Animals Used for Food in the Past

Elizabeth S Wing. Cambridge World History of Food. Editor: Kenneth F Kiple & Kriemhild Conee Ornelas. Volume 1. Cambridge, UK: Cambridge University Press, 2000.


Animal remains excavated from archaeological sites are, to a large extent, the remnants of animals that were used for food. These remains include the fragmentary bones and teeth of vertebrates, the shells of mollusks, the tests of echinoderms, and the exoskeletal chitin of crustacea and insects. As with all archaeological remains, they represent discarded fragments of a previous way of life.

Organic remains are particularly subject to losses from the archaeological record, through the destructive nature of food preparation and consumption, the scavenging of refuse by other animals, and the deterioration that results from mechanical and chemical forces over time. Other losses come through excavation with inappropriate sieving strategies in which the remains of smaller individuals or species are lost. Nonetheless, despite all of these opportunities for the loss and destruction of organic material, animal remains constitute a major class of the archaeological remains from most sites and in some sites, such as shell mounds, they are the most obvious of the remains. However, even among those remains that are preserved, care must be taken in evaluating the extent to which they may represent a contribution to the prehistoric diet.

One reason for such caution is that all remains recovered are not necessarily those of animals that were consumed. For example, along the Gulf coast of Mexico, dogs were definitely eaten and probably even raised for food (Wing 1978).Their remains were often burned, disarticulated, and associated with those of other food remains. But in the West Indies, complete or nearly complete skeletons of dogs are found in burials and they are rarely associated with midden refuse, suggesting that dogs were not a regular item in the diet (Wing 1991). Dogs probably have played more roles in human culture than any other animal, ranging from food animal to guardian to hunting companion to faithful friend. But other animals, too, such as chickens, cattle, and horses, have likewise played different roles and thus their archaeological remains cannot automatically be assumed to constitute only the remnants of past meals.

Another problem is that some animals that were consumed left few remains. On the one hand, these were small, soft-bodied animals such as insect larvae, and on the other hand, they were very large animals, such as sea mammals or large land mammals that were too heavy to be brought back to the habitation site. In the case of small animals, mandibles of shrimp have recently been identified in some southeastern sites in the United States through the use of fine-gauge sieves (Quitmyer 1985). This find (which was predicted) gives encouragement that other hard parts of otherwise soft-bodied animals can be found if they are carefully searched for.At the other end of the size scale, many very large animals were butchered at the kill site and only the meat was brought back to the home site, leaving little or no skeletal evidence at the latter site of this hunting enterprise.

Such a phenomenon has been termed the “schlepp effect” (Perkins and Daly 1968), which expresses the commonsense but nonetheless laborious practice of stripping the flesh off the carcass of a large animal and carrying it to the home site but leaving most of the heavy supporting tissue, the skeleton, at the kill site. Kill sites of single large prey species such as mammoths (Mammuthus) (Agenbroad 1984) and mass kills of bison (Bison bison) (Wheat 1972) have been excavated and the strategy of the kill and processing of the carcasses reconstructed. Good evidence points to the selection of prey based on its fatness and carcass use to maximize caloric intakes of the hunters (Speth 1983).

Human Adaptability

Human technology applied to food procurement and preparation is one of the factors responsible for the broad diet that sets humans apart from other animals and has made their worldwide distribution possible. Prehistoric sites with long sequences of occupation are located on every continent except Antarctica, and almost every island (Woodhouse and Woodhouse 1975). Such a wide distribution encompasses a great range of ecosystems with different potentials and limitations for human subsistence exploitation.

By contrasting the animal resources used in sites located in some of the major landforms, both the similarities and differences in the potentials of these ecosystems can be demonstrated. Some of the differences to be examined are between subsistence exploitation in sites located along rivers and on the marine coast. Other comparisons have to do with sites on continents as compared with those on islands, and subsistence in exploiting the arctic as opposed to the humid tropics.

Levels of Extraction

Different levels of extraction of animal foods from the environment obviously can affect the composition of the diet. The hunting, fishing, and gathering of resources are procurement enterprises that result in diverse food composition and variation throughout the year as different resources become available. During the great majority of human history, beginning several million years ago and extending to the time animals were controlled by domestication, the subsistence economy was based upon hunting, fishing, and gathering. Increased control over animals, whether through the maintenance of captive animals or managed hunting, culminated in animal domestication about 10,000 years ago. A comparison of economies dependent on domestic animals with those relying on wild animals for subsistence reveals a range in diversity and dependability of resources.

The Nature of Animal Remains

Nature of Material

As already noted, the remains of animals used for food consist of the supporting tissue such as bone and teeth of vertebrates, shell of mollusks, and chitin of crustaceans. These tissues are composed of inorganic and organic compounds. The relatively rigid structure of the inorganic portions predominate in bone, constituting 65 percent. In tooth enamel the inorganic portion is 99.5 percent by weight. The inorganic portions of these supporting tissues are composed of compounds of calcium.

By their very nature, skeletal remains are attractive to scavengers: Some meat and other soft tissue may have adhered to them and, in addition, bone itself is sought by many animals as a source of calcium.

Other losses of archaeological remains come about through the influence of forces of nature, termed taphonomic factors. Such natural changes include all types of site disturbances such as erosion or stream washing, land movement, alternating freezing and thawing, and acidic soil conditions. The soil conditions are particularly critical for the preservation of bone, which as a calcium compound can be dissolved in acidic conditions. Destruction of bone under acidic conditions is further complicated by the problem of bone loss being greatest in the least well calcified bones of young individuals. In contrast, losses are the smallest in the enamel portion of teeth (Gordon and Buikstra 1981). Alternating freezing and thawing is the other taphonomic factor that is particularly damaging to organic remains. Bones exposed to the sun develop cracks. When these cracks fill with moisture and then freeze, they enlarge and, ultimately, the bone will fragment into pieces that have lost their diagnostic characteristics.

If losses from the archaeological record can complicate reconstruction of the past, so too can additions to the faunal assemblage. Such additions were often animals that lived and died at the habitation site. These are known as commensal animals, the most well known being the black and the Norway rats (Rattus rattus and Rattus norvegicus) and the house mouse (Mus musculus). Burrowing animals such as moles (Talpidae) and pocket gophers (Geomyidae) may also dig into a site and become entombed. In addition, middens and habitation sites were occasionally used for burial, and the remains of both people and their animals (i.e., dogs) were thus inserted into earlier deposits.

Other ways in which commensal animals can be incorporated in a midden is by association with the target species. For example, many small creatures, such as mussels, snails, crabs, and barnacles, adhere to clumps of oysters and were often brought to the site on this target species. Occasionally, too, the stomach contents of a target species may contain remains of other animals, which are thus incorporated in the site.

Recovery and Identification

Optimum recovery of faunal material is clearly essential for reconstruction of past diets. In some cases, sieving the archaeological remains with 7-millimeter (mm)-gauge screens will be sufficient to recover the animal remains. But as more and more sieving experiments are conducted and archaeologists gain more and more experience using fine-gauge (2 mm and 4 mm) sieves, it is becoming increasingly obvious that fine-gauge sieves are essential for optimal recovery at most sites. A good example of this importance has to do with what was long thought to be the enigma of the preceramic monumental site of El Paraiso on the Pacific coast of Peru. This site was an enigma because early excavations there uncovered no vertebrate remains, suggesting that the local ancient residents had no animal protein in their diets. It was only when sieving with 2-millimeter-gauge sieves was undertaken that thousands of anchovy (Engraulidae) vertebrae were revealed. At last, the part this small schooling fish played in providing fuel for the construction of the monumental architecture was understood (Quilter et al. 1991).

