Tuesday, April 17, 2012

My attempt to teach the controversy: adaptationist vs stochasticist

This month, Tennessee enacted a law that would require education officials at the state and local level to

assist teachers to find effective ways to present the science curriculum as it addresses scientific controversies…[and] help students understand, analyze, critique, and review in an objective manner the scientific strengths and scientific weaknesses of existing scientific theories [such as] biological evolution, the chemical origins of life, global warming, and human cloning.

These “teach the controversy” laws are a nod and a wink to creationists and global warming deniers, denizens of that parallel world I visited earlier. Meanwhile, in the reality-based community, we have some really fun studies that engage real questions about evolution.

One of the long-standing tussles in evolutionary theory is about what pushes evolutionary change. In one corner we have the adaptationists: folks such as Richard Dawkins (and many microbiologists) who see good old-fashioned Darwinian natural selection as being the main force behind change. In the other corner we have the stochasticists: folks who argue that chance fluctuations and genetic drift in small populations are the main way that changes get established within a population (the blog Sandwalk is an example; to some degree, Stephen Jay Gould was of this school).

Evolutionary biologists of both stripes have long known that islands are prolific incubators of unique species. The Galapagos Islands are the classic example, but pretty much any island chain will have its oddities. My favorites are California’s Channel Islands, with the world’s second cutest fox species and the oxymoronic pygmy mammoth. Adaptationists will say that the unique, isolated environment of any island will drive evolution of founder species in idiosyncratic directions, producing species perfectly and uniquely adapted to every specific island. Stochasticists will say that island species are typically “founded” by a very small number of individuals, and since no individual perfectly represents a species, the differences between island species are due to these “founder effects.”

If you were a god, the question of who is correct here would be easy to solve—simply find some islands with distinctive (but related) species, and reverse time’s arrow until you arrived at the common ancestor of all the different types. Being merely human, though, you could do what a group of researchers from Harvard, Duke, and UC Davis did, and act like a god to a bunch of lizards.

When I was a kid, I was a god to a lizard; it was a Carolina Anole that we bought at the Los Angeles County Fair, where it was sold as a “chameleon” because of the species’ ability to change color from muddy brown to green. Anoles are native to the islands of the Caribbean, where, over millions of years and hundreds of islands, they have diverged into many different species. Why so many different species? Are the individual islands all so different that they select for unique adaptations, or are we seeing the effects of chance events in small populations?

To play god for these lizards, the researchers took advantage of an "act of god". A hurricane had eliminated the lizard population of seven tiny islands in the Bahamas. The researchers repopulated each of these lizard-free islands with an anole Adam and Eve, chosen at random from a large and well-established population on a nearby island.

For the stochasticists, this is a promising start. Given that each island’s Adam and Eve were randomly chosen, some of the islands would start with larger-than-normal founders, and others would start with smaller-than-normal founders. Stochasticists would believe that subsequent generations would bear the signatures of their founders, and that you could tell many generations later whether an island’s population had been founded by a relatively large or small Adam and Eve.

What about the adaptationists? The godlike researchers had something to satisfy them as well. It’s pretty well established that small shrubs—like those found on the new islands—provide a selective advantage to anoles with shorter limbs. The Adams and Eves chosen to populate these islands came from a population that had evolved with big trees, and so tended to have long limbs. Thus, all the new populations were subject to selection pressure to make their limbs shorter with every generation. Adaptationists would believe that, after a few generations, all the anoles on all the islands would have the same short limbs, regardless of which Adam and Eve founded their island’s population.

Part of being godlike is being patient, so the researchers gathered data on these lizards for four years; the populations of the islands grew from their founding couples to thirty or forty lizards, all of whom were subject to annual measurements of limb length and other characteristics. All the lizards also gave DNA samples to be sequenced, giving the researchers a certain godlike omniscience about these beasties.

So, after four years, what did the researchers find? Were the stochasticists right, and the lizards descended from larger Adams and Eves remain larger? Or, were the adaptationists right, and all the lizards got smaller and smaller limbs?

Of course, as one of my teachers would say, absolutists are always wrong! After four years, all the lizard populations had clearly evolved by selection. On every island, the population’s limbs were all shorter than the limbs of that island’s Adam and Eve. However, after four years, the imprint of the founders was still visible—the larger the founding couple’s limbs, the larger their descendents’ limbs. In one potent, information-packed graph, we have this result about the last years of the study:

(The islands have poetic names like "N1" and "N15"; all populations evolved towards shorter limbs, and generally speaking, larger-limbed founders lead to larger-limbed descendants. Just to be sure, the researchers followed the large population that provided each island’s Adam and Eve, and it barely changed over this time.)

There’s two other things that the lizard gods noticed in their study. One thing fit in very nicely with the stochasticists’ world-view. The researchers measured the amount of variation in some of the lizards’ DNA sequences (they looked specifically at non-coding DNA, a sort of filler that is not essential or even useful to the lizard; all animals have lots of this, and by definition it is not affected by selection, only by stochastic events). In the generations of Adam and Eve and their first offspring, the lizards on each island had a lot of variation in their DNA sequences. However, in succeeding generations, this sequence variation collapsed. The populations experienced a genetic bottleneck, and all the subsequent generations showed the effect—essentially, they became highly inbred. This fits with the stochasticist view of how island populations evolve and become different from one another, as each population’s genetic variation collapses around a different focus.

Another observation bolsters the adaptationist viewpoint. Most of the lizard populations peaked after three or four generations, and were actually in a significant decline as the study ended. This is a phenomenon that would be unsurprising to the Rev. Thomas Malthus: a small population in a rich environment expands, and eventually exceeds carrying capacity, leading to a crash. Under such circumstances, the selection pressure gets turned up to eleven.

The authors of this paper are not godlike enough to see into the future. Really, they studied just the very beginning of the island evolution process, and nothing like speciation has happened yet. The authors suggest that the adaptationist trend will dominate in the short term (especially given the pressure of overcrowding), driving all the populations to a similar conformation. However, beyond that they are agnostic: there aren’t any significant environmental differences between the islands (at least, any that they can perceive), so selection will no longer work as a force to make the island’s populations divergent. Also, the islands’ populations are really small, so it’s relatively easy for a chance event to have a profound effect. Retreating from the role of omnipotent and omniscient gods, they assume the mantle of Solomon, and judge the case of stochasticist v adaptationist by cutting the baby in half.

Kolbe, Jason J., Manuel Leal, Thomas W. Schoener, David A. Spiller, and Jonathan B. Losos (2012). Founder Effects Persist Despite Adaptive Differentiation: A Field Experiment with Lizards. Science 355: 1086-1089.

No comments:

Post a Comment