So, is Darwinian natural selection optimal?
Nair and I had an argument about whether natural selection optimizes resource utilization. We decided to settle it the adult way, by resorting to published research. The paper is this one. (If you don’t have access, I can send you the paper). I originally meant to only post this as a comment on the earlier post, but the paper was interesting enough that I decided to extend this into blog-post length.
The paper is called ‘Plant Height and Evolutionary Games’, and is a review of game-theoretic approaches to evolutionary processes in TRENDS in Ecology and Evolution, by Daniel Falster and Mark Westoby from Macquarie University, Sydney, Australia. The authors review 14 papers that analyse the problem of how tall a plant should be.
The first thing to be noted is that the best strategy for height is frequency dependent, in that the best strategy for an individual or species depends on what its environment is and on what its competition is doing. This makes the problem to be analysed a ‘game’, in a game-theoretic sense of the term.
The authors quote five main themes in the scientific literature on plant height. I’ll list three here, and leave interested people to look the paper up for themselves.
1) There is a definite cost-benefit relationship to be considered in analyzing plant-height strategies.
2) The analysis should also ultimately come down to looking at which strategy is more successful at reproduction.
3) Evolutionary strategies are individual-centric, not species/population centric.
The last one is significant. The optimum for vegetation would have been to stick to the ground, where every plant gets sunlight, all the resources go into either photosynthesis or reproduction, and none goes wasted into building huge, tall stems. The reason this doesn’t happen is because ‘the tragedy of the common’ kicks in and an arms race for resources begins. With the result that the genes that are selected aren’t the most cooperative genes*; the selected genes aren’t those that impose uniform height on the entire population, for example. The genes selected are those that are strongest individually. What this means is this: natural selection prefers a dwindling population with an increased proportion of stronger individuals to a growing population of weaker individuals.
If the resource to be divided is limited (only so much sunlight falls on a unit area of ground), has some internal scale (bright, not so bright, dark), and the costs involved in obtaining more of the resource also increase with the amount of resource required, it has also been shown that there will be a mixture of strategies of height.
The paper contains more than I have written down here. I leave the reader to make what they will of the remainder of the paper, and point this out about my argument with Nair:
a) Natural selection is not optimal. It isn’t optimal in one species, to say nothing about in a whole ecosystem. b) The reason trees don’t grow to as tall as they could, physically, has to do with (what appear to be) game-theoretic constraints (which can be said to be part of natural selection, just so Nair doesn’t bite my head off).
Daniel S. Falster, & Mark Westoby (2003). Plant height and evolutionary games Trends in Ecology & Evolution, 18 (7), 337-343 DOI: 10.1016/S0169-5347(03)00061-2