2004.N16.Amazing Animals

COPYRIGHT 2004 SCIENTIFIC AMERICAN, INC.

http://www.sciam.com/index.cfm?ref=digitalpdf

TABLE OF CONTENTS

ScientificAmerican.comexclusive online issue no. 16A M A Z I N G A N I M A L S

We humans tend to think of ourselves as Nature’s ultimate creation. A quick survey of the creatures with which we share the planet sug-gests that’s a myopic view of things. Billions of years of evolution has produced a rich diversity of animals, each exquisitely adapted to itsecological niche. This exclusive online issue celebrates some of the more spectacular results.

In the pages that follow, leading biologists share their insights into beasts both familiar and foreign. Some are remarkable for their physi-cal characteristics. Take, for example, the basilisk lizard, best known for its ability to walk on water. Or the star-nosed mole, whose stellaraccessory works uncannily like an eye in its dark, damp environs. And then there’s the komodo dragon, a rare reptile whose stealth, powerand supersized proportions have earned it a fearsome reputation indeed.

Other creatures amaze with their behavior. Some ants conduct warfare that would have given Genghis Khan pause. Parrots match wits withdolphins and nonhuman primates. Lions cooperate, but only when they stand to benefit. And chimpanzees pass social customs down fromgeneration to generation--in other words, they have culture.

Animals fascinate us with the ways in which they resemble and differ from our kind, yet they are neither mirror nor measuring stick.Perhaps American author Henry Beston put it best: "They are other nations, caught with ourselves in the net of life and time, fellow prison-ers of the splendor and travail of the earth." —The Editors

Beasts in Brief• Fido Found to Be Wiz with Words • The Cultured Orangutan• King of Beasts Suffers to Be Beautiful • Dolphin Self-Recognition Mirrors Our Own• Crafty Crow Rivals Primates in Toolmaking • How Geckos Get a Grip• How Bears Power-Nap • Brainy Bees Think Abstractly

Running on WaterBY JAMES W. GLASHEEN AND THOMAS A. MCMAHON; SEPTEMBER 1997The secret of the basilisk lizard's strategy lies in its stroke

The Nose Takes a Starring RoleBY KENNETH C. CATANIA; JULY 2002The star-nosed mole has what is very likely the world's fastest and most fantastic nose

The Komodo DragonBY CLAUDIO CIOFI; MARCH 1999On a few small islands in the Indonesian archipelago, the world's largest lizard reigns supreme

Slave-Making QueensBY HOWARD TOPOFF; NOVEMBER 1999Life in certain corners of the ant world is fraught with invasion, murder and hostage-taking. The battle royal is a form of socialparasitism

Divided We Fall: Cooperation among LionsBY CRAIG PACKER AND ANNE E. PUSEY; MAY 1997Although they are the most social of all cats, lions cooperate only when it is in their own best interest

Talking with Alex: Logic and Speech in ParrotsBY IRENE M. PEPPERBERG; SCIENTIFIC AMERICAN PRESENTS: EXPLORING INTELLIGENCE 1998Parrots were once thought to be no more than excellent mimics, but research is showing that they understand what they say.Intellectually, they rival great apes and marine mammals

The Cultures of ChimpanzeesBY ANDREW WHITEN AND CHRISTOPHE BOESCH; JANUARY 2001Humankind's nearest relative is even closer than we thought: chimpanzees display remarkable behaviors that can only bedescribed as social customs passed on from generation to generation

1 SCIENTIFIC AMERICAN EXCLUSIVE ONLINE ISSUE AUGUST 2004

20

8

6

2

14

26

34

39


Fido Found to Be Wiz with Words

Dogs may be able to understand far more words than a typical ownerteaches them during obedience training. Scientists experimenting with anine-and-a half-year-old border collie in Germany have discovered thatthe dog knows more than 200 words for different objects and can learn anew word after being shown an unfamiliar item just one time. The dog'sability shows that advanced word recognition skills are present in animalsother than humans, and probably evolved independently of language andspeech.

Rico, the border collie, was taught to retrieve different objects by his own-ers, who placed various balls and toys around their apartment and askedRico to fetch specific ones. Rico gradually increased his vocabulary toabout 200 words that he could match to objects. To make sure Rico'sowners weren't giving him subconscious cues that helped him find theright item, Julia Fischer and her colleagues at the Max Planck Institute forEvolutionary Anthropology in Leipzig, Germany, tested Rico's knowledgein a lab, where he retrieved 37 out of 40 items correctly. "Rico's 'vocabu-lary size' is comparable to that of language-trained apes, dolphins, sealions, and parrots," the authors write in their report, published in the June11, 2004, Science.

The team then tested Rico's ability to employ fast mapping, a neurologi-cal process that toddlers use to quickly guess the meaning of new words. The researchers put an unfamiliar object in a roomwith other things he did know and, without teaching Rico the name of the novel item, asked him to get it. Seven times outof 10 he returned with the correct object.

Four weeks later, the scientists tested Rico's ability to recall what he had learned. The objects that he had seen only once dur-ing the previous experiment were placed among eight other things, some familiar and some completely new. In this trial, Ricoretrieved the correct item three out of six times, a feat of learning never before seen in a dog. Rico's performance was com-parable to that of a three-year-old toddler, the scientists observe.

Fischer and her collaborators note that they're not sure whether Rico is exceptionally smart or exceptionally well trained, butthey hope they can use this experiment to further probe how the brain learns to understand words. Rico's powers of com-prehension, they say, show that the processes the brain uses to discern words are not the same as those used to producespeech. Says Fischer: "You don't have to be able to talk to understand a lot." --Elizabeth Querna

King of Beasts Suffers to Be Beautiful

It's not easy being beautiful, especially when you're a male lion. New research suggests that what lady lions love most andwhat other males fear most is a leo with a long, dark mane--which is precisely the worst sort of 'do to have in Africa's oftensweltering environs.

Biologists have long pondered the purpose of the lion's hot, conspicuous mane, which seems at first glance like more troublethan it's worth. Evolutionary theory holds that there should be some benefit gained from it, but what might that be? Twohypotheses have been put forth. The first holds that the extra fur protects the lion from injuries to the neck and shoulders.The second posits that the mane makes the lion more attractive to lionesses and more intimidating to other males.

In recent work, Peyton M. West and Craig Packer of the University of Minnesota studied the reactions of male and femalelions to various types of manes, using dummy lions to model the various coifs. They found that whereas females fell fordummy males with dark manes (as opposed to blondes), males avoided the brunettes. Males also avoided dummies sporting

BEASTS IN BRIEF

COUR

TESY

OF

SUSA

NN

E BA

US

2 SCIENTIFIC AMERICAN EXCLUSIVE ONLINE ISSUE AUGUST 2004 COPYRIGHT 2004 SCIENTIFIC AMERICAN, INC.

long manes.

As it turns out, mane color and length may be pretty good indicators of a lion's health and fitness. "Dark color tends to befound in high-testosterone males," West observes. "Therefore, it isn't surprising that females would prefer darker manes andmales would be intimidated. But there is no correlation between testosterone and mane length." Males pay more attention tomane length because recently injured lions have shorter manes, she explains.

The males' dark allure comes at a significant cost. "A male with a dark mane may have to work harder to stay cool, behav-iorally or physiologically, and is advertising that toughness, along with his toughness in battle," West remarks. (Longermanes, on the other hand, do not appear to retain additional heat.) In some especially hot regions, however, all that fur coststoo much, and the males go maneless.

"As climate changes, things like lion manes, the brightness of bird plumage and the size of deer antlers may be sensitivebioindicators, Packer muses. "They can tell you how well an animal is doing in the environment." The team's findings werepublished in the August 23, 2002, Science. --Kate Wong

Crafty Crow Rivals Primates in Toolmaking

The ability to make tools was once thought to lie solely within the purview of humans.Then in the 1960s Jane Goodall discovered that chimpanzees, too, fashion implementsto perform certain tasks. Since then, researchers have observed tool use in a variety ofanimals. Nonhuman primates are widely thought to be the most sophisticated tool-users after us. Observations of an innovative New Caledonian crow named Bettycould alter that view.

New Caledonian crows are known to make hook tools with natural materials in thewild. But Oxford University zoologists writing in the August 12, 2002, Science report

that Betty, a captive crow, spontaneously performed an unexpected variation on this theme, coaxing a piece of straight wireinto a hook to retrieve a small bucket of food. In a subsequent experiment, the clever crow repeated the feat in nine out of10 valid trials. To bend the wire, Betty anchored one end either in the sticky tape holding the experimental apparatus togeth-er or between her feet, and then manipulated the other end with her beak.

Remarkably, although Betty had previously used supplied wire hooks, she had never seen the process of bending and had noprior training with pliant material. "Purposeful modification of objects by animals for use as tools, without extensive priorexperience, is almost unknown," the team writes, noting that even our primate kin often fail to show such talent in theabsence of explicit coaching.

"Our finding, in a species so distantly related to humans and lacking symbolic language, raises numerous questions about thekinds of understanding of 'folk physics' and causality available to nonhumans, the conditions for these abilities to evolve, andtheir associated neural adaptations," the authors conclude. Birdbrain might not be an insult after all. --Kate Wong

How Bears Power-Nap

For most people, staying in shape means getting regular exercise. Take a vacation from the gym and your hard-earned six-pack goes soft. But imagine if you could sleep for five months and still wake up fit as a fiddle. According to research describedin the February 23, 2001, Nature, this is in fact just how bears emerge from hibernation.

Henry J. Harlow of the University of Wyoming and his colleagues found that hibernating black bears lose less than 23 per-cent of their strength during their 130-day winter slumber. Humans, in contrast, would experience a 90 percent strength lossif they were immobile for so long. Incredibly, when the team took muscle biopsies from denned bears in early and late win-ter, they found that the skeletal muscle cells did not dwindle in size or number-nor did they lose their protein content or oxida-tive capacity.

The researchers suggest that the bears may be maintaining their muscles by drawing on protein reserves from elsewhere inthe body, and by shivering. "Understanding these processes in hibernating bears," the team writes, "may provide new insightinto treating muscle disorders and into the effects of prolonged hospital bed confinement, antigravity and long-distance spacetravel on humans." --Kate Wong

COUR

TESY

OF

SCIE

NCE


The Cultured Orangutan

For humans, the sputtering sound known as a raspberry is com-monly considered a contemptuous gesture. But among someorangutans, the expression seems to simply signify that the utter-er is turning in for the night. Not all orangutan groups mark theend of the day this way, however. In fact, according to a reportpublished in the January 3, 2003, Science, a raspberry before bedis one of 24 socially transmitted behaviors that scientists say mayrepresent cultural variation in the great apes. If the findings areconfirmed by additional field observations, they could push theorigin of culture back nearly seven million years to the last com-mon ancestor of orangutans and the African apes.

Man's closest living relative, the chimpanzee, provides the bestevidence for the existence of nonhuman culture: scientists haveidentified 39 behavior patterns that vary culturally among the ani-mals. To investigate whether similar conduct exists in orangutangroups, Carel P. van Schaik of Duke University and his colleaguesassessed previously collected data on six different wild popula-tions in Borneo and Sumatra. "Culture requires more than just amother-infant bond, but also extensive social contact, and orangutans are at the low end of the sociability spectrum," vanSchaik says. Nevertheless, the team identified two dozen behaviors that fall into three of the four categories of cultural ele-ments: labels, signals and skills (The fourth category is symbols, which only humans employ).

Van Schaik notes that the group found "the biggest behavioral repertoires within sites that showed the most social contact,thus giving the animals the greatest opportunity to learn from one another." Examples of culturally-based behaviors that thescientists distinguished include using leaves as napkins, using leafy branches to ward off attacking insects and riding "snag"(dead trees that are falling toward the ground) for sport. --Sarah Graham

Dolphin Self-Recognition Mirrors Our Own

Whether we're assessing our physiques or checking for food stuck in our teeth, most of us consult a mirror regularly to makesure we appear the way we expect. Though it may seem an unremarkable feat, the ability to recognize oneself in the mirroris actually exceptionally rare among animals. Indeed, only humans and their closest kin, the great apes, have shown thiscapacity, suggesting that factors specific to great apes and humans drove its evolution. Findings announced May 1, 2001,online edition of the Proceedings of the National Academy of Sciences, however, indicate that we and our primate relativesare not alone. According to the report, dolphins, too, exhibit mirror self-recognition.

To test for dolphin self-awareness, Diana Reiss of Columbia University and Lori Marino of Emory University exposed twobottlenose dolphins to reflective surfaces after marking the dolphins with black ink, applying a water-filled marker (sham-marking) or not marking them at all. The team predicted that if the dolphins-which had prior experience with mirrors-rec-ognized their reflections, they would not show social responses; they would spend more time in front of the mirror whenmarked; and they would make their way over to the mirror more quickly to inspect themselves when marked or sham-marked. The experiments bore out all three predictions in both dolphin subjects. Moreover, the animals even selected the bestreflective surface available to view their markings.

Intriguingly, whereas chimpanzees take interest in marks on fellow chimps in addition to marks on their own bodies, the dol-phins focused on themselves. "Dolphins may pay less attention to marks on the bodies of companions because, unlike pri-mates, they do not groom each other," the researchers write. "This difference makes our findings even more interestingbecause dolphins clearly are interested in marks on their own body despite the fact that they do not have a natural tendencytoward social grooming."

The extent of dolphin self-awareness remains to be explored. But the fact that they have passed the mirror test means thatself-recognition may result from large brains and advanced cognitive ability, as opposed to being a by-product of primate-specific factors. That dolphins and primates-which differ profoundly in their brain organization and their evolutionary his-tories-should both exhibit this unusual ability, the authors note, represents "a striking case of cognitive convergence." --KateWong

COUR

TESY

OF

DUKE

NEW

S SE

RVIC

E


How Geckos Get a Grip

In the movie Spider-Man, Peter Parker looks down at thepalms of his hands to find that they have sprouted thick,black hairs, giving him a firm grip on walls and ceilings. Agrowing body of evidence indicates that gecko lizards, too,cling to surfaces with the help of hairlike projections. Thegecko hairs are so tiny, however, that they operate not bycatching on substrate irregularities, but by facilitating theformation of molecular bonds that create electrodynamicattraction between the gecko's feet and the surface uponwhich it is walking. As a result, the charismatic creatures cancrawl upside down even on polished glass.

Previous research, conducted by Kellar Autumn of Lewis and Clark College and his colleagues, had suggested that the gecko'sfoot-hairs, or setae, stick to surfaces by virtue of these so-called van der Waals forces. But the team had been unable to rejecta competing hypothesis, which holds that the adhesion arises from water-based forces. The results of the scientists' subse-quent study, detailed in a report published on August 27, 2002, by the Proceedings of the National Academy of Sciences, dis-proves that theory.

The researchers reasoned that if water-based forces, such as capillary adhesion, were the secret to gecko grip, then the ani-mal's toes--each of which bears hundreds of thousands of setae--should not stick to hydrophobic ("water-fearing") surfaces.In subsequent experiments, however, the gecko toes clung equally well to hydrophobic and hydrophilic ("water-loving") sub-strates. Single, isolated setae were likewise effective on both types of surfaces.

Autumn and his collaborators further determined that the size of the setal tips--the hundreds of spatulae that branch fromeach hair, increasing surface density--is remarkably close to what one would expect if van der Waals forces are the principlemechanism underlying the gecko's sticking power. That implied that it is the size of the setal tip, not the nature of the setalmaterial, that gives the lizard its toehold. Verification of this idea came when the researchers fabricated setal tips from twodifferent materials and found that both adhered to surfaces as predicted.

According to the investigators, the finding not only provides insight into the function of setal structures in geckos and othercreatures, it hints at how synthetic dry adhesives could be improved: subdividing their surfaces into small, setal tip-like pro-trusions, thus increasing surface density, might enhance stickiness. --Kate Wong

Brainy Bees Think Abstractly

The capacity for abstract thinking does not belong to humans alone, as studies of other vertebrates, such as primates, pigeonsand dolphins, have shown. Researchers have found that invertebrates, too, possess higher cognitive functions. A report in theApril 32, 2001, Nature indicates that the humble honeybee can form "sameness" and "difference" concepts-an ability thatmay help them in their daily foraging activities.

To probe the honeybee's mental prowess, Martin Giurfa of the Free University of Berlin in Germany and his colleagues firsttrained the insects to associate certain stimuli with a reward: sugar. For example, in one experiment bees saw the color blueat the entrance to a so-called Y-maze. The entrance led to a decision chamber, where the bees could choose between two paths:one carried a blue target, the other carried yellow. The bees received a reward only if they chose blue, the same color as thatseen at the entrance.

The team then tested whether the bees could apply what they had learned to a new situation. Blue and yellow patches werereplaced with black and white patterns of vertical and horizontal bars. The bees passed with flying colors, heading straightfor the pattern that matched what they saw at the entrance. Moreover, other experiments revealed that the insects could eventransfer their knowledge across the senses: bees that learned about sameness through olfactory training were able to applythat concept to situations involving visual stimuli.

These results, the authors conclude, demonstrate that "higher cognitive functions are not a privilege of the vertebrates."Moreover, because the honeybee nervous system is relatively simple, they write, "there is a realistic chance of uncovering theneural mechanisms that underlie this capacity." --Kate Wong

COUR

TESY

OF

THE

PROC

EEDI

NGS

OF T

HE

NAT

ION

AL A

CADE

MY

OF S

CIEN

CES


The basilisk lizards of Central America are re-nowned for their seemingly miraculous flightacross water. When startled, these green or

brown reptiles scamper over ponds or lakes on theirhind legs—the younger ones appearing virtually air-borne, the larger ones sinking down somewhat. Byvideotaping seven Basiliscus basiliscus captured in aCosta Rican rain forest and by constructing mechanicalmodels in order to understand the underlying physics,we have been able decipher the mystery of these lizards’magnificent movements.

