Nature keeps evolving crabs and the internet is obsessed. What’s going on here?

As XEC, the latest COVID variant takes hold, we are watching viral evolution play out on a time scale short enough to follow, with different strains of the SARS-CoV-2 virus independently acquiring similar or functionally similar mutations that improve its ability to infect us or to evade existing vaccines.

This is the same process that occurs over tens of thousands, even millions of years in living creatures, from slugs to dogs to you and me, producing the incredible diversity we see in the tree of life along with startling replays of the same idea, known formally as convergent evolution. This is simply when nature finds similar solutions to similar problems in evolutionarily distant groups — think about how dolphins and bats each evolved echolocation, despite being unrelated. 

One of the most prominent — and pinchy — ways this manifests is known as carcinization, the idea that nature keeps evolving crabs. Indeed, a crab-like body shape, or morphology, has evolved numerous times independently throughout evolutionary history. From an outsider’s view, it seems like crabs appear so often because Mother Nature “loves” crabs. In the immortal words of English zoologist Lancelot Alexander Borradaile, who coined the term, carcinization is “one of the many attempts of Nature to evolve a crab.”

The concept is so intriguing and delightful, it has spawned the crab meme, which swept some little nerdy part of the internet a few years ago and with it, a wacky speculation that we are all going to evolve into crabs one day. But all bizarre fantasies aside, what’s really happening here is far more interesting.

Cancer the crab

It goes without saying, nature is not consciously trying to evolve anything. Even human intelligence arose through the randomness of natural selection. With all due respect to Borradaile and his fans, that’s not how this works. Rather, if the same sort of thing is evolving over and over, it’s probably because that sort of thing is a trait that offers a survival advantage to species existing in similar situations.

Convergent evolution is what we see when we observe that bats and birds have similar development of their arms into gliding wings, which initially allowed them to glide, and then to fly. Or when we notice that the extinct ichthyosaurs, prehistoric fish, have a very similar body outline, down to the bottlenose shape and tiny teeth, as the modern dolphin — which is not a fish at all, but a mammal. In either case, the hydrodynamic body shape lets them both swim rapidly over long distances.

Another example can be seen among the marsupial mammals of Australia, an island where animal evolution diverged from the rest of the world far back in evolutionary time, we see creatures with eerie parallels to mammals from other continents, creatures that occupy the same ecological niche or role and have evolved similar body shapes or abilities to cope.

Yellow crab on the beach (Getty Images/Bob Stefko)We see this as well in the existence of many types of decapods, which is the technical term for crustaceans with a crab-like body shape. This includes crabs, which evolved from the common ancestor of all crabs, and it includes other kinds of crustaceans with completely different ancestors, occupying a different branch of the evolutionary tree — yet a total plagiarism of the idea of crab.

Carcinization is “really only applicable to that one animal group that we call decapods,” Sebastian Groh, a paleontologist at Cardiff Metropolitan University, told Salon. “And that’s the only group that it’s actually limited to. It doesn’t really occur anywhere outside it. I think that was sort of a misunderstanding.”

Groh’s area of study is the evolution of crocodiles and their relatives from two hundred million years ago to now, looking for example at the convergent evolution of long, narrow snouts in various different branches of their family tree.

Want more health and science stories in your inbox? Subscribe to Salon's weekly newsletter Lab Notes.

The idea that anyone (or any random evolutionary group) might evolve into a crab is a misunderstanding, sure, but makes for a striking meme — and maybe a fruitful opportunity to explain how evolution actually works.

Small changes, big impact

The evolution of any particular trait depends on a huge number of tiny changes. You don’t just have a gene for “looking like a crab.” Rather, for evolution by natural selection to produce crab-like appearance and behavior you would need to have creatures living in environments that give a survival advantage to having these crab-like traits. And you’d need to acquire all genes that code for the many proteins that produce such traits. To understand this, we need to look way closer.

The development of a claw or a flipper requires many particular different genes. And once again, no, the human genome is not a few mutations away from carcinization.

“You’re not going to find a mammal becoming a crab because maybe [the ancestors of crab-like organisms have] got a lot of other genes that would predispose them to that morphology and to that sort of behavior. They’ve got the set up. They’ve got the background,” James McInerney, who holds the chair in Evolutionary Biology at the University of Liverpool, told Salon in a video interview.

Then again, we have to wonder whether that set up of genes evolved through natural selection — Nature “attempting” to evolve a crab — or just random luck. As the authors of one 2016 paper on convergence at both the molecular and the more observable, morphological level put it, “convergence is caused by either repeated adaptations of different evolutionary lineages to similar environmental challenges or chance.”

