Cannibal Crickets Can Control a Crowd: How Eating Your Friends Aids Collective Motion
On the wide, open plains of the American West, it’s more than the buffalo and the antelope that roam. Mormon crickets (Anabrus simplex) also sweep across the land in huge migratory swarms that can stretch six miles long and three miles wide. The crickets (a misnomer, they’re actually flightless katydids) can march up to a mile and a quarter a day in these groups, devouring every scrap of vegetation in their path and devastating agriculture in areas they pass through.

It sounds like a Biblical plague*, but consider the poor crickets. While the swarming behavior isn’t completely understood, entomologists think that it’s partly a strategy to avoid being eaten. Observations and experiments have shown that crickets that become separated from the group are easy prey and a big, cohesive group minimizes the risk of predation for any individual cricket.

Not that life in the swarm is any easier. In addition to consuming any and all plants they come across, the crickets often eat each other. One reason for this should be obvious, says a new study, “huge, concentrated numbers of crickets require huge, concentrated amounts of food. If the landscape doesn’t provide it, a fellow cricket will.”

A not so obvious side effect of this crickety cannibalism is that it might be helpful, even necessary, in keeping the swarm moving as a unit. A swarm of insects is simply the sum of its parts. The group’s movement, coordination, cohesiveness and persistence are the simple decisions and interactions of millions of individuals scaled up to the population level, and some of those decisions and interactions happen to involve one insect eating another. Another swarming insect, the desert locust (Schistocerca gregaria), tends to cannibalize traveling companions that have stopped moving or can’t keep pace with the group, and the threat of cannibalism influences their marching behavior. Individuals keep moving and maintain proper direction and pace to keep from becoming lunch for the guy behind them, and this helps maintain coherent swarm motion.

Snacks for the Road

Sepideh Bazazi and colleagues from the U.S. and Australia thought that the same would be true of Mormon crickets. To test their hypothesis, they used a unique natural “laboratory:” two sites in Daggett County, Utah where lingering swarms of crickets provided a continuous stream of test subjects.

They captured four crickets, hot-glued them to pieces of pieces of wood and placed them, facing in four different directions, in a row in the path of the approaching swarms. As the swarm approached the sacrificial lambs, the researchers filmed things from above. They then repeated the catch-glue-set out-film process with different crickets for a total of 24 times.

After reviewing the films, Bazazi and his co-authors tallied the number of times an approaching cricket from the swarm stopped within one antenna-length of a glued cricket. They also categorized and counted the outcomes of these encounters, whether it was the approaching cricket moving away, making a successful attack and biting the trapped cricket or making an unsuccessful attack and getting kicked away by the bait.

They counted 2,056 total encounters over 24 trials, an average of 4.3 encounters per minute per immobilized cricket. Fifty-nine percent (1,258) of the encounters resulted in attacks, and 58% (734) of those were successful.

Blood on the Sand

The biologists came away with three observations. First, female Mormon crickets are more likely than males to engage another cricket that isn’t moving, even when there’s an even sex ratio in the swarm, but aren’t more likely to be successful if they attack (on the other hand, the sex of the immobilized crickets had no effect on the number of crickets that approached or attacked them). The researchers hypothesize that females engage potential victims more often because they have more demanding nutritional requirements.

They also noted that crickets whose bodies were perpendicular to the direction of swarm movement were most vulnerable to encounters, since a side-facing cricket exposes a larger surface area. The right and left sides of the immobilized crickets were more likely to be approached than the head and back end, but all body areas were equally likely to be attacked and attacked successfully. Crickets who don’t want to become dinner minimize their risk of being bothered by protecting their flanks, which means moving along with the swarm. Like with the locusts, the threat of cannibalism aids the smooth flow of traffic.

Lastly, they found that the duration of a moving cricket’s encounter with an unmoving cricket and the likelihood of that cricket successfully attacking and cannibalizing another are strongly affected by the behavior of the other crickets around it. The number of moving crickets already in contact with an immobilized one increased the probability that an approaching cricket would also attack the victim and that that attack would be successful. While moving crickets were already highly likely to approach and attack even in the absence of others, that likelihood almost doubled once there were nine or more other crickets already on the attack. The more crickets attacking a victim, the longer each new approaching cricket stuck around, too.

Bazazi suggests that this happens because of “social facilitation,” an increase in the frequency of a behavior in response to others engaged in that same behavior. This could work in two ways with the crickets. First, the crickets already interacting with and cannibalizing a stationary cricket could provide approaching crickets with social information about the location of food. Second, cues from the wounded cricket could create a “blood in the water” effect, spurring approaching crickets to attack and feed.

Either way, the result is that crickets move towards others in the swarm, ultimately helping to maintain cohesive movement and momentum for the group, making both the threat of cannibalism and the desire to grab a meal on the go driving factors of swarm movement.

*The insect’s name actually comes from a mutually destructive encounter with a religious group. The wheat crop at first Mormon settlement in Utah was supposedly set upon by the bugs and rescued by the miraculous appearance of a flock of California gulls, which devoured the entire swarm.

References: Bazazi S, Ioannou CC, Simpson SJ, Sword GA, Torney CJ, Lorch PD, & Couzin ID (2010). The social context of cannibalism in migratory bands of the Mormon cricket. PloS one, 5 (12) PMID: 21179402

Bazazi S, Buhl J, Hale JJ, Anstey ML, Sword GA, Simpson SJ, & Couzin ID (2008). Collective motion and cannibalism in locust migratory bands. Current biology : CB, 18 (10), 735-9 PMID: 18472424

Sword GA, Lorch PD, & Gwynne DT (2005). Insect behaviour: migratory bands give crickets protection. Nature, 433 (7027) PMID: 15716941

Images: Cricket photo Whitney Cranshaw, Colorado State University; cricket swarm photo Steve Jurvetson cricket drawing by Art Cushman, USDA, Property of the Smithsonian Institution Department of Entomology. Both via and used under a Creative Commons License


One Comment, Comment or Ping

Reply to “Cannibal Crickets Can Control a Crowd: How Eating Your Friends Aids Collective Motion”