It all begins with a slap of the foot. The basilisk lizardstrikes the water to create upward force. This force, inturn, provides a medium-size, or 90-gram, lizard withas much as 23 percent of the support it needs to stay onthe water surface. Then, a split second later, comes thestroke. As the foot crashes down, it pushes water mole-cules aside and creates a pothole of air. In addition to

the forces generated by accelerating water out of thefoot’s way, the lizard obtains support from forces creat-ed by the difference in pressure between the air cavityabove the foot and the hydrostatic pressure below. To-gether the slap and subsequent stroke can produce 111percent of the support needed to keep an adult lizardstriding along the surface. Smaller lizards, those weighingtwo grams or less, should be able to create 225 percentof the support they need—and consequently, their runsacross the water appear freer and less cumbersome.

All these gains would be lost, however, if the lizarddid not pull its foot out of the hole before the waterclosed in around it. By slanting its long-toed foot back-wards and by slipping it out while it is surrounded onlyby air, the creature avoids the drag that would resultfrom pulling its foot through water. A tiny fringe thatsurrounds the basilisk’s five toes may facilitate this mo-tion. Like a parachute, the fringe flares out as the foot is

Running The secret of the basilisk lizard’s by James W. Glasheen and Thomas A. McMahon

Originally published in September 1997


slapped down, thus creating more surface area—all the betterto hit the water with. Then, as the foot is pulled up, the fringecollapses, and the long toes are withdrawn just before thehole closes.

Although their secret is now unveiled, the lizards are likelyto remain alone on top of the water. Some web-footed birdscan achieve similar runs on water, but their dynamics are

slightly different and not well understood. As for humans,they have nothing to learn from the lizards except to stayashore: an 80-kilogram person would have to run 30 metersper second (65 miles an hour) and expend 15 times more sus-tained muscular energy than a human being has the capacityto expend. The basilisks bask singularly in the liminal worldbetween water and air.

on Water strategy lies in its stroke

ADULT BASILISK LIZARDS usually run on water onlywhen startled; young ones, however, will do so simply toget from one place to another. A medium-size lizard takesabout 20 steps a second when running (sequence below);with each of these steps the lizard’s foot creates an airpocket from which the foot is withdrawn before waterrushes back in. Tiny collapsible fringes around the basilisk’sfoot (right) may help in this process.

The Authors: JAMES W. GLASHEEN and THOMAS A.MCMAHON worked on the watery capabilities of the basiliskat Harvard University, where McMahon is a professor in the

division of engineering and appliedscience. Glasheen, now a con-

sultant with McKinsey andCompany, was a doctoralstudent at the time oftheir collaboration.

SA

RA

CH

EL T

AYLO

R


TheNOSETakes aSTARRINGRoleBy Kenneth C. Catania

THE STAR-NOSED MOLE HAS WHAT IS VERY LIKELY THE WORLD’S FASTEST

Originally published in July 2002


PINK STAR makes this mole’s nose unmistakable. It also makes it one of the most sensitive touch organs observed in the animal

kingdom—one that works uncannily like an eye.

The renowned physicist John Archibald Wheeler once suggested, “Inany field, find the strangest thing and then explore it.” Certainly it ishard to imagine an animal much stranger than the star-nosed mole,a creature you might picture emerging from a flying saucer to greeta delegation of curious earthlings. Its nose is ringed by 22 fleshy ap-

pendages that are usually a blur of motion as the mole explores its environ-ment. Add large clawed forelimbs, and you’ve got an irresistible biological

mystery. How did this creature evolve? What is the star?How does it function, and what is it used for? These aresome of the questions that I set out to answer about this

unusual mammal. It turns out that the star-nosed mole has more than an in-teresting face; it also has a remarkably specialized brain that may help answerlong-standing questions about the organization and evolution of the mam-malian nervous system.

It may comfort you to know that star-nosed moles (Condylura cristata)are small animals, tipping the scales at a mere 50 grams, about twice theweight of a mouse. They live in shallow tunnels in wetlands across much ofthe northeastern U.S. and eastern Canada and hunt both underground andunderwater. Like the other roughly 30 members of the mole family (Talpi-dae), the star-nosed mole is part of the mammalian order Insectivora, a groupknown for its high metabolism and voracious appetite. So the tiny star-nosedmole with its big appetite must locate enough prey to survive cold northernwinters. It finds earthworms in soil, as other moles do, but in addition it hasaccess to a host of small invertebrates and insect larvae found in the rich mudand leaves of its wetland habitat and in the ponds and streams where it swimsalong the murky bottom to root out prey. And seeking prey is where the starcomes into play. The star is not part of the olfactory system—which governssmell—nor is it an extra hand used to gather food. Instead the star is a touchorgan of unsurpassed sensitivity.

AND MOST FANTASTIC NOSE


Getting Close to the StarWHEN I BEGAN to explore the anato-my of the star with a scanning electronmicroscope—an instrument that revealsthe microscopic structure of the skin sur-face—I thought I would see touch recep-tors here and there in various placesacross the skin. Instead I was surprised tofind that the star, like the retina in the hu-man eye, is made up entirely of sensoryorgans. The surface of each of the 22 ap-pendages that ring the nostrils is com-posed of an aggregation of microscopicprotuberances, or papillae, called Eimer’sorgans. Each Eimer’s organ, in turn, ismade up of an array of neural structuresthat signal different aspects of touch.

Three distinct sensory receptors ac-company each Eimer’s organ. At the verybottom of the organ is a single nerve end-ing that is encircled by many concentricrings, or lamellae, of tissue formed by aSchawann cell, a specialized support cell.This lamellated receptor transmits rela-tively simple information about vibra-tions or about when an individual organfirst contacts an object. Above this re-ceptor is another nerve fiber that makescontact with a specialized cell called aMerkel cell. Unlike the lamellated vari-ety, the Merkel cell-neurite complex sig-nals only the sustained depression of the

skin. Both of these receptors are com-monly found in mammalian skin.

At the top of each Eimer’s organ,however, lies a receptor unique to moles.A series of nerve endings forms a circularpattern of neural swellings in a hub-and-spoke arrangement just below the outerskin surface. Our recordings from thebrains of star-nosed moles suggest thatthis latter sensory component providesthe most significant aspect of touch per-ception: an index of the microscopic tex-ture of various surfaces.

More than 25,000 Eimer’s organsform the star, although it has a surfacearea of less than one square centimeter.Together these sensory organs are sup-plied by more than 100,000 nerve fibersthat carry information to the central ner-vous system and eventually to the high-est mammalian processing center, theneocortex. With this formidable array ofreceptors, the mole can make incrediblyfast sensory discriminations as it prowlsits haunts looking for prey.

The star moves so quickly that youcan’t see it with your naked eye. A high-

speed camera revealed that the star touch-es 12 or more areas every second. Scan-ning its environment with a rapid series oftouches, a star-nosed mole can find andeat five separate prey items, such as thepieces of earthworm we feed them in thelaboratory, in a single second.

Acting Like an EyeEVEN MORE SURPRISING than this as-tonishing speed is the manner in whichthe mole uses the star. The star functionslike an eye. Try reading this sentencewithout moving your eyes, and you willsoon appreciate that your visual system isdivided into two distinct functional sys-tems. At any given time only a small por-tion of a visual scene (about one degree)is analyzed with the high-resolution cen-tral area of your retina, the fovea. Themuch larger low-resolution area of yourretina locates potentially important areasto analyze next. The characteristic rapidmovements of the eyes that reposition thehigh-resolution fovea are called saccades.

Just as we scan a visual scene with oureyes, star-nosed moles constantly shift thestar to scan tactile scenes as they travelthrough their tunnels, quickly exploringlarge areas with the Eimer’s organs of all22 appendages. But when they comeacross an area of interest—such as poten-tial food—they always shift the star sothat a single pair of appendages can carryout more detailed investigations. Humanshave a fovea for sight, and star-nosedmoles have a fovea for touch. The mole’sfovea consists of the bottom pair of shortappendages, located above the mouth,each designated as the 11th appendage.Like the retinal fovea, this part of the starhas the highest density of sensory nerveendings. Moreover, the rapid movementsof the star that reposition this tactile fo-vea onto objects of interest are analogousto saccades in the visual system.

The analogy goes even further. In ourvisual system it is not only the movementsof the eyes and the anatomy of the retinathat revolve around the high-resolutionfovea; human brains are specialized toprocess information predominantly fromthis part of the visual scene.

A characteristic feature of informa-tion processing in mammalian sensory K

EN

NE

TH C

. C

ATAN

IA (

pre

ced

ing

pa

ges

); P

OR

TIA

SLO

AN (

ab

ove)

KENNETH C. CATANIA first encountered star-nosed moles when he worked at the NationalZoo in Washington, D.C., many years ago. Catania, who is assistant professor of biologicalsciences at Vanderbilt University, studies the sensory system of mammals and the orga-nization of the neocortex. He received his undergraduate degree in zoology from the Uni-versity of Maryland and his Ph.D. in neuroscience from the University of California, SanDiego. Catania has received the Capranica Foundation Award in Neuroethology and the In-ternational Society of Neuroethologists/ Young Investigator Award and is a Searle Scholar.

THE

AU

THO

R

bPREY IDENTIFICATION by the star-nosed moletakes place in less than half a second. Whenthe longer appendages touch an interestingobject (a), the nose moves so that theshortest and most sensitive appendage canrapidly touch and identify the item (b), whichis immediately consumed.

aFront view

Top view


systems is the topographic organizationof information from sensory receptors.Visual areas contain maps of the retina,auditory areas provide maps of the coch-lea (the receptors in the ear, which aremaps of tones), and touch areas containmaps of the body’s surface. Such mappingis perhaps nowhere better illustrated thanin the somatosensory system of the star-nosed mole.

Charting TouchWORKING WITH my Vanderbilt Uni-versity colleague Jon H. Kaas, I was ableto explore the organization of the star-nosed mole’s neocortex. By recording theactivity from neurons that compose dif-ferent cortical areas, we charted the neu-ral representation of the star, showingwhere and how neurons in the cortex re-spond to tactile stimulation of the Eimer’sorgans. We identified three separate mapsof the star where the responses of neuronsreflect the anatomy of the nose on the op-posite side of the face. (In all mammals,the left half of the body is represented pre-dominantly in the right side of the cortex,and vice versa.) To our amazement, wealso found that these maps are visible insections of the brain that were stained forvarious cell markers—we could literallysee a star pattern in the cortex.

When we compared the sizes of cor-tical brain maps with the appendages ofthe star, we noticed an obvious discrep-ancy. The 11th appendage, which is oneof the smallest parts of the star, had by farthe largest representation in the cortex.The discrepancy is a classic example ofwhat has been termed cortical magnifica-tion: the most important part of the sen-sory surface has the largest representationin the brain, regardless of the actual sizeof the sensory area on the animal.

The same phenomenon occurs in thevisual system, in which the small retinalfovea has by far the largest portion in vi-sual cortex maps. We also discovered thatneurons representing the 11th appendageresponded to tactile stimulation of verysmall areas, or receptive fields, on the 11thappendage, whereas neurons representingthe other appendages responded solely tostimulation of larger areas. The smaller re-ceptive fields for the 11th appendage re-M

ICR

OG

RAP

HS

BY

KE

NN

ETH

C.

CAT

ANIA

; IL

LUST

RAT

ION

S B

Y P

OR

TIA

SLO

AN

APPENDAGES of the star aremade up entirely of sensoryorgans. These Eimer’s organshave elements that are commonto many animals’ skin receptors:a single nerve ending at the verybase (a), which relays informationabout vibrations and initialcontact with an object, andanother nerve fiber that recordssustained pressure (b). But thevery tip of the Eimer’s organ isfound only in moles: neuralswellings arrayed just below theouter skin, which are amazinglysensitive to the details ofsurfaces (c).

CORTICAL MAPS of the star-nosed mole reveal theimportance of the 11th appendage. As thisschematic shows, the most sensitive appendagegets the most space in the cortex (above). Thesame is true for the most sensitive part of thehuman eye. The organization of the cortex alsobeautifully mirrors the position of the appendages(right) and their relative importance.

1

23

4 5 6 7 8 9

10

11

11

Appendages

b

c

a

1

1

3

4

5

6

7

8

9

1011

2

3

4

5

6

7

89

10

Star-nosed mole

Schematic of anatomy in cortical space

Right cortex

Eimer’s organs

How the Nose Knows2


flect a greater acuity for this region andmirror the organization of visual systems.

The discovery of a somatosensory fo-vea in the star-nosed mole suggests thatthis organizational scheme is a generalevolutionary solution to constructing ahigh-resolution sensory system. Visualsystems with a fovea are the most famil-iar, but auditory systems can have anacoustic fovea as well, as has been ele-gantly demonstrated by Nobuo Suga ofWashington University in mustached bats.Many bats emit an echolocation call thatcontains a narrow frequency range andthen analyze returning echoes to navigateand to detect prey. A large proportion ofthe bat’s auditory receptors (hair cells inthe cochlea) and large areas in the bat’sbrain are devoted to analyzing a narrowfrequency range corresponding to a singleharmonic of the returning echo. This is anexample of an acoustic fovea.

Although it is hard to imagine, batshave an auditory version of a saccade aswell. This is necessary because returningechoes are Doppler-shifted to differentfrequencies—depending on the speed ofthe bat and its target, usually an unfortu-nate insect—and often fall outside the fre-quency range of the acoustic fovea. Be-cause the hunting bat cannot change itsacoustic fovea, it constantly changes thefrequency of its outgoing pulses so thatthe Doppler-shifted returning echo will beat the frequency of its acoustic fovea. Thebehavior is called Doppler-shift compen-sation and is the acoustic equivalent ofmoving the eyes, or the star, to analyze astimulus with the high-resolution area ofthe sensory surface and the correspond-ing computational areas of the brain.

The presence of a sensory fovea in themammalian visual system, auditory sys-tem and somatosensory system is a dra-matic case of convergent evolution andpoints to common constraints in the wayevolution can construct a complex brain.After all, why not just wire the entire sen-sory system for high-resolution input andeliminate the need to constantly shift theeyes, star, or echolocation frequency?One reason, of course, is that it wouldtake a massive enlargement of the brain—

and the nerves carrying sensory inputs toit—to accomplish this task.

It is staggering to consider just howmuch larger the human brain would haveto be if the entire retina were to have thesame resolution as the fovea. To accom-plish this, the human brain would have tobe at least 50 times bigger. Your headwould no longer fit through a doorway.Clearly, it is more efficient to devote alarge part of the computational resourcesof the brain to a small part of the sensorysystem and then to move that area aroundlike a spotlight to analyze important as-pects of the world.

Space Race in the BrainAS OFTEN OCCURS, our observationsabout the star-nosed mole’s sensory sys-tem raised as many questions as they an-swered. How does part of a sensory sur-face acquire such a large section of thebrain’s map in the first place? The tradi-tional understanding has been that eachsensory input acquires the same averageamount of area in a cortical map duringdevelopment, and thus the enlarged rep-resentation of a sensory fovea simply re-flects the greater number of neurons col-lecting information from the foveal region.This theoretical framework, suggestingthat each input has equal squatter’s rightsin the brain, is appealing in its simplicity.But a number of studies have recentlychallenged this democratic assessment ofcortical parcellation in the primate visu-al system by showing that inputs fromthe fovea are allocated more cortical ter-ritory than those containing peripheralinformation.

To see what was happening in thestar-nosed mole, we decided to measurethe cortical representations of the 22 ap-pendages and to compare those areaswith the number of nerve fibers collect-ing information from each appendage. Itwas obvious (after counting more than200,000 nerve fibers!) that sensory neu-rons collecting information from the 11thappendage are granted far more corticalterritory in the brain maps than inputsfrom the other appendages. This is anoth-er parallel between the mole’s somatosen-sory system and primate visual systems,and it shows not only that important ar-eas of a sensory surface can have the high-est number of sensory neurons collecting

information per unit area but also thateach of these inputs can be allocated extracomputational space in the brain.

This observation, however, does notexplain how these sensory inputs manageto take the most territory in cortical maps.The question belongs to one of the mostfascinating areas of current research inneuroscience, because changes to corticalmaps could be a critical component oflearning complex skills and recoveringfrom brain injuries or strokes. Severalstudies indicate that a combination of in-trinsic developmental mechanisms andexperience-dependent plasticity affectsthe shape and maintenance of brain maps.

These findings are especially intrigu-ing in the case of the star-nosed mole, be-cause the pattern of use of the nose—asmeasured by how the mole touches prey

with the different appendages—veryclosely matches the pattern of magnifica-tion for the appendage representations inthe cortex. The correspondence suggeststhat behaviors may shape the way the cor-tex is organized. Alternatively, intrinsicdevelopmental mechanisms may matchthe size of cortical maps to their behav-ioral significance. It is the classic questionof nature versus nurture.

The Developing StarLOOKING AT HOW the star developsin mole embryos can help clarify thismatter. Because the star develops beforeits representation in the cortex, sensoryinputs from the star have an opportunityto influence the way that the corticalmaps form during potentially critical pe-riods of development. M

ICR

OG

RAP

HS

BY

KE

NN

ETH

C.

CAT

ANIA


Star-nosed mole embryos come inabout the strangest-looking varietiesimaginable. Although most embryoslook odd, star-nosed moles appear espe-cially weird because the embryonic handsare gigantic—all the better to dig with lat-er—and the nose is obviously unique.

Studies of the embryos revealed thatappendage 11 was the largest appendageduring early development, despite its rel-atively small size in adults. It also becameclear that Eimer’s organs on the star, andthe neural structures within each Eimer’sorgan, matured first on the 11th ap-pendage. It is as if this appendage gets ahead start compared with all the otherones, which later overtake it in size andnumber of Eimer’s organs. As it turnsout, the retinal fovea in the visual systemalso matures early.

When we examined the correspond-ing patterns in the somatosensory cortex,we found that markers for metabolic ac-tivity appear first in the representation ofthe 11th appendage. This suggests thatthe early development of the fovea resultsin greater activity in the developing cor-tical representation of this area, whichcould allow these inputs to capture thelargest area in the cortical map. Strongevidence from the developing visual sys-tem of primates indicates that sensory in-puts with the greatest level of activity areable to capture the largest areas in the

cortex during development. But it is alsopossible that early behavioral patterns instar-nosed moles—which use the 11thappendage to suckle—contribute to ac-tivity-dependent expansion of the foveain the cortical maps. Sorting out the rel-ative contributions of these different in-fluences is one of our goals.

How the Mole Got Its StarONE CAN’T HELP but wonder how thestar-nosed mole evolved. Examining theembryos provided a road map to star-nosed mole evolution, or at least to thatof its enigmatic nose. The appendagesthat form the star develop unlike any oth-er known animal appendage. Ratherthan growing directly out of the bodywall, the star appendages form as cylin-ders, facing backward and embedded inthe side of the mole’s face. In the courseof development, these slowly emergefrom the face, break free from the skinand then, about two weeks after birth,bend forward to form the adult star. Thebackward developmental sequence sug-gests that ancestral star-nosed molesmight have had strips of sensory organslying flat against the sides of the snout.These might have been slowly raised upover many generations until the star wasformed.

Of course, without further evidence,this might remain a “Just So” story. Butthere exist two mole species—the coastmole (Scapanus orarius) and Towns-end’s mole (S. townsendii)—that haveshort strips of sensory organs lying flatagainst the upper side of their noses, andthese adult noses bear an uncanny re-semblance to the embryonic star. Theseintermediate forms strongly suggest thatsuch an ancestor gave rise to the full-fledged star we see today. However theycame to be, these unlikely noses mayhelp reveal much about the influence ofinnate developmental mechanisms andbehavioral patterns on the organizationof the cortex.

The Natural History of Moles. Martyn L. Gorman and R. David Stone. Cornell University Press, 1990.

Sensory Exotica: A World beyond Human Experience. Howard C. Hughes. MIT Press, 1999.

A Nose That Looks Like a Hand and Acts Like an Eye: The Unusual Mechanosensory System of theStar-Nosed Mole. K. C. Catania in Journal of Comparative Physiology, Vol. 185, pages 367–372; 1999.

M O R E T O E X P L O R E

STAR-NOSED EMBRYO providesclues to the animal’s evolutionaryhistory. The appendages start astubes embedded in the mole’sface. They break free of the skinbefore birth. Two weeks afterbirth, they begin to bend forward.Perhaps these unusual nosesbegan as organs that lay flatagainst the snout, just as they doin the adult coast mole (left).



THE

KOMODODRAGON

On a few small islandsin the Indonesian archipelago,

the world’s largest lizard reigns supreme

by Claudio Ciofi

Originally published in March 1999


Adeer nimbly picks its way down a path meanderingthrough tall savanna grasses. It is an adult male of its

species, Cervus timorensis, weighing some 90 kilo-grams (about 200 pounds). Also known as a Rusa deer, the ani-mal knows this route well; many deer use it frequently as theymove about in search of food. This Rusa’s home is the Indone-sian island of Komodo, a small link in a chain of islands separat-ing the Flores Sea from the Indian Ocean. Most wildlife find sur-vival a struggle, but for the deer on Komodo, and on a few of thenearby islands, nature is indeed quite red in tooth and claw. Thisdeer is about to encounter a dragon.

The Komodo dragon, as befits any creature evoking a mytho-logical beast, has many names. It is also the Komodo monitor, be-ing a member of the monitor lizard family, Varanidae, which to-day has but one genus, Varanus. Residents of the island of Komo-do may call it the ora. Among some on Komodo and the islandsof Rinca and Flores, it is buaja darat (land crocodile), a name thatis descriptive but inaccurate; monitors are not crocodilians. Oth-ers call it biawak raksasa (giant monitor), which is quite correct;it ranks as the largest of the monitor lizards, a necessary logicalconsequence of its standing as the biggest lizard of any kind nowliving on the earth. (A monitor of New Guinea, Varanus sal-vadorii, also known as the Papua monitor, may be longer thanthe lengthiest Komodo dragons. The former’s lithe body andlengthy tail, however, leave it short of the thickset, powerful drag-on in any reasonable assessment of size.) Within the scientific

community, the dragon is Varanus komodoensis. And most ev-eryone also calls it simply the Komodo.

The Komodo’s Way of Life

The deer has wandered within a few meters of a robust maleKomodo, about 2.5 meters (eight feet) long and weighing

45 kilograms. The first question usually asked about Komodosis, How big do they get? The largest verified specimen reached alength of 3.13 meters and was purported to weigh 166 kilo-grams, which may have included a substantial amount of undi-gested food. More typical weights for the largest wild dragonsare about 70 kilograms; captives are often overfed. Althoughthe Komodo can run briefly at speeds up to 20 kilometers perhour, its hunting strategy is based on stealth and power. It hasspent hours in this spot, waiting for a deer, boar, goat or any-thing sizable and nutritious.

Monitors can see objects as far away as 300 meters, so visiondoes play a role in hunting, especially as their eyes are better atpicking up movement than at discerning stationary objects.Their retinas possess only cones, so they may be able to distin-guish color but have poor vision in dim light. Today the tall grassobscures the deer.

Should the deer make enough noise the Komodo may hear it,despite a mention in the scientific paper first reporting its existencethat dragons appeared to be deaf. Later research revealed this be-

JOSE

AZ

EL A

uror

a

KOMODO DRAGON flickshis foot-long, yellow forkedtongue to taste the air.


lief to be false, although the animal doeshear only in a restricted range, probablybetween about 400 and 2,000 hertz.(Humans hear frequencies between 20and 20,000 hertz.) This limitation stemsfrom varanids having but a single bone,the stapes, for transferring vibrationsfrom the tympanic membrane to thecochlea, the structure responsible forsound perception in the inner ear. Mam-mals have two other bones working withthe stapes to amplify sound and transmitvibrations accurately. In addition, thevaranid cochlea, though the most ad-vanced among lizards, contains far fewerreceptor cells than the mammalian ver-sion. The result is an animal that is insen-tient to such sounds as a low-pitchedvoice or a high-pitched scream.

Vision and hearing are useful, but theKomodo’s sense of smell is its primaryfood detector. Its long, yellow forkedtongue samples the air, after which thetwo tongue tips retreat to the roof of themouth, where they make contact withthe Jacobson’s organs. These chemicalanalyzers “smell” the deer by recognizingairborne molecules. The concentrationpresent on the left tongue tip is higherthan that sampled from the right, tellingthe Komodo that the deer is approachingfrom the left. This system, along with anundulatory walk in which the headswings from side to side, helps the drag-on sense the existence and direction ofodoriferous carrion from as far away asfour kilometers, when the wind is right.

The Komodo makes its presenceknown when it is about one meter fromits intended victim. The quick movementof its feet sounds like a “muffled machinegun,” according to Walter Auffenberg,who has contributed more to our knowl-edge of Komodos than any other re-searcher. Auffenberg, a herpetologist atthe University of Florida, lived in the fieldfor almost a year starting in 1969 and re-turned for briefer study periods in 1971and again in 1972. He summed up the

bold, bloody and resolute nature of theKomodo assault by saying, “When theseanimals decide to attack, there’s nothingthat can stop them.” That is, there isnothing that can stop them from their at-tempt—most predator attacks world-wide are unsuccessful. The difficulties inobserving large predators in dense vege-tation turn some quantitative recordsinto best estimates, but it is informativethat one Komodo followed by Auffen-berg for 81 days had only two verifiedkills, with no evidence for the number ofunsuccessful attempts.

For the sake of instructive exposition,the Komodo that has ambushed the deerreaches its target. It attacks the feet first,knocking the deer off balance. Whendealing with smaller prey, it may lungestraight for the neck. The basic strategy issimple: try to smash the quarry to theground and tear it to pieces. Strong mus-cles driving powerful claws accomplishsome of this, but the Komodo’s teeth areits most dangerous weapon. They arelarge, curved and serrated and tear fleshwith the efficiency of a plow parting soil.

Its tooth serrations harbor bits ofmeat from the Komodo’s last meal, ei-ther fresh prey or carrion. This protein-rich residue supports large numbers ofbacteria, which are currently being in-vestigated by Putra Sastrawan, onceAuffenberg’s student, and his colleaguesat the Udayana University in Bali andby Don Gillespie of the El Paso Zoo inTexas. They have found some 50 differ-ent bacterial strains, at least seven ofwhich are highly septic, in the saliva.

If the deer somehow maneuvers awayand escapes death at this point, chancesare that its victory, and it, will nonethe-less be short-lived. The infections it in-curs from the Komodo bite will probablykill it within one week; its attacker, ormore likely other Komodos, will thenconsume it. The Komodo bite is notdeadly to another Komodo, however.Dragons wounded in battle with their

comrades appear to be unaffected bythese otherwise deadly bacteria. Gillespieis searching for antibodies in Komodoblood that may be responsible for savingthem from the fate of the infected deer.

Should the deer fail to escape immedi-ately, the Komodo will continue to rip itapart. Once convinced that its prey is in-capacitated, the dragon may break off itsoffensive for a brief rest. Its victim is nowbadly injured and in shock. The Komodosuddenly launches the coup de grâce, abelly attack. The deer quickly bleeds todeath, and the Komodo begins to feed.

The muscles of the Komodo’s jawsand throat allow it to swallow hugechunks of meat with astonishing rapidi-ty: Auffenberg once observed a femalewho weighed no more than 50 kilo-grams consume a 31-kilogram boar in17 minutes. Several movable joints, suchas the intramandibular hinge that opensthe lower jaw unusually wide, help inthe bolting. The stomach expands easily,enabling an adult to consume up to 80percent of its own body weight in a sin-gle meal, which most likely explainssome exaggerated claims for immenseweights in captured individuals.

Large mammalian carnivores, such aslions, tend to leave 25 to 30 percent oftheir kill unconsumed, declining the in-testines, hide, skeleton and hooves. Ko-modos eat much more efficiently, for-saking only about 12 percent of theprey. They eat bones, hooves andswaths of hide. They also eat intestines,but only after swinging them vigorouslyto scatter their contents. This behaviorremoves feces from the meal. Becauselarge Komodos cannibalize young ones,the latter often roll in fecal material,thereby assuming a scent that their big-ger brethren are programmed to avoidconsuming.

More Komodos, attracted by the aro-mas, arrive and join in the feeding. Al-though males tend to grow larger andbulkier than females, no obvious mor-

KOMODO ISLAND has an area of about 340 squarekilometers (130 square miles) and is clearly hilly. Thehighest points are about 735 meters above sea level.Komodo dragons tend to stay below 500 meters butare found at all elevations. The creatures live only ona few Indonesian islands. As shown on the map, Aus-tralia is 900 kilometers southeast, with Java some500 kilometers to the west and New Guinea 1,500kilometers to the northeast.

LAU

RIE

GRA

CE

AUSTRALIA

NEW GUINEA

JAVA

SULAWESI

TIMOR

BORNEO

KOMODO HABITAT


phological differences mark the sexes.One subtle clue does exist: a slight differ-ence in the arrangement of scales just infront of the cloaca, the cavity housing thegenitalia in both sexes. Sexing Komodosremains a challenge to researchers; thedragons themselves appear to have littletrouble figuring out who is who. With agroup assembled around the carrion, theopportunity for courtship arrives.

Most mating occurs between Mayand August. Dominant males can be-come embroiled in ritual combat intheir quest for females. Using their tailsfor support, they wrestle in upright pos-tures, grabbing each other with theirforelegs as they attempt to throw theopponent to the ground. Blood is usual-ly drawn, and the loser either runs or re-mains prone and motionless.

The victorious wrestler initiates court-ship by flicking his tongue on a female’ssnout and then over her body. The tem-ple and the fold between the torso andthe rear leg are favorite spots. Stimula-tion is both tactile and chemical, throughskin gland secretions. Before copulationcan occur, the male must evert a pair ofhemipenes located within his cloaca, atthe base of the tail. The male then crawlson the back of his partner and inserts oneof the two hemipenes, depending on hisposition relative to the female’s tail, intoher cloaca.

The female Komodo will lay her eggsin September. The delay in laying mayserve to help the clutch avoid the brutallyhot months of the dry season. In addi-tion, unfertilized eggs may have a secondchance with a subsequent mating. The fe-male lays in depressions dug on hillslopes or within the pilfered nests ofMegapode birds. These chicken-size landdwellers make heaps of earth mixed withtwigs that may reach a meter in heightand three meters across. While the eggsare incubating, females may lie on thenests, protecting their future offspring.No evidence exists, however, for parentalcare of newly hatched Komodos.

The hatchlings weigh less than 100grams and average only 40 centimeters inlength. Their early years are precarious,and they often fall victim to predators,including their fellow Komodos. Theyfeed on a diverse diet of insects, smalllizards, snakes and birds. Should they livefive years, they can weigh 25 kilogramsand stretch two meters long. By this time,they have moved on to bigger prey, suchas rodents, monkeys, goats, wild boarsand the most popular Komodo food,deer. Slow growth continues throughout

their lives, which may last more than 30years. The largest Komodos, three metersand 70 kilograms of bone, teeth andsinew, rule their tiny island kingdoms.

The Komodo’s Past

Komodos, as members of the classReptilia, do have a relationship

with dinosaurs, but they are not de-scended from them, as is sometimes be-lieved. Rather Komodos and dinosaursshare a common ancestor. Both monitorlizards and dinosaurs belong to the sub-class Diapsida, or “two-arched reptiles,”characterized by the presence of twoopenings in the temporal region of theskull. The earliest fossils from this groupdate back to the late Carboniferous peri-od, some 300 million years ago.

Two distinct lineages arose from thoseearly representatives. One is Archo-sauria, which included dinosaurs. Theancestor of monitor lizards, in contrast,stemmed from primitive Lepidosauria atthe end of the Paleozoic era, about 250million years ago. Whereas some di-nosaurs evolved upright stances, themonitor lineage retained a sprawlingposture and developed powerful fore-limbs for locomotion. During the Creta-ceous, and starting 100 million yearsago, species related to present-day vara-nids appeared in central Asia. Some ofthese were large marine lizards that van-ished with the dinosaurs, about 65 mil-lion years ago. Others were terrestrialforms, up to three meters in length, thatpreyed on smaller animals and probablyraided dinosaur nests. About 50 millionyears ago, during the Eocene, these spe-cies dispersed throughout Europe andsouth Asia and even into North America.

Wolfgang Böhme of the museum ofnatural history in Bonn has contributedmuch to our understanding of the riseand evolution of the Varanus genus,based on morphological data. DennisKing of the Western Australian Museumand Peter Baverstock and his colleaguesat Southern Cross University are contin-uing research into the evolutionary his-tory of the genus through comparisonsof DNA sequences and chromosomalstructure of varanid species and relatedfamilies. They have concluded that thegenus originated between 40 and 25 mil-lion years ago in Asia.

Varanids reached Australia by about15 million years ago, thanks to a colli-sion between the Australian landmassand southeast Asia. Numerous smallvaranid species, known as pygmy moni-

tors, quickly colonized Australia, fillingmultiple ecological niches. More thantwo million years later a second lineagedifferentiated and spread throughoutAustralia and the Indonesian archipela-go, which was at the time far closer toAustralia than it is today, because muchof the continental shelf was above wa-ter. V. komodoensis is a member of thatlineage, having differentiated from itabout four million years ago.

The Indo-Australian varanids couldtake advantage of their unique faunal en-vironment. Islands simply have fewer re-sources than large landmasses do. Be-cause reptilian predators can subsist onmuch lower total energy requirementsthan mammals can, a reptile will havethe advantage in the race for top preda-tor status under these conditions.

In such a setting, reptiles can alsoevolve to huge size, an advantage forhunting. A varanid called Megalaniaprisca, extinct for around 25,000 years,may have reached a length of six metersand a weight of 600 kilograms; the lateextinction date means that humans mayhave encountered this monster. Komo-dos adopted a more moderate giantism.Reasons for the Komodo’s current re-stricted home range—the smallest ofany large predator—are the subject ofdebate and study. Various researcherssubscribe to alternative routes that thedragons’ ancestors may have taken totheir present locale of Komodo, Flores,Rinca, Gili Motang and Gili Dasami.

Komodo has a different paleogeogra-phy from its neighbors. According toworldwide sea-level changes over thepast 80,000 years and bathymetric dataof the study area, Flores and Rinca werejoined until 10,000 years ago. GiliMotang was connected several times totheir combined landmass. Komodo waslong isolated but appears to have joinedits eastern neighbors about 20,000 yearsago, during the last glacial maximum.That association may have lasted 4,000years. (This scenario is based on my cal-culations of the effect of sea-level varia-tions of about 130 meters during the lastPleistocene glaciation, combined withavailable bathymetric data for the area.)