McInerney was lead author on a study on convergent evolution in bacteria that helps us understand how this background might work, and to distinguish these two ways convergence may occur. He and his team used machine learning to look at the genomes of a whole bunch of different strains of Escherichia coli, a bacteria in which different strains repeatedly evolve in convergent ways, to see whether this occurs by chance or by a process of natural selection. 

Now, in bacteria — which are a type of prokaryote, or single-celled organism — a lot of evolution happens by horizontal gene transfer. This occurs when genetic material is incorporated into an organism’s genome in some way other than through reproduction. Bacteria pick up new genes from various sources, keeping their evolution interesting. This isn’t necessarily good for us, though: antibiotic resistance has become a massive problem in large part due to horizontal transfer of genes from drug-resistant species into bacteria species that were once routinely killed by medications.

In E.coli, this means that its pangenome — the totality of genes that are found across all strains — has a huge amount of variability. What McInerny’s team found was that despite all this variability, you could actually predict a lot of what genes you’d find in a particular strain if you knew some of the other genes. To simplify a bit, if Bacteria A acquires gene 1 by horizontal transfer and also acquires gene 2, and then we notice that Bacteria B has also acquired gene 1, we might correctly predict that Bacteria B has gene 2 as well — because in this scenario, genes 1 and 2 tend to stick together in a genome.

In the authors’ words, “at least part of the pangenome can be understood as a set of genes with relationships that govern their likely cohabitants, analogous to an ecosystem’s set of interacting organisms.”

What this means is that evolution in E. coli, despite the serendipity of horizontal gene transfer, isn’t just a matter of chance. Rather, there are deterministic relationships in which certain genes go well or badly with other particular genes, and so similar patterns of genes reoccur, resulting in repeated patterns of evolution. As a result, where you see convergently similar bacteria that evolved from very different ancestors, you actually see the same types of genes in these very distant relatives. Despite their different evolutionary history, they come up with the same or very similar genetic recipes when faced with similar survival challenges.

The details of convergent evolution might of course be more complex in the eukaryotes, multicellular organisms like humans or decapods with large genomes, the total genetic material of an organism. Most traits you can actually observe — what’s called the organism’s phenotype — result from a unique combination of genes and how those genes are expressed.

“I suspect that in eukaryotes, it won’t be just point mutations, either,” McInerney said, referring to small changes in a genome. “It’ll be changes in expression and genetic changes that influence other genetic changes to make them more likely or less likely. I think that’s a really productive field of research right now.”

Tim Sackton, director of Bioinformatics for the FAS Informatics Group at Harvard University, notes that while we can easily look at a sequenced genome and identify where the genes that code for proteins are, the other parts that control where and when these genes are expressed is something we’re still trying to figure out.

”We don't really know the code for these in the same way for these regulatory regions,” he told Salon. And those regions may be very important in understanding convergence. Take for example the puzzle of the flightless birds. Evidence suggests that the loss of flight evolved independently as many as six times, rather than just once, in the ancestors of different ratites — the group of flightless birds including the extinct moa and elephant birds as well as the ostrich, kiwi, cassowary, emu and rhea. In a 2019 study, Sackton and colleagues found that regulatory elements — the ones that determine when a gene is expressed as a protein, like an on-off switch — were where the action was.

“We’d see the same elements would get altered in multiple of these independent transitions to these flightless birds and it’s not only the same element,” Sackton explained. Rather, certain genes tend to accumulate in flightless birds, like clusters of modified elements of the genome.

The loss of flight — which plays out at the phenotype level in changes in the forelimb, for example — thus resulted from way more than a single change in the genome. Instead, various genes work together, with regulatory genes playing the more significant role.

“There’s a lot going on in these ratites, there’s many changes, both in terms of skeletal morphology but also in terms of feather structure and a lot of other things. So it’s a very complicated phenotype, but not every aspect of it is necessarily convergent,” Sackton noted.

There’s still so much to figure out about how convergent evolution works, not just in crabs but in all organisms. And the results can be striking.

“The crabs are showing that’s a very overt phenotype, right?” McInerney said. He emphasized that the convergences we see — so many things that look like crabs, the camera-like eyes that evolved independently in both squids and humans, the emergence of opposable thumbs in giant pandas, chameleons and us — emerge from similarly astonishing, if less obvious, evolution at the molecular level: in the once-hidden world of genes.

So why do crabs keep coming up, again and again? Not because nature chooses it, but because the crab “design” just works so well at keeping certain species alive and passing down those genes. It’s the intricate beauty of evolution in action. As McInerney put it, “Genome evolution favors particular outcomes, and we see it in bacteria. We see it in crabs.”