by Marlene Zuk
Franks and his colleagues demonstrated this ability of ants to plan by showing that they can distinguish between various qualities of nests and perform reconnaissance when they evaluate the possibilities. The official definition of reconnaissance, according to the Oxford English Dictionary, is "an examination of a region to ascertain strategic features through a preliminary survey," and the ants manage this by retaining information about the different potential nests and using it later to recruit for different sites. Like a person who sees an ice cream store while jogging in the morning and files away the information for dessert, the ants can remember the landmarks near a potential nest site or the odors left there, even if they are not actively house hunting at the time. If these cues are removed by an investigator, and the ants are then required to find a new home by virtue of the scientist destroying the old one, the ants can no longer distinguish among the sites.
Ants can also arrive at the best group decision almost as a by-product of individual behavior, without the need for extensive communication. Say that a tasty bit of food can be found at the other side of a deep crack in the ground from the colony. A leafy branch lies across the crevice, and the ants can either take a shorter, direct path to the food or a longer, more convoluted one, depending on which twig they use to cross. The shorter path would be more efficient, and it turns out that this is the one favored by the ants. But how did they arrive at the decision? Even for a devoted myrmecophile, it defies reason to imagine the ants testing out one path and then the other, timing both, and then sending the message to the rest of the colony that they can save their exoskeleton some wear by taking the shorter journey.
It turns out they don't have to. Franks and several other scientists determined that a much simpler process is at work. As an ant returns from a food source, she lays down an odor trail that attracts her nest mates. The more ants that have been back and forth from the colony, the stronger the attraction of a particular pathway. Thus, the shorter trip over the twig gets more use and builds up more odor, simply because it takes less time to go to and from the food, and the ants themselves reinforce the easier path as the best choice. Others follow and, voilà, the colony as a whole has made the right decision. Similar behavior allows the ants to select the easiest sites to excavate when the possible nest entrances are blocked with sand.
At least one other species of ant, the delightfully named gypsy ant, can make collective decisions about which kinds of food can be harvested singly and which require enlisting reinforcements. A group of French and Spanish researchers presented the gypsy ants with dead crickets, which could be moved by a cooperating group of ants but not by a single individual; dead shrimps, which are five hundred times heavier than a worker and must be butchered into individually transportable pieces; or sesame seeds, which as any picnicker knows can be easily borne aloft by an ant acting alone. Making off with food in a timely and efficient manner is important, because other ant species are potentially lurking nearby, ready to snatch any food left unattended. The ants were able to gauge the number of workers necessary to lift and carry the crickets depending on the size of the prey, with small crickets requiring about a dozen workers but large ones fifteen or more, and they quickly recruited an even greater number of ants to carry out the dismemberment of the shrimp before it could be detected by competitors.
Not all group decisions by insects have such a happy outcome. Although the social ants and bees get most of the attention, scientists have also examined collective behavior in forest tent caterpillars, which live in groups until they spin their cocoons and become adult moths. The caterpillars move in munching hordes through the treetops and may either linger on a particularly succulent tree or move quickly through it in search of a more nutritious set of leaves. Because of the caterpillars' discerning tastes, forests that have been attacked by the caterpillars are often a patchwork of ravaged and intact trees. In nature, experiments have shown that they prefer carbohydrate-enriched or untreated aspen leaves, rather than leaves with a high protein content, a kind of anti-Atkins diet. Offered a choice between diets concocted in the laboratory that differ in their nutrient composition, an individual caterpillar will make the "right" selection and eschew an unbalanced low-carbohydrate food in favor of one with the natural blend of proteins, carbohydrates, and fiber.
In groups, however, the caterpillars, like schoolchildren egging each other on to eat Doritos and Twinkies instead of carrot sticks, will often end up choosing the less nourishing offering. The problem seems to occur because, like ants, the caterpillars follow odor trails left by their companions. The initial decision to taste one or another of the foods is made at random, but once a caterpillar has started eating, its odor trail encourages others to follow, and then the entire gang gets trapped by heeding the message that went before it. The caterpillars thus follow each other to their collective nutritional doom. Unlike the bees or ants, the caterpillars lack any capability of communicating their state to each other, so they cannot indicate that they have arrived at a less tasty branch and warn others of their folly. Nietzsche's pessimism about groups seems to be better illustrated by the caterpillars than the bees in this regard, which makes you wonder whether we are so close to the social insects after all. Luckily, the caterpillars differ in their tendency to move around, and if a large proportion of the group was of a more active predilection, the group itself was less cohesive and managed to escape the poor decision.
Flying versus Walking, and the Lead-up to Language
DECISIONS about food or nest sites are closely tied to the success or failure of any given insect colony, and the way that different species get help when an individual finds a food item that is too big for it, or needs to get everyone else on board with a decision has important implications for social behavior. The way that an insect recruits is, in turn, constrained by its own biology. Ants, as I've discussed, lay down an odor trail that becomes stronger and stronger as more workers use it, but to each ant that traverses the trail, its end point is a mystery—she simply follows her nose, so to speak, until she reaches the goal. In the case of establishing a new nest, ants can be carried by their nest mates to the new site, and again those being unceremoniously tucked under a leg need have no idea of where they are being taken.
