Friday, January 4, 2013

Elephant Walks into a Smell Lab . . .



Professor Matthias Laska has made a big name for himself in the tiny world of comparative olfactory psychophysics. He has done this by measuring the olfactory ability of all sorts of non-human species, including mice, honey bees, squirrel monkeys, pigtail macaques, spider monkeys and South African fur seals. His specialty is designing behavioral tests in which the animals can demonstrate their olfactory talents, such as ability to discriminate between test odors or to detect very low concentrations of a single smell.

By testing human volunteers in similar ways, Laska is able to quantitatively rank our skills against those of other species. I relied on his work when, in What the Nose Knows, I argued against the old Greco-Freudian myth that humans have a relatively poor sense of smell. Laska’s work shows humans to be extremely competitive with other mammals.

Laska has recently turned his attention to elephants. How do you test an elephant’s sense of smell? [Very carefully.—Ed.] Laska and his colleagues use a “food-rewarded two-choice instrumental conditioning paradigm.” For those of you who want to try this at home, here’s how. The elephants were trained, upon verbal command,
to sniff at 2 odor sampling ports and then to indicate which of them held the rewarded odor stimulus by placing the tip of their trunk onto a defined position above the corresponding odor port.
Easy peasy. Correct choices are rewarded with a carrot.

In a newly published paper, Laska et al., take an extensive look at the Asian elephant’s ability to discriminate among odorants with similar molecular structure. Laska’s standard tactic is to test “homologous series” of simple alcohols, aldehydes, ketones, and carboxylic acids, i.e., he compares molecules of a given carbon chain length but with different functional groups. He also tests the discriminability of enantiomeric odor pairs, e.g., (-)-carvone and (+)-carvone. These are mirror image versions of the same molecular structure and often have different odor character.

The new results are easily summarized: elephants successfully discriminated all the odor pairs, both the homologous series and the enantiomers. Compared to other species tested similarly by Laska,
the Asian elephants performed at least as well as mice and clearly better than human subjects, squirrel monkeys, pigtail macaques, South African fur seals, and honeybees.
So there! Laska et al. mention some unpublished results showing that elephants are not particularly good at telling apart different concentrations of the same odor. Still, this is a pretty impressive performance.

The study discussed here is “Olfactory discrimination ability of Asian elephants (Elephas maximus) for structurally related odorants,” by Alisa Rizvanovic, Mats Amundin, and Matthias Laska, published online in Chemical Senses, December 16, 2012.

UPDATE January 5, 2013
Professor Laska emailed me regarding elephant’s inability to distinguish different concentrations of the same odor. Apparently, once an elephant figures out that odor X is rewarded he has trouble learning in another test that a different concentration of X is not rewarded. This problem also occurs with most of the other species Laska has studied. It makes sense: in nature the reward value of a particular smell is unlikely to vary with its concentration.

No comments: