Cops with Six Legs

Law and order among insects

In a colony of tree wasps, workers on nursemaid duty crawl this way and that along the bottom of their nest, tending the youngsters in the comb. Most of the workers dutifully look after the queen’s offspring, stopping only to spit a runny meal into the mouth of a pale, lumpy larva snug in its cell. But one of these workers is up to no good. This selfish worker stays still for a minute or two in a suspiciously crouched position. She’s laying her own egg in an empty cell.

ANT AUNT. A worker in the Brazilian ant species Pachycondyla inversa (red dots were painted on for identification) carries eggs as part of her regular nursemaid duties. Such workers routinely check out eggs and kill suspicious ones. M. Patricelli
POLICE ACTION. A honeybee scrutinizes a suspicious egg (arrow). This one had been laid illicitly by another worker instead of by the queen, and the investigating bee subsequently ate it. Ratnieks
WHACK A QUEEN. Two sizes of queens emerge in Schwarziana stingless-bee nursery cells. A standard, big queen grows in a supersize cell (lower right corner), and two dwarf queens (darker tinge) form in cells the same size as those of the surrounding worker larvae. Police typically kill the dwarf queens. Wenseleers
WILD WEST WASPS. Saxon wasps, like tree wasps, are far less law-abiding than honeybees. The wasps lay illicit eggs, but this worker has found one and is about to kill it. Foster

Such rogue egg laying is a crime against insect society. The wheels of justice, however, don’t require a special caste of investigators and prosecutors. Punishment among insects is meted out by ordinary workers—and sometimes the queen herself—says biologist Tom Wenseleers, who has watched dozens of hours of black-and-white videos from infrared security cameras that he’s trained on nests of tree wasps.

In the most dramatic episodes, the egg sneak finds herself surrounded by a posse of vigilante workers. “They’re grabbing on to her; they try to sting her,” says Wenseleers of the Institute for Advanced Study in Berlin.

Insect criminologists initially reported punitive action, among honeybees, in 1989, but researchers have found examples in 15 species since then, 5 of them documented last year. With the recent surge of police reports in insects, researchers see a broader range of both crimes and responses. For instance, queens do some of the rough stuff, such as killing illicitly laid eggs, even though they’re her kin. What’s more, queen wannabes can end up as police targets themselves.

Most biologists who have considered insect societies see them as models for studying altruism, with the workers looking out for the common good. But according to Wenseleers, the new work suggests that the more appropriate image is that of oppressed workers in a police state.

Criminal intent

Honeybees are extremely law-abiding, so when Francis L.W. Ratnieks, of the University of Sheffield in England, wanted to see whether colonies would police a crime, he had to commit the offense himself.

To obtain illicit eggs that he could sneak into his test colony, Ratnieks enlisted accomplices in a neighboring hive. He and his colleague, P. Kirk Visscher, who’s now at the University of California, Riverside, divided a bee colony with a screen that trapped the queen in one zone but still let the colony odor waft throughout. In the queenless zone, workers soon began laying eggs prolifically, as they do when their queen dies. The researchers collected eggs laid by the deceived workers along with bona fide queen-laid eggs from the other part of the colony.

The team then created a major crime wave by planting this booty in the test colony. After 24 hours, Ratnieks and Visscher could find only 1 percent of the transplanted worker-laid eggs, yet 59 percent of the introduced queen-laid eggs had survived. Thus, workers weren’t killing eggs just because they had the scent of another colony. The vigilantes were especially keen on eradicating other workers’ eggs.

In 1989, Ratnieks and Visscher published this first report of insect policing. However, Ratnieks had predicted such a phenomenon in a mathematical model 3 years before, arguing that honeybee workers have incentives to destroy the eggs of other workers.

In honeybees, as in wasps and ants, workers don’t mate but can develop working ovaries and lay eggs. Their unfertilized eggs hatch and grow, but only into males.

The females in these species, the vast sisterhood that does the work of the colony, come only from eggs laid by the queen. These eggs get fertilized with sperm that the queen stores when she’s young. In honeybees, that youthful bout often involves 10 males. All workers are therefore sisters or half-sisters.

Occasionally, the honeybee queen also releases an unfertilized egg. It grows into a royal son. Ratnieks pointed out that a royal son is more closely related to the average female honeybee worker—he’s her brother or half-brother—than the son of some other worker would be. Thus, for the best chance of spreading her own genes to the next generation, the worker honeybee should favor her brother and kill the other workers’ sons. From her point of view, the less-related male just drains colony resources without much genetic payoff.

