Aren't chimps supposed to be plant eaters?
The Predatory Behavior and Ecology of Wild Chimpanzees
When Jane Goodall first observed wild chimpanzees hunting and eating meat nearly 40 years ago, skeptics suggested that their behavior was aberrant and that the amount of meat eaten was trivial. Today, we know that chimpanzees everywhere eat mainly fruit, but are also predators in their forest ecosystems. In some sites the quantity of meat eaten by a chimpanzee community may approach one ton annually. Recently revealed aspects of predation by chimpanzees, such as its frequency and the use of meat as a political and reproductive tool, have important implications for research on the origins of human behavior. These findings come at a time when many anthropologists argue for scavenging rather than hunting as a way of life for early human ancestors. Research into the hunting ecology of wild chimpanzees may therefore shed new light on the current debate about the origins of human behavior.
One of the most important and intriguing questions in human evolution is when meat became an important part of the diet of our ancestors. Physical anthropologists and archaeologists have been using a number of techniques to try to answer this question. The presence of primitive stone tools in the fossil record tells us that 2.5 million years ago early hominids were using stone implements to cut the flesh off the bones of large animals that they had either hunted or whose carcasses they had scavenged. The pattern of obtaining and processing meat by more recent people has been studied by examining archaeological sites in Europe and elsewhere, and also by studying the hunting and meat-eating behavior of modern foraging people, the so-called hunter-gatherers. Before 2.5 million years ago, however, we know very little about the foods that hominids ate, or the role that meat may have played in their diet. We know that the earliest upright-walking (bipedal) hominids, the australopithecines, evolved in Africa about 5 million years ago, and that they shared a common ancestor with modern chimpanzees shortly before that time. Modern people and chimpanzees share an estimated 98.5% of our DNA sequence, making us more closely related to each other than either is to any other animal species. Therefore, understanding chimpanzee hunting behavior and ecology may tell us a great deal about the behavior and ecology of those very earliest hominids. This is the approach I have taken in my field study of the hunting behavior of wild chimpanzees, and especially their relationship with the animal that is their major prey, the red colobus monkey. What are the social and ecological factors that predict when chimpanzees will hunt and whether they will be successful ? What is the effect of chimpanzee predation on the populations of their prey animals, such as the red colobus ? What are the likely similarities in meat-eating patterns between chimpanzees and the earliest hominids ?
In the early 1960's, when Dr. Jane Goodall began her now famous study of the chimpanzees of Gombe National Park, Tanzania, it was thought that chimpanzees were strictly vegetarian. In fact, when Goodall first reported this behavior, many people were skeptical and claimed that meat was not a natural part of the chimpanzee diet. Today, hunting by chimpanzees at Gombe has been well documented (Teleki 1973; Goodall 1986), and hunting has also been observed at most other sites in Africa where chimpanzees have been studied, including Mahale Mountains National Park (Uehara et al. 1992) (also in Tanzania) and Tai National Park in Ivory Coast in West Africa (Boesch and Boesch 1989). At Gombe, we now know that chimpanzees may kill and eat more than 150 small and medium sized animals such as monkeys, wild pigs and small antelopes each year. Chimpanzee society is called fission-fusion, to indicate that there is little cohesive group structure apart from mothers and their infants; instead, temporary subgroupings called parties come together and separate throughout the day. These parties vary in size, in relation to the abundance and distribution of the food supply (Wrangham 1975) and the presence of estrous females (who serve as a magnet for males, Goodall 1986), so the size and membership of hunting parties vary greatly, from a single chimpanzee to as many as 35. The hunting abilities of the party members as well as the number of hunters present can thus influence when a party hunts as well as whether it will succeed in catching a colobus.
Chimpanzee Predatory Behavior
After three decades of research on the hunting behavior of chimpanzees at Gombe, we already know a great deal about their predatory patterns. We know that although chimpanzees have been recorded to eat more than 35 types of vertebrate animals (Uehara 1997), the most important vertebrate prey species in their diet is the red colobus monkey. At Gombe, red colobus account for more than 80% of the prey items eaten. But Gombe chimpanzees do not sellect the colobus they will kill randomly; infant and juvenile colobus are caught in greater proportion than their availability (Stanford et al. 1994a, 1998a); 75% of all colobus killed are immature. Chimpanzees are largely fruit eaters, and meat composes only about 3% of the time they spent eating overall, less than in nearly all human societies. Adult and adolescent males do most of the hunting, making about 90% of the kills recorded at Gombe over the past decade. Females also hunt, though more often they receive a share of meat from the male who either captured the meat or stole it from the captor. Although lone chimpanzees, both male and female, sometimes hunt by themselves, most hunts are social. In other species of hunting animals, cooperation among hunters may lead to greater success rates, thus promoting the evolution of cooperative behavior. Such cooperation has also been posited as important in our own evolution (Washburn and Lancaster 1968). In both Gombe and in the Tai forest in the Ivory Coast, there is a strong positive relationship between the number of hunters and the odds of a successful hunt (Boesch and Boesch 1989; Stanford et al. 1994b). At Tai, Christophe Boesch has documented highly cooperative hunting behavior by the chimpanzees there, and meat-sharing behavior after a kill that rewards those chimps who participated in the hunt.
