!

hat Did The" Eat# Su$sistence and Diet

A meal as a funerary offering: the rich and elaborate food remains found in the 2nddynasty tomb of a noblewoman at Saqqara, Egypt. These include a triangular loaf of bread made from emmer wheat!" nab# berries similar in appearance to cherries!" cut ribs of beef" a coo#ed quail" two coo#ed #idneys" stewed fruit, probably figs" ca#es" pigeon stew" a coo#ed fish. $mpressi%e as such remains are, they are unli#ely to be representati%e of the e%eryday diet of the ancient Egyptians.

from funerary contexts are unlikely to be representa- from a single day. The only way in which we can really tive of everyday diet. Even the meals found so wonder- study what people ate habitually is to examine actual fully preserved at Pompeii are merely a tiny sample human remains.
ASSESSING REMAINS DIET FROM HUMAN

reveal about general health will be examined in !hapter "".

The only incontrovertible evidence that something was consumed by humans is its presence in either stomachs or feces. Both kinds of evidence give us invaluable information about individual meals and short-term diet. The study of human teeth also helps us reconstruct diet, but the real breakthrough in recent years in understanding longterm diet has come from the analysis of bone collagen. hat human bones

#tomachs survive only rarely in archaeological contexts, except in bog bodies. $t is sometimes possible to retrieve food residues from the alimentary tract of decomposed bodies - the anthropologist %on Brothwell achieved this, for example, by removing the grave earth from the lower abdominal area of some British %ark &ge skeletons, and extracting the organic remains by means of
Stomach Contents.

49

flotation' and colon contents have also been obtained from an &nasa(i burial of the ")th century &%. #ome mummies also provide dietary evidence* the overweight wife of the +ar,uis of %ai from -nd-century B! !hina, mentioned above, seems to have died of a heart attack caused by acute pain from her gallstones an hour or so after en.oying a generous helping of watermelon /")0 melon seeds were found in her stomach and intestines1. hen stomachs survive in bog bodies, the dietary evidence they provide can be of the greatest interest. Paleobotanist 2ans 2elbaek3s pioneering studies of the stomach contents of %anish $ron &ge bogmen showed

50

PART II Discove%in& the 'a%iet" o( Human E)*e%ience

that 4rauballe +an, for instance, had consumed over 56 species of wild seeds, together with one or two cereals and a little meat /as shown by some small bone splinters1, while Tollund +an had eaten only plants. But it should be borne in mind that these results, while fascinating, are not necessarily linked to annual diet, since these victims were possibly executed or sacrificed, and thus their last meal - apparently consisting of dense chaff, larger plant fragments, and weed seeds, the residues from sieving in the latter stages of crop processing - may have been out of the ordinary. #uch waste crop cleanings were often used as animal feed, as famine food, or were given to condemned criminals. 2owever, as noted in the section on plant remains, the British 7indow +an /box, pp. 880-891 had consumed a griddle cake before his death, and this rough bread, made of the primary product of crop processing, was nothing out of the ordinary for the period - certainly not a recogni(ably :ritual: dish. Fecal Mate%ial. Experiments have been done to assess the survival properties of different foodstuffs relevant to the study of ancient diet, and it has been found that many organic remains can survive surprisingly well after their .ourney through the human digestive tract, to await the intrepid analyst of coprolites, or fossili(ed feces. !oprolites themselves survive only rarely, in very dry sites such as caves in the western ;nited #tates and +exico, or very wet sites. But, where they are preserved, they have proved to be a highly important source of information about what individuals ate in the past. The first step in any study is to attempt to check that the coprolites are indeed of human origin - this can sometimes be done by analysis of fatty molecules such as coprostanol, and of steroids. <nce this has been done, what can coprolite contents tell