Bees are different. Flying instead of walking means that you can't easily haul your sister workers around, which means that the bees need some other way to convey information to the rest of the colony. And although some species of bees do place dabs of odor on plants and other objects as signposts on the way to a food source, pheromones are not nearly as satisfactory a method for indicating directions for flying insects as they are for crawling ones; the bees have to continually dart down to the vegetation, and the odors fade without continual reinforcement by a stream of workers.
What's more, at least some kinds of bees have to worry about eavesdroppers on their odor cues. James Nieh at the University of California at San Diego has been studying tropical stingless bees in Brazil, Panama, and other parts of Latin America for many years. The stingless bees are social, like honeybees, and Nieh noticed that the species he was studying left scent marks near good food sources. The problem was that the scent marks were easily detected by a larger and more aggressive species of stingless bee, and when the bullies found the food, they dispatched their victims with what Nieh describes as "a range of forms of aggression from threats to intense grappling followed by decapitation." The victim species avoids the odor marks left by the aggressor species and sticks to its own signals, but the aggressor does the opposite, preferring the odor marks of the victim species to its own.
What's a bee to do? One possibility is to encrypt your directions. Instead of setting out an odor that broadcasts "Tasty morsel here!," whisper your findings only to those for whom the message is intended: your nest mates. In other words, evolve a symbolic language with which you can convey what you know to others in the privacy of the hive, without fear of being overheard. Nieh suggests that the famous dance lang
uage of the honeybee, and its counterparts in a few other species, evolved under pressure to hide indications of the location of rich food sources from any competitor bees in the area. Ideally, of course, one would have a code able to be read only by the members of one's own colony, but that degree of encryption seems to be beyond the bees, and so they have had to settle with having species-specific, or at least population-specific, signals. Combined with the other advantages of such communication to a flying rather than walking insect, for example, the inability to carry other workers and the inconvenience of odor trails over long distances, the dance doesn't seem like an anomaly, but like an obvious solution to a problem.
Ants and bees differ in a few other respects: ants are much slower than bees at redirecting their efforts to a newly introduced rich food source, and the members of an ant colony act almost like neurons in the brain when responding to stimulation. Ants also exhibit something called stigmergy, which sounds like either an eighties band or what happens when the recipients of social stigma gather in groups, for example, smokers outside a building, but is the way that the ants coordinate each other's movements by changing the odor trails that convey activity patterns. This too means that the ants can make decisions without resorting to the direct exchange of information among individuals.
Bee Spoke
REGARDLESS of the waggle dance's evolutionary origin, the idea that bees could possess a symbolic language has never been simply relegated to an incidental by-product of their flying existence, a serendipity of evolution. Anthropologists endlessly debate whether it is possible to have thinking without language, whether one has to be able to formulate thoughts into something resembling words to be truly sentient. And they take enormous pains to define what makes our language special, and how it can be the one holdout in making humans different from all other animals. But the bees make us ask instead whether it is possible to have language without thinking, since even the most ardent admirers of the waggle dance do not maintain that the bees' cognitive capacities mirror our own. So do the bees speak? And if so, does it mean we have to admit them into a special club, unlike any other animal?
Although many beekeepers had noticed that single foraging bees seem to advertise the location of nectar-rich patches of flowers to the rest of the hive, the first detailed description of the forager's performance was made in 1919 by the Austrian scientist Karl von Frisch, who shared the 1973 Nobel Prize in physiology and medicine with the ethologists Konrad Lorenz and Niko Tinbergen for his accomplishment. He was able to carefully track the movements of individual bees by placing his colonies in glass-walled observation hives and marking the bees with either dabs of paint or tiny numbered circles that he glued to their backs.
Von Frisch noted that when a worker bee returned to the hive after visiting a rich food source, she performed a stereotyped series of movements on the surface of the honeycomb. If the food source is close by, less than 50 yards or so, she did a rather simple "round dance," in which the forager runs in narrow circles. More distant food patches warranted a "waggle dance," which contains information about both the distance of the food from the hive and the direction in which it lies. The waggle dance consists of a straight run followed by a semicircle first to one side and then another, in a rough figure eight, with the bee waggling her abdomen energetically during the straight run.
The length of the run is correlated with the distance of the food from the hive, while the angle of the bee's body relative to vertical indicates the angle between the sun and the food source. The vibrating wings of the dancing bee also convey auditory information to the rest of the hive; silenced bees do not recruit others to the food source, and it makes sense that sound would be needed, since the inside of the hive is dark and the other workers cannot simply watch what the dancer is doing. Once the dancer has completed her performance, other bees venture out of the hive and go, more or less directly, to the location she indicated.
