Friday, March 23, 2012

You look just like.... well, almost like

Female hoverfly on cistus flower; Wikimedia

Ken mused a few weeks ago about mimicry, the traits that many species have evolved as defense against predators, according to Darwinians -- the butterfly with an eye spot on the back of its wing, the moth that has taken on the coloring of tree bark.  A commentary in this week's Nature, accompanying a paper about mimicry ("A comparative analysis of the evolution of imperfect mimicry", Penney et al.), quotes the British naturalist, Henry Walter Bates, who described these kinds of traits 150 years ago as "a most powerful proof of the theory of natural selection". 

Ken's point in his post was that, yes, Darwinians have explained these as examples of very clever adaptation, a way to outwit predators and increase the odds of survival, but that in fact these ruses aren't always all that successful.  As he said, after describing some examples of very effective mimicry in butterflies that he had come across in his travels, "...while I did see the effectiveness of protective coloration in these two instances, I also did, after all, see the butterflies.  I wasn't completely fooled."

Indeed, if it's so effective, why haven't all species evolved protective coloration?  And, the Darwinian's answer is that it's only one kind of adaptation, and there are many others.  Each trait has its own adaptive purpose, and it is the job of science to uncover what that is.  But, as Ken also pointed out, most adaptive explanations can't be confirmed, no matter how plausible they seem.  Furthermore, whether an organism is eater or eatee it may typically largely be due to chance, and the genetic contribution is usually very slight, and essentially undetectable (even industrial melanism in moths, recently confirmed statistically with new data, was a lot of work).  And, the assumption that natural selection detects all functional genetic variation is simply an assumption, but it makes Just-so stories about adaptation unfalsifiable.

We read the Nature pieces on mimicry, then, with this in mind.  The question posed in the Penney et al. paper is why some harmless hoverflies are such good mimics of stinging Hymemoptera (wasps or bees), and others are much less convincing.  They point out that evolutionary theory about mimicry assumes that some copies are pretty exact but that examples of inexact copying abound (though, there does come a point where one would ask how we're sure it's in fact a mimic or what 'exact' means in this kind of situation). 

Explanations of poor mimicry include that it may look poor to us, but it's good enough to fool a predator, or imperfect mimicry is even more adaptive than perfect mimicry, or that imperfect mimicry benefits co-specifics (this is a kin selection argument), or that there are constraints on the evolution of a more perfect copy, or relaxed selection, whereby selection for mimicry becomes weak enough that it is "counteracted by weak selection or mutation", that is, that there's no selective benefit to refining the mimic further. 

To try to determine which of these explains the poor hoverfly mimics, Penney et al. used "subjective human rankings of mimetic fidelity...across a range of species" of two different taxa (Syrphidae and Hymenoptera) and compared them for consistency against a statistical analysis of trait values. They found a strong positive relationship between body size and mimetic fidelity, and suggest that "body size influences predation behaviour and thereby the intensity of selection for more perfect mimicry."

The idea is that the larger the prey, the more benefit to the predator, and thus the more urgently the prey needs to figure out a way to avoid the predator.  Smaller or more abundant hoverflies needn't spend so much energy trying to fool a predator, as each insect is less likely to be preyed upon because there are more of them to choose from, or because they are less of a mouthful, and so less disirable. 

So, they explain imperfect mimicry as the relaxation of selection on mimicry, though they do not find this counteracted by weak selection or mutation, and they do not reject the constraints hypothesis.  They conclude that "reduced predation pressure on less profitable prey species limits the selection for mimetic perfection."

The same explanation always, but always different
Notice that each of the 5 possible explanations they offer assume that selection of some sort must be the explanation, if only they knew it.  This is the Darwinist assumption, that the ground-state of life is competitive natural selection.  If one selection story is shown to be wrong, then it must be another, as we've seen in this case.  This explanatory tack is very widely accepted, indeed, assumed without question.  But the assumption of selective adaptation is not itself ever questioned.  Is it true?