Another methodological problem that needs further consideration at this site and at others is the relative contribution of vertebrates and invertebrates to the prehistoric diet. Mollusk shells are relatively more durable and less likely to be lost to scavengers than are the bones of vertebrates of similar size, which results in a bias favoring mollusks.

In evaluating the potential contribution of vertebrates and mollusks to the prehistoric diet, one must also keep in mind biases in their preservation and recovery. For example, molluskan shells are more massive relative to their edible soft tissues than are the skeletons of vertebrates to their soft tissues. Consequently, shells will be the most visible component of a shell midden even before taking into account the greater durability of shell and the fewer losses to scavengers.

One way some of these problems can be addressed is by making estimations of potential meat weight using allometric relationships between measurable dimensions of the shell or bone and meat weight. Such relationships, of course, can only be developed by using the accurate weights and measurements of modern specimens comparable to the species recovered from archaeological deposits. Knowledge of modern animals and a reference collection are the most important research tools of the faunal analyst.

First, an attempt is made to estimate the minimum weight of the meat represented by the animal remains. Next, this estimate is contrasted with the maximum estimate of the meat weight that could have been provided by the minimum number of individual animals calculated to be represented. If the two estimates are approximately the same, this implies that all of the meat from each animal was consumed at the source of the refuse.Yet if the two estimates differ substantially, it suggests that only portions of some animals were consumed at the site, with other portions distributed within the community.

The Meaning of Animal Remains Assemblages

Animal Exploitation in Different Ecosystems

Riverine versus coastal. Many similarities exist between faunal assemblages from riverine sites and those located along the coast. In each case, both vertebrate and invertebrate aquatic animals were important food sources. Furthermore, these two aquatic situations were subject to influxes of animals, either in breeding congregations or migrations, that augmented the resident fauna. Of course, aquatic species used by riverine fishermen and gatherers were different from those extracted from the sea.

Aquatic organisms, fishes and mollusks, are important in the faunal assemblages of both riverine and coastal sites. Shell mounds are more typically associated with coastal sites, although they do occur along rivers (Weselkov 1987). Riverine shell mounds are accumulations typical of the Archaic time period (about 7000 to 3000 B.P.) in eastern North America, exemplified by those found on the St. Johns River in Florida and the Green River in Kentucky (Claasen 1986). Along coastal shores, shell mounds date from Archaic times to the present.

Gathering mollusks is invariably accompanied by fishing, but the relative contribution of fish and shell-fish to the diet follows no pattern. Many shell mounds are visually very impressive and, in fact, are so large that in many places these mounds have been mined for shell material to be used in modern road construction. Less visible components of these archaeological sites are the vertebrate, crustacean, and material cultural remains.

As indicated in the discussion of methods, those for reconstructing the dietary contributions of the two major components of a shell mound are still being perfected. One exciting development has been a greater understanding of the importance (as food) of very small animals in the vertebrate, predominantly fish, components of these mounds.

As already mentioned, the use of fine-gauge screen sieves in the recovery of faunal remains has provided a more accurate understanding of the size range of fishes consumed in the past. Catches of small fishes are being documented in many parts of the world. For example, at four sites in the Darling Region of Australia, otoliths of golden perch (Maguaria ambigua) were preserved. These could be used to extrapolate lengths of fishes caught up to 24,000 years ago (Balme 1983). The range in their estimated lengths is 8 to 50 centimeters (cm) and the mean is approximately 20 cm. The range in estimated lengths of sardines (Sardina pilchardus) from a fourth-century A.D. Roman amphora is between 11 and 18 cm (Wheeler and Locker 1985). Catfish (Ariopsis felis) and pinfish (Lagodon rhomboides) from an Archaic site on the southeastern Gulf coast of Florida are estimated to range in length from 4 cm to 25 cm, and the means of the catfish and pinfish are 10 cm and 12 cm, respectively (Russo 1991).

An important question has to do with how these small fishes could be eaten and yet leave skeletal remains behind. Many contemporary diets include small fishes such as sardines and anchovies in which the entire body of the fish is consumed. The answer may lie in the fact that small fishes are used in many parts of the world in the preparation of fish sauces that are made without skeletal tissue. In other words, the well-preserved, intact skeletal remains of small fishes suggest that the fishes might have been employed in sauces.

In addition to mollusks and fishes in the protein portion of a coastal and riverine diet, migratory or breeding congregations would have added significantly to the diet. A well-known example of this phenomenon is the seasonal migration of anadromous fishes such as salmon (Salmonidae) and herring (Clupeidae) (Rostlund 1952); methods of preserving and storing this seasonal surplus would also have developed.

Other examples of such exploitation include the breeding colonies of sea bird rookeries, sea turtle nesting beaches, and seal and sea lion colonies. Many of these colonies are coastal phenomena and are strictly confined to particular localities. During the breeding cycle, most animals are particularly vulnerable to predation, and people through the ages have taken advantage of this state to capture breeding adults, newborn young, and, in the cases of birds and turtles, eggs.

Unfortunately, only some of the evidence for this exploitation can be demonstrated in the archaeological remains. Egg shells are rarely preserved. Some of the breeding animals in question, like the sea mammals and the sea turtles, are very large and may have been butchered on the beach and the meat distributed throughout the community. Thus, it is difficult to assess the relative importance of these resources within a particular refuse deposit and by extension within the diet of the humans being studied.

Continental versus island. A continental fauna differs from an island fauna in its diversity of species. There is a direct relationship between island size and distance from the mainland and species diversity (MacArthur and Wilson 1967). Human exploitation on a continent can range from catches of very diverse species in regions where different habitats are closely packed, to dependence on one or two species that form herds, typically in open grasslands. On islands other than very large ones, prehistoric colonists found few species and often augmented what they did find with the introduction of domestic species as well as tame captive animals.

This kind of expansion seems to be a pattern in many parts of the world. For example, several marsupials (Phalanger orientalis, Spilocuscus maculatus, and Thylogale brunii), in addition to pigs (Sus scrofa) and dogs (Canis familiaris), were deliberately introduced into the Melanesian Islands between 10,000 and 20,000 years ago (Flannery and White 1991). Similarly, sheep (Ovis sp.), goats (Capra sp.), pigs (Sus sp.), and cats (Felis sp.) were all introduced into the Mediterranean Islands at a time when the domestication of animals was still in its initial stages.A variety of wild animals, such as hares (Lepus europaeus), dormice (Glis glis), foxes (Vulpes vulpes), and badgers (Meles meles), were also introduced into this area (Groves 1989).

Likewise, in the Caribbean Islands, domestic dogs as well as captive agouti (Dasyprocta leporina), opossum (Didelphis marsupialis), and armadillo (Dasypus novemcinctus) were introduced from the South American mainland, whereas the endemic hystricognath rodent locally called “hutia” (Isolobodon portoricensis) and an endemic insectivore (Nesophontes edithae) were introduced from large to small islands (Wing 1989).

Although tame animals were doubtless kept by people living on the mainland, they are not easily distinguished from their wild counterparts. But this problem is only part of an increasingly complex picture as human modifications of the environment, either overtly through landscape changes resulting from land clearing or more subtly through hunting pressure, have altered the available species on both islands and continental land masses.

Because of the generally lower species diversity on islands, exploitation of terrestrial species was augmented by marine resources. These were primarily fishes and mollusks. In the Caribbean Islands, West Indian top shell (Cittarium pica) and conch (Strombus gigas), and a whole array of reef fishes including parrotfishes (Scaridae), surgeonfishes (Acanthuridae), grouper (Serranidae), and jacks (Carangidae), were of particular importance.

Arctic versus humid tropics. The Arctic has long, cold, dark winters but also short summers, with long spans of daylight that stimulate a brief period of extraordinarily high plant productivity. By contrast, the humid tropics have substantially more even temperatures and lengths of daylight throughout a year that in many cases is punctuated by dry and rainy seasons. Needless to say, these very different environmental parameters have a pronounced effect on the animal populations available for human subsistence within them.