In other words, the bees seem to have symbolic representations for the distance and direction of the food, which fits many if not all of the criteria for an actual language. This was big news. Historian of science Tania Munz points out that during the 1960s, bee language was "the most widely studied form of animal communication and some deemed it the most complex second only to human speech." Even Carl Jung took note, musing that we would interpret the bees' behavior, if it occurred in humans, "as a conscious and intentional act and can hardly imagine how anyone could prove in a court of law that it had taken place unconsciously.... Nor is there any proof that bees are unconscious." Those with a yearning to see the waggle dance for themselves need look no further than YouTube, of course; one video of a dancing bee had nearly eighty thousand hits, and enthusiastic if sometimes inadvertently ironic comments such as, "I couldn't do that. Bees are smarter than me," "Why would you shake your butt as communication, weird," or, even better, "Wow. Their [sic] smart."
Although von Frisch's discovery was mainly greeted with amazement and rather uncritical acceptance by both the general public and other scientists, a few remained skeptical that the bees were truly capable of using the sophisticated information encoded in the dances. Foremost among these was Adrian Wenner, a professor at the University of California, Santa Barbara—and, in the spirit of full disclosure, my former teacher and mentor as an undergraduate. A soft-spoken but determined man, Wenner did not dispute the information contained in the waggle dance; he could observe a returning forager and calculate the distance and direction of the food patch perfectly well himself. He just didn't think the bees were using the information.
Wenner claimed that a much simpler explanation for how the bees found the food existed: the other workers simply smelled the odor that lingered on the recruiter's body, left the hive and flew, sniffing the air, until they perceived the same scent emanating from a patch of flowers. The experiments that von Frisch and other scientists performed demonstrated merely that the bees found the food, he said, not how they did so. His hypothesis was much more parsimonious, and hence, Wenner concluded, scientists were obliged to use it rather than the more elaborate explanation that required talking bees. Why, then, did the bees dance, and why did the dance contain information that was interpretable by humans, if the bees didn't use it? Wenner would always smile an impish smile when asked that question, and point out that nature did not evolve for a purpose—to suggest that it did was teleological and unscientific. The dance didn't have to be used by the bees in the way we could use it; a cricket's call can be used to calculate the temperature because he sings more quickly when it is warmer, but no one has ever suggested that the crickets evolved their chirps so that they could act as thermometers.
Wenner's iconoclastic views were not particularly popular, which he also attributed to an unscientific bias toward wanting to believe the more dramatic and exciting story of an insect language. Eventually, however, scientists began to pit the two ideas against each other. To some extent, it is unfair to claim that von Frisch dismissed the use of odor cues by the bees, since he did acknowledge its role in some of his papers, and indeed most researchers acknowledge that the bees in the hive do not ignore the information contained in the scent of a returning worker.
Many biologists were convinced that the bees do indeed use the information in the dance by some experiments published in 1975 by James L. Gould, in which he manipulated the dancer to "lie" about where the food was located using a flashlight to mimic the sun and, hence, alter the angle at which the dance was produced relative to the sun's actual position. Wenner was unconvinced, suggesting that the experiment was never replicated, and he and a few other scientists also claimed that Gould did not fully control for the bees' use of odors as an alternative explanation.
Several scientists have tried to manufacture artificial bees that could be made to dance inside the hive to further test the hypotheses, and one of these was able to recruit at least some bees to the food source it was programmed to dance about to the rest of the colony. Wenner once again dismissed th
ese findings as inconclusive, and it is certainly the case that the mechanical bees didn't do the job nearly as well as a real one.
The conclusive set of experiments, at least in the majority of scientists' view, came from H. Esch and colleagues, who were able to manipulate something called the optic flow perceived by the bees. Bees measure distance by gauging the way images in the environment move across their eyes as they fly, rather like clocking the trees that tick by the windows of a moving train. The scientists trained the bees to fly through a tunnel lined with a black and white pattern that presented an optical illusion to the insects, making it appear that they had flown a longer distance than they actually had. When the fooled bees got back to the hive, they produced a dance that indicated the food was farther away than it was. The recruits promptly flew to the wrong site, indicating that they had indeed been misled by the dance itself.
Yet other studies used harmonic radar to track individual bees and the flight paths they took to the feeder or flower patch; these showed that most of the bees recruited by a dancer took a straight path to the food, rather than zigzagging back and forth the way they would be expected to if they were simply using the odors in the air to find the patch that smelled like the dancer inside the hive.
Finally, my friend Kirk Visscher and a former student of his, Gavin Sherman, demonstrated that the waggle dances help the bees survive in nature. In a clever experiment, they used lighting to mimic the sun and misdirect the bees, so that the dances didn't help the members of the hive to find a food source. They allowed a control set of colonies to dance appropriately. At the end of the season, the deceived colonies had accumulated significantly less honey than those in the control group, an important consideration in the well-being of the hive.