Animal adaptations and defences — The World, with the Evidence

Old plate with several rows of butterflies that look very much alike. — Plate published by Henry Walter Bates in 1862: Dismorphia butterflies and Ithomiini butterflies from the Amazon valley that look very much alike, an example of so-called “Batesian” mimicry. Henry Walter Bates, Public domain, source

The porcupine did not want its quills, nor the cuttlefish its ink, nor the chameleon its colour.

None of them decided anything, and no force decided it for them.

This is the hardest idea on this page, and our language does not help.

The problem is in the grammar

Take the following sentence: this frog developed a poison to protect itself.

It seems perfectly normal. It is wrong.

“To” implies an intention. It implies that someone, somewhere, wanted this result. There is no one.¹

The most awkward part is that there is almost no escaping it. Anyone who tries to describe an animal defence without saying what it is for finds that the French language keeps pulling them back to it.

The word manufactures a belief, and this time the push comes from the grammar itself.

The solution we have adopted

This site eventually found a way round it, and we share it because it can be useful well beyond these pages.

Do not say what a trait is for, but what happened to those that did not have it.

Let us go back to the frog.

Wrong version: this frog developed a poison to protect itself.

Right version: among these frogs, those whose skin was the most toxic were eaten less, and left more descendants.²

That is three more words. It attributes intention to no one. Above all, it describes the mechanism instead of naming it: we see what happened, generation after generation, without needing to believe in a plan.

The test can be applied to any textbook sentence: the long version almost always turns out to be clearer.

What this really changes

This is not a linguistic affectation. It has three real consequences.

First consequence: there is no perfecting. One defence is not “better” than another in absolute terms. It worked there, at that time, against those particular predators. If the predators change, it can become useless — or a burden.

Second consequence: there is no lagging behind. An animal that has no venom has not “failed” to develop it. It simply did not follow that path. This is the ladder misconception, which we take apart in our page on classification.

Third consequence: nothing is planned. An adaptation does not prepare for the future. It is the result of a past.

Mimicry: deceiving without knowing it

Mimicry is often presented as the height of animal cunning. Yet the rule remains the same: no butterfly decided to look like a leaf, or like a toxic species.

Two major mechanisms that should not be confused

Batesian mimicry: a harmless species takes on the appearance of a dangerous or toxic species.
Müllerian mimicry: several genuinely toxic or unpalatable species adopt similar visual patterns (the yellow and black stripes of wasps and bees, for example). The signal is shared, which speeds up predators’ learning at the cost of fewer individuals.³

Testing the hypothesis: the field experiment

For a long time, mimicry remained an appealing story that people told without testing it. How can one show that a visual pattern really protects an animal?

First, the idea has to be turned into a prediction. This one is elegant, because it is risky: if the protection comes from predators having learned to fear the venomous model, then it should disappear where that model does not exist. No model, no learning; no learning, no protection.⁴

The idea thus becomes testable — and refutable.

To test it, researchers made fake snakes out of plasticine: some painted with the red, yellow and black rings of the coral snake, others with neutral patterns. Placed in the forest, these replicas retain the marks of predators’ beaks or teeth.⁵

The result is striking. Where the coral snake actually lives, about 8% of the ringed replicas are attacked. Where it is absent, about 65%.⁶

Eight versus sixty-five. The same pattern, on the same object, with an almost reversed effect — depending on whether the local predators have encountered the model or not.

The researchers went further still: the protective effect diminishes as one moves north and gains altitude — that is, as the coral snake becomes rare. It is not just a matter of present or absent: it is a question of frequency.⁷

(Researchers took up the idea of plasticine replicas, with fake caterpillars, to measure predator attacks on six continents.) The signal works only where predators have already learned to fear it.⁸

“Useless” organs are not useless

One sometimes hears of “vestigial” organs, presented as useless leftovers that evolution forgot to remove.

That is not what the word means.

In biology, a vestigial organ is an organ that has lost its ancestral function — which is not the same as being useless: it may perfectly well retain another.⁹

That is where the whole difference lies, and the two were long confused.

The case of the appendix

The appendix long served as the textbook example of a useless organ.

Work published in 2007 proposes a function: it may serve as a reservoir for beneficial bacteria, making it possible to recolonise the intestine after an episode that has emptied it.¹⁰

The hypothesis rested in particular on a clinical comparison, whose figures circulated widely.¹¹

This page at first refused to give them, because they could only be found in activist retellings. On checking them, we discovered something else: they have not been confirmed.

A 2012 study of 257 patients found the opposite of what was expected: people who tested positive for the bacterium Clostridioides difficile had more often kept their appendix. Another, of 1,580 patients, found no difference in mortality according to the presence or absence of the organ — but its authors report more severe forms in those who had had the operation, which points in yet a third direction. Finally, a 2023 systematic review, bringing together eight studies and more than four thousand patients, concludes that a previous appendectomy increases neither severity, nor recurrence, nor mortality — while noting more frequent recourse to removal of the colon among those who had had the operation.¹²

Three studies, three results that do not point in the same direction. This is what a genuinely open question looks like — and that is why none of them can settle it alone.

This, then, is the honest state of the question. The bacterial reservoir hypothesis remains an interesting and debated idea. The clinical support attributed to it, however, does not hold up.

What remains true nonetheless: “useless” is no longer the accepted description of this organ. It contains immune tissue, and a 2017 study rejects the idea of a functionless vestige. But between “it is not useless” and “here is its function, demonstrated”, there is a gap that this page will not bridge.¹³

The reversal

This hypothesis implies something unexpected.

If the appendix serves as a reserve of bacteria for cases where the intestine is emptied, then its usefulness depends on how frequent these episodes are. In countries with good water supply and sanitation, these episodes have become rare.

In other words: the appendix may perhaps be becoming vestigial — now. Not in a distant past.

An organ is not useful or useless in absolute terms. It is useful in a given world; let the world change, and the answer changes.

Our editorial safeguards

Because this subject touches on notions that are often misinterpreted, this page imposes three strict safeguards on itself:

No medical advice: this page gives no health recommendations and does not discuss clinical indications (such as appendectomy). These are subjects for a health professional, not for an encyclopaedia.
Refusing polemical opportunism: the subject of vestigial organs is regularly exploited in debates outside biology. This page is not structured in reaction to these controversies. Writing in order to answer an opponent means letting them choose the subject.
Refusing reintroduced finalism: not falling into the trap of replacing the animal’s will with a “will of nature” or “of selection”. Natural selection is not a conscious force that chooses: it is a statistical result.

What fossilisation hides from us

One last thing, about what we can know of ancient defences.

Hard parts are preserved: bones, teeth, shells, carapaces. Soft parts, almost never: skin, venom, colours, behaviours.¹⁴

The defences we know from the distant past are therefore mainly hard defences.

This does not mean that ancient animals had no venom, camouflage or alarm calls. It means that we would not see them.

It is the same caution we apply to objects made by humans ten thousand years ago: what remains is not what existed. It is what has survived.

Note 1

[…] the same ridicule that Voltaire heaped on Dr. Pangloss: […] Everything is made for the best purpose. Our noses were made to carry spectacles, so we have spectacles. Legs were clearly intended for breeches, and we wear them.

Gould, Stephen Jay and Lewontin, Richard C., “The Spandrels of San Marco and the Panglossian Paradigm”, Proceedings of the Royal Society B, vol. 205, 1979, p. 581-598

See the full note ↓

Note 2

Mayr, Ernst, What Evolution Is, Basic Books, 2001

See the full note ↓

Note 3

Batesian mimicry evolves when a palatable species, the "mimic," resembles a dangerous species, the "model," because both receive protection from predation.