Rat Attack

Killer Instinct
by Peter Tyson

Why is infanticide so widespread among primates,
including the great apes, and what does that say
about us?

By almost any standard, lemurs are adorable, like a child's stuffed animal come alive. "Yellow Silver," a wild male lemur I once had the pleasure of observing close up in a Madagascar rain forest, epitomized this cuteness. Reclining on a hammock-like vine just six feet away, he was so close I could smell his bushy black fur and see his human-like fingernails and eyelashes. Totally ignoring me, he draped one hand lazily over a twig and now and then curled up his two-foot tail chameleon-style then let it unfurl. Yellow Silver was dozing off a hot afternoon—a picture of idle contentment.

Years later I learned that Yellow Silver, like many of his kind, was an infant-killer. Patricia Wright, the primatologist who named him and is studying his species, the Milne-Edwards' sifaka, told me recently what it's like to witness an infanticide. A male will distract a female holding a baby by biting her on the elbow or shoulder. "The baby's exposed, and then the young male"—Wright sucks in her breath—"just gets the baby and slits the belly. The baby lets go and drops to the ground. We're all screaming and crying because it's so awful. The baby doesn't die right away, so it's very, very painful."

Polar-opposite images of Yellow Silver: easygoing fur ball, savage killer. How did such behavior, which to us humans is both shocking and puzzling, arise in nature? That is, how could a species bent on passing on its genes possibly benefit by killing its own members? Infanticide is widespread in the animal kingdom, occurring in rats and lions, wasps and tree swallows, sand sharks and sea lions. But it is its presence among primates, including our closest relatives, the chimps and gorillas, that I find most compelling, in an unsettling way. To wit: What does its existence among the great apes say about us?

Challenging an old bias

The realization that infant-killing occurs widely in nature is relatively recent. In large part this had to do with our repellence of the act. Charles Darwin set the tone when he wrote, in The Descent of Man, "Our early semi-human progenitors would not have practiced infanticide, for the instincts of the lower animals are never so perverted as to lead them regularly to destroy their own offspring." As late as the mid-1960s, classical ethology—the study of animal behavior in the wild—held that creatures rarely killed members of their own species, except in captivity or unnaturally crowded conditions.

A sea change in this view began to sweep over biologists in the early 1970s. It was triggered by a Harvard graduate student named Sarah Hrdy, who went to India at the time to study Hanuman langurs. Fed by Indians, who consider them sacred, these black-faced monkeys live in dense populations near cities. Hrdy went to investigate reports of violent behavior among the langurs, including infanticide. "My working hypothesis was that I was going to be looking at pathological behavior brought about by crowding," Hrdy told me. "Within a year, I had changed my mind."

Infanticide is not pathological in Hanuman langurs, Hrdy decided. Rather, it appears to be a strategy males can use to help them achieve reproductive success. When a male langur enters a new group to breed, it's to his reproductive advantage to kill off any unweaned infants and thus bring their mothers more quickly back into fertility so he can mate with them and pass on his genes. (A mother producing milk cannot get pregnant.)

In the four decades since Hrdy's seminal work, biologists have documented infanticide in some two dozen species of primates and prosimians (lemurs and other lower primates). And it's not only males. Female primates also kill babies, from Milne-Edwards' sifaka females that kill others' infants in an act Wright says may be simple competition, to the chimpanzee mother and her daughter whom Jane Goodall famously observed killing and eating baby chimps on several occasions. (Altogether, Sarah Hrdy has identified five categories of infanticide in nature—see sidebar.) But among primates, mostly it occurs when young males like Yellow Silver enter a new group to breed.

Ongoing debate

Not everyone agrees that Hrdy's hypothesis—known to scientists as sexually selected infanticide by males—is correct. Robert Sussman, an anthropologist at Washington University in St. Louis, is one of the more vocal critics. Sussman notes that infanticide among primates remains a statistically uncommon event, and while more than two dozen primate species documented as infanticidal may sound like a lot, the number of recognized primate species tops 300, he says.

Sussman also feels that all the conditions needed for Hrdy's hypothesis—among them, that infanticidal males are more reproductively successful than males that don't practice infant-killing—are difficult to show in the wild. Finally, any genetic link, which Hrdy's hypothesis presumes, is even more challenging to demonstrate. "There's no evidence whatsoever that there are males that carry genes for infanticide," Sussman told me.