But what if the queen mated with one male only? That might change everything, Ratnieks predicted. Given that males come from unfertilized eggs, a worker of a singly mated queen would on average share more genes with a nephew—the son of one of her sisters—than with a brother. Workers of the saxon wasp, Dolichovespula saxonica, for example, tend not to police their coworker’s eggs if the queen mates once, but tend to do so if the queen has multiple mates, Ratnieks and his colleague Kevin Foster reported in 2000.

Relatedness isn’t the whole story, however. Last year, another lab’s survey of policing reports from a broad range of species highlighted ones with singly mated queens that don’t conform to the simple relatedness model (SN: 8/28/04, p. 142: Available to subscribers at Policing egg laying in insect colonies). Ratnieks argues that the survey doesn’t disprove his original model; it had predicted that such factors as colony efficiency might change the incentives for policing. For example, a worker that starts laying eggs often stops working.

Trial by jury

Among Earth’s great societies, bees and wasps have no monopoly on misbehavior. Patrizia D’Ettorre wondered whether there’s justice among ants.

She has focused on Pachycondyla inversa, an ant collected in Brazil from tree knotholes and cacao pods. Working with Ratnieks, D’Ettorre and Jürgen Heinze, both of Regensburg University in Germany, settled the colonies in plastic nest boxes with ample meals of honey and cockroaches.

The researchers applied the same approach that Ratnieks had used. They sneaked eggs from an accomplice community into their test colony to simulate a crime spree.

The nursery setup demanded rethinking for ants. They don’t array their young in a honeycomb of cells as social bees and wasps do but instead make an egg heap. Piling may sound lazy, but as nest temperature and humidity shift, nursemaids once or twice each day move the whole pile to the most favorable spot.

To keep track of the illicit eggs, D’Ettorre had planned to dot each egg with paint before presenting it to the nursemaids. But, she says, “the surface is slippery, and they clean it very efficiently.” She finally did the experiment the hard way, one egg at a time. Each time she transplanted an egg, she had to keep her eyes on it until the ants decided its fate. In the test, she tracked nearly 300 eggs.

When D’Ettorre offered a queen-laid egg, the workers typically reacted within minutes. They stacked 80 percent of these eggs on the heap and killed the other 20 percent.

Offered one of the cheater eggs laid by workers, the nursemaid ants hesitated. It’s hard not to anthropomorphize the scene she describes as consultation: “They have it in their mandibles, and they pass it to each other. In the end, usually one will take it and crush it.” In these cases, the workers killed up to 90 percent of worker-laid eggs.

What forensic evidence led the workers to close in on the bad eggs? All ant eggs are coated with a mix of waxy hydrocarbons. The researchers determined that the worker-laid eggs have a suspicious profile, with only a touch of one compound that abounds on the eggs of queen ants.

In the July 7, 2004 Proceedings of the Royal Society of London B, the team reported its work, calling it the “first well-documented study of ants with distinct [queens] to show that workers kill worker-laid eggs in preference to queen-laid eggs.”

The Brazilian ants, in sum, police their eggs according to the same rules followed by bees and wasps. That’s noteworthy because bees, wasps, and ants each evolved social life independently, Ratnieks and Wenseleers point out in the Jan. 7 Science. Therefore, policing must also have arisen several times independently.

Special victims

If the honeybee and ant colonies suggest tightly regulated, Orwellian worlds, researchers have now found insects whose society seems more like a smuggler’s cove on a foggy night.

Blatant, common cheating among tree wasps (Dolichovespula sylvestris) has given Ratnieks and Wenseleers the opportunity to study punishment without having to perpetrate the crimes themselves. To collect the shady characters, the researchers ran requests for tree wasps in newspapers around Sheffield. It turns out that a lot of backyard birdhouses are actually wasp houses.

To keep track of who’s who, the research team briefly anesthetized eight colonies and then glued a tiny, numbered tag on each of hundreds of wasps. Wenseleers set up security cameras, like those used by human guard services, and observed that the workers behaved selfishly left and right. Overall, half of the male eggs were laid by workers rather than the queen.

Crime usually didn’t pay, though. Within a day, police action had destroyed 91 percent of illicit eggs and only 4 percent of the legitimate ones, the researchers report in an upcoming Behavioral Ecology and Sociobiology.

The wasps didn’t seem to be hiding their illicit egg laying. The videos showed wasps backing into cells to lay eggs even when other wasps were nearby. Sometimes they ignored the misbehaving worker, but other times they attacked her.