us about food intake= +acroremains can be extremely varied in a human coprolite, in fact this variety is an indication of human origin. Bone fragments, plant fibers, bits of charcoal, seeds, and the remains of fish, birds, and even insects are known. #hell fragments - from molluscs, eggs, and nuts - can also be identified. 2air can be assigned to certain classes of animals by means of its scale pattern, visible under the microscope, and thus help us to know which animals were eaten. Eric !allen analy(ed prehistoric coprolites from Tehuacan, +exico /the valley studied and excavated intensively by >ichard +ac?eish in the "956s1, and identified hair from gophers, whitetailed deer, cottontail rabbit, and ringtailed cats. 2e also managed to ascertain that some millet grains in the feces had been pounded, while others had been rolled on a metate /grinding stone1. +icroremains such as pollen are of less help since, as we have already noted, most of the pollen present is inhaled rather than consumed. Pollen does, however, provide data on the surrounding vegetation, and on the season when the coprolite was produced. The fecal material from the 4reenland Eskimo mummies /see pp. 888-8@1 contained pollen of mountain sorrel, which is only available in Auly and &ugust. Bungal spores, remains of the nematode worm plant parasites, algal remains, and other parasites have also been identified in coprolites. Exceptional conditions in 7ovelock !ave, ?evada, have preserved @666 coprolites dating from -@66 to "@6 years ago, and >obert 2ei(er3s study of their contents yielded remarkable evidence about diet, which seems to have comprised seeds, fish, and birds. Beather fragments were identified from waterfowl such as the heron and grebe' fish and reptile scales, which pass through the alimentary canal unaltered, also led to identification of several species. Bish remains were abundant in some of the coprolites'
51

one, for example, from "666 years ago, contained @.0 g /6.- o(1 of fish bone which, it was calculated, came from "6" small chubs, representing a total live weight of -60 g /C.) o(1 the fish component of a meal for a single person. Even where feces have not been preserved, we are now sometimes able to detect and analy(e residues of digested food by studying sewers, cesspits, and latrines. Biochemical analysis of ditch deposits near latrines at the >oman fort of Bearsden, #cotland, revealed an abundance of coprosterol, a substance typically found in human sewage, as well as a bile acid characteristic of human feces. & low amount of cholesterol showed that there was little meat in the diet. ?umerous fragments of wheat bran in the deposit probably formed part of the feces, and no doubt came from defecated bread or some other floury food. !oprolites and fecal residues represent single meals, and therefore provide short-term data on diet, unless they are found in great ,uantities, as at 7ovelock !ave, and even there the coprolites represent only a couple of meals a year. Bor human diet over whole lifetimes, we need to turn to the human skeleton itself.
Human Teeth as Evidence (o% Diet

Teeth survive in extremely good condition, made as they are of the two hardest tissues in the body. PierreBranDois Puech is one of a number of scientists to have studied teeth from many periods in an attempt to find some evidence for the sort of food that their owners en.oyed. The method involves a microscopic examination of the abrasions on certain dental surfaces.

52

53

To study food intake, one should examine the lateral or side surfaces which best escape abrasion from inorganic particles in the food, and whose wear instead ought to reflect the movement of the organic food itself within the mouth. & replica is made by pouring a thin film of resin onto the tooth-face' when this is peeled off after solidification, it forms a faithful imprint which can be examined under the light microscope or in the scanning electron microscope. Puech began by studying recent teeth from individuals with known dietary habits. The principle involved was that abrasive particles in food leave striations on the enamel whose orientation and length are directly related to the meat or vegetation in the diet and its process of cooking. +odern meat-eating 4reenland Eskimos were found to have almost exclusively vertical striations on their lateral surfaces, while largely vegetarian +elanesians had both vertical and hori(ontal striations, with a shorter average length. hen these results were compared with imprints from fossil teeth, it was found that from the late 7ower Paleolithic onward, there is an increase in hori(ontal and a decrease in vertical striations, and a decrease in average striation length. $n other words, less and less effort was needed in the mastication of food, and meat may have decreased in importance as the diet became more mixed* early people crushed and broke down their food with their teeth, but less chewing was re,uired as cooking techni,ues developed and improved. There are exceptions, such as a Homo erectus individual who seems to have been mainly vegetarian, eating thin, chewy vegetable foods, but on the whole the generali(ation seems sound. The biting /occlusal1 surfaces of human teeth are of limited help in Puech3s techni,ue, since much of the wear here is due to the method of

food preparation - meat can be exposed to windborne dust, for example, or food may have been cooked on ashes, and the result is the incorporation of extraneous abrasive particles in the food. Burthermore, our ancestors often used teeth not simply for chewing but as a third hand, for cutting, tearing, and so on. &ll these factors add striations to the biting surfaces. The lower .awbone of the Homo erectus /or :archaic: Homo sapiens1 individual from +auer, near 2eidelberg in western 4ermany, dating back some half a million years, has marks suggesting that meat was held in the front of the mouth and cut off with a flint tool that left its traces on six front teeth. ear on ?eanderthal teeth reveals that here too teeth were often used in the same way. Tooth decay as well as wear will sometimes provide us with dietary information. >emains of the !alifornia $ndians display very marked tooth decay, attributed to their habit of leaching the tannin out of acorns, their staple food, through a bed of sand which caused excessive tooth abrasion. %ecay and loss of teeth can also set in thanks to starchy and sugary foods. %ental caries became abundant on the coast of 4eorgia /;#&1 in the "-th century &%, particularly among the female population. $t was in this period that the transition occurred from hunting, fishing, and gathering to mai(e agriculture. The anthropologist !lark 7arsen believes that the rise in tooth decay over this period, revealed by a study of hundreds of skeletons, was caused by the carbohydrates in mai(e. #ince the women of the group were more sub.ect to the caries than were the men, it is probable that they were growing, harvesting, preparing, and cooking the corn, while the men ate more protein and less carbohydrate. 2owever, not all scientists accept these conclusions, pointing out that women may have suffered from more
54

