Why did butterfly metamorphosis evolve?
May 31, 2006 1:48 AM   Subscribe

How does evolutionary theory explain the caterpillar-butterfly metamorphosis?

I've been thinking about this and it kind of blows my mind. I sort of took it for granted that caterpillars become butterflies, but actually, why should this metamorphosis have arisen?
posted by evariste to Science & Nature (34 answers total) 2 users marked this as a favorite
Evolutionary theory dosn't have to explain everything, you know.

That said, metamorphosis is a repeating motif in the lifecycle of insects and other 'lower' life-forms. Think tadpoles -> frogs, and those are vertebrates. I know a lot of other animals do these sorts of things (also maggots to flies).

In fact, now that I think about it, don't pretty much all insects go through a 'larvae' phase and an 'adult' phase? Flies are maggots before they become flies, and wasps and bees are larvae who live in those little cells before they become bees/wasps. I think most insects go through this metamorphosis.

Since the genes that are expressed in each phase are somewhat separate, each one can evolve separately. So natural selection can operate on larvae and adult phases separately.


How life split into the larvae/adult structure is the essence of your question, and I don't know off the top of my head, but remember whole chromosomes can get duplicated and repurposed, something that happens every once in a while in our evolution
posted by delmoi at 2:28 AM on May 31, 2006

On top of my head one possible reason for metamorphosis could be that the various stage don't compete with the resources avaiable to the others. For instance, afaik, the caterpillars feed on vegetation while the adult butterfly feeds on nectar. All other conditions equal, if all the butterflies were immediately born as such, they would have a limited amount of food avaiable (considering how long they live and how far they can go flying) and this competition would likely decrease the number of bflies an ecosystem could support at any given time.

I remember reading about caterpillars being not palatable or toxic in some instance, which would give addictional protection for a while.

Also we shouldn't fall into the fallacy of single cause : probably there is more the one cause for the developement of metamorphosis.
posted by elpapacito at 2:32 AM on May 31, 2006

The metamorphosis arose because the first homometabolous insects (much more than just butterflies) outsurvived their hemimetabolous competitors in a particular evolutionary niche, and were able to keep surviving over millions of years to the present day.

Why were they able to outsurvive their competitors? Probably a combination of random chance and the new traits being more adaptive for their environment.

Why were their new traits more adaptive? Well, you can't really tell. In fact, you can't really be sure that they were more adaptive—it might have been entirely due to random chance.

Even accepting the likelihood that the trait was more adaptive somehow, you don't know why. You can guess (and I'm willing to bet that this page will soon be full of guesses) but your guess won't be science, since you won't be able to test it. It will just be yet another unprovable "evolutionary just so story."
posted by grouse at 2:34 AM on May 31, 2006 [1 favorite]

I mean holometabolous. D'oh!
posted by grouse at 2:35 AM on May 31, 2006

But I prefer your typo Grouse, homometabolous insects, Hmn...;-))
Any ideas out there from the fossil record of the first such insects?
posted by Wilder at 2:58 AM on May 31, 2006

The problem with explantions of things like this is they are all "just so" stories, it is relatively easy to make up explanations for something when you know the results. You would need proper fossil evidence to be sure.

That said, I guess you could say that the grub form is like the "embryo" for the bugs, and just happened to start hatching a bit early, and the grub was able to survive better than an insect which hatched straight into its adult form. After this, selection could act on the two seperate forms, giving each a specialised life cycle.
posted by scodger at 3:12 AM on May 31, 2006

Not that I have a specific answer to this very specific question or anything...

but this query appears a variant of an underlying question of the mechanics of evolution, in general. I always imply skepticism in such a query, a knee-jerk reaction, I'll admit.

There are an infinite number of such questions that can be posed, so the more appropriate response is to suggest that evariste become more intimate with evolution. Illustrating a specific pathway from a progenitor life form to a current one necessarily incorporates a lot of speculation, but comprehending the process, in general, doesn't.

Life is a (regular+random) response to the (periodic+random) environment. Uncountable trillions of these transactions yeild some truly inexplicable results, especially when one picks a single life form at a specific point in time. It's like asking the question, what exact series of dice rolls preceded the one I just made? I may not be able to do that, but I CAN tell you how a dice roll works in general.

So the 'answer' to this question is to enhance the question to the form 'How can evariste better understand the process of evolution', which is answerable.

I'm not being condescending, incidentally, even though it may sound so... I get the chicken/egg dichotomy thrown at me constantly from the fundamentalists on the fringe of my life and this is the way I respond.
posted by FauxScot at 3:53 AM on May 31, 2006

The OP raises a good point. If one simplistically models the evolution of an organism on say, a 2-dimensional XY grid (two biological features subject to change), and then look at fitness as the third axis (Z), then we can see how moving around the XY grid affects fitness (Z).

For example, moving up on X (more likely to cocoon) -- how does this affect the height, Z?

If one believes that a given organism, say one that doesn't have the caterpillar/butterfly metamorphosis cycle, is resting at a pretty high level of fitness Z (perhaps a "valley" or a "pit" in this XY diagram), then the question is: how does the organism "climb" over local maxima by varying traits X and Y (resulting less fitness Z -- a high peak) in order to find a deeper local minimum (greater fitness Z than before -- an even deeper valley.)

This concludes my butchering of Dawkins' explanation in Extended Phenotype. Thanks all!
posted by NucleophilicAttack at 4:09 AM on May 31, 2006

Speculation follows. It seems that there is a significant cost to carry a butterfly to adulthood. By spawning a bunch of lower cost grubs you get a buch of shots at reaching adulthood, not putting all the eggs in one basket, so to speak.
posted by bystander at 4:20 AM on May 31, 2006

A short answer from talk.origins to the "how" question:

Growth patterns intermediate to full metamorphosis already exist, ranging from growth with no metamorphosis (such as with silverfish) to partial metamorphosis (as with true bugs and mayflies) complete metamorphosis with relatively little change in form (as with rove beetles), and the metamorphosis seen in butterflies. It is surely possible that similar intermediate stages could have developed over time to produce butterfly metamorphosis from an ancestor without metamorphosis. In fact, an explanation exists for the evolution of metamorphosis based largely on changes in the endocrinology of development (Truman and Riddiford 1999).


1. Truman, J. W. and L. M. Riddiford, 1999. The origins of insect metamorphosis. Nature 401: 447-452.
posted by martinrebas at 4:24 AM on May 31, 2006

martinrebas, thanks for posting that. you beat me to it, i was just about to reference that same paper (which i'd be happy to email you as a PDF, evariste, if you email me - addy on my profile).

I'm not being condescending, incidentally, even though it may sound so
FauxScot, sorry but you are being condescending... the OP asked a specific question, sending them away to learn more about evolution doesn't help or answer that question, just adds noise. Especially when there is an answer (or, at the very least, evidence supporting insights) to that question.
posted by tnai at 4:54 AM on May 31, 2006

tnai, I diagree, sorry to derail slightly. I was in exactly this position in High school biology when studying the eye. I simply could not comprehend from its complex structure how such a thing could have developed. (In Catholic High School at the time they were leery of teaching evolution, it wasn't forbidden but certainly not encouraged)
I really had to get into the theory behind evolution to find it was completely understandable from basic principles for a light-sensitive cell over millions of years of random variations to develop into the eye.
So I think FauxScot has a good point, YMMV
posted by Wilder at 6:22 AM on May 31, 2006

If you want to be really freaked out consider that the caterpillar digestive tract is pretty much inverted compared to our own. Nutrients are extracted out up from through a couple of stages before the remainder ends up in a WAY high pH zone to kill off any pathogens. We dump food into a low pH zone then schlep it through a compost heap.

One thing to keep in mind is that caterpillars are not especially mobile compared with butterflies and moths. This lets them spread into other areas and proliferate more easily.
posted by plinth at 6:51 AM on May 31, 2006

Here's one interesting theory...
posted by curtm at 7:11 AM on May 31, 2006

If I'd place a bet on the driving force behind metamorphosis, it would be on seasonal survival. This is because radical changes in morphology in order to survive hostile changes in environment has evolved multiple times in very distantly related species and across kingdoms. Plants do it, bacteria do it, fungi do it, protozoans do it, so it's reasonable that some animals will hit on it as a strategy.

Another driving force behind metamorphosis is the ability to contract many sexual partners in a short period of time, and again, radical changes in morphology to take maximum advantage of a short mating season has evolved independently in plants and fungi to my knowledge. (There might be a colonial bacterium that acts like a fungi in this regard, but I'm not certain.)
posted by KirkJobSluder at 7:53 AM on May 31, 2006

slight derail
While tnai has a point, I think it's already been addressed. Perhaps the question could be more satisfactorily asked as "Does evolutionary biology posit an adaptive theory for the metamorphic lifecycle of the butterfly?"

enormous derail
tnai, the asker may imply skepticism, but the reader infers it

posted by mzurer at 8:12 AM on May 31, 2006

mzurer: Well, when I was studying biology, the questions went the other way. It wasn't so much, "how do we explain the metamorphic life cycle of the butterfly?" instead it was, "how can the evidence of variations in metamorphosis in Lepidoptera improve generalized theories of animal development?"