One of the main recent findings about hunting by chimpanzees was its seasonality (Stanford et al. 1994a). At Gombe, nearly 40 % of the kills of colobus monkeys occur in the dry season months of August and September. This is apparently a time of food shortage in the forest, since the chimpanzees' body weights do decline (Wrangham 1975). This is actually less strongly seasonal than in the Mahale Mountains, where 60% of kills occur in a 2 month period in the early wet season. Why would chimpanzees hunt more often in some months than in others ? This is an important question, because studies of early hominid diets have shown that meat-eating occurred most often in the dry season, at the same time that meat-eating peaks among Gombe chimpanzees (Speth 1989). And the amount of meat eaten, even though it composed a small percentage of the chimpanzee diet, is substantial. I estimate that in some years, the 45 chimpanzees of the main study community at Gombe kill and consume more than 1500 pounds
of prey animals of all species. This is far more than most previous estimates of the weight of live animals eaten by chimpanzees. A large proportion of this amount is eaten in the dry season months of August and September. In fact, during the peak dry season months, the estimated per capita meat intake is about 65 grams of meat per day for each adult chimpanzee. This approaches the meat intake by the members of some human foraging societies in the lean months of the year. Chimpanzee dietary strategies may thus approximate those of human hunter-gatherers to a greater degree than we had imagined.
Several other aspects of hunting by Gombe chimpanzees are noteworthy. First, although most successful hunts result in a kill of a single colobus monkey, in some hunts from 2-7 colobus may be killed. The likelihood of such a multiple kill is tied directly to the number of hunters in the hunting party. Interestingly, the percentage of kills that are multiple kills has rose markedly in the late 1980's and early 1990's, which in turn meant that many more colobus overall were being eaten in the late 1980's compared to five years earlier (Stanford et al. 1994a, 1998a). This is most likely due to changes in the age and sex composition of the chimpanzee community. The number of adult and adolescent male chimpanzees in the study community rose from 5 to 12 over the 1980's, due to a large number of young males who were maturing and taking their places in hunting parties. One could therefore say that the fate of the Gombe red colobus monkeys is in the hands of the chimpanzee population; this is reflected in the colobus mortality rate in relation to the number of hunters available in a given era.
Throughout her years of research, Jane Goodall has noted that the Gombe chimpanzees tend to go on "hunting crazes," during which they would hunt almost daily and kill large numbers of monkeys and other prey (Goodall 1986). The explanation for such binges has always been unclear. My own research has focused on the causes for such spurts in hunting frequency, with unexpected results. The explanation for sudden changes in frequency seems to be related to whatever factors promote hunting itself; when such factors are present to a high degree or for an extended period of time, frequent hunting occurs. For example, the most intense hunting binge we have seen occurred in the dry season of 1990. From late June through early September, a period of 68 days, the chimpanzees were observed to kill 71 colobus monkeys in 47 hunts. It is important to note that this is the observed total, and the actual total of kills that includes hunts at which no human observer was present may be one-third greater. During this time the chimpanzees may have killed more than 10 % of the entire colobus population within their hunting range (Stanford et al. 1994).
To try to solve the binge question my colleagues and I examined the database of hunts recorded over the past decade to see what social or environmental factors coincided with hunting binges. Knowing that hunting was seasonal helped, in that I expected binges to occur mainly in the dry season, and this proved to be the case. But other interesting correlations leapt out as well. Periods of intense hunting tended to be times when the size of chimpanzee foraging parties was very large; this corresponded to the direct relationship between party size and both hunting frequency and success rate. Additionally, hunting binges occurred especially when there were females chimpanzees with sexual swellings (the large pink anogenital swellings that females exhibit during their periods of sexual receptivity, or estrus) travelling with the hunting party. When one or more swollen females was present, the odds of a hunt occurring were substantially greater, independent of other factors (Stanford et al. 1994b, 1998a). This co-occurrence of party size, presence of swollen females and hunting frequency led me to ask the basic question, "why do chimpanzees hunt ?"
Why Do Chimpanzees Hunt ?
