I must disagree with Larry Moran, who accuses the field of evo-devo of animal chauvinism — not that it isn’t more or less true that we do tend to focus on metazoans, but I disagree with an implication that this is a bad thing or that it is a barrier to respectability. Larry says we need to cover the other four kingdoms of life in greater breadth, which I agree is a fine idea. I would like to have a complete description of the genome of every species on earth, a thorough catalog of every epistatic interaction between those genes during development, a hundred labs working on each species, and a massive collection of papers for each one documenting every step and every protein and every variation in their development. I would like it tomorrow.

I think we all agree that that would be impractical. The question is how we will focus our research to maximize our use of limited resources, and get us useful answers that will lead us in productive directions. Larry is advocating maximizing our phyletic breadth by following organisms representative of the greatest amount of diversity. He is proposing this in opposition to the proposal from Jenner and Wills, who suggest a different strategy — and I find myself agreeing more with Jenner and Wills than with Moran.

Their suggestion is that instead of picking our model organisms by where they are on the phylogenetic tree, we ought to select them by their suitability for answering specific questions in evo-devo. If we want to “discover law-like generalities (nomothetics)” we need to emphasize the pursuit of conceptual themes rather than haring off to collect another organism in the stamp book. That sounds like a good path to intellectual respectability to me.

Along those lines, Jenner and Wills recommend several species that deserve expanded study; they all happen to be animals, but they all also happen to present illustrations of interesting concepts in evo-devo. They even offer a handy list of these central evo-devo concepts that we should be pursuing.

  • Developmental programming. Allometry of horns in the beetle Onthophagus nigriventris.

  • Developmental bias. Variation in body size in C. elegans.

  • Developmental constraint. Shell morphology in the gastropod Cerion.

  • Redundancy. Anterior-posterior axis development in Drosophila melanogaster.

  • Modularity. Sense organs in the cavefish Astyanax mexicanus.

  • Evolvability. In silico cell-lineage evolution.

  • Origin of evolutionary novelties. The sea anemone Nematostella vectensis (bilateral symmetry, triploblasty).

  • Relationship between micro- and macroevolution. The three-spined stickleback and Heliconius butterfly wing patterns.

  • Canalization and cryptic genetic variation. D. melanogaster phenotypic variation increase during HSP90 impairment.

  • Developmental and phenotypic plasticity, polyphenism. Ant caste polyphenism and caste determination by primordial germ cells in the parastic wasp Copidosoma floridanum.

Larry mentions that there has been a significant foundation of research in gene regulation that is of great importance in understanding development, work that has been done in viruses and bacteria. I agree that all of that is important — I’ve read Ptashne, too — and that understanding those seminal ideas from people like Jacob and Monod ought to be a required element of beginning workers in evo-devo. But the real question ought to be whether we can better address the concepts listed above with a test tube full of bacteria, or with the animal species described. I suspect the latter, but Larry can explain to me otherwise; I fear, though, that understanding modular interactions between fields of tissues requires an experimental subject that has tissues.

One complaint I would have with Jenner and Wills, though, is that while their concept-focused approach is laudable, it is a little peculiar that plants are so poorly represented in their list of model organisms, and that is an unnecessary exclusion. Plants ought to be excellent subjects for studying modularity and the origin of novelties, for instance, so I suspect some bias towards animal models that isn’t entirely based on their applicability to specific questions.

Still, it’s not enough to tell us that bacteria exhibit properties of gene regulation just like metazoans — the question is whether they exhibit higher-level properties of multicellularity, tissue organization, and organismal patterning that are the subjects of interest to people doing evo-devo. Try as one might, I just don’t see bacteria as being entirely useful subjects for studying the evolution of body plans, except maybe as an outgroup that lacks a body plan.

Jenner RA, Wills MA (2007) The choice of model organisms in evo-devo. Nat Rev Genet. 8:311-314. Epub 2007 Mar 6.


  1. #1 Torbjörn Larsson
    March 31, 2007

    Ah, the never ending argument about research strategies with finite resources.

    Yes, concentrating resources on some areas may make us miss things and distort the picture for a while. I remember Pharyngula’s posts on the non-general, compressed and tightly orchestrated development in Drosophila.

    But concentration, competition and feedback are good things (to use a PZism, for once) in science, at least in a developing field.

    Being a layman, I can only assume that Jenner and Wills review means that the field has matured enough to start to ask specific questions. It sounds like there are exciting times ahead here. I suspect PZ would wish he had more arms. Ooops, I forgot, he already does!

  2. #2 Torbjörn Larsson
    March 31, 2007

    Ooh, a fight!

    You know that the public always cheers on the small guys, don’t you? Animals may not be the cuddly crowd pleaser that it seems at first sight. :-)

  3. #3 Torbjörn Larsson
    March 31, 2007

    He practically invented the idea of patenting sequences to keep academia from doing research.

    I’m not well versed in the background of this, but off hand it seems patenting is done to attract funding and venture capital. Wouldn’t that, and the added market pressure, lead to faster, albeit differently constrained, results? When you do an agreement with a devil, willingly or not, it isn’t often only the devil gets something.

    Another question is if what is mostly (today) discovery should be grouped under a system concerning development and, traditionally, construction. Perhaps it has some moral problems. But it seems to work.

New comments have been temporarily disabled. Please check back soon.