Is sex in the details?

The architect Mies van der Rohe is supposed to have said `God is in the details'. I have always taken this to mean that life, substance and satisfaction are to be found (according to van der Rohe) in the concrete execution of a plan (that is, in the ways the particulars ®t together and interact), rather than in the grand conception itself, which is necessarily abstract and therefore vague. Sex has been the grand problem of evolutionary biology for two decades. West, Lively and Read (1999) bring to full consciousness a long-standing tension in thinking about sex. This is not the familiar tension between ecological and mutational theories of sex. Instead, it is a tension between purist and integrationist approaches to the whole problem. West et al. propose to change the terms of the debate in ways that could have interesting and therapeutic consequences. Surely many of us have long accepted that the Red Queen and the Grim Mutator both seem likely to play signi®cant roles in the maintenance of sex, yet we have also looked forward to a Decisive Answer in which one actor would prevail over the other. West et al. call attention to the inconsistency in this view. Those who obsess about sex tend to be zoologists. We easily forget that plants de®ned the problem. Many angiosperms are self-compatible hermaphrodites that can self-fertilize a little, or a lot, or any level in between. In addition, many perennials can reproduce vegetatively. For such species there are no qualitative developmental or genetic barriers to incremental (and in the end, profound) retreats from sex (see Bell, 1982). Thus, many species that remain fairly sexy must do so in the face of easy access to greater asexuality. Their addiction to varying but signi®cant levels of outcrossing should force even hopelessly unreconstructed zoocentrists to admit that ecology must explain some of the variance in rates of outcrossing and vegetative reproduction, and that for many species, sex isn't needed every generation (see Hurst & Peck, 1996). A smaller number of self-compatible hermaphroditic animals (West et al. mention the nematode Caenorhabitis elegans) illustrate the same point. This `balance argument' (Williams, 1975; Maynard Smith, 1978) was advanced to show that sex must be advantageous in the short term. It also shows that ecology must be part of the explanation, because populations or closely related species that differ greatly in effective outcrossing rates, as some do, cannot plausibly do so (at least not in general) because they differ greatly in their underlying mutation rates, which must usually be similar. Unconditionally deleterious mutations must also be important, and West et al. review several lines of evidence that support this view. An additional line of evidence derives from well-established differences between the ®xation probabilities of synonymous and nonsynonymous mutations (see Kondrashov & Crow, 1993; Crow, 1995; Drake et al., 1998; Eyre-Walker & Keightley, 1999). Synonymous nucleotide substitutions are typically about ®ve times more likely to ®x than nonsynonymous substitutions, on average, as estimated from comparisons between hundreds of orthologous genes in various taxa, especially rats and mice (e.g. Makalowski & Boguski, 1998). This implies that at least 4/5 of all mutations that change an amino acid must be deleterious. If mammals have about 50 000 genes averaging 2000 bp in length, then a typical mammal has around 10 functional base pairs. If even half of these nucleotides (5 10) were capable of mutating to deleterious states, and if the average mutation rate were around 4 10 per nucleotide per generation, then there would be 20 10 ˆ 0.2 deleterious substitutions per haploid genome per generation. This number may substantially underestimate the overall deleterious mutation rate in most mammals because the per generation nucleotide substitution rate is undoubtedly larger than 4 10 in many species (Drake et al., 1998), especially those with long lifespans (e.g. Eyre-Walker & Keightley, 1999), and there are other classes of mutations (e.g. indels, including transposon hops). So deleterious mutations must go at least some distance toward supporting sex in many taxa, even if (perhaps) they do not do so, by themselves, in very many cases. Given these well-known facts, how can anyone not be a pluralist? Why should West et al. feel compelled to argue the case? For one thing, it is necessary to establish that mutational and ecological factors may interact cooperatively to favour sex, and West et al. discuss this issue at length. At another level, it may also be necessary to relieve some physics envy. Simple, general, cleanly testable theories are beautiful. A universal process that could explain a pervasive pattern and that could be tested by a single decisive experiment would be a kind of Correspondence: Dr J. Seger, Department of Biology, University of Utah, 257 South 1400 East, Salt Lake City, UT 84112±0840, USA. Tel: +1 801 581 4758; fax: +1 801 581 4668; e-mail: seger@bionix.biology.utah.edu