Not just "stamp collecting", or Something fishy about that poison

The Times reports that Venom Runs Thick in Fish Families:

The study, published in June in The Journal of Heredity, analyzed and compared DNA sequences from 233 species and used the results to create a new family tree for spiny-rayed fishes. That group includes many types of toadfish, scorpionfish (lionfish are a type of scorpionfish), surgeonfish, rabbitfish, jacks, stargazers and saber-toothed blennies.

The family tree shows how the species are related, and which evolved from the same ancestor. Based on the tree, the researchers predicted which species should be venomous. Then, to test their predictions, Dr. Smith dissected 102 specimens, looking for venom glands and delivery systems like spikes, fangs or sharp fins.

Of the 102 species he examined, previous research had suggested that 26 were venomous. But the new analysis predicted that 61 would be venomous -- and the dissections bore that out.

This is the sort of result that made my experiment with the Evolution Project so much fun. I'd find oddball research which had, without ceremony or todo, shown how vital and central evolution is to all of biology.

A common critique of biology that I'd hear when I told people I planned to major in biology is "I always liked it, but there was too much memorization." What evolution teaches us is that it's only about memorization if you're doing it wrong.

Click through for guidance on doing it right.

Before Darwin and Mendel and their synthesis in the 20th century, biology was largely about collecting disparate facts. Without some unifying theory, it was impossible to synthesize facts, so all one could do is memorize.

The image at right is a phylogeny of the poisonous fishes, stolen shamelessly from the original paper (the image above is based on two pictures in the original paper as well). Blue branches on the tree are those with poisonous spines. Green, magenta, turquoise, and orange areas represent other poison apparatus, and representative members of each poisonous clade are illustrated.

What is obvious even at low resolution is that possessing poisonous spines is not randomly distributed across the entire tree. If one individual in a clade (all of the descendants of a common ancestor) have poisonous spines, it's almost certain that other members of the clade do as well. I can only find one poisonous species that doesn't have a poisonous sister species.

And as the Times article pointed out, the only reason you'd bother looking in those particular species for poisonous spines is an obvious prediction from evolution.

Drs. Smith and Wheeler present their research not just as an insight into the history of these fish, but as "A Phylogenetic Road Map for the Bioprospecting of Piscine Venoms." They write:

To date, most venom bioprospecting has focused on snakes, resulting in six stroke and cancer treatment drugs that are nearing U.S. Food and Drug Administration review. Fishes, however, with thousands of venoms, represent an untapped resource of natural products.

Knowing where to look relies on solid science, and evolution is the science they and pharmaceutical researchers will be relying on.

They will be doing the same thing I tell to students I TA. Don't memorize the traits that every individual group of organisms has. Learn the pattern of evolution and understand how the various characters you need to know map onto that pattern. You will memorize less, you'll be able to guess at more things, and you'll be better able to integrate new information about those groups and different groups as it comes up.

That's how a science should work, by making testable predictions, then testing them.

Like a good scientific discovery, this also raises new questions. Sticklebacks, eel pouts, stablefish, sculpins and greenlings all fall into a clade nestled between the scorpionfish/lionfish clade and the stonefish/waspfish clade. The parallel groups are poisonous, but the middle one isn't. Did poison evolve separately in each poisonous clade, or did the non-poisonous clade lose the ability to make poison? What about other non-poisonous groups with poisonous relatives? Tracing the evolution of the poison-making genetics in other fish might yield clues about the ways that these poisons and the fish that have them evolved long ago.