caries in a period of high population growth because of greater loss of calcium with the higher number of pregnancies. Binally, as mentioned above /p. -C81, direct evidence of diet can be obtained from phytoliths extracted from the surface of human teeth.
Isoto*ic Methods+ Diet ove% a ,i(etime

!hat Did The" Eat# Su$sistence and Diet

>ecently, a revolution has taken place in dietary studies through the reali(ation that isotopic analysis of human tooth enamel and bone collagen can reveal a great deal about long-term food intake. The method relies on reading the chemical signatures left in the body by different foods - we are what we eat. Plants can be divided into three groups - two groups of land plants and one of marine plants - based on their differing ratios of the carbon isotopes ")! and "-!. !arbon occurs in the atmosphere as carbon dioxide with a constant ratio of ")!*"-! of about "*"66'") in ocean waters, the amount of ! is slightly higher. hen atmospheric carbon dioxide is incorporated into plant tissues through photosynthesis, plants use relatively more "-! than ")! and the ratio is altered. Plants that fix carbon dioxide initially into a three-carbon molecule /called !) plants1 incorporate slightly less ")! into their tissues than do those using a fourcarbon molecule /!8 plants1. By and large, trees, shrubs, and temperate grasses are !) plants' tropical and savanna grasses, including mai(e, are !8 plants. +arine plants photosynthetically fix carbon differently from most land plants, and have a higher ")!E"-! ratio. &s plants are eaten by animals, these three different ratios are passed along the food chain, and are eventually fixed in human and animal bone tissue. The ratio found in bone collagen by means of a mass

55

PART II Discove%in& the 'a%iet" o( Human E)*e%ience

spectrometer thus has a direct relation to that in the plants which constituted the main foods. The ratios can show whether diet was based on land or marine plants, and whether on !) or !8 land plants. <nly archaeological evidence, however, can provide more detail about precisely which species of plants or animals contributed to the diet. 2enrik Tauber applied this techni,ue to collagen from prehistoric skeletons in %enmark, and found a marked contrast between +esolithic people and those of the ?eolithic and Bron(e &ge. $n the +esolithic, marine resources were predominant even though fish bones were very scarce in the excavated material whereas in the later period there was a change to reliance on land foods, even in coastal sites. &t coastal sites in other parts of the world, the techni,ue has confirmed a heavy reliance on marine resources. $n prehistoric sites on the coast of British !olumbia, Brian !hisholm and his associates found that about 96 percent of protein had come from marine foods' little change was apparent over five millennia, and it was noted that adults seemed to eat more food from the sea than did children. >ecently, isotopic analysis of tooth enamel from four Australopithecus africanus individuals from +akapansgat, #outh &frica, revealed that they ate not only fruits and leaves, as had been thought, but also large ,uantities of carbon-") enriched food such as grasses or sedges, or the animals which ate those plants, or both. $n other words, they regularly exploited fairly open environments /woodlands or grasslands1 for food' and since their tooth wear lacks the characteristic scratches of grasseaters, it is possible that they were indeed already consuming meat, by hunting small animals or scavenging larger ones.

-one Colla&en Studies and the Rise o( Ne. !o%ld A&%icultu%e.