Traditional contemporary subsistence activities as well as evidence from archaeological faunal remains in the Alaskan Arctic indicate that important contributors to the human diet have been caribou (Rangifer tarandus), sea mammals—particularly seals (Callorhinus ursinus and Phoca vitulina) and sea lions (Eumatopius jubata)—seabirds, and marine fishes, primarily cod (Gadus macrocephalus) (Denniston 1972; Binford 1978;Yesner 1988).

Although the species of animals differ in different parts of the Arctic regions, many characteristics of a northern subsistence pertain. Foremost among these is a marked seasonal aspect to animal exploitation correlated with the migratory and breeding patterns of the Arctic fauna. Moreover, the length of time during which important animal resources, such as the anadromous fishes, are available becomes increasingly circumscribed with increased latitude (Schalk 1977). To take full advantage of this glut of perishable food, people need some means of storage. And fortunately, in a region where temperatures are below freezing for much of the year, nature provides much of the means.

Another characteristic of northern subsistence is the generally low species diversity; but this condition is counteracted by some large aggregations of individuals within the species. Still, the result is that heavy dependence is placed on a few species. At Ashishik Point an analysis of the food contribution of different animals to the prehistoric diet revealed that sea mammals and fishes predominated (Denniston 1972). Of the sea mammals, sea lions are estimated consistently to have provided the greatest number of calories and edible meat. This estimation agrees with the observation by David Yesner (1988) that dietary preference among the prehistoric Aleut hunter-gatherers was for the larger, fattier species.

The fauna of the humid tropics is much more diverse than that of the Arctic, although the tropical animals that people have used for food do not generally form the large aggregations seen in higher latitudes. Exceptions include schools of fishes, some mollusks, bird rookeries, and sea turtle breeding beaches. Many of the animals that have been exploited are relatively small. The largest animals from archaeological sites in the New World tropics are adult sea turtles (Cheloniidae), deer (Mazama americana and Odocoileus virginianus), and peccary (Tayassu pecari and Tayassu tajacu) (Linares and Ranere 1980). Some of the South American hystricognath rodents, such as the capybara (Hydrochaeris sp.), paca (Agouti paca), and agouti (Dasyprocta punctata), were used and continue to be used widely as food. Fish and shellfish were also very important components of prehistoric diets (Linares and Ranere 1980) and typically augmented those based on terrestrial plants and animals.

The kinds of land mammals that have been most frequently consumed in much of the tropics since the beginning of sedentary agriculture prompted Olga Linares (1976) to hypothesize a strategy for their capture she called “garden hunting.” She suggested that many of these animals were attracted to the food available in cultivated fields and gardens and were killed by farmers protecting their crops. Objective ways of evaluating the importance of garden hunting have been proposed, which are based on the composition of the faunal assemblage (Neusius 1996).

Different Levels of Extraction

Hunting, fishing, and gathering. It has been estimated that fully 90 percent of all those who have lived on earth have done so as hunter-gatherers (Davis 1987). The animals that were procured by these individuals varied depending upon the resources available within easy access of the home site or the migratory route. But in addition to these differences in the species that were obtained and consumed, certain constraints governed what was caught, by whom, how the meat was distributed, and how it was prepared. Certainly, technology played an important part in what kinds of animals were caught on a regular basis and how the meat was prepared. Many specialized tools such as water craft, nets, traps, weirs, bows, arrows, and blowguns and darts all permitted the capture of diverse prey.

A limitation imposed on all organisms is that of energy, which meant that food procurement generally took place within what has become known as the “catchment” area (Higgs and Vita-Finzi 1972). This area is considered to be inside the boundaries that would mark a two-hour or so one-way trip to the food source. Theoretically, travel more distant than this would have required more energy than could be procured. Once animal power was harnessed, trips to a food source could be extended. However, another solution to the problem of procurement of distant resources was by adopting a mobile way of life. This might have taken the form of periodic trips from the home site or it might have been a migratory as opposed to a sedentary method of living.

