One of the most striking features of the most common insects is the really remarkable changes that happen when the larva forms a pupa that completely restructures itself into the adult. The metamorphosis is often so radical that there is no obvious similarity between the larva and the adult at all.

Not all insects do this. The silverfish, true bugs, mantids, cockroaches, dragonflies, damselflies, mayflies, stoneflies, lice, and a bunch of other mostly-obscure groups develop directly from nymphs to adults, with no pupa stage. The young just become more similar to the adults with every molt until they finally mature (by developing genitalia, and usually also by growing wings).

This was apparently the original state of the ancestor to all insects, and so these various insects that do not have a pupa stage are only distantly related to each other, having diverged from one another sometime back in the Devonian period, as much as 400 million years ago

Complete metamorphosis (egg – larva – pupa – adult) evidently only evolved in insects once, with all the insects that do it being derived from a common ancestor sometime in the Permian, approximately 100 million years later. This gave them a huge advantage in that the different life stages could now optimize themselves for what they were doing at the time: the larvae could devote all their efforts to eating and growing and avoiding being eaten, while the adults could completely restructure into a form better suited for finding mates, dispersing, and finding good places to lay eggs. As a result of this ability to optimize to conditions, the groups that do undergo complete metamorphosis are some of the most widespread and prolific:

Coleoptera (beetles)
Diptera (flies)
Hymenoptera (ants, bees, wasps, and sawflies)
Lepidoptera (butterflies and moths)
Neuroptera (lacewings and ant lions)
Trichoptera (caddisflies)
Mecoptera (scorpionflies)
Siphonaptera (fleas)

While complete metamorphosis is clearly a very successful strategy, it is also a rather confusing one. It is not immediately obvious how an insect could evolve the ability to do this[1]. It basically boiled down to two general hypotheses:

1. Perhaps the differences between the final stage nymph and the adult simply became so large that the last nymphal molt just turned into an immobile thing that stayed still while it got completely restructured. In this case, the young nymphs would correspond to the larva, and the old nymph to the pupa.

2. Perhaps the egg hatched prematurely, and a “pronymph” (basically a still-developing embryo) developed the ability to eat and grow. In this case, the larva is essentially an extremely extended embryonic stage. All of the nymph stages ended up being collapsed into one stage, the pupa.

Up until recently, there was considerable argument over which of these hypotheses was correct. But, as of about 1999, endocrine studies have indicated that the second hypothesis is the better fit to what is observed about insect development [2]. So, the various grubs, maggots, and caterpillars[3] that you see around are better regarded as eggs that can eat than anything else. And, since they do nothing other than eat, grow, and hide, they can become very good at all of these things.

For some insects, they actually pretty much treat the entire larval and pupal stage as an extended egg stage. For example, bees, ants, and a lot of wasps will just build a nest, seal them up with a lot of food, and then just wait for them to “hatch” as fully-formed adults.

Once this extended embryo stage is over, they need to change completely for the adult phase, which needs to be able to disperse, mate, and lay eggs. Which means that they benefit from being able to fly (or at least run fast), somehow get the attention of mates, and produce lots of eggs before they get eaten. So, during the pupal stage, the features of the larval stage get almost completely broken down and replaced, and they need to create legs, wings, antennae, better eyes, etc. from practically nothing.

Only practically nothing, though. It turns out that the grub/maggot/caterpillar has more features than are easily seen by the naked eye. There are thickened spots in their skins, called Imaginal Discs. These are parts of their epidermis that are programmed to eventually develop into more complex body parts. They don’t look like much, but during pupal development they kind of extend out, a bit reminiscent of a collapsable radio antenna. What body part they will become is set very early on in development. So, for example, an imaginal disc in one caterpillar that is destined to become a wing, if it is transplanted onto another caterpillar, will turn into an additional wing. This mostly happens under the skin of the pupa, although you can usually see a pattern on the pupal surface showing where the newly-grown body features will be (this is most pronounced in beetles,

but often almost invisible in flies)

One huge advantage of the pupal stage, with the legs, wings, and antennae all developing under the skin, is that the insect can pop right out of the pupa in a very short time without worrying about withdrawing all of the fiddly bits. Insects that don’t have a pupal stage, like our walking stick from last summer, take a long time in their final molts, and sometimes run into issues that result in losing legs. Meanwhile, butterflies emerge from their pupae so fast that if you don’t look sharp they will pop out while you are momentarily distracted. One minute they are all sealed up, and the next they are hanging on to the outside and pumping up their wings.

So, all in all, complete metamorphosis looks to be extremely advantageous. Probably the only reason that all insects don’t do it, is because it was a low-probably evolutionary development. Insects apparently only developed the ability once, and then their most successful orders all radiated from there.

—————–
[1] Complete metamorphosis is so extreme that it has sometimes lead to extreme explanations. To the point that Donald A. Williamson has gone so far as to suggest that maybe the first insect to undergo a complete metamorphosis was actually two species (one wormlike, and one more leggy) that somehow hybridized into a single species, with the wormlike species providing the larva body plan and the leggy species providing the adult body plan. Which is pretty much like proposing hybridizing, say, rats with salmon. And if you think that sounds crazy, well, you’re not the only one.

[2] James W. Truman & Lynn M. Riddiford (1999) “The origins of insect metamorphosis”, Nature, vol. 401, pp. 447-452. This is available at http://www.nature.com/nature/journal/v401/n6752/full/401447a0.html, but this link may not work for everybody – I think I only have access to it because our university library has a subscription to Nature.

[3] While there is some slop in the usage, in general grubs (distinct head capsule, only six legs) turn into beetles; maggots (pretty much no head capsule and no legs) turn into flies; and caterpillars (distinct head capsule, six “true” legs, and some number of “prolegs) turn into butterflies, moths, and sawflies. I’ve seen the generally-featureless larvae of ants, bees, and wasps mostly referred to as grubs, not usually as maggots, even though they don’t have a head capsule or legs.