The researchers noticed a new twist on worker policing. At least five times in the study, the police force punishing a wrongdoer by destroying her egg included a worker who later laid an egg herself.

Ratnieks notes that tree-wasp colonies have few cells available for eggs, so a wasp that foils a coworker’s bid for motherhood could free a cell for her own egg. If further work verifies this behavior, Ratnieks says, these wasps may turn out to be acting largely for personal gain. Or, as he puts it, tree-wasp police could be “corrupt.”

Queen for a day

It’s not only workers that appear in the annals of crime, especially in the less law-abiding species. Anecdotal evidence for policing by queens has turned up in a range of species, including an ant and a bumblebee, but until the tree-wasp experiment, it had been documented in detail only in the ant Dinoponera quadriceps, Ratnieks says.

In the recent experiment, the security cameras showed that the queen tree wasp, which can rove around the colony, herself nailed 49 percent of the worker eggs that were destroyed.

The colonies where scientists have seen queens walking the beat have been small ones, Ratnieks points out. A tree-wasp colony has only a few hundred inhabitants, but colonies of common wasps number in the thousands, so a single queen can’t keep a nest under control.

Queens figure, dramatically at times, in a different kind of policing scenario, arising from what Ratnieks calls caste-fate conflict. The development of eggs into queens or workers is not determined purely by genetics. In honeybees, for example, the same egg could go either way, depending on how it’s raised. For other species, no one yet knows all the factors, be they genetic predisposition or nurture, that set certain eggs on the royal road.

Some insect colonies produce many queens, most of which are killed by workers before they have an opportunity to form colonies of their own. Ratnieks and Andrew F.G. Bourke of the Zoological Society of London propose that these extra queens are merely “selfish” rebels, seeking to spread their own genes rather than to be part of some scheme that would benefit the colony as a whole. Policing in insect societies would then have evolved to quash such uppity young females, says Ratnieks.

He, Wenseleers, and their colleagues in Brazil are examining Brazilian bees of the genus Schwarziana. Although stingless, these bees have their own style of aggression against people. “They grab your hair,” says Wenseleers, “and they try to get in your ears.”

The Schwarziana royalty arises in two sizes in the presence of a reigning queen. When a full-size queen matures, workers encircle her and quickly build a dome-shaped prison of wax around her. They regularly feed her by biting a little hole in the wax and passing through food. They then seal the hole. Such a spare queen remains under control.

On the other hand, what Wenseleers calls the dwarf queens challenge the status quo of the colony. They’re the same size as workers and emerge from the same kind of cells, yet they have well-developed ovaries. Unchecked, they could mate and take over colony reproduction. These little queens don’t get wax boudoirs. Instead, workers typically kill them, Wenseleers and Ratnieks report in their Jan. 7 Science article.

Police and potential queens also clash in colonies of another stingless bee, Melipona beecheii. It lives in the Yucatán, where people have collected its honey for centuries. The queen lays all her eggs, whether they’ll end up as workers or queens, in cells of the same size, which are half to two-thirds full of liquid food. The egg “floats like a wobbly man on top of the pasty goo,” says Ratnieks. Workers then seal the cell holding the new egg.

Some 16 percent of female eggs mature as reproductively competent queens instead of as workers, Ratnieks and Wenseleers reported in the September 2004 Ethology. But these ambitious youngsters survived only an average of 27 hours before their nest mates killed them.

To see policing in detail, the researchers put several new queens into an established colony. In four cases, a single worker caught a queen and neatly bit off her head in as little as 4 seconds. Sometimes, five or so workers grabbed a young queen’s legs, antennae, and mandibles and dragged her around the colony for perhaps an hour, gradually pulling her apart.

One queen lay immobile, as if feigning death, for 14 hours, and thus outlived all others of her kind. However, when she finally moved, she acted aggressively toward the workers and the rightful queen, and the workers killed her too.

All that drama sounds as if the interests of an individual insect aren’t necessarily the same as those of her colony, even though she and the other members are all closely related. The Brazilian dwarf queens and the Yucatán queen free-for-all offer routes to the throne—if they weren’t blocked by swift police action.

Wenseleers argues that, overall, “coercion plays a more important role than kinship in favoring cooperation in insect societies.” Regardless of how many legs you count, that’s a police state.


Susan Milius is the life sciences writer, covering organismal biology and evolution, and has a special passion for plants, fungi and invertebrates. She studied biology and English literature.