The carbon isotope bone collagen method is particularly useful for detecting changes in diet, and has revolutioni(ed the study of the rise of food production in the ?ew orld. &nna >oosevelt used the techni,ue to assess the diet of the prehistoric inhabitants of the <rinoco floodplain in Fene(uela. &nalysis of samples from a number of skeletons by her colleagues ?ikolaas van der +erwe and Aohn Fogel revealed a dramatic shift from a diet rich in !) plants such as manioc in 066 B! to one based on !8 plants such as mai(e by &% 866. &lthough the techni,ue cannot specify the actual plants consumed, the abundant mai(e kernels and grinding e,uipment found in the area3s sites from &% 866 confirm the insight provided by isotopic analysis. The techni,ue is even more crucial in ?orth &merica, where the rise of agriculture was signaled by the introduction of mai(e, a !8 food native to +eso- america, into a predominantly !) plant environment /in the ?ear East, where the first domesticated plants were themselves part of the !) plant environment, the techni,ue is of less use to studies of the origins of agriculture1. $n some cases, mai(e3s contribution to a diet can be ,uantified. $n skeletons from southern <ntario, 2enry #chwarc( and his colleagues found that the proportion of !8 plants /i.e. mai(e1 in the diet increased between &% 866 and "5@6, reaching a maximum of @6 percent by about "866. Othe% -one Colla&en Techni/ues. #ome scholars have attempted to extend the carbon isotope techni,ue to apatite, the inorganic and ma.or constituent of bone, in the hope that it could be applied even in cases where collagen has not survived /it often degrades after "6,666 years1' others, however, have found this method unreliable, so that the collagen method is the only

56

one whose validity is confirmed for the present. ?evertheless there are collagen techni,ues available involving isotopes of elements other than carbon. >atios of nitrogen isotopes in collagen, for example, can reflect dietary preferences in the same way as carbon. The "@? isotope increases as it passes up the food chain from plants to animals* a low ratio of "@? to "8? points to an agricultural subsistence, while a high ratio points to a marine diet. <ne anomaly here is caused by coral reef resources such as shellfish, which, because of the way nitrogen is fixed by plants in reefs, give a low nitrogen value. Thus, in cases where a seafood diet seems likely, the carbon isotope method needs to be employed for confirmation. The two methods have also been applied together to historic and prehistoric material in East and #outh &frica by #tanley &mbrose and +ichael %e?iro. They found it possible to distinguish marine foragers from people using land resources, pastoralists from farmers, camel pastoralists from goatEcattle pastoralists, and even grain farmers from non-grain farmers. 4roups that depended on the meat, blood, and milk of "@domestic animals had the highest ? values, those dependent mainly on plant foods had the lowest. The results agreed well with ethnographic and archaeological evidence. !omparison of "@? and ")! levels in Preclassic +aya burials and animal bones from the early village site of !uello, Beli(e /"-66 B!-&% -@61, excavated by ?orman 2ammond and analy(ed by him, ?ikolaas van der +erwe, and >obert 2. Tykot, has also produced interesting results /see diagram1. +easuring the amounts of ")! and "@ ? in fossili(ed ?eanderthal bones from the cave of +aurillac, !harente, has led Brench researchers to the conclusion that
57

their diet was almost exclusively carnivorous. The same carbon and nitrogen isotopes have also been analy(ed in other kinds of tissue, such as the skin and hair of mummies from the ?ubian %esert, dating from )@6 B! to &% )@6, and suggest that the population ate goats and sheep, cereals and fruit. #ince isotopes show up in hair only two weeks after they are consumed /whereas bone shows what was eaten over a lifetime1, different segments of the same hair can show changes in diet, the segments closest to the scalp even indicating the season at the time of death. 7ocks of hair from -666- yearold Peruvian and !hilean mummies have even been found to contain traces of cocaine consumption from the chewing of coca leaves. #cientists have also found that concentrations of strontium, a stable mineral component of bone, can provide data on diet. +ost plants do not discriminate between strontium and calcium, but when animals eat plants, strontium is discriminated against in favor of calcium' most of the strontium is excreted, but a small constant percentage enters the blood stream and becomes incorporated into bone mineral. The contribution of plants to the diet can therefore be assessed through the proportions of strontium and calcium /#rE!a1 in human bone - the bigger the contribution /e.g. in a vegetarian1, the higher the #r*!a ratio, whereas a meat-eater3s diet gives a low ratio. #outh &frican

!hat Did The" Eat# Su$sistence and Diet

anthropologist &ndrew #illen has discovered by this techni,ue that Australopithecus robustus, formerly thought to have been a vegetarian because of its powerful grinding .aws, did eat some meat and was therefore probably omnivorous. &nalysis by +argaret #choeninger of strontium levels in bones from the eastern +editerranean has shown that the proportions of plant and animal foods in the diet did not change radically from the +iddle Paleolithic until the +esolithic, when there was a shift toward a greater use of plant foods. 2er results show that people here had a plant-rich diet a considerable time before cereals were domesticated. #choeninger has used the same techni,ue to study skeletal material at !halcat(ingo, an <lmec site in central +exico at its peak around C66-@66 B!, where a combination of strontium results and an assessment of gravegoods indicates a ranked society with a differential consumption of meat. #he found that the highest- ranked people buried with .ade had the lowest bone strontium /and therefore ate plenty of meat1' those buried with a shallow dish had a higher strontium level /and thus ate less meat1' while a third group lacking any grave-goods had the highest strontium level /and probably ate very little meat1. & different picture emerges where shellfish contributed to diet, because strontium concentrations are far