Among the great apes (the gorilla,the orangutan,the bonobo, and the chimpanzee) and ourselves, only humans and chimpanzees hunt and eat meat on a frequent basis. Since neither humans or chimpanzees are truly carnivorous - most traditional human societies eat a diet made up mostly of plant foods - we are considered omnivores. The important decisions about what to eat and when to eat it should therefore be based on the nutritional costs and benefits of obtaining that food compared to the essential nutrients that the food provides. However, as I discussed above, there are social influences such as party size and composition that seem to play an important in meditating hunting behavior as well. Understanding when and why chimpanzees should choose to undertake a hunt of colobus monkeys rather than simply continue to forage for fruits and leaves, even though the hunt involves risk of injury from colobus canine teeth and a substantial risk of failure to catch anything, has been a major goal of my research (see also Stanford 1998b, 1999).
In his study of Gombe chimpanzee predatory behavior in the 1960's, Geza Teleki considered hunting to have a strong social basis (Teleki 1973). Some early researchers had said that hunting by chimpanzees might be a form of social display, in which a male chimp tries to show his prowess to other members of the community (Kortlandt 1972). In the 1970's, Richard Wrangham conducted the first systematic study of chimpanzee behavioral ecology at Gombe and concluded that predation by chimps was nutritionally based, but that some aspects of the behavior were not well explained by nutritional needs alone (Wrangham 1975). Toshisada Nishida and his colleagues in the Mahale Mountains chimpanzee research project reported that the alpha there, Ntilogi, used captured meat as a political tool to withhold from rivals and dole out to allies (Nishida et al. 1991). And William McGrew (1992) has shown that those female Gombe chimps who receive generous shares of meat after a kill have more surviving offspring, indicating a reproductive benefit tied to meat-eating.
My own preconception was that hunting must be nutritionally based. After all, meat from monkeys and other prey would be a package of protein, fat and calories hard to equal from any plant food. I therefore examined the relationship between the odds of success and the amount of meat available with different numbers of hunters in relation to each hunter's expected payoff in meat obtained. That is, when is the time, energy and risk (the costs) involved in hunting worth the potential benefits, and therefore when should a chimp decide to join or not join a hunting party? And how does it compare to the costs and benefits of foraging for plant foods ? These analyses are still underway because of the difficulty in learning the nutritional components of the many plant foods in the chimps' diverse diet , but the preliminary results have been surprising. I expected that as the number of hunters increased, the amount of meat available for each hunter would also increase. This would have explained the social nature of hunting by Gombe chimpanzees. If the amount of meat available per hunter declined with increasing hunting party size (because each hunter got smaller portions as party size increased), then it would be a better investment of time and energy to hunt alone rather than join a party. The hunting success rates of lone hunters is only about 30 %, while that of parties with 10 or more hunters is nearly 100 %. As it turned out, there was no relationship, either positive or negative, between the number of hunters and the amount of meat available per capita. This may be because even though the likelihood of success increases with more hunters in the party, the most frequently caught prey animal is a one kilogram baby colobus monkey. Whether shared among 4 hunters or 14, such a small package of meat does not provide anyone with much food.
Effects of Chimp Predation on the Colobus Population
I estimate that from 1990 through 1995, the colobus kills made by the male chimpanzee Frodo alone have eliminated about 10% of the colobus monkeys in the home range of the Gombe chimpanzees. One chimpanzee hunting party can decimate a group of red colobus in a matter of minutes. What is the likely long term effect of intensive chimp predation on the colobus population ? Using information on the size and age and sex composition of red colobus group, combined with knowledge of the hunting patterns of Gombe chimps, it is possible to estimate the impact of predation on the colobus. Based on my monitoring of five colobus groups over the past four years, plus censusing of a number of other groups that occupy the 18 square kilometers of the chimpanzees' hunting range, I estimate there are about 500 (plus or minus 10 %) in the chimpanzees' range. I estimate that from approximately 75 to 175 colobus are killed by chimpanzees annually; I base this estimate on those kills that have been observed, plus the expected number of kills per day in which no human observer was following them in the forest. The annual mortality rate in the colobus population that is due to chimpanzee predation is thus between 15 and 35 % , depending on the frequency of hunting that year (Stanford et al. 1994a, 1998a). While 15 % mortality due to predation has been recorded for other species of mammals, it must be remembered that this figure represents predation by chimpanzees only, and does not include death at the hands of other predators (leopards and eagles occur at Gombe and are known predators of monkeys) or mortality due to disease, infanticide or other factors. A 35 % mortality rate would mean, if it happened every year, that the red colobus population would almost certainly be in sharp decline. The explanation here is that the average annual mortality due to chimp predation, taken over the past decade, is about 20 % of the colobus population killed by chimpanzees each year (Stanford et al. 1994a).
