Evolution of Toxicity

How does defensive toxicity evolve? Being as I am completely ignorant of scientific orthodoxy on the topic, I am free to wildly speculate regarding this mystery.

Consider the monarch butterfly for a well-known example. As caterpillars, monarchs acquire a store of cardenolide aglycones from eating milkweed. These cardenolides are steroids that, in high enough doses, can stop the heart and thus are toxic to most animals. The monarchs then are permanently toxic to most predators from cardenolides ingested as a caterpillar and are thus protected from predation.

Storing cardenolides is a trait that must have evolved. Any investigation into the evolutionary history of a trait must begin with the Occam's Razor of evolutionary theory: the trait arose through random genetic mutation; it was beneficial to the reproductive viability of an individual monarch; this increase in viability eventually led to universal trait ownership in the global monarch population.

But how can this explanation be correct? Let us try to explicitly paint this story of the cardenolides-storing "gene" [standard euphemism for "gene or set of genes"]. We will start with a single monarch since identical genetic mutation in two distinct individuals is probabilistically impossible:

• A single monarch is endowed through genetic mutation with the cardenolides-storing gene. Until the first interaction with a predator, this gene has no effect on the reproductive viability of the single butterfly.
  
• A bird eats the butterfly and dies or is injured from the toxin.
  
• The gene disappears.

The gene conferred no survivability advantage to its host since its first effect occurred only after the monarch was killed. If the monarch had offspring before it was killed, those offspring would have no survivability advantage over their cousins without the special gene. Genes that confer no reproductive viability advantage do not become universal.

Consider further the role of the predator, in this example a bird. If not killed, the bird might learn not to eat monarchs again. Through cultural transmission the bird could possibly teach other birds not to eat monarchs. Even if such cultural transmission is likely, this benefits both the monarchs endowed with this gene and not, giving no evolutionary advantage to the gene itself [using a Dawkinsian gene-level evolutionary argument which is clearly valid trait-level as well]. If the bird was killed, on the other hand, then birds with a gene instructing them to not eat monarchs might replace the old ones after many generations. This scenario is less likely than the first; it requires a long period of large-scale predator/prey interactions with a monarch population significantly endowed with the gene, while simultaneously never giving any viability advantage to the cardenolides-storing monarchs. Neither explanation can account for selective pressures favoring the toxic gene.

Is there still some way to invoke basic adaptive natural selection for monarch toxicity? There is a candidate solution: taste. Cardenolides taste horrible. Is it possible that the bird tasted our monarch and rejected it? The butterfly then has received the ultimate survivability boost from the gene: it lives where its cousins would certainly have died. This is certainly possible, but I have strong doubts about this hypothesis. For one thing, I wonder about the state of a butterfly after being tasted by a bird. Is it likely that a tasted butterfly can survive and have offspring? It is true that this tasting could possibly leave the butterfly alive; this is why beak-marks on the wings of butterfly specimens are common. But what are a monarch's chances of reproducing after this injury? Additionally, any lesson the bird learned from the tasting the monarch--that all monarchs or even all butterflies taste terrible and are possibly toxic--will confer no specific advantage to the monarchs with the mutation. Remember that our one monarch, in order to pass on beneficial genes, will need to live in a situation where a monarch without the genes would die. Unless birds habitually sample butterflies in a non-fatal way before deciding to eat them, I find it extremely unlikely that it is taste that conferred the viability advantage for the monarchs with the toxic gene.

It is clear to me that the standard adaptive natural selection argument can't explain the toxicity trait of monarch butterflies. But exaptation--my old friend of herons and voles--could explain it well. Here's my explanation:

Cardenolides are not only toxic to birds, but insects as well. In fact, the toxins are used by plants specifically as a defense against being eaten by insects. Let us paint a different evolutionary story:

• A single monarch is endowed through genetic mutation with some kind of immunity to cardenolides
  
• This monarch has access to food plants [as a caterpillar] that other monarchs do not. The immunity gene confers a strong advantage to this monarch at its most vulnerable stage of life.
• Because of increased survival rates as caterpillars, monarchs with the mutant genes proliferate and replace those without it

This sounds very probable. But where does the toxicity fit in? I have only explained how monarchs might have evolved tolerance to cardenolide. Here is the "wild speculation" I promised:

Suppose the gene for "immunity" is actually a gene for removing the cardenolides from the plant matter before digestion. This is one of several plausible methods for tolerating a toxin. The toxin thus must be excreted or stored. If the toxin is stored--or even if it persists a significant amount of time before being excreted--the monarchs will be toxic to birds. This toxicity is a by-product of the natural selection of the cardenolide-tolerant gene.

The benefit of this approach is that it allows the monarchs to acquire the toxic gene as a species. The downfall of the adaptive natural selection argument in the first place was that the gene had to be beneficial to the first butterfly in order to propagate. If the mutation had an initial evolutionary benefit--allowing access to new abundant food sources--then it will prevail throughout the species. Only then will the pressures of predation influence the evolution of these monarchs. Remember that all of our explanations for how the presence of the gene in the first individual gave it an evolutionary edge failed because all of those explanations gave the same advantage to the monarchs not possessing the mutation as to the monarchs with it. This failing is erased when the evolutionary impetus for the gene's survival is not dependent on being eaten or "tasted" to be expressed.

Now, as a species, the monarchs evolve better and better storage mechanisms for the toxins to be more effective at deterring predators. The predators evolve in parallel to learn not to eat monarchs.

This is exaptation.

This is also a "just-so story." But I would be surprised if toxicity in most prey animals is not a result of exaptation [there is a tantalizing genetic correlation between toxicity and coloration that broadcasts toxicity--like the red/black monarch wings--which seems to support my argument].


Update: I just found a paper [Ecological factors influencing the evolution of insects' chemical defenses, J. Skelhorn et al., Behavioral Ecology, doi:10.1093/beheco/arm115 ] discussing the evolution of defensive toxicity in insects. It says that the subject is relatively unexplored but gives this insight:

One potential explanation [for the evolution of defensive toxicity] is that chemically defended individuals suffer less from predation than those that do not invest in costly chemical defenses. However, chemical defense often cannot be detected prior to attack, meaning that in order for chemically defended individuals to suffer less from predation than visually similar undefended individuals, they must be more likely to survive predatory attacks. Although there is now some evidence that aposematic insects often survive predatory attacks relatively unharmed and that predators selectively reject prey based on their chemical content, it is currently unclear under what ecological circumstances such differences in survival would allow costly chemical defenses to evolve...
...[A]lthough sequestered chemicals may be stored systemically in body tissues, many species store a large proportion of the chemicals in the integument and wings. This may increase the speed with which predators perceive an individual to be defended and as a result reduce damage to the insect.

This vindicates the premise for my theory but sheds some doubt on the need for it [the paper nowhere discusses the transmission of the gene from the first individual to further generations, but points to evidence that birds taste and release butterflies. And I was right about storing the toxic chemicals, it seems]. I still think exaptation is a cleaner explaination, but pure adaptative natural selection is more plausible in light of this. The paper is really good, it's worth a read.

Update 2: I just realized that toxicity can be manufactured by the prey species, not just acquired through feed. This strongly suggests that the adaptive route is possible.