Octopus (21 page)

The other source of our attitude toward animals is a holdover from seventeenth-century philosophy, specifically that of French philosopher and scientist René Descartes, who stated that animals are no better than automata (pieces of machinery) and that therefore we may treat them as we like. He separated humans, who have souls and minds, from animals, which are just bodies. Animals were assumed to be simple and their reactions automatic, and lowly invertebrates were even simpler. This attitude made conducting research on animals easier to justify.

From two different perspectives, psychology and ethology, students of animal behavior in the mid-twentieth century continued this tradition of seeing animals as things. Psychology was dominated by behaviorism. The animal was a black box into which we didn't need to peer, with no mind to complicate the learning process. Scientists put stimuli in and responses came out; they could make and predict learning curves and learning laws but not worry about individuals. Personality was just “noise,” variation that could be eliminated by getting a large sample size. Ethologists of the time were equally uninterested in individual variation. They wanted to study the species-typical patterns of behavior that happened in the animal's natural environment.

When scientists studied octopuses as simple organisms, they didn't fare well using either of these approaches. For instance, psychologists had the simple belief that with more training, the choice-reward link for the subjects would get stronger. As we know, in octopuses there's no clear choice of prey. In the lab, in a simple choice-rewarded experiment, an octopus spent some of its time checking the unrewarded stimulus and never got to the accepted criterion of eight out of ten correct choices. It stayed curious about what else might be going on and checked the alternate regularly. Cephalopods have so far shown so little stereotyping of action sequences that their behavior doesn't fit the fixed action pattern model of ethology, either.

When and why did scientists abandon pursuit of the average output for the group? For one thing, the ideal of animal brains as stimulus-response relays was fading fast in the second half of the twentieth century. Clearly, learning was not the same in all species, and the idea of constraints on learning was an exciting new one. People began to wonder whether it was a bit limiting to regard mind and consciousness the domain of only humans. We have always looked for exclusive abilities of humans, from language to tool use to consciousness, only to discover that we share these abilities with other animals. If we share 99.5 percent of our genes with chimpanzees, it's not surprising that they share many of our abilities. Donald Griffin's claim (in 1981) that animals must have minds although simpler ones than ours, and so they could have awareness, was scorned for years, in part because it's impossible to prove. But this new approach made a dent in the view of animal as machine.

It became clear that intelligent animals, and some we would class as not very intelligent ones, didn't just give automatic reactions to external situations. Otters play, rats explore mazes before they make their choices, even bees seem to have memorized a map of their neighborhood. Scientists started examining different strategies for getting variable reactions to the same goal, and a whole new area—game theory—grew as part of a new direction of behavioral ecology. Given a difficult but important choice, like aggressive competition, an animal might choose an evolutionarily stable
strategy, the same one each time. Or it might vary the strategy depending on the actions of others. In aggressive contests, one animal might choose a hawk strategy (attack first) or a dove strategy (hang back and make peace).

Similar variations in sexual strategies result in mostly male animals of many species responding to sexual selection and having two basic choices. They either dominate because of size and ability and hold territories or defend females, or, being smaller and weaker, hang around as satellites or even “sneakers” that look like females, waiting for a quick chance at reproduction. Giant cuttlefish (Sepia apama) use some of these strategies, with smaller males staying near consort pairs and pestering females, while larger males ward them off. Given the chance, females of that species will take the quick mating opportunity with a sneaker male—they like variety in their mates. Octopuses are less obvious about competition, but males have been seen fighting at females' dens. Female octopus sexual strategies center around choice—deciding which and how many males to accept, and when and for how long.

Studying animal strategies has helped us identify variability in behaviors, such as the octopus's food choice. Handy as the average is, it seldom matches the dynamic nature of what's going on in the real world. Behavioral extremes may matter a lot, especially if you have a variable environment, such as the subtidal areas of Alaska where some giant Pacific octopuses live, and selection on the individual level may batter the mean from year to year. Bottom trawling for oysters can drastically change a marine habitat, for example, wiping out several food sources so only generalists like the octopus will survive.

To study individual differences among octopuses, we realized we had to find variation, not minimize it. So we needed to bring octopuses into the lab to give them all the same environment of an aquarium tank and put them in the same simple situations. We settled on three areas: Alerting—opening the tank lid, Threat—touching the octopus with a test-tube brush, and Feeding—dropping a crab into the tank.

Our choice of octopus species to test for personality stemmed from convenience rather than deliberate selection; we worked with what Roland had available at the Seattle Aquarium. We chose the small red octopus rather than the giant Pacific octopus because of size: you can keep about fifty little red octopuses in the same volume of water as one giant Pacific octopus. We would have liked to choose our test species based on which would give us more variability. Every aquarium keeper or researcher who has worked with more than one species knows they are different, and Roland and James have been compiling the trends. Besides being great escape artists, common octopuses are notably active and feisty. Other species, such as the red octopus and the giant Pacific octopus, are less so, though one member of Jennifer's pygmy octopus species would kill another if two were kept in a small space. Still, we needed a large number of octopuses to do these tests, so we chose the red octopus and tested forty-four of them.

Roland conducted the tests for the study. He kept three octopuses at a time, each in its own tank, in the basement of the aquarium, and he tested every second day over two weeks. When finished with each phase, he took the octopuses back to the bay, let them go, and caught another three from a different place. About once a month, Roland mailed Jennifer the numbers to turn into data sets. It took three years to test all those octopuses and record the presence or absence of nineteen likely behaviors when each animal was exposed seven times to three different situations: Threat, Alerting, and Feeding.