Most theories of evolutionary biology I've read will propose a relationship between generalized evolutionary pressures, and generalized types of adaptations, and then point to a variety of different variations on that theme. So for example, there are generalized theories of mimicry, kin selection, and number of offspring that can be applied to examples across entire groups, or even between kingdoms.
posted by KirkJobSluder at 8:40 AM on May 31, 2006

That is a fascinating detail, plinth. I wonder why the caterpillar does it that way,hmm --could it be that the caterpillar relies on microbes for its ability to digest plant material, so it sterilizes its waste in order that competitors can't culture these same microbes to their advantage, and so that the microbes themselves can't eat any of the plant, by being deposited freely on their food source, that the caterpillar might get around to eating later?
posted by jamjam at 8:58 AM on May 31, 2006 [1 favorite]

It's likely your question re metamorphosis can't be explained by evolution. Not many things can be.

There are an infinite number of such questions that can be posed, so the more appropriate response is to suggest that evariste become more intimate with evolution. Illustrating a specific pathway from a progenitor life form to a current one necessarily incorporates a lot of speculation, but comprehending the process, in general, doesn't.

This, apart from sounding quite condescending, is blanket doublespeak from the evolutionary camp. It might help, evariste, to remember that evolutionary theory is pretty weak scientifically; the reason it enjoys so much support is because so far we've failed to come up with anything better. Monkeys > humans seems plausible if you stand back and squint at it, but even that is missing huge blocks of critical evidence which cause the theory to stop far short of being definitive. FauxScot's answer is another way of saying "evolution can't explain everything [or indeed very much at all], so it's up to you broaden your mind."

Interestingly, if you ask a Christian to prove the existence of God, they will say almost exactly the same thing. Evolution is as much a religion as religion.
posted by BorgLove at 9:01 AM on May 31, 2006

Be sure to check out BorgLove's intelligent design web site, linked from his profile!
posted by Hildago at 9:07 AM on May 31, 2006

Respectfully BorgLove, that is bullshit. I, along with others here have given an explanation of how holometabolous insects evolved according to evolutionary theory. I'm sorry you don't like it.

Saying that humans come from "monkeys" is just an illustration of how little you know about (a) primate evolution, and (b) what biologists believe about primate evolution.
posted by grouse at 9:14 AM on May 31, 2006

Hildago, the use of the word 'intelligent' is coincidental. I'm not a creationist.

Grouse, it has nothing to do with what I like or don't like. Evolution is a weak scientific theory. It lacks solid, conclusive evidence in many areas, and the way it polarizes people, as exemplified by your comment, makes it as much a religion as anything else people might choose to believe. Also, the humans > monkeys comment is of course an oversimplification; rather than taking it literally, try to understand the point I'm making. Actually forget, it's a derail.

posted by BorgLove at 9:33 AM on May 31, 2006

Yeah, I can't resist at this point. The Heliocentric Theory also polarized people, with science on one side and religion on the other. It was polarizing because the religious side was flat wrong and had its monopoly on truth-giving threatened, not because the theory was as religious as religion itself. You'rll also find that if you look outside of certain regions of the US with heavily Christian Fundamentist populations, there's not even really much debate.

BorgLove, It is not clear if you use such mushy-mouthed reasoning out of a somewhat admirable attempt at tolerance, or because you really believe what you're saying.
posted by mzurer at 10:17 AM on May 31, 2006

Apparently the new definition of "religion" is whatever polarizes people. Extreme political parties, "breeders" vs. people who don't want to have children, and those who like the Da Vinci Code vs. those who hate it are all religions now. Not just strongly held beliefs.

It is an important point that humans did not descend from monkeys, nor does any evolutionary biologist believe this. The sooner we stop talking about this, the sooner I can answer the stop-beating-your-wife-yet question of "how did humans evolve from monkeys?" They didn't. If you don't understand this, or that no organism on the planet evolved in any significant way from any other organism that is still extant, it will be impossible to understand evolution.

