Which came first, the rattle or the rattling?

Whether or not you’ve seen one in person, you’re probably familiar with the way a rattlesnake uses its rattle. When cornered, a rattler takes a ready-to-strike posture, stiffens its muscles, and quickly vibrates its tail. Of course on the end of that tail is a rattle, or as the authors describe it in a recent paper published in The American Naturalist, “interlocking segments of keratin that fit loosely within one another.” (Keratin is a protein that also makes up your hair, nails, outer layer of skin…and a snake’s scales.) The distinctive rattling sound that’s made when these segments rapidly vibrate against one another causes many animals (including humans) to instinctively freeze and slowly back away hoping to avoid a nasty bite. It’s a very effective warning signal.

western diamondback

A western diamondback rattlesnake (Crotalus atrox) in Arizona, rattling its tail to scare away a perceived threat – presumably the photographer.
Photo from Allf et al. (2016)

Warning Signals

It turns out the behavior of tail rattling is much more common than the rattle itself. Rattlesnakes belong to the Viparidae family of snakes, all of which have long hollow fangs that inject proteolytic (protein-degrading) venom into their prey victims or some other animal stupid enough to mess with a rattlesnake. The rattle on the tip of a rattler’s tail helps it send a clear warning to stay away to avoid those fangs. A second family of snakes called Colubridae contains about two-thirds of all snake species, and they’re generally harmless to anything that’s too big to strangle. This is why you’ll find many kinds of small colubrids—kingsnakes, corn snakes, garter snakes, etc.—in any pet store that happens to be into the whole reptilian thing.

I owned a California Kingsnake (for 21 years, she passed away this past April) and I had seen her rapidly vibrate her tail a few times when she seemed to feel threatened. As someone who studies marine invertebrates, I’m not much of an expert on the biology of snakes. But I did know that kingsnakes don’t have rattles on their tails like rattlesnakes*, so I was a bit puzzled when I saw my kingsnake “rattle” her tail. It made me wonder whether the behavior had evolved before rattles themselves, but my wondering pretty much stopped there and I went on with my own non-snake research.

IMG_1022

The late Irma (1995–2016), my California kingsnake (Lampropeltis getula californiae).

Now researchers at University of North Carolina have done much more than just wonder about tail-vibrating behavior in non-venomous snakes. They subjected the idea to some hypothesis testing, showing that the action of a snake “rattling” its tail did indeed evolve before the rattle itself.

Reconstructing the Past

To figure this out, they took a phylogenetic tree (like a family tree, but representing the evolution of a bunch of different species) of snakes from a previous study and mapped the behavioral trait onto the tree. This statistical method, also known as ancestral character state reconstruction, uses some fancy math to calculate the likelihood that a given trait existed in the common ancestors of living species.

So let’s take a look at the phylogenetic tree in Figure 2 of Allf et al. (2016)1. First of all, each name on the right represents a different species of snake, and the circle next to the name shows whether that species vibrates its tail (white) or does not (black). Notice that all but three species included in the tree exhibit the behavior of tail vibration; even vipers that don’t have rattles and the non-venomous colubrids. Only the group of rattlesnakes has a rattle on its tail. At first glance it seems obvious that the behavior is older than the morphology, but it’s important to test these things statistically… 

fig2_tree

Figure 2. Phylogenetic tree of snakes species in the families Colubridae and Viperidae, showing how tail-rattling behavior has evolved.
As shown in Allf et al. (2016)

Now follow the lines (branches) next to each species, from right to left. Every time two branches come together you’ve reached a node, which represents the common ancestor of those two species. A pie chart on each node shows the probability of that ancestor possessing the trait of tail vibration. You should quickly notice that almost every ancestor has a completely white pie (a single black line in it as opposed to a black wedge) indicating a nearly 100% chance that that ancestor vibrated its tail. These likelihood values provide statistical support for what already seemed obvious, that the behavior of tail vibration evolved long before the physical rattle of rattlesnakes.

But there’s more to this study than simply confirming that rattling evolved before rattles. Seeing how a particular morphological trait functions often makes perfect sense at first glance: wings are great for flying, eyes great for seeing, teeth great for biting, etc. But how do these highly specialized structures arise in the first place? In other words, how does evolutionary novelty come about?

Evolution is NOT Random

There is a common misconception that evolution is completely random, causing many people to be skeptical of how such a wide variety of well-functioning organisms could have evolved on Earth without some sort of divine intervention. You wouldn’t expect to build a computer by randomly throwing circuit boards and wires at each other, right? So how could evolution produce such a huge variety of organisms so perfectly suited to their environments, with no predetermined design or direction? It’s because the view that evolution is 100% random is wrong, there’s more to it than that.