"If a male comes into the group and almost all the females are lactating, he should try very hard to eliminate kids."

Hrdy acknowledges that finding genetic links to infanticidal behavior in primates is a long way off. But she points to such evidence in various types of rodents, in which some strains or family lines have been shown to be more prone to infanticide than others. (Sussman counters that these are laboratory rodents in unnatural and stressful conditions.)

Yellow Silver's case

Notwithstanding detractors, most biologists today accept Hrdy's hypothesis, including Pat Wright. Indeed, what Yellow Silver did years after I saw him on that lazy afternoon appears to be a textbook case of male sexually selected infanticide, she says.

When male Milne-Edwards' sifakas reach six years of age, they begin looking to join another group in which to breed. To improve their chances of breeding successfully, they look for a group with no male or only one male. Five years after I saw Yellow Silver, he finally, at age nine, took over another group—and the group's lone baby quickly disappeared. "We didn't actually see him attack," Wright says. "That's good in a way," she adds under her breath. "I'd rather not see it."

Wright didn't need to see the event to guess what had happened—and why. Sarah Hrdy had predicted early on that infanticide would not happen in seasonal breeders like lemurs, because males would still have to wait until the next breeding season for females to become fertile. (In Hanuman langurs, a female becomes fertile just four days after her infant's death.) So why do Yellow Silver and his fellow males kill infants? The answer, Wright says, may be that they still gain a reproductive advantage.

Milne-Edwards' sifaka groups are small, only three to nine members, and females only breed every two years on average. If a group has only one breeding male and two breeding-age females—a common scenario—and an incoming male like Yellow Silver kills any unweaned infants, the next year the resident male may mate with one female and the immigrant male with the other. (Wright has genetic evidence that confirms this paternity outcome in two cases, she says.) Thus, both males sire offspring that next year, which is a year faster for both males than it would be if the unweaned babies had lived.

"It's amazing, but we've seen it enough times and we've got the genetic evidence now, so that seems to be it," Wright told me. "It's a cruel game, though."

Mating with infant-killers

Females also play the game, if only by necessity. In lemurs, females are dominant and get to choose males they mate with, but they appear unable, either singly or collectively, to stop incoming males from committing infanticide. (And they get no help from resident males, which is, Wright says, "the most shocking thing to me," as those males almost certainly fathered the infants being attacked.) Moreover, the females, as in Hrdy's langurs, mate with the very males that killed their infants.

Ryne Palombit, a former student of Hrdy's who studies infanticidal chacma baboons in Botswana, explains a possible reason why. "Students in my classes ask me all the time, 'Why don't the females just not mate with this guy?'"—meaning an infanticidal male. "I call that the Lysistrata strategy," Palombit says, chuckling. In Aristophanes' 5th-century B.C. play of that name, Greek women collectively go on a sex strike and take over the cities in an effort to end the Peloponnesian War.

"In theory, it could work," Palombit says, "but all the females would have to boycott this male, because as soon as one cheats and mates with him, she'll have a higher reproductive advantage." Females are competing with one another as well, and, as it happens, refusing to mate is virtually non-existent among mammals, Palombit says. Wright has seen only one instance among Milne-Edwards' sifakas, and the spurned male finally left the group.

Suppressing the urge

While counterstrategies in some primate species are clearly insufficient to prevent infanticide, in others they seem to work. More than a third of infant deaths among chacmas are by infanticide, but virtually none are among East African olive baboons, a closely related species that Palombit also studies. Palombit thinks it's because of a key difference in behavior.

In chacmas, once an alpha male takes control of a group, he is the only one that mates with the group's females, and thus he can feel confident that he'll be the father of any infants conceived during his tenure. But his tenure as alpha is short, averaging only seven or eight months. So it's to a new alpha's benefit to kill unweaned young. "If a male comes into the group and almost all the females are lactating," Palombit says, "he should try very hard to eliminate kids, because [otherwise] he won't have any reproductive success. Or he should leave the group."

In olive baboons, Palombit says, an alpha's tenure lasts a lot longer, up to two years, so he can afford to wait until mothers wean their infants and resume cycling. Moreover, in olives, two lower-ranking males can join forces and unseat the alpha while females are ovulating, and thus an olive alpha can never enjoy the same degree of confidence regarding paternity that a chacma alpha enjoys. These reasons may be why infanticide is rarely seen among olives, Palombit says.