Tantalizing fossil evidence supportsthe notion that today’s Komodo popula-tions are relics of a larger distributionthat once reached Timor, to the east ofFlores. Fossils of two identical forms of anow extinct pygmy elephant, Stegodon,about 1.5 meters at the shoulder, onboth Timor and Flores suggest thatthose two islands might have been


sufficiently close in the Pleistocene to al-low migration.

The limited resources of an islandcould have driven the evolution of thepygmy elephants, because smaller indi-viduals, with lower food requirements,would have been selected for. In contrast,today’s Komodo dragon may haveevolved from a less bulky ancestor; theavailability of the relatively small ele-phants as prey may have been a drivingforce in the selection of largeness that re-sulted in the modern three-meter Komo-do. (A large reptile still needs far less foodthan a mammal of similar size.) Auffen-berg suggests that the Komodo couldonce “have been a highly specializedpygmy stegodont predator,” althoughprey species similar to modern deer andboars may also have been present beforethe arrival of modern humans within thepast 40,000 years.

Further attempts to reconstruct theKomodo’s evolutionary history requiremore comprehensive fossil finds and ac-curate dating of the islands that harborextant populations. The work of Kingand Baverstock, as well as the integra-tion of paleogeographic data and ge-nome analysis, should shed more lighton the origin of the species.

The World Discovers a Dragon

The West was unaware of the Komo-do until 1910, when Lieutenant van

Steyn van Hensbroek of the Dutch colo-

nial administration heard local storiesabout a “land crocodile.” Members of aDutch pearling fleet also told him yarnsabout creatures six or even seven meterslong. Van Hensbroek eventually foundand killed a Komodo measuring a morerealistic 2.1 meters and sent a photo-graph and the skin to Peter A. Ouwens,director of the Zoological Museum andBotanical Gardens at Bogor, Java.

Ouwens recruited a collector, whokilled two Komodos, supposedly mea-suring 3.1 and 2.35 meters, and capturedtwo young, each just under one meter.On examination of these specimens,Ouwens realized that the Komodo wasin fact a monitor lizard. In the 1912 pa-per in which Ouwens introduced the Ko-modo to the rest of the world, he wrotesimply that van Hensbroek “had re-ceived information . . . [that] on the is-land of Komodo occurred a Varanusspecies of an unusual size.” Ouwensended the paper by suggesting the crea-ture be given the name V. komodoensis.

Understanding the Komodo to be rareand magnificent, local rulers and theDutch colonial government institutedprotection plans as early as 1915. AfterWorld War I, a Berlin Zoological Muse-um expedition roused worldwide interestin the animal. In 1926 W. Douglas Bur-den of the American Museum of NaturalHistory undertook a well-equipped out-ing to Komodo, capturing 27 dragonsand describing anatomical features basedon examinations of some 70 individuals.

The Komodo’s Future

More than 15 expeditions followedBurden’s, but it was Auffenberg

who performed the most comprehensivefield study, looking at everything frombehavior and diet to demographics andthe botanical features of their territory.Auffenberg determined that the Komo-do is, in fact, rare. Recent estimates sug-gest that fewer than 3,500 dragons livewithin the boundaries of Komodo IslandNational Park, which consists of the is-lands of Komodo (1,700 individuals),Rinca (1,300), Gili Motang (100) andPadar (none since the late 1970s), andsome 30 other islets. A census on GiliDasami has never been done. About an-other 2,000 Komodos may live in re-gions of the island of Flores. The Komo-do is now officially considered a “vul-nerable” species, according to the WorldConservation Union; it is also protectedunder the Convention on InternationalTrade in Endangered Species of WildFauna and Flora.

The Komodo dragon has faced majorchallenges during the past 20 years thatthreaten its survival in part of the nation-al park and on Flores. The disappear-ance of dragons on Padar probablystems from poaching of their primaryprey, deer. Policing this rugged andsometimes inaccessible habitat isdifficult; two days after I finished a cen-sus of the island in 1997, 10 deer were

Ibecame interested in Komodos as a graduate student at the Universityof Kent at Canterbury in England. My doctoral thesis, in conservation bi-

ology, required me to perform field research on a rare or endangered spe-cies. I wished to work with reptiles, and I wanted to combine fieldworkwith state-of-the-art molecular biological techniques, which are useful indetermining genetic relationships and divergences between populations.Such studies require collecting blood from a study specimen. Based onthese parameters, the creatures that would have most benefited fromstudy were limited to two species.

The first was a tortoise, Testudo hermanni, that is distributed through-out southern Europe. I instead chose the Komodo both for the challengeand because it is still one of the world’s least studied large predators. Iwould discover many of the reasons for this continuing ignorance. Allthe materials needed for fieldwork must be shipped in or created fromscratch; building Komodo dragon traps is arduous and time-consuming;while rare, attacks by Komodos on humans are not unheard of; and thenthere is the smell.

I wanted mobile traps and immobilized Komodos. I therefore built de-vices along the lines of humane mousetraps, only my mice might reachlengths of three meters. I made the devices with local timber and iron-mesh fencing material. Each trap measured three meters by a half meterby a half meter and had a closable door. Goat served as both bait and asrations for me and a local ranger assistant. Komodos would force them-

selves into the trap as far as they could to get to the meat at the otherend. Once they touched the bait, which was connected to a trigger mech-anism, the entrance to the trap closed.

At this point, we would hang the entire trap on a balance, thus de-termining the weight of the captured individual. Then we would openthe door at the tail end and pull the Komodo out. Komodos smellquite intense to begin with, what with their oral bacterial factories andtheir frequent association with carrion. The rotting goat meat adds tothe aroma, and punctuating the olfactory experience is the habit ofthe threatened Komodo to immediately vomit and defecate, in prepa-ration for fight or flight. Once the rear legs were free, we would tiethem together. We would then continue to pull the Komodo from thetrap until the front legs appeared, and we tied those. Finally, we wouldtape the mouth shut, allowing us to do a quick physical examinationand take blood. We went through this routine on animals smaller thanabout 2.5 meters in length. When we happened to trap any of thelargest individuals, we contented ourselves with drawing blood whilethe Komodo remained ensnared. Using these techniques, I was able toget blood samples from 117 Komodos over five months in 1994 and1997, and I am currently analyzing them. Also in 1997 I attached trans-mitters to eight Komodos to obtain information about movement andhome-range size. —C.C.

From Grad Student to Dragon Wrangler


poached. Nevertheless, a trend towardless poaching overall on Padar hasmoved officials to discuss a reintroduc-tion program.

Padar covers an area of only about 20square kilometers and supports no morethan 600 deer, in turn limiting the num-ber of Komodos. Consequently, geneticdiversity, as insurance against inbreed-ing, would be highly desirable among anew, small Komodo population.

To assist this plan, I started a geneticstudy of the remaining Komodo popula-tions in 1994 to determine the degree ofgenetic similarity within and between theexisting groups. I am currently analyzingDNA from blood samples of 117 drag-ons drawn in 1994 and 1997 [see box onopposite page]. The findings shouldeventually allow the authorities to choosethe most appropriate source populationsfor restocking Padar, based on genetic di-versity. Sex ratio and age structure willalso be factors in the choice of individu-als.

Komodos on Flores face the twinthreats of prey depletion and habitat en-croachment by humans. New settlersslash and burn the monsoon forest, andKomodo dragons are among the firstspecies to disappear. In 1997 I set up abiotelemetric study to look at movementand home-range size of adult dragons inareas with differing degrees of humanpresence, both inside and outside the na-tional park. A data collection covering anumber of consecutive years can showconclusively whether human interfer-ence drives Komodos simply to migrateto different areas or to extinction.

I also initiated a long-term survey toobtain information on the distributionand level of threat to Komodo popula-tions throughout Flores. The survey re-lies on traps set in localities chosen onthe basis of habitat and on sighting re-ports by local people. Over the past 20years, habitat loss has caused the speciesto vanish from an area stretching for

150 kilometers along Flores’s northwestcoast. Populations on the north and westcoasts are also threatened by deforesta-tion and indirectly through deer hunting.

The fortunes of the Komodo dragonare inexorably linked with those of nu-merous other species of fauna and flora,and measures to protect this giant lizardmust take into account the entirety of itsnatural habitat. For example, althoughcentral Flores is inhospitable to dragons,the southern and eastern regions of theisland may harbor scattered populations,still unknown to researchers, that couldact as “umbrellas” to protect the ecosys-tem as a whole. The charismatic dragonalready draws some 18,000 visitors ayear to the area, and patches of forestcontaining Komodos could be the cor-nerstone of an economically viable pro-tection plan for the entire habitat, basedon ecotourism.

In addition, I hope to save the extantpopulations of Komodos by altering thecurrent usage patterns of natural re-

sources, in a transition to sustainableland use. Local officials have already ex-pressed interest in such a plan. For ex-ample, slash-and-burn agriculture couldbe superseded by the cultivation of plantspecies that do not require clearing ofthe canopy to be economically useful. Atechnique as simple as instruction in themanufacture and laying of brick couldsave hardwood now harvested for houseconstruction.

The fate of the world’s few thousandKomodos, living out their lives in a tinycorner of the earth, is probably nowin human hands. Policy decisions, as inso many wildlife conservation issues, willbe as much aesthetic as scientific oreconomic. We can choose to create ahomogeneous world of stultifying same-ness. Or we can choose to maintain aremnant of the mystery that provokedmedieval cartographers to mark theunexplored territories of their maps withthe exhilarating warning, “Here there bedragons.”

POSSIBLE ROUTES (right) by which Komodoancestors traveled to their current island habi-tat are still the subject of debate. Whether theycame from Asia directly or through Java or Aus-tralia first is not clear. Certainly the lower sealevels of the past made more routes possiblethan are obvious today. The more recent re-search in the region has updated the decades-old knowledge that we had of the Komodo’scurrent territory (below).

LAU

RIE

GRA

CE

FLORES

KOMODO

PADAR

GILI MOTANGGILI DASAMI

REO

LABUANBAJO

WAE WUUL RESERVE

POTA

RIUNG

? ?

?

?

?

KOMODO DRAGON RANGE FROM 1970 DATAKOMODO DRAGON RANGE FROM 1997 DATAAREAS TO BE EXPLORED

RINCA

BORNEOSULAWESI

The Author

CLAUDIO CIOFI received his undergraduate education at the Uni-versity of Florence. In 1998 he completed his Ph.D. at the Durrell Insti-tute of Conservation and Ecology at the University of Kent at Canter-bury in England. He is now based at the Zoological Society of London.Ciofi has worked in collaboration with the University of Gadjah Madain Java and with Udayana University in Bali. His Komodo project, orig-inating as a population genetic study, has broadened to include behav-ioral ecology and demography and the consequent protection of habitatand involvement of indigenous people. His research has been supportedby the Zoological Society of London, the Wildlife Conservation Society,the Smithsonian Institution, Earthwatch Institute and British Airways.

Further Reading

A Modern Dragon Hunt on Komodo. L. Broughton in Nation-al Geographic, Vol. 70, pages 321–331; 1936.

Zoo Quest for a Dragon. David Attenborough. LutterworthPress, 1957. Reprinted by Oxford University Press, 1986.

The Behavioral Ecology of the Komodo Monitor. WalterAuffenberg. University of Florida–University Presses of Florida,1981.

Komodo: The Living Dragon. New edition. Dick Lutz and J.Marie Lutz. Dimi Press, Oregon, 1996.

A lecture by Walter Auffenberg is available (in RealAudio) atwww.si.edu/natzoo/hilights/lectures.htm on the World Wide Web.

SA


In the animal world, both predators and parasites survive at the expense of oth-er species. Nevertheless, they don’t get the same press. I am besieged with mailcontaining pleas for money on behalf of wolves and killer whales, but I have yet to

see a T-shirt with the slogan “Long Live the Hookworm.” The problem is, of course,that humans associate a parasitic lifestyle with disease. Our perception is of a furtive or-ganism that insinuates itself inside us and, unlike a decent predator, intends to destroy usever so slowly.

But there exists a form of parasitism considerably less macabre. Social parasitism, as it iscalled, has evolved independently in such diverse creatures as ants and birds. A female cuck-oo, for instance, lays her egg in the nest of another species, such as a warbler, and leaves it forthe host to rear. The brown-headed cowbird does the same. Each bird has evolved so that it pro-duces eggs that match those of its chosen baby-sitter.

Even more varied than these avian parasites are the slave-making ants. The unusual behaviorof the parasitic ant Polyergus breviceps—which I have been studying for 15 years in Arizona atthe American Museum of Natural History’s Southwestern Research Station—offers a perfect ex-

SLAVE RAID by Polyergus ants on a Formica ant colony can be a complex undertaking. The Polyergusworkers charge into a Formica nest (1), stealing pupae that they will take back home and enslave.Meanwhile young Polyergus queens mate amid this warring frenzy, and then each one sets off alone to estab-lish her own colony (2). The newly mated queen battles Formica workers at the entrance of their burrowsome distance from her natal nest (3), fights her way to the Formica queen in the chamber below and kills her(4). By licking the dying queen, the Polyergus queen acquires chemicals that win over the Formica workers,and so they tend to her eggs as well as those of their deceased queen (5).

Slave-Making QueensLife in certain corners of the ant world is fraught

with invasion, murder and hostage-taking. The battle royal is a form of social parasitism

ILL

UST

RA

TIO

N B

Y R

OB

ER

TO

OST

I

2

2 1

by Howard Topoff

Originally published in November 1999



3

4

5


ample. Like the other four species of Polyergus found through-out the world, these ants have completely lost the ability tocare for themselves. The workers do not forage for food, feedthe young or the queen, or even clean up their own nest. Tosurvive, Polyergus ants must get workers from the related antgenus Formica to do their chores for them. Thus, Polyergusworkers periodically undertake a slave raid in which about1,500 of them travel up to 150 meters (492 feet), enter aFormica nest, expel the Formica queen and workers, and cap-ture the pupae.

Back at the Polyergus nest, slaves rear the raided brood un-til the young emerge. The newly hatched Formica workersthen assume all responsibility for maintaining the mixed-species nest. They forage for nectar and dead arthropods, re-gurgitate food to colony members, remove wastes and exca-vate new chambers. When the population becomes too largefor the existing nest, it is the 3,000 or so Formica slaves thatlocate another site and physically transport the approximate-ly 2,000 Polyergus workers, together with eggs, larvae, pu-pae and even the queen, to the new nest.

This master-slave arrangement is not unique. Of the ap-proximately 8,800 species of ants, at least 200 have evolvedsome form of symbiotic relationship with one another. At oneend of the behavioral continuum are facultative parasites,such as the ant Formica wheeleri. These ants are capable ofcaring for themselves but undertake periodic slave raids ondifferent ant species to supplement their labor force. In con-trast, Polyergus and other dulotic (from the Greek word for“servant”) ants are obligatory social parasites. The workersand queen cannot survive without the assistance of slaves.

My field research on Polyergus has been guided by onestubborn objective: to determine the most important adapta-tions in the evolution of obligatory social parasitism. Accord-ingly, I have homed in on the one behavior that is truly spe-cific to Polyergus ants: the capacity of a queen to take over a

Formica nest single-handedly. Because, in addition to thelarge slave-capturing raids that can be seen in some other antspecies, Polyergus has developed an unusual way for a newqueen to establish her own colony.

In most ant species, the process of setting up a new nest isstraightforward. After flying away from her natal colony andmating, a queen tears away her wings, excavates a chamber,lays a few eggs and nourishes her larvae with stored nutrients.When the brood matures, the adult workers immediately as-sume the job of colony maintenance. But a parasitic queenlike Polyergus is incapable of rearing her first brood withoutslaves. So she is confronted with a seemingly impossible mis-sion: to invade a colony of Formica, kill the resident queenand become accepted by the workers. Moreover, she must ac-complish all this without the assistance of a single soldier ant.

For several weeks every year, a few hundred eggs laid by anestablished Polyergus queen develop into males and intoqueens that leave the parent colony and attempt to form anew one. In Arizona the young queens of P. breviceps havewaived even the most traditional sexual ritual among ants:the mating flight. Instead of soaring off, winged Polyergusqueens embark with workers in a well-timed slave raid.Amid the tumult of the advancing swarm, a primed queenwill stop running, attract a male with a pheromone from hermandibular gland, mate with him and then pull off herwings. (Clandestine it’s not.)

Two strategies of colony founding are now available to thisjust-inseminated queen. First, she may continue in the slaveraid and arrive in an invaded colony of Formica whoseworkers and queen are scattered across the terrain. Such dis-organization could facilitate the queen’s mission, but successis usually short-lived. The problem is that colonies of Polyer-gus are extremely territorial and will not tolerate other colo-nies of the same species within their raiding turf. The nexttime this (now usurped) nest is raided by the parent Polyer-

FORMICA PUPA (left) is carried by a Polyergus worker to the Polyergus nest, where it will become part of a brood that the pre-viously captured Formica slaves attend to. Formica worker (right) emerges from its pupal state and views itself as a Polyergus,because that is the only life it knows. It cares for the Polyergus workers and queen, feeding them, cleaning the nest and even mov-ing the nest if it becomes too crowded. A colony of 2,000 Polyergus may have as many as 3,000 Formica slaves. Without them,the colony would perish.

CO

UR

TE

SY O

F H

OW

AR

D T

OPO

FF

These ants have completely lost the ability to care for themselves. The workers do not forage for food, feed the young or the queen, or even clean up their own nest.


gus colony, the new Polyergus queen, along with any work-ers she has produced, will be destroyed.

The alternative tactic for an up-and-coming Polyergusqueen is to bolt from the raiding column and on her own lo-cate a more distant colony of Formica. Although there are noguarantees, this behavior at least increases the likelihood offinding an appropriate host nest outside the raiding territoryof a resident Polyergus population.

