Arseholes! Systematics, phylogenetics and HPS

There’s been some developments this day.

First of all a defunct blog on history and philosophy of science has revived with a new skin and as a group blog: AmericanScience: A Team Blog. I keep wanting to say “F&*k yeah!” It used to be the Forum for the History of Science in America. Some nice pieces there already (but horrible horrible font choices: I say this as a quandam graphic artist and compositor, okay?)

Second we have, courtesy of Softpedia (which turns out to be a useful science news feed) some developments in systematics and phylogeny. Or perhaps I should say, some phylogeny FAIL.

The first one is that – as the headline screams – ‘Mosaic of Life’ May Replace ‘Tree of Life’ - why? because it turns out genes cross taxonomic boundaries. Which we knew already. Gene trees are not species trees, and it is a category error to think that they should be the same. In fact, in order to identify lateral transfer of genes, we already need to have a species tree, or (to be consistent with my prior arguments) a relationship statement of taxa, which turn out to be something you can represent as a tree. This is on a par with the awful New Scientist article from a while back.

The second one is that a cricket Species Found Unchanged After 100 Million Years - except that the species are not the same, only the whole group, which is only the same if you ignore changes at the species level. The genus has remained the “same” for values of the “same” that include species evolution. And “genus” is an artificial construct at any rate, defined largely by traits that do not change much. Hence, the title of the press release should have read: “Things which do not change much by definition haven’t changed much, if you squint”.

Finally Bjørn Østman at Pleiotropy points out problems with a recent reassignment of Acoels (arsehole lackers) from a basal node on the evolutionary tree to a within-deuterosome node. The problem? Deuterosomes have arseholes, so this guy, which everybody wanted to be a kind of surrogate for the common ancestor of all bilaterans, is in fact a vastly reduced version of us. But…

They used microRNAs as their character set. They ignored, in other words, morphology and development, the traditional criteria for inferring relationships, in favour of a single test criterion. And this is because RNA is a “magic molecule”.

But there’s even a problem with Bjørn’s objection: why should we think that the morphology and development of these guys should be placed at the base anyway? The answer is simple: it makes evolutionary sense. No-arseholes come before arseholes, right? Well now this is an interesting question. How do we know that? Are we using hypotheses as evidence here? Why can’t a group be greatly reduced, evolutionarily? Even Darwin noted this among his barnacles. The presumption of what Hennig called transformation series is not supported by direct evidence, and so wanting these guys to fall anywhere in an evolutionary tree is fraught, although not, I think, beyond overcoming. It’s just that we had better be testing things as closely to the evidence as we can, and not rely on “what everybody knows”, or as J. B. S. Haldane called it, Aunt Jobisca’s Theorem.

<end of rant>

13 Comments

Filed under Evolution, History, Links, Philosophy, Rant, Science, Systematics

13 Responses to Arseholes! Systematics, phylogenetics and HPS

  1. Jim Thomerson

    One wonders how important horizontal transfer of genes is in eukaryotes. Do we have more than one example? A properly delimited genus is monophyletic and therefore quite real. The two I have erected are single species genera and therefore really real. What percentage of eukaryote genera are in fact monophyletic? Anyway, all very interesting.

  2. Horizontal gene transfer does not appear to be anywhere near as prevalent in eukaryotes as in prokaryotes, but it’s not completely unknown. There are indications that it has happened on a number of occasions in plants, for instance, which I suspect may be related to the existence of plants of plasmodesmata, open channels in cell walls allowing the passage of cytoplasm.

  3. Gunnar

    I have heard that Ascideans produce some fibers that may have originated from cellulose. Also, chironomid midges and certain oligochaetes in low-oxic environments share a haemoglobin-like molecule which to my knowledge is unknown from other protostomes-

  4. Gunnar

    Using horizontal gene transfer to explain these disjunct biochemistries is mere speculation, though, and should not be taken as anything else.

  5. Sam C

    On FoNtS (& speLLng):

    I say this as a quandam graphic artist and compositor, okay?

    To get full pseudo-pscientific respeck, you surely have to have been a quantum compositor?

    Sorry, of course, a Quantum Compositor with big initials.

  6. John Harshman

    So what you’re saying is that the assholes we save by investing in acoel flatworms can be spent commenting on evolutionary biology?

  7. Michael Fugate

    I think the rule is that if you get small enough, you can lose lots of things.

  8. John Harshman

    I see that Bjørn is being educated in the comments section of his blog. For one thing, it isn’t just acoels, but *all* “platyhelminths” that lack an anus, and even his favored phylogeny shows the other platyhelminths as protostomes (and lophotrochozoans, presumably), and so descended from animals with assholes. I put it down to a case of molecular-envy. Rejecting good molecular analyses because they contradict some preferred set of morphological characters.

  9. John, you’re most kind (about the being educated part, I mean).

    I don’t have molecular envy, but rather a fear of relying on microRNA for phylogenetic reconstruction. I don’t they they are as good as you imply. Also, I’m not closed to the idea that arseholes can get lost, but I am rather bent on explanations for it, rather than just saying that the molecular evidence implies it. [Full disclosure: I work in part on bacterial evolution using molecular phylogenetics.]

    • John Harshman

      I don’t know squat about micro-RNAs, but a character is a character. I’m suspecting that homoplastic gain of a micro-RNA is unlikely, which is good. Homoplastic loss is probably much more likely. And they’re working with few characters there. But they don’t have just the micro-RNA data, right? They have both nuclear and mitochondrial sequence data that point to the same thing. Three independent sources make it pretty convincing for me.

      I’m glad to hear that you aren’t suffering from molecule-envy. But be aware that you created quite a strong impression otherwise. We would all like explanations for evolutionary transformations, but rejecting a tree because it doesn’t fit a preferred scenario doesn’t help. And assholes appear to have been lost many times.

Leave a Reply