Friday, 7 June 2013

George Williams contradicts Red Queen hypothesis [updated]

[Update: 15.07.2017: See also this article on the implications this ignored challenge by George Williams of the parasite red-queen theory has for science and its historiography: Joachim L. Dagg: How counterfactuals of Red-Queen theory shed light on science and its historiography. Studies in History and Philosophy of Biological and Biomedical Sciences 64: 53-64. The article will be freely available at this link up to 22 August 2017.]  

The Edge has a post with many evolutionary biologists remembering George C. Williams. In it Robert Trivers described a incident at an earlier memorial session commemorating William D. Hamilton suggesting a disagreement between Williams and Hamilton about the maintenance of sexual reproduction:
We [Williams and Trivers] last saw each other at the William Hamilton memorial session at Amherst in 2000 during the meetings of the Human Behavior and Evolution Society, at which both of us spoke. He was sitting behind me while others preceded us and I could hear Doris saying, "Now, George, don't do what you are thinking of. Just tell the stories you have about Bill, don't do it." So I was full of anticipation when George got up because I knew he was surely going to do exactly what his wife thought was a bad idea. George gets up and says "I wish Bill were here today, because I have a bone to pick with him".

And then he went and picked that bone for the entire talk. It had to do with the evolution of sex and patterns of evidence that George had pointed out years ago that contradicted (so George said) aspects of Bill's parasite approach. I thought it was wonderful. There were those that said it was inappropriate and why didn't he tell stories, but I thought it was perfect for the occasion, both vintage George Williams — no wasted motion with that organism! — and a tribute to the enduring importance of Bill's ideas.
Curious to find out what exactly that bone was, I fortunately found the proceedings of the Meeting for the Year 2000 [scroll down to the 12th conference, click on the "Program" link and a pdf should open] of the Human Behaviour and Evolution Society. Williams's contribution starts at page 128, but the crucial passage is on page 130:

No real scientists ever agree on everything, and Bill and I had a brief conflict last year at the Stony Brook conference. I am not convinced that adaptation by local pathogens to parental genotypes need be the major problem solved by sexuality. I think that the general unpredictability of offspring environments is what provides the main advantage. This issue is most appropriately settled not by modeling or data gathering but by consulting authorities. For a reliable insight on the significance of sexuality there are many appropriate authorities, but one that is especially clear is the strawberry plant (Fragaria). Offspring that develop immediately in the parents’ environment, with pathogens adapted to those parents’ genotypes, will not be sexually produced; whereas those that develop at variable times in the future, over a large range of habitats will be. The allocation of resources to sexual and asexual reproduction must be that which balances the two-fold cost of meiosis by the advantage of genetic diversity among widely dispersed seeds. 
[Update, 8.6.2013: The above quoted piece has also been re-published as Williams (2000) "Some thoughts on William D. Hamilton." Trends in Ecology and Evolution 15(7): 302.] 

The above, of course, relates back to Williams's book of 1975 (Sex and Evolution, Princeton University Press), in which the third chapter is devoted to organisms with a life-history that contradicts the Red Queen hypothesis. While the chapter is titled "The Strawberry-Coral Model," it is not about a mathematical or computer model, but about a purely conceptual (verbal) model. It could also be called a life-history pattern or scenario.

Strawberries, corals and other organisms corresponding to that life history pattern produce runners, polyps or the like asexually. This clonal progeny necessarily lives close to their parents and their supposedly co-adapted parasites , whereas the dispersal stages are produced sexually. This is the exact opposite of what the parasite Red Queen hypothesis predicts. The progeny staying close to the parents should be produced sexually, if recombination was a protection against co-adapting parasites.

Thinking about strawberry-coral type organisms, the potato, Solanum tuberosum, should be even worse for the Red Queen hypothesis than the strawberry or corals. Its sexually produced fruits are not only dispersed by fruit eating Vertebrates, but they are also rich in fungicidal and pesticidal Solanine. The asexually produced tubers, however, are not rich in Solanine as long as they stay underground and do not get green.

P.S.: Unfortunately, Hamilton was already dead when Williams picked this bone, and by now Williams is dead too. It would have been nice and worthwhile to see them duke this out.


  1. Nice post. I wasn't aware that Hamilton had formulated a specific parasite Red Queen hypothesis. That certainly explains why the current understanding of Red Queen is so different from its original formulation by Van Valen.
    I am not sure you can compare the runners and the seeds of strawberries, since they are very different structures. Constraints may dictate the dispersal capacity of the asexual and sexual progeny, rather than adaptive optima.

    1. Hi Corneel,
      You are right about the ambiguous use of the term Red Queen hypothesis.

      If I'm not mistaken, Bell (1982 "The Masterpiece of Nature") was the first to call Hamilton's hypothesis "Red Queen," because it takes a subset of the biotic factors that Van Valen included, namely parasites and pathogens. That is, it's a sub-hypothesis of Van Valen's. (But see my two posts 'On Testing the Red Queen hypothesis' here: and link at the bottom there.)

      It has caught on despite Van Valen's priority,and Hamilton himself comments on this situation in his autobiographical essays somewhere.

      Concerning runners and seeds, Williams does not say that these structures are similar. He only compares them insofar as they are offspring and either disperse or not.

      The parasite Red Queen hypothesis would have us expect that clonal offspring should disperse away from the parasites that have been adapting to the parental genotype and that the offspring staying close to the parents should be produced sexually. All he says is that strawberries, corals and many other organisms with that life-history patterns contradict this prediction of the parasite Red Queen.

    2. I have been wondering about the change in emphasis of the Red Queen hypothesis. Thanks for clearing it up.

      Re. runners and seeds. I understood that the different dispersal modes of sexual and asexual progeny are used as a test of the parasite Red Queen hypothesis. My concern is whether these comparisons are valid. Asexual reproduction is often accomplished by splitting, budding and the like. The resulting individuals may not evolve high dispersal rates as easily as seeds, even if this would be favorable. Other models may be more suitable (cyclical parthenogens maybe?).