Killing Machines

After locating a suitable Formica nest, the serious businessof colony usurpation begins. Because this takes place un-

derground, where direct observations are not possible, mygraduate students and I conducted our studies using trans-

parent laboratory nests. Before each test we placed 15 Formi-ca gnava workers in a nest with 15 pupae and one queen. (Incontrast, a wild colony of F. gnava contains about 5,000workers.) We then placed a newly mated Polyergus queenjust outside the nest.

In most cases the Polyergus queen quickly detects the en-trance and erupts into a frenzy of ruthless activity. She boltsstraight for the Formica queen, literally pushing aside anyFormica workers that attempt to grab and bite her. Our earli-er studies had shown that the Polyergus queen’s two main de-fensive adaptations are powerful mandibles for biting her at-tackers and a repellent pheromone secreted from the Dufour’sgland in her abdomen. With worker opposition liquidated,the Polyergus queen grabs the Formica queen and bites herhead, thorax and abdomen for an unrelenting 25 minutes. Be-tween bouts of biting she uses her extruded tongue to lick thewounded parts of her dying victim. Within seconds of thehost queen’s death, the nest undergoes a most remarkabletransformation. The Formica workers behave as if sedated.They calmly approach the Polyergus queen and start groom-ing her—just as they did their own queen. The Polyergusqueen, in turn, assembles the scattered Formica pupae into aneat pile and stands triumphantly on top of it. At this point,colony takeover is a done deal.

The Royal Feast

Our next goal was to find the key to this remarkablebrainwashing of the Formica workers. One hypothesis

was “chemical acquisition,” whereby the Polyergus queenacquires Formica queen chemicals during the act of killingand licking her. To test this idea, my student Ellen Zimmerli

and I added a twist to our original study: we killed the Formi-ca queen—by freezing and defrosting her—immediately priorto introducing the Polyergus female. Our hypothesis predictedthat the Polyergus queen would still have to attack the deadhost queen, pierce her exoskeleton and ingest her body fluids.

The results were exactly as we had anticipated. On enter-ing the nest, the Polyergus queen pounced on the motionlessFormica queen and started to bite and lick her for about 25minutes, just as if she were alive. As soon as she finished“killing” the lifeless Formica queen, the Polyergus queen wasgroomed by the Formica workers and permanently acceptedas their new ruler.

A second prediction of the “chemical acquisition” hypoth-esis is that it would be difficult for a Polyergus queen to bewelcomed by Formica workers if no Formica queen were

present in the nest. So in our next series of laboratory studieswe simply removed the Formica queen before introducingthe Polyergus female. Sure enough, this act provoked theproverbial battle royal. Fighting between the Formica work-ers and the Polyergus queen was relentless. Although neitherspecies has a stinger, the mandibles of the workers are suffi-ciently formidable to pin the queen down by the legs and biteher until she dies.

Wanted: Single Queen with Workers

Because mature Formica nests often have many queens—unlike Polyergus colonies—we were also curious to see

what would happen when a newly mated Polyergus queeninvaded a polygynous nest. We established a series of Formi-ca colonies that contained between two and 25 queens. Sur-prisingly, the number of Formica queens was of no conse-quence to the Polyergus queen. Because she is accepted as theroyal party once she dispatches the first Formica queen, she isin no rush. Hour by hour, day by day, she methodically lo-cates and kills every Formica queen, sometimes taking sever-al weeks to clear out all remnants of the opposition.

Although our tests had uncovered this suite of behavioraladaptations on the part of Polyergus queens, we also discov-ered that success was not routine. Sometimes the phalanx ofFormica workers in the queen’s path was simply too power-ful, and she was ripped to pieces. Perhaps, I thought, queenshave another strategy. Because the mating seasons of Polyer-gus and Formica queens overlap, it seemed reasonable topostulate that they must, at least occasionally, encounter oneanother on the ground shortly after mating. Suppose a Poly-ergus queen killed a Formica queen in the field? Having thus

To unravel the mystery, it became clear that we had to

start thinking like an ant.


acquired the relevant chemicals, the victorious Polyergusqueen should immediately be able to enter any Formica col-ony with impunity. Without worker attacks to contend with,this queen could leisurely embark on her killing spree, elimi-nating any resident Formica queens.

Not so. When we introduced a newly mated Polyergus queento a newly mated queen of F. gnava, the result was a nonevent.The two spent a few seconds checking each other out withtheir antennae, but we never witnessed a single aggressive in-teraction between any of the opposing queens we tested.

These results suggested that young Formica queens do notpossess the same chemical signature found in more maturequeens. To figure out when a Formica queen takes on the aura,or “aroma,” of an established queen, we set up another exper-iment. Christine A. Johnson and I kept newly mated Formicaqueens in laboratory nests until they laid their eggs. Then werepeated the earlier test. Still no interest on the part of Polyer-gus. After several weeks the eggs hatched into larvae, and weconducted yet another round of tests. The Polyergus queenscontinued to ignore their Formica counterparts.

To unravel the mystery, it became clear that we had to startthinking like an ant. Suppose, we reasoned, a newly matedPolyergus queen entered a new nest and killed a Formicaqueen that was raising her first brood of eggs or larvae. Theinvading queen would be unable to feed herself or the broodand would soon starve. The earliest time that killing theFormica queen would be effective is when her brood had de-veloped into workers eager to assist in foraging and nestmaintenance.

Johnson discovered that it took almost two months fromthe time a Formica queen was inseminated to the stage whenher first brood completed development. As soon as these firsteight workers were functional, Johnson introduced a Polyer-gus queen that had been ruling her own nest of Formicaworkers.

Surprisingly, the Polyergus queen remained passive. In-

deed, it took an additional five months, by which time 19young Formica workers were present, before the Polyergusqueen assaulted the Formica queen. Despite her previous dis-interest, it was clear that the Polyergus queen had retainedher regicidal inclination and aptitude. The Formica queenwas killed, and the handful of newly emerged workers ac-cepted the Polyergus as their new queen. We concluded thatthe chemistry of the Formica queen must change dramatical-ly between the moment of fertilization and the time she hasan established nest. But it appears that the change is a matu-rational one brought about by internal processes, not bymerely having a brood. Elucidating the nature of this chemi-cal transformation should prove a fruitful path for futurestudies.

A Dangerous Living

Having determined that invading an established Formicanest is the key to successful colony takeover, we were

faced with one final thorny issue. Linda Goodloe, a graduatestudent working with the eastern species Polyergus lucidus,had discovered that new queens go after only the Formicaspecies that they grew up with—in other words, the ones thathad been enslaved by their parent queen’s colony. But socialparasitism requires intricate behavioral interactions betweentwo species: parasite and host. Clearly, the evolution from afree-living to a parasitic way of life required that a newlymated queen occasionally invade the nest of an unfamiliarspecies. Although a risky business at best, a successful adop-tion by the foreign workers would enable the invading queento lay many more eggs than she could possibly raise on herown. Rapid colony development, in turn, would set the stagefor the debut of slave raids and the chemical imprinting nec-essary for the slave ants to care for their captors.

So we decided to set up a situation in which this chancehappening could occur. To do so, we traveled to a habitathigher in elevation, one where Polyergus conducts slave raidson Formica occulta instead of F. gnava. We captured coloniesof F. occulta, installed them in our laboratory nests and intro-duced a newly mated Polyergus queen from a colony foundat the lower elevation—a colony that therefore contained F.gnava slaves.

As expected, attempts by Polyergus queens to take overcolonies of Formica containing unfamiliar workers andqueens were only partially successful. Five Polyergus queensshowed no interest in attacking the F. occulta queens; three ofthese nonaggressive Polyergus queens were killed by F. occul-ta workers, and two others evaded attack by abandoning thenest. But the most significant outcome was that two of theseven Polyergus queens did seize and kill the foreign Formicaqueen. And when they had finished licking the dead ruler,both Polyergus queens were promptly adopted by the F. oc-culta workers.

What’s in a Name?

In Origin of Species, Charles Darwin’s description of Polyer-gus shows that he was keenly aware of the numerous conun-

drums raised by the evolution of social parasitism—one ofthose being the issue of what is in it for the slaves. After all, acolony of Formica can forfeit more than 14,000 pupae during asingle raiding season. Their only evolutionary “defense” seemsto be brood replacement, thanks to the Formica queen’s enor-

FORMICA NESTS (red) that are attacked often lie within 150meters (492 feet) of a Polyergus nest (center). This particularPolyergus colony conducted 14 raids on 12 Formica colonies inthe course of about six weeks between July 15 and August 24.

8/9

8/24

7/15

8/6

8/5

7 METERS

7/20

8/12

7/25

7/21

8/8

8/108/177/26

7/30

TO

M M

OO

RE


mous reproductive ability. Although defenseless Formica pupaeare unable to thwart their capture, it is unclear why adultslaves don’t abandon the Polyergus colony and rejoin theirfree-living nestmates.

The answer lies in imprinting. Newly hatched slaves viewthe Polyergus workers, brood and queen as their family.Even though they do not participate in raids on other Formi-ca nests, the slaves respond aggressively to any Formica theymeet while foraging. The imprinting process is similar to thatoccurring between a duckling and its mother—except that inants the stimuli are chemical rather than visual and auditory.The fact that parasitism develops from olfactory bonds be-tween Formica and Polyergus suggests that “slavery” is aninappropriate term for these insects. A more accurate humananalogy would be adoption, because the Polyergus nest is theonly “home” ever known by Formica workers.

Early experience can promote social bonds between differ-ent species of ants, but the process is not open-ended. Theability to form such interspecific attachments declines as theevolutionary relatedness of the creatures decreases. This factexplains why parasite-host relations invariably conform tothe rule identified by Italian entomologist Carlo Emery: so-cial parasites are taxonomically close to their hosts. Not sur-prisingly, this genetic relatedness is connected to ecologicalsimilarity—the quintessence of a successful parasite-host rela-tionship. The Polyergus-Formica association works well pre-cisely because Formica workers in a Polyergus nest need onlyconduct the same foraging and nest maintenance activitiesthat they do in their own colonies. Having been reared in aPolyergus nest does not change Formica’s species-specific be-

haviors of foraging, feeding or fighting. (Fortunately, work-ers don’t mate.)

Since the publication of Origin of Species, scientists have rec-ognized social parasitism in insects and birds as a classic exam-ple of convergent evolution. My field and laboratory researchon the most salient adaptations for parasitism by Polyergus re-veal the depth of this convergence. In England, cuckoos para-sitize four species of host, but any given individual female cuck-oo specializes in one particular host species. And how does thisfemale cuckoo select the appropriate host species? Simple: sheuses the “Polyergus principle” of imprinting. Just as a Polyer-gus queen selects the same species of Formica present in hernest when she emerged, a female cuckoo opts for the hostspecies in whose nest she hatched.

When I first heard the term “social parasitism” as a collegestudent, it sounded like an oxymoron. After all, the term“social” denotes communication, cooperation and evenaltruism—all diametrically opposed to the patently selfishhabits of parasites. As I learned, however, the termis appropriate because a social parasite’s infiltration intothe host’s life is based on the same developmental andcommunicative processes that both parasite and hostuse for interacting with members of their own species.Nevertheless, 15 years of research have reinforced my empa-thy with Darwin as he struggled to incorporate socialparasitism into his theory of natural selection. As usual, Dar-win put it best: “I tried to approach the subject in a skepticalframe of mind, as any one may well be excused for doubtingthe existence of so extraordinary an instinct as that ofmaking slaves.”

BATTLE ROYAL between the Formica queen (left) and thePolyergus queen (right) can take 30 minutes or more. Roughly thesame size and often evenly matched, the queens repeatedly bite

each other with their strong jaws. If the Polyergus queen wins, shelicks the wounds of the Formica queen, thereby gathering thechemicals that make the Formica workers view her as their leader.

The Polyergus queen is out for blood—or other body fluids.

CO

UR

TE

SY O

F H

OW

AR

D T

OPO

FF

SA

The Author

HOWARD TOPOFF became interested in social insects as an undergradu-ate, when he studied army ants in the department of animal behavior at theAmerican Museum of Natural History in New York City. After receiving hisPh.D. in 1968 from a joint program of the museum and the City University ofNew York, he joined the museum as a curator and continued his field researchon the social behavior of army ants. Though a professor in the department ofpsychology at Hunter College of C.U.N.Y., his research is field-oriented andbased primarily at the museum’s Southwestern Research Station, located in theChiricahua Mountains of Arizona. His interest in the evolution of behavior insocial insects led to his more recent studies of slave-making ants. When notteaching or “slaving” away in the field, he develops multimedia science presen-tations for schoolchildren, college students and the adult public. He invitesquestions and can most easily be reached by e-mail: [email protected].

Further Reading

Colony Takeover by a Socially Parasitic Ant, POLYER-

GUS BREVICEPS: The Role of Chemicals Obtained dur-ing Host-Queen Killing. Howard Topoff and Ellen Zim-merli in Animal Behaviour, Vol. 46, Part 3, pages 479–486;September 1993.

Queens of the Socially Parasitic Ant POLYERGUS DoNot Kill Queens of FORMICA That Have Not FormedColonies. Howard Topoff and Ellen Zimmerli in Journal ofInsect Behavior, Vol. 7, No. 1, pages 119–121; January 1994.

Adaptations for Social Parasitism in the Slave-Mak-ing Ant Genus POLYERGUS. Howard Topoff in Compara-tive Psychology of Invertebrates. Edited by Gary Greenbergand Ethel Tobach. Garland Publishing, 1997.


YOUNG FEMALE LIONS, shown here, band together in groups of six to 10,called prides. Such togetherness does not always make them more successfulhunters, as scientists once presumed; loners frequently eat more than individualsin a pride do. Instead communal living makes lions better mothers: pridematesshare the responsibilities of nursing and protecting the group’s young. As a result,more cubs survive into adulthood.

Divided We Fall:Cooperation among Lions

Although they are the most social of all cats, lions cooperate only when it is in their own best interest

by Craig Packer and Anne E. Pusey

In the popular imagination, lions hunting for food present a marvel ofgroup choreography: in the dying light of sunset, a band of stealthy catssprings forth from the shadows like trained assassins and surrounds its

unsuspecting prey. The lions seem to be archetypal social animals, risingabove petty dissension to work together toward a common goal—in this case,their next meal. But after spending many years observing these creatures inthe wild, we have acquired a less exalted view.

Our investigations began in 1978, when we inherited the study of the lionpopulation in Serengeti National Park in Tanzania, which George B. Schallerof Wildlife Conservation International of the New York Zoological Society

Originally published in May 1997


PH

OTO

GR

AP

HS

BY

CR

AIG

PA

CK

ER


began in 1966. We hoped to discoverwhy lions teamed up to hunt, rear cubsand, among other things, scare off rivalswith chorused roars. All this together-ness did not make much evolutionarysense. If the ultimate success of an ani-mal’s behavior is measured by its life-time production of surviving offspring,then cooperation does not necessarilypay: if an animal is too generous, its com-panions benefit at its expense. Why, then,did not the evolutionary rules of geneticself-interest seem to apply to lions?

We confidently assumed that wewould be able to resolve that issue intwo to three years. But lions are su-premely adept at doing nothing. To thelist of inert noble gases, including kryp-ton, argon and neon, we would addlion. Thus, it has taken a variety of re-search measures to uncover clues aboutthe cats’ behavior. Indeed, we have ana-lyzed their milk, blood and DNA; wehave entertained them with tape record-ers and stuffed decoys; and we havetagged individuals with radio-trackingcollars. Because wild lions can live up to18 years, the answers to our questionsare only now becoming clear. But, as weare finding out, the evolutionary basisof sociality among lions is far more com-plex than we ever could have guessed.

Claiming Territory

Male lions form lifelong allianceswith anywhere from one to eight

others—not out of any fraternal good-will but rather to maximize their own

SISTERHOOD makes it possible for pridematesto protect their cubs against invading males (top).Angry groups can ward off lone males, which areon average nearly 50 percent larger than females(middle). And they will frequently attack and killless powerful trespassing females (bottom).

SERENGETINATIONAL PARK


chances for reproducing. Most compan-ions are brothers and cousins that havebeen reared in the same nursery group,or crèche. Others consist of nonrelativesthat teamed up after a solitary nomadicphase. Once matured, these coalitionstake charge of female lion groups, calledprides, and father all offspring born inthe pride during the next two to threeyears. After that, a rival coalition typi-cally moves in and evicts them. Thus, amale lion’s reproductive success dependsdirectly on how well his coalition canwithstand challenges from outsidegroups of other males.

Male lions display their greatest ca-pacity for teamwork while ousting in-vaders—the situation that presents thegreatest threat to their common self-in-terest. At night the males patrol theirterritory, claiming their turf with a se-ries of loud roars. Whenever we broad-cast tape recordings of a strange maleroaring within a coalition’s territory, theresponse was immediate. They searchedout the speaker and would even attacka stuffed lion that we occasionally setbeside it. By conducting dozens of theseexperiments, our graduate student JonGrinnell found that unrelated compan-ions were as cooperative as brothersand that partners would approach thespeaker even when their companionscould not monitor their actions. Indeed,the males’ responses sometimes borderedon suicidal, approaching the speakereven when they were outnumbered bythree recorded lions to one.