As a TOTAL but necessary derail: I think it's hysterical in the context of this discussion that BorgLove's web site prominently features "intelligent design" and a big lowercase T that looks like a cross. But it really has nothing to do with the cryptocreationist "intelligent design" movement nor Christianity. I'm totally serious.
posted by grouse at 10:18 AM on May 31, 2006

Grouse, quit being a jackass. I've clearly stated I'm not a creationist; nor am I Christian. Attributing these characteristics to me because of coincidental symbolism is just silly. My opinion diverges from yours, but I don't see the need to ridicule you because of it; you should extend me the same courtesy, if you can.

My apologies to evariste for the derail on an interesting topic. I shall say no more.
posted by BorgLove at 10:38 AM on May 31, 2006

I think the question has been substantially answered, but here's another way of looking at it. Insects typically go through a number of developmental phases (called 'instars') as they grow. These instars are separated by molting off the exoskeleton.
Why would the different instars need to be the same as one another? Presumably, only the sexually active form would need genitalia, etc.
If you think about it that way, it's not at all surprising that some stages would be specialized for eatin', and some for mate-finding and egg-dispersing.
posted by nowonmai at 11:20 AM on May 31, 2006

jamjam - what I read about that suggested it as a way of sterilizing pathogens. I don't have that book handy, but I'll check it later. here's an interesting little blurb about it.
posted by plinth at 11:29 AM on May 31, 2006

I wouldn't know, but this is my guess:

Some context: We grow from embyro to child to sexually mature adult. Our metamorphosis is puberty, but since the physical changes of puberty are minor (compared to gaining wings), they are neither (much) of a cost nor a survival liability to us while they're happening.

So I suspect the distinct metamorphosis of a butterfly is a highly advantagous refinement that has been hammered into the similar path to get from an egg to a flying insect. It is the most advantagous refinement of that path because of the vast energy required to grow such massive organs as batterfly wings, and the vast handicap those organs place on an insect while being formed but not capible of flight, and their extreme vulnerability to damage while being formed. However, once the wings work, they more than pay for that investment.

A catepillar is a feeding machine that eats constantly to store away massive amounts of energy/fat in order to make the transition to butterfly and emerge ready to fly. In terms of energy required, having fully developed wings and the power to use them is no mean feat.

The only way a butterfly straight from the egg could be ready to fly would be if the parent only laid one or two eggs, and even then, that might not be enough, and the cost of those eggs to the parent in terms of energy would be so massive the extra energy/fat required might preclude flight either because of added weight, or because flight is too energy intensive to build up energy reserves.

Flight is useful - it means a species can spread out and thus not compete against itself for energy. It also means you can find a mate that isn't your brother/sister.

Are there any flying insects today that do not have a larval stage?

In other words, if you imagine proto-butterflies, you can pretty easily see how the path from egg to adult would lend itself to growing/transitionary phases, or a less active phase, that over time would get refined into a cocoon stage, which has some pretty obvious advantages.

I can imagine a viable life cycle for every intermediate step, with advantage over the previous. Hell, maybe those imaginings might even bear some resemblence to what happened :)
posted by -harlequin- at 12:50 PM on May 31, 2006 [1 favorite]

Uh, really no intent to ridicule. I don't know how to say that anymore than "I am totally serious." I guess the more I emphasize how serious I am the more it looks like I am snarking, so it's kind of a lost battle. I still think it is a funny coincidence in this context, but not a telling one.
posted by grouse at 4:21 PM on May 31, 2006

BorgLove, I apologize if I framed you as an intelligent design advocate and you aren't one.
posted by Hildago at 7:44 PM on May 31, 2006

Flying insects w/o a larval stage? How about praying mantids? Although how do you distinguish from a nymph and a larva? Mantids do not fly until the late instars, but there is no pupation.
posted by plinth at 7:44 PM on May 31, 2006

Thanks to everyone! That was quite educational.
posted by evariste at 9:17 PM on May 31, 2006

In case anybody's still peeking in here, but this turned up in the news today:

Maine Researchers at the University of Southern Maine have identified a "starter hormone" responsible for initiating the process by which caterpillars transform themselves into moths.

...though this probably speaks more to the how than the why.
posted by SteveInMaine at 1:54 PM on June 1, 2006

Interesting, thanks Steve!
posted by evariste at 2:31 PM on June 1, 2006

« Older indian take out in Los Angeles   |   Living In Someone Else's Shoes Newer »
This thread is closed to new comments.