« This year in intelligent design | Main | Sustainable Darwinian Agriculture and Organic Tomatoes »

Ants en't ents

Science advances by disproving previously-tenable hypotheses. For example, "The earth is <10,000 years old" was disproved by annual sediment layers long before we were able to estimate the actual age. Actually, Tom Kinraide and I argued in "Strong inference -- the way of science" that a hypothesis needs to be explanatory as well as falsifiable. So for a young earth to ever have qualified as a hypothesis, it would first have needed to explain at least some real world observations. Right off hand, I can't think of any actual data that an unbiased person would look at and say, "Well, these data would make sense, but only if we assume the earth is <10,000 years old."

Similarly, if someone wanted to convert "intelligent design" from religious whining into a scientific discipline, we'd need some falsifiable hypotheses. Suppose, for example, we hypothesized that current features of plants and animals (not just their single-celled, distant ancestors) were supernaturally-imposed designs to maximize their success. That hypothesis is consistent with the many examples of sophisticated adaptations (err, "design"), but what can we conclude from the many examples of maladaptation ("bad design")? Maladaptation is predicted by evolutionary theory (when current conditions don't match those under which past selection occurred, for example) but if some design team is continuously intervening in evolution, do maladaptations imply that they had a busy week? If so, should we expect the problem to instantly disappear, once they get around to it?

This week's paper is another example of the pattern we see repeatedly in biology: many sophisticated adaptations, but also serious "design flaws." In particular, Acacia trees can be fooled into feeding and housing ants that are harming them.

Breakdown of an Ant-Plant Mutualism Follows the Loss of Large Herbivores from an African Savanna was published this week in Science by Todd Palmer and five coauthors, three of whom I know from my years at UC Davis.

Some Acacia trees provide food (nectar from leaf bases, not flowers) and housing (hollow thorns) to ants. The ants, to varying extents, protect the trees. These researchers have previously shown that one ant species "protects" their tree by pruning it to make it harder for ants in nearby trees to invade, even if the invading ant species would be better at protecting the tree from animals that might eat the leaves. Fortunately, these "cheaters" are rare, at least where tree-eating animals are common. But what would happen if the large browsing animals died out?

Rather than kill the animals to find out, the researchers used fences to keep large animals away from the trees. Within ten years, most trees were providing less food and housing for ants. Interestingly, this was not true of the trees whose own ant "bodyguards" pruned them. ("Someone's eating my leaves, I'd better keep paying the bodyguards!") Incompetent design team, or poor adaptation to a new environment?

The trees sampled were >40 years old, so the decreased support for ants was a response of individual trees, shaped by past evolution, rather than an example of current evolution (change in the genetic composition of populations) in action.

Where browsing animals were excluded, the less-protective, tree-pruning ant species became more common, while the more-protective, nonpruning species lost "territory," dropping from 50% of trees to about 35%.

Even this good ant species wasn't as good as it had been. As the trees provided less nectar, some of the ants switched from arboriculture, or tree husbandry, to herding of insects that suck sap from the trees.

These negative changes in ant populations would reduce, and perhaps reverse, the benefits of keeping browsing animals out. The article doesn't have direct comparisons of tree survival in browsed and browser-exclusion treatments, because detailed survival measurements have only been made in browsed treatments, so far (Stanton, personal communication). But they do show that trees have lower survival when they are occupied by ant species that increase with browser exclusion, so the indirect negative effects of removing browsers may turn out outweigh the direct positive effects.


Thanks for reminding me of your great Kinraide and Denison article in the ABT.
Rereading it, I recall my puzzlement that you didn't mention Popper in reference to falisificationism; if you cared to go further toward the disproof node in the phylogeny of science, you could have mentioned C.S.Peirce, of course. One of my favorites in this genre is, Boulding, K 1980. Science: Our common heritage. Science 22 February 1980, Volume 207, Number 4433; he does not cite Popper either. Curious. And, thanks for reminding me of Platt's strong inference paper, which shaped my thinking in grad school about what we are up to in this business. The paper is so old that Science charges for it, but I did find an html version on line (http://256.com/gray/docs/strong_inference.html) and as I recalled found that Platt did advocate Bacon and Popper as important nodes. One thought that lingers from the paper and debates of those days is that “taxonomy" was not up to strong inference (although he did deem that it “has its place"). Well, Willi Hennig had injected the roots of strong inference into systematics by that time, but it would be a decade before Henig’s ideas would take hold. Lots more to say about the logic and practice of science since Platt, including how to deal with historical phenomena, Gould, evolution, the problems of loosely integrated systems, such as ecosystems, but I have to get back to work.

Thanks for your comments, interesting as usual. I haven't actually read Popper, just summaries of his ideas, but Platt really helped focus my thinking, despite his annoying, nearly explicit assumption that all scientists are male. I remember being impressed by Boulding, but not the paper you cite. I'll add it to my list. I forget whether it's Strong Inference or Tragedy of the Commons (both big influences on me) that's been ripped out of both copies of Science in the UC Davis library. Jstor has both, of course, but just bitmaps. John Dennehy's summaries of classic papers on Evilutionary Biologist are great!

Post a comment

(If you haven't left a comment here before, you may need to be approved by the site owner before your comment will appear. Until then, it won't appear on the entry. Thanks for waiting.)

Type the characters you see in the picture above.