Canadian researchers have for the first time successfully conditioned free-living juvenile salmon to recognize a novel odor. The work was published in the Journal "Animal Behaviour". This fantastic feat was accomplished by graduate student Antoine Leduc in the lab of Grant Brown bravely diving into the current to condition the animals in their natural environment. Antoine Leduc chose an approx. 500m long stretch of a little river called Catamaran Brook and trained 40 juvenile salmon which he had individually marked earlier. To condition the animals, he located them one by one in the stream by snorkeling. Once he had found one, he'd release an alarm pheromone together with lemon essence just upstream of the animal. Half of the animals received this treatment and half of the animals got the lemon scent only with stream water. The idea was that the fish would associate the lemon odor with the alarm pheromone and hide, whereas the other half of the fish wouldn't care too much about the lemon odor. All fish were videotaped before and after the substances were released into the water. At least 24h later, the animals were tested for single-trial classical conditioning: Antoine Leduc again located each animal in the stream, videotaped it, but this time only released the lemon odor, and taped the animals again.
Once back in the lab, the researchers evaluated the videos for evidence of predator avoidance: how much did the animal feed, move or hide? It turned out that after training, the two groups of salmon differed substantially in that the animals which had smelled the lemon together with the alarm pheromone were feeding less, moving less and hiding more compared to the group which had only received lemon and stream water.
These results serve to indicate that juvenile salmon can in principle learn to avoid an odor from a nearby predator if the predator releases such an odor and feeds on conspecific salmon nearby. This capacity is probably extremely valuable if visibility is low in the stream. The more disturbing are recent findings by the same group of researchers that a shift in acidity (due to acid rain) completely abolishes the effectiveness of the alarm pheromone, leaving predator avoidance only to vision.