In general, large groups dominatesmaller ones. In larger coalitions, themales are typically younger when theyfirst gain entry into the pride, their sub-sequent tenure lasts longer and theyhave more females in their domain. In-deed, the reproductive advantages of co-operation are so great that most solitarymales will join forces with other loners.These partnerships of nonrelatives, how-

MALES are quick to challenge lions they do notknow—real or not. When the authors playedtape recordings of strange males roaring within acoalition’s turf, representatives from that coali-tion immediately homed in on the sound. More-over, they often took the offensive, pouncing ondecoys placed nearby.

SERENGETI NATIONAL PARK in Tanzaniahouses a population of lions that has beenstudied by scientists since 1966.


PREY CAPTURE is usually done by a single lion,when the group is hunting warthog and wilde-beest (photographs). Because she will very likelysucceed in capturing such easy prey, her sisterswill probably eat even if they refrain from thechase. Thus, the pride will often stand back at asafe distance, awaiting a free meal. But when asingle lion is less likely to make a kill—say, if sheis stalking zebra or buffalo—her pridemates willjoin in to pursue the prey together (charts).

FOR WARTHOG

FOR WILDEBEEST

0 0.2 0.4 0.6 0.8 1

PROBABILITY THAT AN INDIVIDUAL LIONWILL BEHAVE AS DESCRIBED

FOR ZEBRA

PURSUEALONE

REFRAIN PURSUEWITH OTHERS

HOW INDIVIDUAL LIONS ACT WHEN HUNTING

FOR BUFFALO


ever, never grow larger than three. Co-alitions of four to nine males are alwayscomposed of close relatives. Why donot solitary males recruit more partnersuntil their groups also reach an insuper-able size? The reasons again come downto genetic self-preservation and, in par-ticular, weighing the odds of gaining ac-cess to a pride against those of actuallyfathering offspring.

Although large coalitions produce themost offspring on a per capita basis, thisaveraging assumes fair division amongcompanions—a form of cooperationthat does not happen in the Serengeti.In fact, the first male to find a female inestrus will jealously guard her, matingrepeatedly over the next four days andattacking any other male that mightventure too close. Dennis A. Gilbert, inStephen J. O’Brien’s laboratory at theNational Cancer Institute, performedDNA fingerprinting on hundreds of ourlion samples and found that one maleusually fathered an entire litter. More-over, reproduction was shared equallyonly in coalitions of two males. In thelarger coalitions, a few males fatheredmost of the offspring. Being left child-less is not too bad from a genetic stand-point if your more successful partner isyour brother or cousin. You can still re-produce by proxy, littering the worldwith nephews and nieces that carryyour genes. But if you are a lone lion,joining forces with more than one ortwo nonrelatives does not pay off.

Hunting

Traditionally, female lions werethought to live in groups because

they benefited from cooperative hunt-ing. (The females hunt more often thanthe resident males.) But on closer exam-ination, we have found that groups ofhunting lions do not feed any betterthan solitary females. In fact, largegroups end up at a disadvantage becausethe companions often refuse to cooper-ate in capturing prey.

Once one female has started to hunt,her companions may or may not joinher. If the prey is large enough to feedthe entire pride, as is the usual case, thecompanions face a dilemma: although

a joint hunt may be more likely to suc-ceed, the additional hunters must exertthemselves and risk injury. But if a lonehunter can succeed on her own, herpridemates might gain a free meal. Thus,the advantages of cooperative huntingdepend on the extent to which a secondhunter can improve her companion’schances for success, and this in turn de-pends on the companion’s hunting abil-ity. If a lone animal is certain to suc-ceed, the benefits of helping could neverexceed the costs. But if she is incompe-tent, the advantages of a latecomer’s as-sistance may well exceed the costs.

Evidence from a wide variety of bird,insect and mammalian species suggeststhat, as expected, cooperation is mostwholehearted when lone hunters doneed help. The flip side of this trend isthat species are least cooperative whenhunters can most easily succeed on theirown. Consistent with this observation,our graduate student David Scheel foundthat the Serengeti lions most often work

together when tackling such difficultprey as buffalo or zebra. But in takingdown easy prey—say, a wildebeest orwarthog—a lioness often hunts alone; hercompanions watch from the sidelines.

Conditions are not the same through-out the world. In the Etosha Pan of Na-mibia, lions specialize in catching one ofthe fastest of all antelopes, the spring-bok, in flat, open terrain. A single lioncould never capture a springbok, and sothe Etosha lions are persistently cooper-ative. Philip Stander of the Ministry ofEnvironment and Tourism in Namibiahas drawn an analogy between theirhunting tactics and a rugby team’s strat-egy, in which wings and centers move inat once to circle the ball, or prey. Thishighly developed teamwork stands insharp contrast to the disorganized hunt-ing style of the Serengeti lions.

All female lions, whether living in theSerengeti or elsewhere, are highly coop-erative when it comes to rearing young.The females give birth in secrecy and

KILLS are shared by the entire pride. If kills aremade close to home, mothers bring their cubs tothe feast. But they deliver nourishment frommore distant kills in the form of milk.


keep their litters hidden in a dry river-bed or rocky outcrop for at least amonth, during which time the cubs areimmobile and most vulnerable to preda-tors. Once the cubs can move, though,the mothers bring them out into theopen to join the rest of the pride. If anyof the other females have cubs, theyform a crèche and remain in near-con-stant association for the next year and ahalf before breeding again. The moth-ers lead their cubs to kills nearby butdeliver nourishment from more distantmeals in the form of milk. When theyreturn from faraway sites, the motherscollapse, leaving their youngsters tonurse while they sleep. We have studiedover a dozen crèches, and in virtuallyevery case, each cub is allowed to nursefrom each mother in the group. Com-munal nursing is a major component ofthe lion’s cooperative mystique.

And yet, as with most other forms ofcooperation among lions, this behavior

is not as noble as it seems. The membersof a crèche feed from the same kills andreturn to their cubs in a group. Some aresisters; others are mother and daughter;still others are only cousins. Some haveonly a single cub, whereas a few have lit-ters of four. Most mothers have two orthree cubs. We milked nearly a dozenfemales and were surprised to discoverthat the amount of milk from each teatdepended on the female’s food intakeand not on the actual size of her brood.

Because some females in a pride havemore mouths to feed, yet all produceroughly the same amount of milk, moth-ers of small litters can afford to be moregenerous. And in fact, mothers of singlecubs do allow a greater proportion oftheir milk to go to offspring that are nottheir own. These females are most gen-erous when their crèchemates are theirclosest relatives. Thus, milk distributiondepends in large part on a pattern of sur-plus production and on kinship. These

factors also influence female behavioracross species: communal nursing ismost common in those mammals—in-cluding rodents, pigs and carnivores—that typically give birth to a wide rangeof litter sizes and live in small kin groups.

Although female lions do nurse theoffspring of other females, they try togive milk primarily to their own cubsand reject the advances of other hungrycubs. But they also need sleep. Whenthey doze for hours at a time, they pre-sent the cubs with an enormous temp-tation. A cub attempting to nurse froma lioness who is not its mother will gen-erally wait until the female is asleep orotherwise distracted. The females musttherefore balance the effort needed to re-sist the attentions of these pests againsttheir own exhaustion.

Generosity among female lions, then,is largely a matter of indifference. Fe-males that have the least to lose sleepbest—owing either to the small size of

0 0.1 0.2 0.3 0.4 0.5 0.6

0 0.1 0.2 0.3 0.4 0.5 0.6

PROPORTION OF ATTEMPTS

NUMBER OF ATTEMPTS(PER FEMALE PER HOUR)

NURSING ATTEMPTS BY CUBS

ONMOTHERS

ON OTHERFEMALES

ONMOTHERS

ON OTHERFEMALES

ONMOTHERS

ON OTHERFEMALES

WHEN OTHER CUBS ARE ALREADY NURSING

WHEN CUB GREETSTHE FEMALE

NURSING is a job shared by all mothers in apride, not out of generosity but, rather, fatigue.Cubs feed when their mothers return from hunt-ing (top). If the mothers stay awake, they will notlet cubs other than their own, such as the largeadolescent shown, take milk from them (bottom).Although cubs try to nurse most often from theirown mothers, they can be quite cunning in theirattempts to nurse from other females (charts).


their own litter or to the company ofclose relatives. Female spotted hyenashave resolved this conflict by keepingtheir cubs in a well-protected den.Mothers return to their cubs for shortperiods, feed their brood and then sleepsomewhere else in peace. By watchinghyenas at the den, we found that moth-er hyenas received as many nursing at-tempts from the cubs of other females asdid mother lions, but the hyenas weremore alert and so prevented any otherthan their own offspring from nursing.

Surviving in the Serengeti

As we have seen, female lions are most gregarious when they have depen-

dent young; the crèche is the social coreof the pride. Childless females occasion-ally visit their maternal companions butgenerally keep to themselves, feedingwell and avoiding the social complexi-ties of the dining room or nursery. Moth-ers do not form a crèche to improve theircubs’ nutrition. And gregarious mothersmay actually eat less than solitary moth-ers; they have no system of baby-sittingto ensure a more continuous food supply.Instead mother lions form a crèche onlyto defend themselves and their cubs.

A female needs two years to rear hercubs to independence, but should hercubs die at any point, she starts matingwithin a few days, and her interval be-tween births is shortened by as much asa year. Male lions are rarely affection-ate to their offspring, but their territori-al excursions provide effective protec-tion. Should the father’s coalition be

ousted, however, the successors will bein a hurry to raise a new set of offspring.Any cubs left over from the previousregime are an impediment to the newcoalition’s immediate desire to mateand so must be eliminated. More than aquarter of all cubs are killed by invad-ing males. The mothers are the ultimatevictims of this never-ending conflict,and they vigorously defend their cubsagainst incoming males. But the malesare almost 50 percent larger than the fe-males, and so mothers usually lose inone-on-one combat. Sisterhood, on theother hand, affords them a fightingchance; in many instances, crèchematessucceed in protecting their offspring.

Male lions are not their only problem.Females, too, are territorial. They defendtheir favorite hunting grounds, denningsites and water holes against other fe-males. Large prides dominate smallerones, and females will attack and killtheir neighbors. Whereas most malescompress their breeding into a few shortyears, females may enjoy a reproductivelife span as long as 11 years. For this rea-son, boundary disputes between prideslast longer than do challenges betweenmale coalitions, and so the females fol-low a more cautious strategy when con-fronted by strangers. Karen E. Mc-

Comb, now at the University of Sussex,found that females would attempt torepel groups of tape-recorded femalesonly when the real group outnumberedthe taped invaders by at least two. Fe-males can count, and they prefer a mar-gin of safety. Numbers are a matter oflife and death, and a pride of only oneor two females is doomed to a futile ex-istence, avoiding other prides and neverrearing any cubs.

The lions’ pride is a refuge in whichindividuals united by common repro-ductive interests can prepare for the en-emy’s next move. The enemy is other li-ons—other males, other females—andthey will never be defeated. Over theyears, we have seen hundreds of malescome and go, each coalition tracing thesame broad pattern of invasion, murderand fatherhood, followed by an inevit-able decline and fall. Dozens of prideshave set out to rule their own patch ofthe Serengeti, but for every new pridethat has successfully established itself,another has disappeared. Lions can seemgrand in their common cause, battlingtheir neighbors for land and deflectingthe unwanted advances of males. Butthe king of beasts above all exemplifiesthe evolutionary crucible in which a co-operative society is forged.

The Authors

CRAIG PACKER and ANNE E. PUSEYare professors in the department of ecology,evolution and behavior at the University ofMinnesota. They conducted their studies atthe Serengeti Wildlife Research Institute, theUniversity of Chicago and the University ofSussex. Packer completed his Ph.D. in 1977at Sussex. That same year, Pusey received herPh.D. from Stanford University.

Further Reading

A Molecular Genetic Analysis of Kinship and Cooperation in African Lions.C. Packer, D. A. Gilbert, A. E. Pusey and S. J. O’Brien in Nature, Vol. 351, No. 6327,pages 562–565; June 13, 1991.

Into Africa. Craig Packer. University of Chicago Press, 1994.Non-offspring Nursing in Social Carnivores: Minimizing the Costs. A. E. Puseyand C. Packer in Behavioral Ecology, Vol. 5, No. 4, pages 362–374; Winter 1994.

Complex Cooperative Strategies in Group-Territorial African Lions. R. Hein-sohn and C. Packer in Science, Vol. 269, No. 5228, pages 1260–1262; September 1,1995.

AFFECTION is common among pridemates,which rely on one another to help protect theiryoung. Male lions present one of the greatestthreats: if one coalition takes over a new pride,the newcomers—eager to produce their own off-spring—will murder all the pride’s small cubsand drive the older cubs away.


TalkingwithAlex:

Parrots were once thought to be no more than excellent mimics, but research is showing that they understand what they say.Intellectually, they rival great apes and marine mammals

Logic by Irene M. Pepperberg

Originally published in Scientific AmericanPresents: Exploring Intelligence 1998


Bye. I’m gonna go eat dinner. I’ll see you tomorrow,” Ihear Alex say as I leave the laboratory each night. What makesthese comments remarkable is that Alex is not a graduate stu-dent but a 22-year-old Grey parrot.

Parrots are famous for their uncanny ability to mimichuman speech. Every schoolchild knows “Polly wanna cracker,”but the general belief is that such vocalizations lack meaning.Alex’s evening good-byes are probably simple mimicry. Still, Iwondered whether parrots were capable of more than mind-less repetition. By working with Alex over the past two decades,I have discovered that parrots can be taught to use and under-stand human speech. And if communication skills provide aglimpse into an animal’s intelligence, Alex has proved thatparrots are about as smart as apes and dolphins.

When I began my research in 1977, the cognitive capacityof these birds was unknown. No parrot had gone beyond thelevel of simple mimicry in terms of language acquisition. Atthe time, researchers were training chimps to communicatewith humans using sign language, computers and special boardsdecorated with magnet-backed plastic chips that representwords. I decided to take advantage of parrots’ ability to pro-duce human speech to probe avian intelligence.

My rationale was based on some similarities between par-rots and primates. While he was at the University ofCambridge, Nicholas Humphrey proposed that primates hadacquired advanced communication and cognitive skillsbecause they live and interact in complex social groups. Ithought the same might be true of Grey parrots (Psittacuserithacus). Greys inhabit dense forests and forest clearingsacross equatorial Africa, where vocal communication plays animportant role. The birds use whistles and calls that they mostlikely learn by listening to adult members of the flock.

Further, in the laboratory parrots demonstrate an abilityto learn symbolic and conceptual tasks often associated withcomplex cognitive and communication skills. During the1940s and 1950s, European researchers such as Otto D. W.Koehler and Paul Lögler of the Zoological Institute of theUniversity of Freiburg had found that when parrots are exposedto an array of stimuli, such as eight flashes of light, some ofthem could subsequently select a set containing the same num-ber of a different type of object, such as eight blobs of clay.

Because the birds could match light flashes with clay blobs onthe basis of number alone means that they understood arepresentation of quantity—a demonstration of intelligence.

But other researchers, including Orval H. Mowrer, foundthat they were unable to teach these birds to engage in referen-tial communication—that is, attaching a word “tag” to a partic-ular object. In Mowrer’s studies at the University of Illinois, aparrot might learn to say “hello” to receive a food reward whenits trainer appeared. But the same bird would also say “hello”at inappropriate times in an attempt to receive another treat.Because the parrot was not rewarded for using the word incor-rectly, eventually it would stop saying “hello” altogether. Someof Mowrer’s parrots picked up a few mimicked phrases, butmost learned nothing at all.

Because parrots communicate effectively in the wild, itoccurred to me that the failure to teach birds referential speechmight stem from inappropriate training techniques ratherthan from an inherent lack of ability in the psittacine subjects.For whatever reason, parrots were not responding vocally tothe standard conditioning techniques used to train other speciesto perform nonverbal tasks. Interestingly, many of the chim-panzees that were being taught to communicate with humanswere not being trained with the standard paradigms; perhapsparrots would also respond to nontraditional training. To testthis premise, I designed a new method for teaching parrots tocommunicate.

Go Ask Alex

The technique we use most frequently involves two hu-mans who teach each other about the objects at hand whilethe bird watches. This so-called model/rival (M/R) protocol isbased, in part, on work done by Albert Bandura of StanfordUniversity. In the early 1970s Bandura showed that childrenlearned difficult tasks best when they were allowed to observeand then practice the relevant behavior. At about the same time,Dietmar Todt, then at the University of Freiburg, independentlydevised a similar technique for teaching parrots to replicatehuman speech.

In a typical training session, Alex watches the trainer pickup an object and ask the human student a question about it:TI

MO

THY

ARC

HIB

ALD

and Speech in Parrots35 SCIENTIFIC AMERICAN EXCLUSIVE ONLINE ISSUE AUGUST 2004


for example, “What color?” If the student answers correctly,he or she receives praise and is allowed to play with the objectas a reward. If the student answers incorrectly, however, thetrainer scolds him or her and temporarily removes the objectfrom sight. The second human thus acts as a model for Alexand a rival for the trainer’s attention. The humans’ interac-tions also demonstrate the consequences of an error: themodel is told to try again or to talk more clearly.

We then repeat the training session with the roles of train-er and model reversed. As a result, Alex sees that communica-tion is a two-way street and that each vocalization is not specificto an individual. In Todt’s studies, birds were exposed only topairs of individuals who maintained their respective roles. Asa result, his birds did not respond to anyone other than thehuman who initially posed the questions. In contrast, Alexwill respond to, interact with and learn from just about anyone.The fact that Alex works well with different trainers suggeststhat his responses are not being cued by any individual—oneof the criticisms often raised about our studies. How could anaive trainer possibly cue Alex to call an almond a “corknut”—his idiosyncratic label for that treat?