To understand the impact of this mortality on the colobus population, it is important to consider certain aspects of the monkey population. First, female colobus appear to give birth about every 2 years, and births occur in every month of the year. Since chimpanzees prey mainly upon young colobus (under 2 years old), female colobus that lose a baby to chimpanzee hunters are able to begin cycling again soon afterward, and to produce a new offspring as soon as 7 months later. These two facts, lack of breeding seasonality and mortality of immatures rather than adults, may well minimize the impact of predation on the colobus, in that a single infant lost is more quickly replaced than an older offspring or adult would be.
To learn whether chimpanzee predation has the potential to be a limiting factor in the size of the colobus population at Gombe, I compared the intensity of hunting by chimpanzees with the size of red colobus groups in each of the valleys of the chimpanzeesÕ hunting range. The central valley of the chimpanzees' range (their so-called core area) is Kakombe Valley; the chimps made about one-third of all their hunts there over the past decade. As one travels away from the center and toward the northern and southern borders of the chimpanzees' range, their use of the more peripheral valleys is much less frequent, and their frequency of hunting there is less also. Only about 3 % of all hunts took place at the northern and southern edges of their range. I found that the size of red colobus groups also varied over the area of the chimps' hunting range. In the core area, red colobus groups averaged only 19 animals, little more than half the average of about 34 at the outer boundaries (Stanford in press). In other words, colobus groups are small where they are hunted frequently, and larger where hunting is infrequent. Moreover, I found that this size difference was due largely to a difference between core area and peripheral groups in the percentage of the groups that was immature colobus. In the core area, only 17 % of each group was infants and juveniles, while fully 40 % of peripheral groups were immature. This is a direct demonstration of the power of predation to limit both group size and population size in a wild primate population. From now on, we must consider the possibility that in addition to their other interesting traits, chimpanzees may be among the most important predators on certain prey species in the African ecosystems where they live.
Chimpanzee Hunting Behavior and Early Hominid Evolution
Did early hominids hunt and eat meat in a pattern similar to one described above for wild chimpanzees ? It is quite probable that they did. Recent discoveries in Ethiopia by Tim White, Gen Suwa and Berhane Asfaw of the fossil remains of very early autralopithecines (Australopithecus ramidus) show that 4.4 million years ago primitive hominids lived in a forest environment that they shared with colobus monkeys and small antelope. A. ramidus was different from chimpanzees in two prominent anatomical features: they had much smaller canine teeth, and a lower body adapted for walking on the ground rather than swinging though trees. They almost certainly continued to use trees, however, for nighttime shelter and for daytime fruit gathering, as do modern ground-living primates such as baboons. In spite of lacking the large canine teeth and tree-climbing adaptations that chimpanzees possess, early hominids probably ate a large number of small and medium sized animals, including monkeys. Large canine teeth are not necessarily important for carnivory; chimpanzees do not use their canine teeth to capture adult colobus; rather, they grab the prey and flail it to death on the ground or against a tree limb. The chimpanzees' superb climbing ability is not essential for hunting monkeys either; once the prey is cornered in an isolated tree crown, group cooperation at driving the monkeys from one hunter to another would have been a quite efficient killing technique.
In addition to the availability of prey in the trees, there were of course both large and small animals to find or capture on the ground. Many researchers now believe that the carcasses of large mammals were an important source of meat for early hominids once they had stone tools to use for removing the flesh from the carcass (Bunn and Kroll 1986). But the evidence for stone tool use dates to only 2.5 million years ago. For the 3 or so million years of human evolution prior to that time, did our ancestors eat meat ? Many researchers feel sure that they did, though the amount and the frequency of meat-eating are open to conjecture. Blumenschine (1987), for example, showed that a scavenging niche was probably available to early hominids during the Pliocene period, and Marean (1989) reasoned that the presence of saber-toothed cats meant there was a ready supply of large ungulate carcasses from which flesh could be gleaned. Speth (1989), while showing that meat-eating in early hominids was probably seasonal, also acknowledged that evidence of stone tool use in the fossil record may indicate only irregular or infrequent use of meat during periods of drought or food scarcity. While scavenging is a frequently posited mode of getting meat for our ancestors, wild chimpanzees (particularly the males who do most of the hunting) show little interest in dead animals as a food source, so scavenging may have evolved as an important mode of getting food as hominids began to make and use tools for getting at meat. Before this time, it seems likely that earlier hominids were hunting mammals as chimpanzees do today, and the role that hunting played in the early hominids' social lives was probably as complex and politically charged as it is in chimpanzees. These early homininds may have been important predators in Pliocene forest ecosystems. When we ask the question "when did meat become an important part of the human diet ?," we must therefore look well before the evolutionary split between apes and humans in our own family tree.