Is a tendency towards infanticide buried in my genes or at least in my distant past as a human being?

Bonobos seem to have found a different strategy for avoiding infanticide—females have sex anytime, all the time, with anybody, even other females and juveniles. "By having so much sex with everybody, including the males, the females are totally confusing paternity in a way," says Frans de Waal, a primatologist at Emory University. "There is no male in the group who can potentially exclude the possibility that he's the father of an offspring." So, again, infanticide may confer few if any reproductive benefits, and, in fact, it has never been reported in bonobos.

The human factor

Not so in humans. In truth, some scholars believe infanticide was an important factor in human evolution. "People have often assumed that the origin of the human pair-bond lay in males providing food for offspring," de Waal says. "But it may have been simpler—they may have just been protecting offspring [from infanticide]."

The practice still occurs today among humans, though, as Hrdy stresses, it's quite different from that in non-human primates. By far the most common form of infanticide among people involves mothers themselves, she says, and it has to do with local resources and how much social support the mother has. Wherever mothers do not have access to other forms of birth control, they sometimes have to resort to infanticide. "Among !Kung hunter-gatherers in the Kalahari, about 1 percent of live-born infants are abandoned at birth," she says. "It usually has to do with birth spacing or perhaps birth defect and is very regretfully done." In Highland New Guinea, on the other hand, infanticide has been, Hrdy says, "off the charts, since mothers will eliminate daughters in hopes of bearing a son sooner."

Even in the United States, infanticide takes place on occasion, and the genetic relatedness that, according to Hrdy's hypothesis, wild males like Yellow Silver must weigh before attempting the act may play a role in some cases. One widely cited study by researchers at McMaster University in Hamilton, Ontario found that a preschool American stepchild is 60 times more likely than a biological child to be the victim of infanticide.

Getting personal

All this leads me to wonder: Just how different am I fundamentally from Yellow Silver? When I first met the woman who would later bear my two children, I never felt an urge to kill off her four-year-old son from a previous marriage. But is a tendency towards infanticide, even if suppressed, buried in my genes or at least in my distant past as a human being?

I don't know, and even experts don't have a definitive answer yet. But I certainly wasn't giving it any thought while watching Yellow Silver do absolutely nothing on that warm day in Madagascar. In fact, the lemur's languor was infectious, and I, too, dozed off, probably a picture of idle contentment.

Enlarge this image
Rat Attack

Milne-Edwards' sifakas, a kind of lemur, are one of more than 20 species of primates and prosimians (lower primates) that biologists have documented practicing infanticide in the wild.

Enlarge this image
Rat Attack

Infanticide is just one of a number of strategies that male Hanuman langurs use to better their chances of reproductive success.

Five Types of Infanticide

The five functional categories of infanticide, as first outlined by Sarah Hrdy in 1979:

(1) exploitation of the infant as a resource, usually cannibalism;

(2) competition for resources, where death of the infant increases resources available to the killer or its lineage;

(3) sexual selection, where individuals improve their own opportunities to breed by eliminating dependent offspring of a prospective mate;

(4) parental manipulation of progeny, where parents on average increase their own lifetime reproductive success by eliminating particular offspring; [and]

(5) social pathology, for cases where infanticide on average decreases the fitness of the infanticidal individual.

—from Hrdy, Sarah Blaffer and Glenn Hausfater, "Comparative and Evolutionary Perspectives on Infanticide: Introduction and Overview." In Infanticide: Comparative and Evolutionary Perspectives, Glenn Hausfater and Sarah Blaffer Hrdy, editors. Aldine Publishing Co., 1984, p. xvi.

Enlarge this image
Rat Attack

Yellow Silver at home in the rain forest of Ranomafana National Park, Madagascar

Enlarge this image
Rat Attack

More than a third of infant deaths in chacma baboons are the result of infanticide by males.

Enlarge this image
Rat Attack

"Bonobo society may be an interesting case of females having found a solution," Frans de Waal says, referring to the problem of infanticide. Above, a bonobo resting.

Enlarge this image
Rat Attack

Why do chimpanzee males and even females sometimes eat the infants they kill? More research may answer this and other outstanding questions about infanticide among wild primates.

Peter Tyson is editor in chief of NOVA Online.

Rat Attack Home | Send Feedback | Image Credits | Support NOVA

© | Created February 2009