In addition to the basic M/R system, we also use supple-mental procedures to enhance Alex’s learning. For example,once Alex begins to produce a word describing a novel item,we talk to him about the object in full sentences: “Here’s thepaper” or “You’re chewing paper.” Framing “paper” within asentence allows us to repeat the new word frequently andwith consistent emphasis, without presenting it as a single,repetitive utterance. Parents and teachers often use such vocalrepetition and physical presentation of objects when teachingyoung children new words. We find that this technique hastwo benefits. First, Alex hears the new word in the way that itis used in normal speech. Second, he learns to produce theterm without associating verbatim imitation of his trainerswith a reward.

We also use another technique, called referential mapping,to assign meaning to vocalizations that Alex produces sponta-neously. For example, after learning the word “gray,” Alex cameup with the terms “grape,” “grate,” “grain,” “chain” and “cane.”Although he probably did not produce these specific new wordsintentionally, trainers took advantage of his wordplay to teachhim about these new items using the modeling and sentence-framing procedures described earlier.

Finally, all our protocols differ from those used by Mowrerand Todt in that we reward correct responses with intrinsicreinforcers—the objects to which the targeted questions refer.So if Alex correctly identifies a piece of wood, he receives apiece of wood to chew. Such a system ensures that at everyinteraction, the subject associates the word or concept to belearned with the object or task to which it refers. In contrast,Mowrer’s programs relied on extrinsic reinforcers. Every correctanswer would be rewarded with a preferred food item—a nut,for example. We think that such extrinsic rewards may delaylearning by causing the animal to confuse the food item withthe concept being learned.

Of course, not every item is equally appealing to a parrot.To keep Alex from refusing to answer any question that doesn’tinvolve a nut, we allow him to trade rewards once he has cor-rectly answered a question. If Alex correctly identifies a key, hecan receive a nut—a more desirable item—by asking for itdirectly, with a simple “I wanna nut.” Such a protocol providessome flexibility but maintains referentiality of the reward.

What’s Different, What’s the Same

I began working with Alex when he was 13months old—a baby in a species in which individuals live up to60 years in captivity. Through his years of training Alex hasmastered tasks once thought to be beyond the capacity of allbut humans and certain nonhuman primates. Not only can heproduce and understand labels describing 50 different objectsand foods but he also can categorize objects by color (rose,blue, green, yellow, orange, gray or purple), material (wood,wool, paper, cork, chalk, hide or rock) and shape (objects hav-ing from two to six corners, where a two-cornered object isshaped like a football). Combining labels for attributes such ascolor, material and shape, Alex can identify, request anddescribe more than 100 different objects with about 80 per-cent accuracy.

In addition to understanding that colors and shapes repre-sent different types of categories and that items can be catego-rized accordingly, Alex also seems to realize that a single objectcan possess properties of more than one category—a green trian-gle, for example, is both green and three-cornered. When pre-sented with such an object Alex can correctly characterize eitherattribute in response to the vocal queries “What color?” or“What shape?” Because the same object is the subject of bothquestions, Alex must change his basis for classification to answereach query appropriately. To researchers such as Keith J. Hayesand Catherine H. Nissen, who did related work with a chim-panzee at the Yerkes Regional Primate Research Center at EmoryUniversity, the ability to reclassify items indicates “abstract apti-tude.” On such tests, Alex’s accuracy averages about 80 percent.

Alex has also learned the abstract concepts of “same” and“different.” When shown two identical objects or two itemsthat vary in color, material or shape, Alex can name whichattributes are the same and which are different. If nothingabout the objects is the same or different, he replies, “None.”He responds accurately even if he has not previously encoun-tered the objects, colors, materials or shapes.

Alex is indeed responding to specific questions and not justrandomly chattering about the physical attributes of the objects.When presented with a green, wooden triangle and a blue,wooden triangle, his accuracy was above chance on questionssuch as “What’s same?” If Alex were ignoring the question andresponding based on his prior training, he might have respond-ed with the label for the one anomalous attribute—“color”—rather than either of the correct answers—“matter” or “shape.”

Alex’s comprehension matches that of chimpanzees anddolphins. He can examine a tray holding seven differentobjects and respond accurately to questions such as “Whatcolor is object-X?” or “What object is color-Y and shape-Z?” Acorrect response indicates that Alex understood all parts of thequestion and used this understanding to guide his search for theone object in the collection that would provide the requestedinformation. His accuracy on such tests exceeds 80 percent.

We also used a similar test to examine Alex’s numericalskills. He currently uses the terms “two,” “three,” “four,” “five”and “sih” (the final “x” in “six” is a difficult sound for a parrotto make) to describe quantities of objects, including groupingsof novel or heterogeneous items. When we show Alex a “con-founded number set”—a collection of blue and red keys andtoy cars, for example—he can correctly answer questions aboutthe number of items of a particular color and form, such as“How many blue key?” His accuracy in this test, 83.3 percent,

36 SCIENTIFIC AMERICAN EXCLUSIVE ONLINE ISSUE AUGUST 2004COPYRIGHT 2004 SCIENTIFIC AMERICAN, INC.

equals that of adult humans who are given a very short time toquantify similarly a subset of items on a tray, according towork done by Lana Trick and Zenon Pylyshyn of theUniversity of Western Ontario.

Alex also comprehends at least one relative concept: size.He responds accurately to questions asking which of twoobjects is the bigger or smaller by stating the color or materialof the correct item. If the objects are of equal size, heresponds, “None.” Next, we will try to get Alex to tackle rela-tive spatial relations, such as over and under. Such a proposi-tion presents an added challenge because an object’s position

relative to a second object can change:what is “over” now could be “under”later.

One last bit of evidence reinforces ourbelief that Alex knows what he is talkingabout. If a trainer responds incorrectly tothe parrot’s requests—by substituting anunrequested item, for example—Alex gen-erally responds like any dissatisfied child:he says, “Nuh” (his word for “no”), andrepeats his initial request. Taken together,these results strongly suggest that Alex isnot merely mimicking his trainers but hasacquired an impressive understanding ofsome aspects of human speech.

Tricks of the Training

What is it about our technique that allowsAlex to master these skills? To address thatquestion, we enlisted a few years ago thehelp of Alo, Kyaaro and Griffin—threeother juvenile Grey parrots. Of the manydifferent variations on our technique wetried with these parrots, none worked aswell as the two-trainer interactive system.We attempted to train Alo and Kyaarousing audiotape recordings of Alex’s train-ing sessions. The birds also watched videoversions of Alex’s sessions while they werein isolation (with an automated systemproviding rewards) or in the presence oftrainers who were slightly interactive.Griffin viewed the same videos in the

presence of a highly interactive human trainer who rephrasedmaterial on the video and questioned the bird directly.Although all three parrots occasionally mimicked the targetedlabels presented in the interactive video sessions, they failedto learn referential speech in any of these situations.

When we then trained these birds using the standard M/Rprotocol, their test scores improved dramatically. In the pasttwo years Griffin, for example, has acquired labels for sevenobjects and is beginning to learn his colors. The parrots’ fail-ure to learn from the alternative techniques suggests thatmodeling and social interactions are important for maintain-

TRANSCRIPTS OF DIALOGUES indicatethat Alex can count objects on a tray.Dialogue 1, recorded in 1986, shows thatAlex can distinguish five objects of twodifferent types—in this case, plant stakesand keys. Dialogue 2, from 1997, revealsthat Alex has become more sophisticatedin his ability: presented with a more com-plex set of objects), Alex can count thenumber of blue blocks and green woolballs without being distracted by the otheritems on the tray.

DIALOGUE 1

Alex is shown two plant stakes and three keys on a tray.

Trainer: How many key?Alex: Wood

Irene (with back to tray, to trainer): Are there any wood?Trainer (to Irene): Yes.

Irene: Try that.Trainer: Okay, tell me, how many wood?

Alex: Two.Irene: Two?

Trainer: Yes.

Alex is given one stake, which he chews apart. It is replaced, and the trayis presented again.

Trainer: Now, how many key?Alex: Key.

Trainer: That’s right, keys. How many?Alex: Two wood.

Trainer: There are two wood, but you tell me, how many key?Alex: Five.

Trainer: Okay, Alex, that’s the number of toys; you tell me, how many key?

Alex: Three.Irene: Three?

Trainer: Good boy! Here’s a key.

DIALOGUE 2

Irene: Okay, Alex, here’s your tray. Will you tell me howmany blue block?

Alex: Block.Irene: That’s right, block…how many blue block?Alex: Four.

Irene: That’s right. Do you want the block?Alex: Wanna nut.

Irene: Okay, here’s a nut. (Waits while Alex eats the nut.)Now, can you tell me how many green wool?

Alex: Sisss...Irene: Good boy!


ing the birds’ attention during training and for highlightingwhich components of the environment should be noted, hownew terms refer to novel objects and what happens whenquestions are answered correctly or incorrectly. All these con-cepts are critical in training birds to acquire some level ofhuman-based communication.

The M/R technique and some variants have also provedvaluable in teaching other species referential communication.Diane Sherman of New Found Therapies in Monterey, Calif.,uses the M/R technique for teaching language skills to devel-opmentally delayed children. Even Kanzi, the bonobo (pygmychimpanzee) trained by Sue Savage-Rumbaugh and her col-leagues at Georgia State University, initially learned to com-municate with humans via computer by watching his motherbeing trained—a variant of our modeling technique. Kanzi’sabilities are probably the most impressive of all primates’trained to date. Chimpanzees have been taught human-basedcodes through a variety of techniques; however, apes thatwere trained using protocols similar to those developed by

Mowrer demonstrated communicationskills that were far less flexible and less“languagelike” than those of apes trainedusing systems that had more in commonwith our techniques.

Bird Brains

Alex continues to perform as well asapes and dolphins in tests of intellectualacuity, even though the structure of theparrot brain differs considerably from thatof terrestrial and aquatic mammals. Unlikeprimates, parrots have little gray matterand thus not much of a cerebral cortex, thebrain region associated with cognitive pro-cessing in higher mammals. Other parts ofAlex’s brain must power his cognitive func-tion.

The parrot brain also differs somewhatfrom that of songbirds, which are knownfor their vocal versatility. Yet Alex has sur-passed songbirds in terms of the relativesize of their “vocabularies.” In addition, hehas learned to communicate with membersof a different species: humans. With eachnew utterance, Alex and his featheredfriends strengthen the evidence indicatingthat parrots are capable of performing com-plex cognitive tasks. Their skills reflect theinnate abilities of parrots and suggest thatwe should remain open to discoveringadvanced forms of intelligence in otheranimals.

About the AuthorIRENE M. PEPPERBERG’s work is for the birds—or so the funding

agencies first thought. “My early grants came back with pink sheetsbasically asking what I was smoking,” she jokes. Pepperberg actuallytrained as a theoretical chemist: as a Ph.D. student at HarvardUniversity, she generated mathematical models to describe boroncompounds. But an episode of Nova featuring “signing” chimps,singing whales and squeaking dolphins drew her to her current work.“I was fascinated to see that people could study animal behavior as acareer,” she says. Now Pepperberg is an associate professor at theUniversity of Arizona at Tucson, a city that brings tears to her eyes—literally. “I’m allergic to everything that grows in Tucson,” Pepperbergsays of the trees, grasses, molds and weeds. In 1997 she used the fundsfrom a John Simon Guggenheim Memorial Foundation fellowship towrite a book on parrot cognition and communication, In Search ofKing Solomon’s Ring: Studies on the Communicative and Cognitive Abilitiesof Grey Parrots (currently in press).

Alex also has a life in publishing—he is the title character in Alexand Friends, a children’s book about the animals that have learned tocommunicate with humans. Through the Internet, you can order aspecial copy—one that Pepperberg has signed and Alex has chewed. Itis available at www.azstarnet.com/nonprofit/alexfoundation/ on theWorld Wide Web.

Wood

2

3

2

4

4

2

Rock

Wool

2

Keys

Wood

Rocks

Keys

Wood

Jacks

Yellow wool

Corks

Rocks

Wood

Object set Target items CorrectAlex's response

SA

ALEX’S ACCURACY in identifying the num-ber of targeted items in 1986 was 70 percent(seven out of 10 questions); now Alex is cor-rect more than 80 percent of the time.

GEO

RGE

RETS

ECK


As researchers quietly approach a clearing in the TaïForest of Ivory Coast, they hear a complex

pattern of soft thuds and cracks. It sounds asthough a small band of people are busy in the forest, apply-ing some rudimentary technology to a routine task. On en-tering the clearing, the scientists observe several individualsworking keenly at anvils, skillfully wielding wooden ham-mers. One or two juveniles have apprenticed themselves tothe work and—more clumsily and with less success—arestruggling to lift the best hammer they can find. All this activ-ity is directed toward cracking rock-hard but nutritious coulanuts. Intermittently, individuals set aside their tools to gathermore handfuls of nuts. An infant sits with her mother, gath-ering morsels of broken nuts.

In many ways, this group could indeed be a family of for-aging people. The hammers and anvils they leave behind,some made of stone, would excite the imagination of any an-thropologist searching for signs of a primitive civilization.Yet these forest residents are not humans but chimpanzees.

The similarities between chimpanzees and humans havebeen studied for years, but in the past decade researchers havedetermined that these resemblances run much deeper thananyone first thought. For instance, the nut cracking observedin the Taï Forest is far from a simple chimpanzee behavior;rather it is a singular adaptation found only in that particularpart of Africa and a trait that biologists consider to be an ex-pression of chimpanzee culture. Scientists frequently use theterm “culture” to describe elementary animal behaviors—such as the regional dialects of different populations of song-birds—but as it turns out, the rich and varied cultural tradi-

tions found among chimpanzees are second in complexityonly to human traditions.

During the past two years, an unprecedented scientific col-laboration, involving every major research group studyingchimpanzees, has documented a multitude of distinct cultur-al patterns extending across Africa, in actions ranging fromthe animals’ use of tools to their forms of communicationand social customs. This emerging picture of chimpanzeesnot only affects how we think of these amazing creatures butalso alters human beings’ conception of our own uniquenessand hints at very ancient foundations for humankind’s ex-traordinary capacity for culture.

Contemplating Culture

Homo sapiens and Pan troglodytes have coexisted forhundreds of millennia and share more than 98 percent

of their genetic material, yet only 40 years ago we still knewnext to nothing about chimpanzee behavior in the wild. Thatbegan to change in the 1960s, when Toshisada Nishida of Kyo-to University in Japan and Jane Goodall began their studies ofwild chimpanzees at two field sites in Tanzania. (Goodall’s re-search station at Gombe—the first of its kind—is more fa-mous, but Nishida’s site at Mahale is the second-oldest chim-panzee research site in the world.)

In these initial studies, as the chimpanzees became accus-tomed to close observation, the remarkable discoveries began.Researchers witnessed a range of unexpected behaviors, in-cluding fashioning and using tools, hunting, meat eating, foodsharing and lethal fights between members of neighboring com-

Culturesthe

ChimpanzeesofHumankind’s nearest relative is even closer than we thought: chimpan-

zees display remarkable behaviors that can only be described as social

customs passed on from generation to generationby Andrew Whiten and Christophe Boesch

Originally published in January 2001


The notion that the great apes—chimpanzees, gorillas,orangutans and gibbons—can imitate one another

might seem unsurprising to anyone who has watchedthese animals playing at the zoo. But in scientific circles,the question of whether apes, well, ape, has become con-troversial.

Consider a young chimpanzee watching his mothercrack open a coula nut, as has been observed in the TaïForest of West Africa. In most cases, the youth will even-tually take up the practice himself. Was this because heimitated his mother? Skeptics think perhaps not. They ar-gue that the mother’s attention to the nuts encouraged theyoungster to focus on them as well. Once his attentionhad been drawn to the food, the young chimpanzeelearned how to open the nut by trial and error, not by im-itating his mother.

Such a distinction has important implications for anydiscussion of chimpanzee cultures. Some scientists definea cultural trait as one that is passed down not by geneticinheritance but instead when the younger generationcopies adult behavior. If cracking open a coula is some-thing that chimpanzees can simply figure out how to doon their own once they hold a hammer stone, then it can’tbe considered part of their culture. Furthermore, if theseanimals learn exclusively by tri-al and error, then chimpanzeesmust, in a sense, reinvent thewheel each time they tackle anew skill. No cumulative cul-ture can ever develop.

The clearest way to estab-lish how chimpanzees learn isthrough laboratory experi-ments. One of us (Whiten), incollaboration with Deborah M.Custance of Goldsmiths Col-lege, University of London,constructed artificial fruits toserve as analogues of those theanimals must deal with in thewild (right). In a typical experi-ment, one group of chimpan-zees watched a complex tech-nique for opening one of thefruits, while a second group ob-served a very different method;we then recorded the extent towhich the chimpanzees hadbeen influenced by the methodthey observed. We also con-ducted similar experiments

with three-year-old human children as subjects. Our re-sults demonstrate that six-year-old chimpanzees show im-itative behavior that is markedly like that seen in the chil-dren, although the fidelity of their copying tends to bepoorer.

In a different kind of experiment, one of us (Boesch),along with some co-workers, gave chimpanzees in theZurich Zoo in Switzerland hammers and nuts similar tothose available in the wild. We then monitored the reper-toire of behaviors displayed by the captive chimpanzees.As it turned out, the chimpanzees in the zoo exhibited agreater range of activities than the more limited and fo-cused set of actions we had seen in the wild. We interpret-ed this to mean that a wild chimpanzee’s cultural environ-ment channeled the behavior of youngsters, steering themin the direction of the most useful skills. In the zoo, with-out benefit of existing traditions, the chimpanzees experi-mented with a host of less useful actions.