Endless forms

It’s difficult for some people to imagine how such a variety of plant, animal, and microbial forms of life have evolved. Behavior-facilitated adaptation is just one way that evolutionary novelty can come about.
Illustration by Ainsley Seago (www.trufflebeetles.com)

We do know that genetic variation—a crucial component of evolution—arises completely at random (for the most part) by the process of mutation. Most mutations are random mistakes made by cells as they make copies of their DNA before dividing, and the vast majority of mutations have effects that are usually either negative or neutral. It’s very rare for a random mutation to actually improve the survival or reproductive abilities of the organism who gets it.

But another critical piece of the evolutionary process is natural selection, which is anything but random. Natural selection is extremely effective at teasing out even the most subtle advantages among individuals. This means that if a trait has any influence (no matter how small) on increasing the survival and reproductive chances of some individuals relative to others, then natural selection is said to favor that trait. This causes a beneficial mutation to become more and more common each generation, until it belongs to every member of the species. Natural selection is so good at doing this that it can favor changes that are imperceptible to even the most observant scientists with the most expensive instruments.

Can Environment Induce New Adaptations?

Despite the extreme non-randomness of natural selection, it can still only work with what it’s given. In other words, while the environment is incredibly effective at causing species to adapt over time by favoring individuals with the best traits, it can do nothing to make those traits appear in the first place. Or can it?

Well, the rattlesnake study says…sorta maybe. This is not to say the environment directly induces beneficial DNA mutations in an organism (except epigenetics now tells us that it sort of can), but it does appear that tail-rattling behavior provided the perfect backdrop for evolution of the rattle itself. The authors suggest that this might have happened in one of two ways.

The first hypothesis is that tail-rattling behavior as a warning signal simply provided a situation that allowed a rattle to evolve. Let’s say genetic variation among individuals who all rattled their tail meant that some snakes produced a little more keratin on the tips of their tails than others. This could cause these individuals to make a bit more noise when they vibrated their tails, making them a bit better at startling predators. These individuals would survive longer and leave more offspring, some of which would produce even a bit more keratin tissue on the tail than their parents. Continued selection for increased keratin could eventually lead to the fully formed rattle as it exists today. If tail vibration hadn’t existed first, then selection never would have favored individuals who happened to grow a bit more tail keratin.

Alternatively (and more remarkably IMO), rattling behavior may have actively facilitated rapid evolution of the rattle. Like, actually caused the rattle to form. In this scenario, tail vibrations is hypothesized to have caused a sort of callus to form on the tip of the tail. Snakes that vibrated their tails more rapidly and for a longer duration would have developed larger calluses. If certain mutations made some snakes more susceptible to callus formation than others, and if these individuals were also better at deterring predators, then selection would favor callus formation associated with tail vibration.

This would mean that the environment helped cause the formation of the rattle as opposed to simply favoring random mutations that gradually made the rattle bigger. It may sound like a subtle distinction, but in evolutionary biology it’s huge.

While the researchers are not yet able to say which scenario is more likely, more results are consistent with the latter hypothesis. When they mapped two aspects of tail vibration onto the snake phylogeny, they found that duration was longest and intensity was highest in rattlesnakes and their close ancestors:

Allf 2016 7

From Allf et al. (2016)

But if tail calluses only developed when snakes behaved a certain way (i.e. rattled their tails), then we wouldn’t expect them to pass those calluses on to their offspring. That would be similar to your kids inheriting the big muscles you earn by hitting the gym every day. This type of evolution where an individual acquires a trait during its lifetime and passes it to offspring is called Lamarckian inheritance, and it’s something evolutionary biologists disproved decades ago (although read more on epigenetic Lamarckism if you really want to fall down a rabbit hole).

giraffes-are-heartless-creatures-Jan-2012-tiny

A classic example of Lamarckism is the giraffe acquiring a longer neck during its lifetime and passing on that trait to its offspring. We now know evolution doesn’t work this way (mostly).

Of course reality is rarely as simple as we’d like it to be. While natural selection may not have directly induced mutations that grow rattles, it could have provided a strong advantage to snakes that rattled their way to callused tails. And if this selection pressure remained strong for a long period of time, the stage was set to favor any mutation that may have caused even a tiny bit of extra tail keratin formation, independent of the behavior itself. We’d call this genetic assimilation, where a trait that originally forms in response to environmental stimulus eventually becomes hard-wired by the genome. In other words, formation of rattles (the noun) became programmed into the development of rattlesnakes even though their ancestors had to work hard for it, by rattling (the verb).

It may seem counterintuitive that the use of a tool could evolve before the tool itself, but future studies may show that’s it far more common in nature than we now realize. While a resurgence in neo-Lamarckism isn’t likely, it is possible that behavior can be an effective precursor, and even facilitator, of the evolution of novel structures.


*The “king” in kingsnake references the fact they often eat other snakes, including rattlesnakes – they’re resistant to rattlesnake venom (which is obviously badass).

1Bradley C. Allf, Paul A. P. Durst, and David W. Pfennig,“Behavioral Plasticity and the Origins of Novelty: The Evolution of the Rattlesnake Rattle,” The American Naturalist (Volume & page numbers not yet available) DOI: 10.1086/688017

Advertisements