More accurate than that assumption is that the ground state of life is existence, or over time, persistence.  Whatever reproduces, reproduces.  This is an assumption, too, and is testable....but isn't very helpful at all in and of itself.  We can go a bit further: There is differential reproduction among differing genotypes, but even in a totally non-selective world this would be the case (in formal terms, genetic drift is inevitable).  Sometimes success may be due to a systematic relationship between the state that succeeds and its success, and that's natural selection, but this need not be the case.  The question is when and to what extent predictable, non-random  change is occurring, and that is not at all easy to show most of the time.

More profoundly, selection need not be (as Darwin seriously thought, and as most modern biologists accept without thinking seriously about) the force-like phenomenon it is usually, if often tacitly, assumed to be.  It can be weak, ephemeral, local, moveable, and even probabilistic.  Even from a purely selectionist point of view, all sorts of species with all sorts of variation are reproducing in all sorts of ways in relation to all sorts of factors -- including each other. There is no reason to expect that single factors, alone, will necessarily motor on through with some clear-cut force-like trajectory of change.  These statements are not at all controversial, yet seem to be in effect ignored when each investigator's favorite trait is being evaluated 'on the margins' as one would say in statistics: that is, evaluated in isolation, on its own.

We have a great parallel here with polygenic causation that is so pervasive and frustrates GWAS as we've said countless times here.  With polyfactorial ecologies, what oozes through the blizzard of factors will not necessarily be simple or explicable in terms of one factor on its own -- say, looking like something else.  This is a very different perspective than trying to analyze everything as if it is the type of selection we have to identify or explain why, surprisingly, it's not perfect.

Think of it in this very simple way.  It is almost always possible to change most traits.  Experimentally, this is reflected by the fact that most traits respond to artificial selection.  In this case, that means that it should always be possible for natural selection to lead to change in ways that make any species that somebody else eats look more like the background of where it lives (even against bacterial predators, some form of camouflage defense at the molecular level should always be possible).  If the selectionist stories are accurate, that camouflage increases your odds of living to frolic another day, then every species should be camouflaged and should normally dwell on its match. 

This is so clearly not how nature is that one wonders why fundamentalistic Darwinism ever took hold, even by Darwin himself.  Why isn't everything camouflaged?  The answer, which we referred to casually as the 'slop' of nature in earlier posts (e.g., here), is that evolution is persistence in the entire ecology of each organism, and sometimes something seemingly so obvious as mimicry clearly seems to happen. But not most of the time.  Or, each trait in each organism can be argued to have some such story. That is so ad hoc, or perhaps post hoc, it has a resemblance to creationism--in the epistemological sense of being something in which every individual observation has the same explanation (selection made it so), no matter what the details.  If we assume selection, we can always adjust our explanations for what's here to fit some selection-story.  One simple component of this, obviously, is that the predators are evolving their detection ability as well.  It's all an endless dance, and this is not controversial biology, but is within what we know very well.

Biology should grow up.  The ground state of life is persistence, however it happens.  And stuff happens.  There are lots of ways to persist.  Selection is one, but it's only one, and it is itself a subtle, moveable feast.


Anne Buchanan said...

So, this story seems to be making a splash today all over the web (here, e.g. -- I love the headline: "Researchers solve Darwin's puzzle".) Everyone is repeating the idea that smaller hoverflies don't have to be such great mimics because they aren't as appealing a mouthful to the predator bird as larger hoverflies. If this argument were true, ants, and even smaller insects, would rule the earth. Every bug, no matter what size, has a predator. And birds come in many sizes.

Maybe the relaxed selection argument is in fact true. I just don't find the reasoning convincing.

Ken Weiss said...

Hmmm, makes me wonder why I like M&Ms as much as a thick steak, but I don't care if the M&Ms are the same color as the plate they're on (or the steak, for that matter).

Yet, a big pile of fries seems so delectable (whether or not they have red ketchup on them), but by the time I get to the last one, it has no appeal whatever. But not because it's so small....

Steve Bates said...

It's said that every flea that flies
Is dined upon by one that spies
It through its native camouflage...
A more successful flea-o-phage!

- SB the YDS

Steve Bates said...

(... with apologies, of course, to Jonathan Swift.)

Ken Weiss said...

and on ad infinitum. Everybody's somebody else's dinner.