Major transitions in evolution
Before going to Portugal to teach I attended the semi-annual meeting of the International Society for the history, philosophy and social studies of biology which was held this year in Montpellier, France. Along with Alejandro Rosas of the National University of Colombia, I organized and took part in a round table on the major transitions in evolution. Here are five questions I asked and what I had to say about them.
1) Should the topic be understood as “transitions” or additions i.e. as a change from one level of selection to another or as an addition of another level of selection?
I would argue for additions i.e. there is an evolving metapopulation of mitochondria in an evolving population of eukaryotic cells, an evolving metapopulation of cells in an evolving population of multicellular animals, an evolving metapopulation of modules in an evolving population of modular organisms and an evolving metapopulation of multicellular individuals in an evolving population of eusocial colonies etc. To employ a metaphor that has been used, such a point of view is neither deflationary nor inflationary but both simultaneously.
2) What is required for the emergence of an (additional) aggregate level of Darwinian individuals?
I agree with Godfrey-Smith that it is the emergence of entities subject to evolution by natural selection. However, rather than B, G and I (i.e. bottleneck, germ line and integration) I would argue that what is required is an additional aggregate developmental cycle, including but not restricted to a life cycle. That in turn requires a minimum of a complementary pair of life history strategies, each of which constructs the ecological conditions that favour, and in the transition or addition, come to induce (or whatever predicts them comes to induce) the other. So niche construction or eco-evolutionary dynamics are central. For example in a renewable environment, a cycle of plentiful resources which favour (and come to induce) growth, which constructs scarce resources, which favour (and come to induce) maintenance in the form of diapause for example, which constructs plentiful resources again – is a developmental cycle subject to selection, albeit not a life cycle, i.e. not progeneration in Griesemer’s sense.
3) Queller’s “fraternal” or what Nowak’s group has called the “staying together” case: is it attributable to kinship or ecology?
Staying together is better described as kin facilitated, because it most fundamentally requires an ecological explanation. If it’s hot and you need to lose heat it is better to be apart, but if it’s cold and you need to retain heat it is better to be together; if parasites are the greatest threat it is probably better to be apart, but if predators are the greatest threat it is probably better to be together; and so on. Such advantages of being small or large and there are many others, pertain to the advantages of having a proportionately greater ratio of surface area to volume for the small relative to the large or volume to surface area for the large relative to the small. The point is that kinship can facilitate either staying together or going apart depending upon ecological conditions. This implies that in the recent great debate between E. O. Wilson and (almost) everyone else in which he emphasizes the importance of ecology over kinship (defensible nests for eusociality for example), I have concluded that he is more right than wrong, including his 2012 description of genomes as chimeric (selected for at different levels) albeit not about some of the more fanciful extensions e.g. that humans are a eusocial species.
On the “staying together” case too, it is worth noting that a new life cycle it is not typically just a “staying together” but a cyclical staying together and going apart. For example cells in a multicellular organism reproducing asexually by multicellular propagules do not just “stay together” – they cyclically stay together and go apart – the propagules go apart from each other and if there are distinct reproductive and somatic cells, the reproductive ones go apart from the somatic.
4) Queller’s “egalitarian” or what Nowak’s group has called the “coming together” case: are sexual families additional Darwinian individuals and if so, what role do cooperation and conflict play in them?
Like Michod, I think sexual families likely are additional Darwinian individuals i.e. there is an evolving metapopulation of individuals in an evolving population of sexual families. However, sexual families too are not just a coming together but a cyclical coming together (fertilization) and going apart (meiosis) in haploids with a zygotic meiosis, or a going apart and coming together in diploids with a gametic meiosis. On the role of cooperation and conflict in these comings and goings, assuming an equal allocation to both at equilibrium, the simplest possibilities would be that in haploids for example, they either come together in conflict and go apart in cooperation, or come together in cooperation and go apart in conflict. The former (conflict then cooperation case) is implied by the traditional PBS theory of the evolution of anisogamy in which at fertilization, microgametes are viewed as reproductive parasites of macrogametes, but then recombination at meiosis is assumed, as it traditionally has been, to be mutually advantageous for whatever reason. The latter (cooperation then conflict case) is implied by the “market” or “trade” theory which Noë and Hammerstein originally suggested and I more recently have revived, and which might then be taken to imply that meiosis is a matter of conflict with genetic recombination a side effect of the mechanics of gene conversion rather than gene conversion a side effect of the mechanics of recombination. In a stunning recent paper, Laurence Hurst and some colleagues from China (Yang et. al. 2012) employed next generation sequencing techniques to determine that the great majority of recombination events in Arabidopsis (over 90% and probably nearer 99%) are gene conversion events. On the other hand, whether they come together in conflict or in cooperation, it is still possible that they go apart in cooperation – sexual families having discovered the advantages of risk reduction under uncertainty (like investing in index funds or dollar cost averaging rather than trying to pick stocks or time markets). What is unlikely I think is that both are a matter of conflict which would not pay back the two-fold cost of sex and would hence be unlikely to out compete asexuals.
5) Can it be said, as Bourke (2011) believes for example, that egalitarian or coming together transitions (or additions) require a “shared reproductive fate.”?
I doubt it. Assuming he means being confined inside the same ‘skin’ as say a vertically-transmitted symbiont, then I doubt it is necessary. I don’t know a lot about multispecies communities but occasional casual mutually positive interactions rather than regular cyclical ones between members of different species for example would not be sufficient. Nor do I think that “differential persistence” is sufficient – a slab of a hardwood persists longer than one of a softwood but are hardly Darwinian individuals (unless viewed as cultural ones). But I think it likely that theorists like Watson and Bouchard are still generally right. Some interspecific “coming togethers and going aparts” without remaining confined in the same ‘skin’ are likely systematic enough to indeed constitute additional Darwinian individuals. Positively non-additive fitness interactions at some stage at least sufficiently large enough to compensate for any negatively non-additive ones which may obtain at another would be a key requirement.