In the past, as today, there was doubtless a division of labor in the food quest. It is generally thought that men did the hunting and fishing (although in many traditional societies today women do the inland, freshwater fishing) whereas the women, children, and older male members of the community did the gathering. Some excellent studies of contemporary hunter-gatherers provide models for these notions about the division of labor. One of the best and most frequently cited is by Betty Meehan (1982) entitled Shell Bed to Shell Midden, in which she describes shellfishing practices in detail and the relative importance of shellfish to the diet throughout the year. This is the case at least among the Gidjingali-speaking people of Arnham Land in northern Australia, whose shell-fish gathering is a planned enterprise entailing a division of labor for collecting molluscan species.

Food sharing is another phenomenon that probably has great antiquity in the food quest, although admittedly, we know more about the patterns of food sharing from contemporary studies than from archaeological remains. A classic study of sea turtle fishing along the Caribbean coast of Nicaragua (Nietschmann 1973) describes the distribution obligations of meat obtained from the large sea turtle (Chelonia midas). Such patterns make certain that meat does not go to waste in the absence of refrigeration and furthermore assure the maintenance of community members who are less able or unable to procure food for themselves.

That a large carcass was shared can sometimes be detected in an archaeological assemblage of animal remains. As we noted earlier, this observation can occur when estimates of meat yield based on the actual remains recovered are compared with estimates of potential meat obtained from the calculated numbers of individual animals represented in the same sample. Disparity between these two estimates could point to the incomplete deposit of the remains of a carcass, which may indicate sharing, either among the households, or through a distribution network, or even through a market system. Plots of the dispersal of parts of deer carcasses throughout a prehistoric community that was entirely excavated provide a demonstration of food distribution (Zeder and Arter 1996). Perhaps sharing also occurred when many small fishes were caught in a cooperative effort. In this case the catch may have been shared equally with an additional share going to the owner of the fishing equipment.

Communal cooperation was probably also involved in the concept of “garden hunting” (Linares 1976), which can be viewed in the broader perspective of deliberately attracting animals. Food growing in gardens or fields, or stored in granaries, was used as bait to entice animals to come close to a settlement.

Clearly this strategy would have been most successful with animals that ate agricultural products. Some of those so trapped were probably tamed and eventually domesticated, suggesting that it is no accident that many of our domestic and tame animals consume crop plants. Today agriculture and animal husbandry are combined to produce the plant and animal foods consumed throughout most of the world. But the cultivation of crops and the husbandry of animals did not arise simultaneously everywhere. In much of the Western Hemisphere crops were grown in the absence of a domestic animal other than the dog. However, the management, control, and domestication of animals eventually led to a different level of exploitation.

Animal domestication. Domestic animals have skeletal elements and teeth that are morphologically distinct from their wild ancestors. The observable changes are the result of human selection, which is why domestic animals are sometimes referred to as man-made animals. Human selection prior to modern animal husbandry may have been unintentional and more a result of isolation and methods of confinement (e.g., whether animals were tethered, or kept in stalls or corrals).

Animals were, of course, tamed first, and paintings on walls and on pottery sometimes provide evidence of domestication. Selection expressed as changes in the morphology of an animal also indicates domestication, but the state of animal “tameness” is difficult to recognize in the fragmentary remains from archaeological sites and consequently rarely possible to document. Moreover, many animals were held captive and tamed but for some reason never became domesticated, meaning that their skeletal remains do not differ morphologically from their wild counterpart. Collared peccaries (Tayassu tajacu), for example, were believed to have been tamed and kept by the Maya for food or ritual purposes (Hamblin 1984), and stone pens found on the island of Cozumel are thought to have been used to keep peccaries for human convenience. At the present time, peccaries are trained by some hunters in Central America to fill the role of watchdogs. Clearly, many motives instigated the taming of animals, but the most important of these undoubtedly was ready access to meat.