Interestingly, some of the results from the experimentsinvolving the artificial fruits converge with this idea. In onestudy, chimpanzees copied an entire sequence of actionsthey had witnessed, but did so only after several viewingsand after trying some alternatives. In other words, theytended to imitate what they had observed others doing at

the expense of their own trial-and-error discoveries.

In our view, these findingstaken together suggest that apesdo ape and that this abilityforms one strand in culturaltransmission. Indeed, it is diffi-cult to imagine how chimpan-zees could develop certain geo-graphic variations in activitiessuch as ant-dipping and para-site-handling without copyingestablished traditions. Theymust be imitating other mem-bers of their group.

We should note, however,that—just as is the case with hu-mans—certain cultural traits areno doubt passed on by a combi-nation of imitation and simplerkinds of social learning, such ashaving one’s attention drawn touseful tools. Either way, learn-ing from elders is crucial togrowing up as a competentwild chimpanzee.

—A.W. and C.B.

Do Apes Ape?Recent studies show that chimpanzees and other apes can learn by imitation

PRACTICE MAKES PERFECT as a juvenile chimpanzeeexperiments with an artificial fruit it has been given to“peel” after watching others do so. Such studies helpscientists determine how chimpanzees learn by imitat-ing others.

SAR

AH

MA

RSH

AL

L A

ND

AN

DR

EW

WH

ITE

N N

gam

ba,

UW

EX

, U

gand

a


ILL

UST

RA

TIO

NS

BY

PA

TR

ICIA

J. W

YN

NE

; MA

P B

Y S

UE

CA

RL

SON

A Guide to the Cultures of Chimpanzees

In an effort to catalogue cultural variations among chimpanzees,we asked researchers working at six sites across central Africa to

classify chimpanzee behaviors in terms of occurrence or absence inseven communities. (There are two communities at Mahale.) Thekey categories were customary behavior, which occurs in most or allmembers of one age or sex class; habitual, which is less common but

which still occurs repeatedly; present; absent; and unknown. Certainbehaviors are absent for ecological reasons (eco): for example, chim-panzees do not use hammers to open coula nuts at Budongo, be-cause the nuts are not available. The survey turned up 39 chim-panzee rituals that are labeled as cultural variations; 18 are illustratedbelow. —A.W. and C.B.

customary

customary

absent

present

customary

absent

present

absent

absent

customary

absent

absent(eco)

absent

customary

customary

habitual

habitual

absent

KIBALE

absent(eco?)

absent(eco?)

absent(eco)

absent

absent

absent

present

absent

absent

absent(eco)

absent(eco?)

absent(eco?)

absent

absent

absent

absent

habitual

absent

absent

absent(eco?)

absent

absent

absent

absent

absent

absent

absent

BOSSOUTAÏ

FOREST BUDONGOMAHALEM-GROUP

Hammering nutsTo crack open nutritious coula nuts, chim-panzees use stones as rudimentary ham-mers and anvils.

Pounding with pestle With the stalks of palm trees acting asmakeshift pestles, chimpanzees can poundand deepen holes in trees.

Fishing for termitesChimpanzees insert thin, flexible strips ofbark into termite mounds to extract theinsects, which they then eat.

Wiping ants off stick manuallyOnce the ants have swarmed almost half-way up sticks dipped into the insects’nests, chimpanzees pull the sticks throughtheir fists and sweep the ants into theirmouths.

Eating ants directly off stickAfter a few ants climb onto sticks insertedinto the nests, chimpanzees bring the sticksdirectly to their mouths and eat the ants.

Removing bone marrowWith the help of small sticks, chimpanzeeseat the marrow found inside the longbones of monkeys they have killed andeaten.

Sitting on leavesA few large leaves apparently serve as pro-tection when chimpanzees sit on wetground.

Fanning fliesTo keep flies away, chimpanzees utilizeleafy twigs as a kind of fan.

Tickling selfA large stone or stick can be used to probeespecially ticklish areas on a chimpanzee’sown body.

MAHALEK-GROUP

absent

absent(eco?)

customary

absent

absent

absent

absent

absent

absent

GOMBE

absent

absent

customary

customary

present

absent

absent

present

habitual


ThrowingChimpanzees can throw objects such asstones and sticks with clear—though ofteninaccurate—aim.

Inspecting woundsWhen injured, chimpanzees touch woundswith leaves, then examine the leaves. Insome instances, chimpanzees chew theleaves first.

Clipping leavesTo attract the attention of playmates or fer-tile females, male chimpanzees noisily tearleaf blades into pieces without eating them.

Squashing parasites on leavesWhile grooming another chimpanzee, anindividual removes a parasite from its part-ner, places it on a leaf and then squashes it.

Inspecting parasitesParasites removed during grooming areplaced on a leaf in the chimpanzee’s palm;the animal inspects the insect, then eats ordiscards it.

Squashing parasites with fingersChimpanzees remove parasites from theirgrooming partners and place the tiny in-sects on their forearms. They then hit thebugs repeatedly before eating them.

Clasping arms overheadTwo chimpanzees clasp hands above theirheads while grooming each other with theopposite hand.

Knocking knucklesTo attract attention during courtship,chimpanzees rap their knuckles on trees orother hard surfaces.

Rain dancingAt the start of heavy rain, adult males per-form charging displays accompanied bydragging branches, slapping the ground,

Bossou,Guinea

Taï Forest,Ivory Coast

Budongo,Uganda

Kibale,Uganda

Gombe,Tanzania Mahale,

Tanzania

customary

absent

customary

absent

absent

absent

absent

present

absent

customary

present

customary

absent

absent

customary

habitual

customary

habitual

KIBALE

present

customary

habitual

absent

absent

absent

customary

absent

customary

present

absent

customary

absent

customary

absent

absent

absent

habitual

customary

absent

customary

unknown

unknown

absent

customary

customary

customary

BOSSOUTAÏ

FOREST BUDONGOMAHALEM-GROUP

MAHALEK-GROUP

absent

absent

customary

unknown

unknown

absent

customary

customary

GOMBE

customary

present

absent

habitual

present

present

absent

habitual

customary


munities. In the years that followed, oth-er primatologists set up camp elsewhere,and, despite all the financial, politicaland logistical problems that can besetAfrican fieldwork, several of these out-posts became truly long-term projects.As a result, we live in an unprecedentedtime, when an intimate and compre-hensive scientific record of chimpan-zees’ lives at last exists not just for onebut for several communities spreadacross Africa.

As early as 1973, Goodall recorded 13forms of tool use as well as eight socialactivities that appeared to differ betweenthe Gombe chimpanzees and chim-panzee populations elsewhere. She ven-tured that some variations had whatshe termed a cultural origin. But whatexactly did Goodall mean by “cul-ture”? According to the Oxford Ency-clopedic English Dictionary, culture isdefined as “the customs . . . and achieve-ments of a particular time or people.”The diversity of human cultures ex-tends from technological variations tomarriage rituals, from culinary habitsto myths and legends. Animals do nothave myths and legends, of course. Butthey do have the capacity to pass on be-havioral traits from generation to gen-

eration, not through their genes but bylearning. For biologists, this is the fun-damental criterion for a cultural trait: itmust be something that can be learnedby observing the established skills ofothers and thus passed on to futuregenerations [see box on page 66].

By the 1990s the discovery of newbehavioral differences among chimpan-zees made it feasible to begin assem-bling comprehensive charts of culturalvariations for these animals. William C.McGrew, in his 1992 book Chimpan-zee Material Cultures, was able to list19 different kinds of tool use in distinctcommunities. One of us (Boesch), alongwith colleague Michael Tomasello ofthe Max Planck Institute for Evolution-ary Anthropology in Leipzig, Germany,identified 25 distinct activities as poten-tial cultural traits in wild chimpanzeepopulations.

The most recent catalogue of culturalvariations results from a unique collab-oration of nine chimpanzee experts (in-cluding the two of us) who pooled ex-tensive field observations that, taken to-gether, amounted to a total of 151years of chimp watching [see box onopposite page]. The list cites 39 pat-terns of chimpanzee behavior that we

believe to have a cultural origin, includ-ing such activities as using sticks to“fish” for ants, making dry seats fromleaves, and a range of social groominghabits. At present, these 39 variants putchimpanzees in a class of their own,with far more elaborate customs thanany other animal studied to date. Ofcourse, chimpanzees also remain dis-tinct from humans, for whom culturalvariations are simply beyond count.(We must point out, however, that sci-entists are only beginning to uncoverthe behavioral complexity that existsamong chimpanzees—and so the num-ber 39 no doubt represents a minimumof cultural traits.)

Multicultural Chimpanzees

When describing human customs,anthropologists and sociologists

often refer to “American culture” or“Chinese culture”; these terms encom-pass a wide spectrum of activities—lan-guage, forms of dress, eating habits,marriage rituals and so on. Among ani-mals, however, culture has typicallybeen established for a single behavior,such as song dialects among birds. Or-nithologists haven’t identified variation

Scientists have been investigating chimpanzee culture forseveral decades, but too often their studies contained a

crucial flaw. Most attempts to document cultural diversityamong chimpanzees have relied solely on officially publishedaccounts of the behaviors recorded at each research site. Butthis approach probably overlooks a good deal of cultural vari-ation for three reasons.

First, scientists typically don’t publish an extensive list ofall the activities they do not see at a particular location. Yetthis is exactly what we need to know—which behaviors wereand were not observed at each site. Second, many reports de-scribe chimpanzee behaviors without saying how commonthey are; without this information, we can’t determinewhether a particular action was a once-in-a-lifetime aberra-tion or a routine event that should be considered part of theanimals’ culture. Finally, researchers’ descriptions of poten-tially significant chimpanzee behaviors frequently lack suffi-cient detail, making it difficult for scientists working at otherspots to record the presence or absence of the activities.

To remedy these problems, the two of us decided to take anew approach. We asked field researchers at each site for a listof all the behaviors they suspected were local traditions. Withthis information in hand, we pulled together a comprehensivelist of 65 candidates for cultural behaviors.

Then we distributed our list to the team leaders at eachsite. In consultation with their colleagues, they classified each

behavior in terms of its occurrence or absence in the chim-panzee community studied. The key categories were custom-ary behavior (occurs in most or all of the able-bodied mem-bers of at least one age or sex class, such as all adult males),habitual (less common than customary but occurs repeatedlyin several individuals), present (seen at the site but not habitu-al), absent (never seen), and unknown.

Our inquiry concentrated on seven sites with chimpanzeeshabituated to human onlookers; all told, the study compiled atotal of more than 150 years of chimpanzee observation. Thebehavior patterns we were particularly interested in, ofcourse, were those absent in at least one community, yet ha-bitual or customary in at least one other; this was our criteri-on for denoting any behavior a cultural variant. (Certain be-haviors are absent for specific local reasons, however, and weexcluded them from consideration. For example, althoughchimpanzees at Bossou scoop tasty algae from pools of waterwith a stick, chimpanzees elsewhere don’t do this, simply be-cause algae are not present.)

The extensive survey turned up no fewer than 39 chim-panzee patterns of behavior that should be labeled as culturalvariations, including numerous forms of tool use, groomingtechniques and courtship gambits, several of which are illus-trated throughout this article. This cultural richness is far inexcess of anything known for any other species of animal. —A.W. and C.B.

The Culture ClubHow an international team of chimpanzee experts conducted the most comprehensive survey of the animals ever attempted



in courtship patterns or feeding prac-tices, for example, to go alongside thedifferences in dialect.

Chimpanzees, though, do more thandisplay singular cultural traits: eachcommunity exhibits an entire set of be-haviors that differentiates it from othergroups [see illustrations on pages 64and 65]. As a result, we can talk about“Gombe culture” or “Taï culture.” In-deed, once we observe how a chim-panzee behaves, we can identify wherethe animal lives. For instance, an indi-vidual that cracks nuts, leaf-clips duringdrumming displays, fishes for ants withone hand using short sticks, and knuck-le-knocks to attract females clearlycomes from the Taï Forest. A chimpthat leaf-grooms and hand-clasps dur-ing grooming can come from the KibaleForest or the Mahale Mountains, but ifyou notice that he also ant-fishes, thereis no doubt anymore—he comes fromMahale.

In addition, chimpanzee cultures gobeyond the mere presence or absence ofa particular behavior. For example, allchimpanzees dispatch parasites foundduring grooming a companion. But atTaï they will mash the parasites againsttheir forearms with a finger, at Gombethey squash them onto leaves, and atBudongo they put them on a leaf to in-spect before eating or discarding them.Each community has developed a uniqueapproach for accomplishing the samegoal. Alternatively, behaviors may looksimilar yet be used in different contexts:at Mahale, males “clip” leaves noisilywith their teeth as a courtship gesture,whereas at Taï, chimpanzees incorporateleaf-clipping into drumming displays.

The implications of this new pictureof chimpanzee culture are many. Theinformation offers insight into our dis-tinctiveness as a species. When we firstpublished this work in the journal Na-

ture, we found some people quite dis-turbed to realize that the characteristicthat had appeared to separate us sostarkly from the animal world—our ca-pacity for cultural development—is notsuch an absolute difference after all.

But this seems a rather misdirected re-sponse. The differences between humancustoms and traditions, enriched andmediated by language as they are, arevast in contrast with what we see in thechimpanzee. The story of chimpanzeecultures sharpens our understanding ofour uniqueness, rather than threateningit in any way that need worry us.

Human achievements have madeenormous cumulative progress over thegenerations, a phenomenon Boesch andTomasello have dubbed the “ratchet ef-fect.” The idea of a hammer—once sim-ply a crude stone cobble—has been mod-ified and improved on countless timesuntil now we have electronically con-trolled robot hammers in our factories.Chimpanzees may show the beginningsof the ratchet effect—some that usestone anvils, for example, have gone astep further, as at Bossou, where theywedge a stone beneath their anvil whenit needs leveling on bumpy ground—but such behavior has not become cus-tomary and is rudimentary indeed be-side human advancements.

The cultural capacity we share withchimpanzees also suggests an ancientancestry for the mentality that must un-derlie it. Our cultural nature did notemerge out of the blue but evolved fromsimpler beginnings. Social learning simi-lar to that of chimpanzees would ap-pear capable of sustaining the earlieststone-tool cultures of human ancestorsliving two million years ago.

Whether chimpanzees are the solespecies on the planet that shares hu-mankind’s capacity for culture is tooearly to judge: nobody has undertaken

the comprehensive research necessaryto test the idea. Early evidence hintsthat other creatures should be includedin these discussions, however. Carel P.van Schaik and his colleagues at DukeUniversity have found orangutans inSumatra that habitually use at least twodifferent kinds of tools. Orangutansmonitored for years elsewhere havenever been seen to do this.

And Hal Whitehead of DalhousieUniversity and his colleagues have be-gun to document the ways in whichpopulations of whales that sing in dif-ferent dialects also hunt in differentways. We hope that our comprehensiveapproach to documenting chimpanzeecultures may provide a template for thestudy of these other promising species.

What of the implications for chim-panzees themselves? We must highlightthe tragic loss of chimpanzees, whosepopulations are being decimated justwhen we are at last coming to appreci-ate these astonishing animals more com-pletely. Populations have plummeted inthe past century and continue to fall asa result of illegal trapping, logging and,most recently, the bushmeat trade. Thelatter is particularly alarming: logginghas driven roadways into the forest thatare now used to ship wild-animalmeat—including chimpanzee meat—toconsumers as far afield as Europe. Suchdestruction threatens not only the ani-mals themselves but also a host of fasci-natingly different ape cultures.

Perhaps the cultural richness of theape may yet help in its salvation, how-ever. Some conservation efforts have al-ready altered the attitudes of some localpeople. A few organizations have be-gun to show videotapes illustrating thecognitive prowess of chimpanzees. OneZairian viewer was heard to exclaim,“Ah, this ape is so like me, I can nolonger eat him.”

The Authors

ANDREW WHITEN and CHRISTOPHE BOESCH have collab-orated since 1998 on the cross-cultural study of chimpanzees.Whiten, a fellow of the British Academy, is professor of evolution-ary and developmental psychology at the University of St. Andrewsin Scotland. Boesch is co-director of the Max Planck Institute forEvolutionary Anthropology in Leipzig, Germany, and a professor atthe University of Leipzig. The chimpanzee field-study directors par-ticipating in the research described here are Jane Goodall, JaneGoodall Institute, Washington, D.C.; William C. McGrew, MiamiUniversity; Toshisada Nishida, Kyoto University, Japan; VernonReynolds, University of Oxford; Yukimaru Sugiyama, TokaigakuenUniversity, Japan; Caroline E. G. Tutin, University of Stirling, Scot-land; and Richard W. Wrangham, Harvard University.

Further Information

Chimpanzee Material Culture. William C. McGrew. Cam-bridge University Press, 1992.

Cultures in Chimpanzees. A. Whiten, J. Goodall, W. C. McGrew,T. Nishida, V. Reynolds, Y. Sugiyama, C.E.G. Tutin, R. W. Wrang-ham and C. Boesch in Nature, Vol. 399, pages 682–685; 1999.

Chimpanzees of the Taï Forest: Behavioral Ecology andEvolution. Christophe Boesch and Hedwige Boesch-Aschermann.Oxford University Press, 2000.

Primate Culture and Social Learning. Andrew Whiten inCognitive Science. Special issue on primate cognition, Vol. 24,pages 477–508; 2000.

Chimpanzee Cultures Web site: http://chimp.st-and.ac.uk/cultures/Wild Chimpanzee Foundation Web site: http://www.wildchimps.org


Documents

2004.N16.Amazing Animals