Kosh

I probably spelled that wrong. Along the lines of the "irreducible complexity" argument, what's the current theory in evolution for explaning the development of metamorphosis in animals like:

Frogs
Salamanders
Butterflies

Thanks.

Duvenoy

These animals simply have a larval stage in their lives. There's really nothing especally complex about it.

Think of it as similar to the development of a placental animals fetus, except that it's all done outside the mother's body. From egg to tadpole, to frog all of the comparable stages are present. The biggest difference is that the larvae is able to take care of it's self, unlike the mammal's fetus / baby.

I've never heard of amphibian or arthropods used in IC arguments, although I'm sure they have been.

Hope this helps.

doov

[ June 15, 2002: Message edited by: Duvenoy ]</p>

lpetrich

Such multistage lifecycles are very common.

Frogs are well-known for having a larval phase: tadpoles, which are somewhat fishlike.

Immature salamanders are also aquatic; they look like adult ones with gills sticking out of the sides of their heads.

Insects are well-known for that; sometimes larvae and adults are drastically different.

Many marine invertebrates have larvae that look very different from the adults. The larvae are always free-swimming, even though the adults often live on the ocean floor, and may even be sessile (fixed in place).

How would such stages develop?

In the amphibian case, it's a case of partial transition to living on the land. Here is what would allow a partial transition to be reasonable: the square-cube law. Total water goes up as the cube of the size, while water leakage goes up as the square of the size. Thus, large animals leak water more slowly than small ones.

So some Paleozoic fish could have developed living-on-land adaptations that only appear when the fish has grown big enough to not leak water too fast.

Completing the move to the land requires laying bigger eggs and pushing the larval phase into the egg -- thus the gill-like structures of amniote embryos.

In the insect case, it's also a case of different phases being adapted to different environments. The earliest insects may have performed an amphibian-like transition; dragonflies have aquatic larvae and wings most likely evolved from gills.

One side effect is that wings would be fully developed only in the adult phase, making the immature ones flightless. This means that immature and mature insects can become adapted in different directions, one for feeding and the other for breeding.

In "incomplete" metamorphosis, immature and mature phases cannot be very different, because of growth continuity. However, if an intermediate growth phase could be set aside for remodeling the insect, the immature and mature phases would be allowed to be much more different. Thus, "complete" metamorphosis, where an insect larva gets remodeled into an adult in the pupa.

Finally, I turn to marine-invertebrate larvae. <a href="http://cas.bellarmine.edu/tietjen/Ecology/early_animal_evolution.htm" target="_blank">This paper on early animal evolution</a> mentions a very interesting hypothesis.

It's that early animals were directly homologous to today's larvae and that today's adult phases are produced by outgrowing the larval-phase organization.

Basically, the ancestral form of development is propsed to be"maximal indirect", which involves a cell-differentiation mechanism ("type I embryogenesis") that does not scale very well, meaning that its user cannot grow very big. In present-day forms, adults originate from special set-aside cells in the larvae; this feature may have started out as a way of growing a "tail" that helped the animal swim. There was a cute illustration of that hypothesis that I had once seen, but I had forgotten to bookmark it.

This "tail" was very likely segmented, with new segments being produced at the far (rear) end. As concluded from how forming new segments at the rear end is a common mode of present-day embryonic growth.

The "tail" had a differentiation mechanism that scaled better than the "head", and longer-growing animals would become mostly or all "tail" as they grow. Sometimes the larval phase would be pushed into the egg, leaving little or no trace.

Kosh

Thanks for the answers guys. This is what I came up with myself as a hyposthesis when I was at the aquarium, but it was based on nothing more than a "hey, they still look like developing fetuses" observation.

Looks like us mammals got the short end of the stick on the evolution of parenting!

MrDarwin

Can't really address the others, but regarding insects, there is no "irreducible complexity" to their metamorphosis. In fact, among the many insect orders, there is a complete range of kinds of metamorphosis, from none (animal just grows gradually, each stage a bigger version of the last), to gradual (wing buds become bigger with each instar, and become functional only after the last molt), to complete (and even here there is a broad range, from mobile pupae with separate appendages but completely immobile pupae as in most butterflies and moths).

In other words, the various orders of insects have retained a record of pretty much all the various stages in the evolution from no metamorphosis to complete and dramatic metamorphosis.

[ June 16, 2002: Message edited by: MrDarwin ]</p>

lpetrich

In this PNAS <a href="http://www.pnas.org/misc/special.shtml" target="_blank">special issue</a>, there is an <a href="http://www.pnas.org/cgi/content/full/97/9/4434" target="_blank">article on some marine-invertebrate development</a> that illustrates the "tail" hypothesis:


Essentially, indirect developers, marine mollusks, annelids, echinoderms, and others, hatch as the yellow sphere in that diagram, which is their larval phase. It eventually grows a cyan tail, which becomes the adult.

And as that tail grows, it adds segments. Hox genes help specify how each part of the tail is to develop; each one specifies rearward identity, meaning that if a Hox gene is knocked out, then a part will develop in the fashion of a more-forward part. This is consistent with Hox genes duplicating to cope with an ever-lengthening tail; since it's the rearmost part that needs some new specification, it is natural for a duplicated Hox gene to specify the current rearmost part.

Peez

Perhaps this would be easier to understand if one keeps in mind how genes work. For example, a gene ‘for' a long tail works by coding for a protein which, when produced, causes the cells in the appropriate location to grow and divide in such a way as to produce a long tail. Thus, genes do not directly create physical features of an organism, they indirectly influence the development of certain features at certain times. Looking at it in this way, it is simple to see that while some genes will be expressed early in development, others may be expressed later (during ‘metamorphosis'). Hope this helps.

Peez

Duvenoy

And just to throw a wrench into the reduction gears of metamorphsis, scorpions and Madagascar hissing cockroaches bear their young alive, as tiny minitures of themselves. Neonate scorps ride on their mother's back for a time (and some few might be eaten if mama gets hungery).

doov

Edited fir really obvious typo.

[ June 17, 2002: Message edited by: Duvenoy ]</p>