Some animals were held in captivity with no effort made to tame them. The captivity of these animals was a means of storing fresh meat. An example of animals still kept this way today are sea turtles maintained in corrals along the shore below low tide. Live animals were also maintained on board ships during long ocean voyages. These practices probably were very old, and animals such as domestic pigs were often released on islands to assure a source of food for the next voyage or for other people who followed.

But control over animals, or even their domestication, did not necessarily mean a sole reliance on them for food. Rather, especially in the early stages of domestication, humans continued to depend on hunted and fished resources. But with the introduction of livestock into new regions, traditional subsistence strategies were modified. An interesting example may be seen in sixteenth-century Spanish Florida, where Spanish colonists, suddenly confronted with wilderness and relative isolation, changed their traditional subsistence in a number of ways. They abandoned many accustomed food resources and substituted wild resources used by the aboriginal people of the region. Moreover, their animal husbandry practices shifted from a traditional set of domesticated animals to just those that flourished in the new environment (Reitz and Scarry 1985).These changes required flexibility. Yet presumably, such changes in subsistence behavior documented for Spanish settlers in Florida were repeated in many places throughout the ages (Reitz and Scarry 1985).

Dependence upon livestock became more complete only after peoples and their agricultural systems were well established. It should be noted, however, that even in industrial societies today, reliance upon domestic animals is not complete. Most contemporary Western diets include fish and other seafood, and wild animal food, such as venison, is viewed as a delicacy and is often the food for a feast.

Accompanying a greater human dependence upon domestic animals for food was an increased human use of the animals’ energy and other products. Animals, used for draft, greatly increased the efficiency of agricultural enterprises, and utilizing them for transportation extended the catchment area.

The employment of animals for dairy products was of crucial importance and can be detected in archaeological remains by the kill-off pattern that characteristically shows male individuals killed as young animals and females maintained to old age for their milk (Payne 1973). When such provisions as milk (and eggs) came into use, they provided an edible resource without loss of the animal, and the latter came to be viewed as capital.


The human diet is characterized by the great variety of plants and animals that it includes. The animal protein portion of the diet depends most heavily on vertebrates and mollusks but also includes crustaceans, arthropods, and echinoderms. Historically, this flexibility has been instrumental in the worldwide distribution of human beings.

Despite this flexibility, selection for certain resources was clearly practiced. Usually, certain species were targeted even though a great variety of species were used. When selection was exercised, the determining factors seemed to have been those of a dependable resource and one high in body fat. By resource dependability we mean, for example, the annual salmon run, stable oyster beds, perhaps a captive flock of pigeons or a domestic herd of goats. Selection for animals with the highest body fat has been documented specifically in the preference for sea lions by the prehistoric Aleuts and generally by archaeological remains of the bison, which revealed selection of the fattest individuals. In this connection it should be noted that domestic animals tend to store more fatty tissue than do their wild ancestors, which may have been a further incentive for the maintenance of domesticates (Armitage 1986).

Food distribution and sharing is another characteristic of human subsistence that provided everyone with a degree of security even if personal catches were not successful. Methods of food preparation and storage doubtless varied greatly. Most likely these were salting, smoking, or drying, or a combination, but none of these methods is clearly visible in archaeological remains. It is only when animal remains are found far outside of their normal range (e.g., codfish remains in West Indian historic sites) that one can be sure some method of preservation was employed.

Similarly, cooking methods can be interpreted from archaeological remains only in a general way. It is likely that meat was boiled or stewed when bone was not burned and roasted when it was burned. Yet such interpretations must be made cautiously because bone can become burned in many ways.

Even though animal remains from archaeological sites are fragmentary and much is often missing from the whole picture, they do provide an important perspective on past human diets.