Columns

I’ve heard all my life that dogs can smell fear. If this has been empirically demonstrated, I don’t know about it. However, science has established that fish can. It may only work within a species, and it would be more accurate to speak of detecting schreckstoff than of smelling fear. Still, it’s a neat trick for a fish. 

Schreckstoff—a wonderful word. It has nothing to do with soft-hearted green-skinned ogres; “shreck” is “fright” or “terror” and “stoff” is, well, “stuff,” “matter.” It’s a substance, whose nature was only recently clarified, emitted by an injured fish that alerts other fish to the presence of a predator. 

The word was coined by Karl von Frisch, the German naturalist who decoded the dance language of the honeybee. (Some diehards, notably Adrian Wenner, still deny that the bees are saying anything to each other. Dancing can convey fairly complex information, though; ask any hula performer. And you may have seen one of those Youtube videos in which graduate students dance their dissertations. I recall one in which belly dancers simulated the currents of the Southern Ocean.) 

Anyway, von Frisch, back in 1938, discovered that minnows released schreckstoff when attacked by a predator. The process is triggered by a break in the victim’s skin and results in other members of the school taking evasive action. Most widespread in the large order of fish that includes carp, catfish, and piranhas, chemical alarm signals were confirmed in rainbow and brook trout in the 1990s. 

I first read about the phenomenon when researching an article about salmon and pollutants for another venue a couple of years ago. Nathaniel Scholz, an ecotoxicologist with NOAA Fisheries, reported that young coho salmon (probably in the smolt stage) respond to alarm chemicals from a piece of coho skin smaller than a grain of rice in 100 liters of water. When skin solution is introduced to a clean tank, a salmon, which had been swimming actively, drops to the bottom and hovers in place. But after three hours’ exposure to copper at a concentration of ten parts per billion, the fish doesn’t react—a good way to become someone’s dinner. 

Then last week I was reading the “Findings” column in Harper’s, the back-page feature that reports oddball scientific discoveries. And there was shreckstoff. New research by Suresh Jesuthasan and Ajay Mathuru in Singapore and Rainer Friedrich in Switzerland has isolated its active ingredient: a sugarlike molecule called glycosaminoglycan chondroitin, GAG for short. It’s a chemical relative of the chondroitin sulfate used as a dietary supplement for treatment of osteoarthritis. 

The organism they used was the guppy-sized zebra fish, a minnow relative and a kind of finned lab rat. Ground-up zebra fish tissue was subjected to a series of chemical tests-by-elimination. The researchers destroyed all the proteins in one sample, but it still elicited escape behavior in live fish. So schreckstoff couldn’t be a protein. But when they broke down GAG chondroitin with enzymes or neutralized it with antibodies, the fish didn’t react. 

Friedrich and his colleagues were also able to locate the zebra fish’s shreckstoff receptors: specialized neurons called crypt cells in the olfactory bulb, the part of the brain that processes odors. These neurons responded when the fish were exposed to schreckstoff but not to other odors. When you consider the size of the zebra fish brain, measuring any kind of neural activity must be quite a feat. My hat is off to these guys. 

The release of schreckstoff can be considered a kind of automatic altruism, warning other fish of imminent danger. This could become fodder for the renewed debate over kin selection versus group selection in evolution (E.O. Wilson has abandoned kin selection for group selection; big scandal), since the fish in a school are not necessarily close relatives. It’s also been hypothesized that that GAG chondroitin has something to do with the skin’s immune system, and that its role as an alarm signal is a secondary exaptation. I’m not taking sides here.