Blute Blog

In 2016 I published an article on “density-dependent selection revisited: mechanisms linking explanantia and explananda” in Biological Theory. Levels of selection was discussed there briefly, but only to argue that when a new level is added, the old does not disappear but becomes part of the mechanism of development of the new. The point I would like to make here is the more general one that density-dependence is an ecological theory which can be applied to multilevel selection which makes it unlike any others that I am aware of.

To summarize briefly, the theory was that (assuming somatic and reproductive functions utilize the same resource and that they interact cooperatively) low density relative to resources i.e. plentiful resources favour consumption (eating and excreting) and producing many small offspring. High density relative to resources (i.e. scarce resources), on the other hand, favour digestion (breaking down and building up for maintenance, motility and mutability – the 3 M’s) and re-producing (producing a few large offspring capable of giving rise to grand-offspring). The reason for these is that the small and/or those with short, fast life cycles (the r selected) have a disproportionate surface area relative to volume while the large and/or those with long, slow life cycles (the K selected) have a disproportionate volume relative to surface area. The former is good for consuming and producing while the latter is good for digesting and re-producing. What does that mean for multi-level selection? Obviously in retrospect, it means that plentiful resources favour lower levels (individuals within groups) and scarce resources favour higher levels (groups themselves).

Viewed from the ecological perspective of density-dependent selection then, it is no wonder that members of flocks of birds, schools of fish, herds of deer etc. commonly hang out together waiting for tough times to improve, or migrate together, or sometimes innovate together – the 3M’s. Indeed, if density relative to resources and those relative to antagonists both matter and are positively correlated in time and space, it should not be surprising to see, as we do, that such flocks, schools, herds etc. repeatedly migrate together between good feeding and good breeding grounds.

On the other hand, major transitions in evolution such as a transitions from unicellular to multicellular organisms are cases in which this approach to multilevel selection may be difficult to justify. If we assume that new more inclusive entities have an outer membrane, the consumption (eating and excreting) of the small would have to take place at some cost to the large, and the digestion (breaking down and building up) of the large would have to take place at some cost to the small – all of which would make them different situations from cases in which the two are separate. However that may not be a fatal flaw – after all, for the low density case, some cells do die during multicellular development without the whole doing so and for the high density case, some large and/or long-lived multicellular organisms do under go senescence without becoming extinct.