the species category"), so the "actual common ancestor is (or was) a species, but it does (did) not belong to any supraspecific subdivision of the descendant group" - again a desctructive [sic] (making classification cripple [sic], with millions - one for each "accepted" non-monotypic taxon! - of species "not belonging anywhere") and illogical "convention" designed only to defend the indefensibly harmful dogma
It contains two claims of interest.

The first is that we have millions of species that under a phylogenetic classification would not belong into any higher taxon. This does not fly because firstly we don't actually have them (do you know which fossil bones belong to the one species that is the most recent common ancestor of all birds? because I don't), and secondly there is no problem with assigning the species to the clade that is descended from it.

But yes, under the unrealistic assumption that we could ever have a 100% certain ancestral species and wanted to confidently treat it as ancestral then it could not be assigned to any of the subclades of the clade whose common ancestor it is. That is how it should be, and there would still be no problem under a rankless phylogenetic classification, although a problem would arise from the Linnean binary species names. The problem is well understood, and it is partly for this reason that some cladists argue for abandoning Linnean ranks.

The second claim is the one that really pushed my button: that the inapplicability of the concept of paraphyly to species is a "subterfuge" and an "illogical convention designed only to defend the indefensibl[e]".

I believe that Willi Hennig's Phylogenetic Systematics is of relevance here, and in particular its chapter II: Tasks and methods of taxonomy. The first figure of that chapter is the original of the following famous diagram which I found on the website of the USA's National Center for Science Education:

Note the difference that Hennig makes between tokogenetic relationships within (sexual) species and phylogenetic relationships between species. Terms ending in -phyly such as monophyly or paraphyly apply to phylogenetic relationships. By their very definition they cannot apply where there is no phylogenetic relationship, in the same way that colour terms like 'red' or 'blue' cannot apply to a sound.

I do not mention this because I believe that whatever Willi Hennig wrote is automatically correct. The uncritical Darwin worship of some 'evolutionary' systematists weirds me out - as if nothing has been discovered and no new argument developed since 1859! Likewise, Hennig's arguments have to be evaluated on their merits, and it would be unsurprising to find that he was wrong about some things.

But this book from 1966 does have some relevance to the claim that cladists only "designed" the non-applicability of paraphyly to species as an "illogical convention" to defend what had become indefensible. Willi Hennig invented phylogenetic systematics, and that convention was there from the start. Indeed calling it merely a convention is utterly misleading; the whole logic of phylogenetic systematics is built on the observation that different classificatory principles apply within and between sexually reproducing species.

Sorry for shouting, but this is important. This distinction is the heart and core of cladism and has been so from the very first minute. Anybody who thinks that the incoherent phrase "but species are paraphyletic" is an argument against phylogenetic systematics operates at the same level as the creationist who asks why, if humans are descended from monkeys, there are still monkeys, or as the physics crank who thinks that his totally unprecedented idea of a flash-light on a train moving at light speed disproves Einstein. One should at least try to obtain a very rough understanding of something before one publicly declares it to be obviously illogical.


  1. I’ve read through your excellent series on species concepts and systematics and am now persuaded to abandon the notion of “paraphyletic species.” However, your framing of the issue here will doubtless be seen by “evolutionary” taxonomists as an attempt to have it both ways: invoking reticulation as a trump card below the species level while ignoring it above the species level.

    Another way to frame the issue is that phylogenetic inferences become increasingly unreliable with larger and more recent gene flow between populations. Is this correct? For example, a population-level phylogenetic analysis using mtDNA found polar bears to be nested within some North American populations of brown bears. Some confused people (fortunately not the authors of the paper) interpreted this as “Ursus arctos is paraphyletic” when it was really an artifact of introgression, as revealed by discordance with trees from nuclear DNA. In this sort of situation are then three ways to interpret the species boundaries that are consistent with Hennig’s terminology, none of which have anything to do with “making species monophyletic”.

    A. Polar bears, North American brown bears, and Eurasian brown bears have a primarily tokogenetic relationship and should be considered part of the same species (in which case the species concept is so restrictive as to be useless for all practical purposes).
    B. The North American and Eurasian brown bears have a primarily tokogenetic relationship and should be considered members of the same species, while polar bears have primarily phylogenetic relationship to brown bears a whole (sister species), and a tokogenetic relationship to the North American subspecies (due to introgression).
    C. The North American brown bears should be considered a separate species, with a primarily phylogenetic relationship to polar bears (their sister taxa) and to Eurasian bears (the outgroup).

    Which taxonomy is most appropriate depends partly on the methods and assumptions used to deal with outliers in the samples and reconcile the discordant trees, which I suppose will always involve some degree of professional judgement.

    As for the “ancestor problem,” yes, this seems to be entirely a shortcoming of Linnean ranks. There was a specimen-level phylogeny of dinosaurs (Tschopp et al 2015) that made headlines for resurrecting Brontosaurus. The authors used apomorphy counts and a phenetic similarity index to determine whether to lump specimens into the same species or the same genus. The famous mount from the American Museum of Natural History fell outside the Brontosaurus + Apatosaurus clade and wasn’t quite similar enough to the closest specimen (B. excelsus) to be conspecific, and so was treated as an “indeterminate apatosaurine.” Based on my limited understanding of paper, this was a good example of a potential ancestor not belonging anywhere. But in a uninomial system, no problem, you’d just give AMNH460 a new species name and group it with the AMNH460 + Brontosaurus + Apatosaurus clade, which MAYBE include its descendents.

    Disclaimer: I’m not a taxonomist, and sorry for using only vertebrate examples.

    1. Thanks!

      I do not ignore reticulation above the species level, I just don't think it happens to the degree that it swamps out the phylogenetic structure of the tree of life. If it did, then that would be what could convince me to abandon phylogenetic systematics. Although sadly not in favour of 'evolutionary' systematics because in that case there wouldn't be paraphyletic taxa either.

      I have not really looked into the situation of the bears, but the first thing I would observe is that all too many people look at a single locus like mitochondrial DNA or, in plants, chloroplast DNA, and then try to conclude something about species relationships. As you pointed out, the next gene might tell a different story, and at least chloroplasts seem to introgress more easily than nuclear DNA.

      I really have no opinion on whether Eurasian and American brown bears are best treated as one or two species. Different situations may require different species concepts. I just think that one should classify by relatedness: If there is a branching order, then by closeness of the branches, and if there is massive reticulation then it would have to boil down to population genetics and phylogeography.

      I also don't have a preferred idea for how precisely do deal with fossil ancestors, which do appear to be reliably infer-able in rare cases such as unicellular plankton organisms with a well fossilising skeleton. I only work on extant stuff...