Thursday, September 29, 2005

We've arrived

Hurrah! Creationists have got round to taking a pop at ecological theory.

What these folks are referring to is a recent paper in Science by Sean Nee at the University of Edinburgh and his colleagues that took issue with a body of theory created by Eric Charnov at the University of New Mexico.

Charnov's ideas are based around a concept called 'life-history invariants'. The concept is that, viewed proportionately, there are features of biology that are roughly consistent across lots of species. Such constants might include the proportion of resources invested in offspring, as a function of body mass; or the age at maturity, as a function of total lifespan.

This seems to show that organisms follow similar rules about how to put their resources into growth, development, reproduction and so on, and that that they have evolved similar responses to trading off say, growth against reproduction. This has an intuitive appeal, particularly to biologists minded to look for generalities.

Now, Nee and his colleagues are saying that these apparent invariants are really just an artifact of the analyses used to compare how one trait, such as age at maturity, varies with another, such as lifespan. They argue that the way these analyses have traditionally been carried out — and then used to support the idea of life-history invariants — makes the link between traits look closer than it really is, and that really species are much more variable in how they allocate their resources.

This does not mean that life-history invariants don't exist — although it certainly looks like a blow to the theory — just that we need to find new ways of looking for them. It also shows the many pitfalls that biologists face in handling their data. For example, there are about half a dozen different ways to do a regression analysis — the technique that the new work accuses the old work of mis-using — and many biologists lack the technical knowledge to pick the most appropriate one.

The new finding is similar to a debate that ran a couple of decades ago around another supposed general ecological principle, the self-thinning law, which claimed to show how the size of plants in an area declined as their population density increased. Here too, errors were made in analysing the data that made the relationship look stronger than it really was. Most plant ecologists now think the old self-thinning law is wrong, although a new version, based on how metabolic rate varies with size, has been suggested in its place — you can't keep generality down.

The (mildly) interesting thing about the attack is that life-history invariants aren't obviously contradictory to a Biblical view of biology in the way of natural selection - any more than gravity, or the double helix, is. But anti-Darwinians are desperate to jump on any hint of dispute among biologists, and hold it up as evidence that the whole edifice is crumbling.

p.s. The Guardian cartoon strip 'Dr Clever's World of Facts' today has the best description of intelligent design I've seen so far - 'Creationism with a website'.