"6

$$$$$$$$$$$$$$>

58

&one collagen analysis of 'reclassic (aya burials and animal bones from the site of )uello, &eli*e, showed that mai*e formed +,--. percent of the diet of humans, and of dogs bred for food. The wide /+ range of both ) and /, 0 for dogs suggests a mi1ed diet. 2orest species, such as deer, and marine turtles ate only )+ plants, and had a lower protein inta#e, indicated by the /,0 figures. Armadillos ha%e high figures due to eating grubs that themsel%es eat the roots of mai*e plants.

HUMANS Mud tu%tle dee% !hit e0tail 1ec ca%" A%madillo

g 5

<D -%oc3et dee%

Do& 2

AGGG$GGG7 0;8

--6

06

i i i iGGGGGGGG$GGG$GGG7 067 0 6668 delta C 9*e% mil:

04

AGGG$GGG7

05

59

PART II Discove%in& the 'a%iet" o( Human E)*e%ience

higher in molluscs than in plants. #keletons from an &rchaic huntergatherer population of around -@66 B! at a northern &labama site proved to have a higher strontium level, thanks to the molluscs in their diet, than those from an agricultural +ississippian population buried at the same site in about &% "866. >ecent studies, however, suggest that due to contamination from sediments and ground water in which some bones are buried, strontium values can be mis
SUMMAR<

FURTHER READING
Most o( the sou%ces &iven at the end o( Cha*te% = a%e a**%o*%iate (o% this cha*te% as .ell. In addition> hel*(ul volumes a%e+

&ll the methods described in this chapter are providing archaeology with new tools, not to say with :food for thought.: The evidence available varies from botanical and animal remains, both large and microscopic, to tools and vessels, plant and animal residues, and art and texts. e can discover what was eaten, in which seasons, and sometimes how it was prepared. e need to assess whether the evidence arrived in the archaeological record naturally or through human agency, and whether the resources were wild or under human control. <ccasionally we encounter the remains of individual meals left as funerary offerings or as the contents of stomachs or feces. Binally, the

Brothwell, D. & P. 1997. Food in Antiquity: A Survey of the Diet of Early Peoples. Johns Hopkins Univ. Press: Baltimore, MD. il!ert, ".#. an$ Mielke, J.H. %e$s.&. 19'(. he Analysis of Prehistoric Diets. )*a$emi* Press: +ew ,ork & -on$on. Harris! D.". #ed.$. %&&'. The 3rigins and Spread of Agriculture and 'astoralism in Eurasia. ()* Press: *ondon. Harris, D.". & Hillman, ... %e$s.&. 19'9. Foraging and Farming. he Evolution of Plant E+ploitation. Unwin H/man: -on$on. Hastorf! ).A. , Popper! -.S. #eds.$. %&... )urrent 'aleoethnobotany. Analytical (ethods and )ultural

leading and one should keep an open mind until possible pitfalls are better understood. $n any case, the techni,ue is only a complement to not a replacement for - the analysis of carbon isotopes. The #r*!a ratio reveals the proportionate amounts of meat and plants in the diet' but isotopic analysis is needed to learn what kinds of plants were being consumed. &rchaeology provides the evidence that permits more precise identification of the plant and animal species involved.

human body itself contains a record of diet in its tooth- wear and in the chemical signatures left in bones by different foods. +any of the techni,ues lie in the domain of the specialist, particularly the biochemist, but archaeologists should know how to interpret the results, because the rewards are enormous for our knowledge of what people ate, how they exploited their resources, and in what proportions. Prospects for future research look good. #ince we understand increasingly what a good or balanced diet entails, we can now begin to examine past diets in terms of nutritional value and assess the state of health of our ancestors.

$nterpretations of Archaeological 'lant 4emains. (niversity of )hicago Press: )hicago. /lsen! S.0. %&12. 3steology for the Archaeologist. Peabody M0se0m: .am!ri$1e, Mass. Pearsall, D.M. 19'9. Paleoethnobotany: A Handboo3 of Procedures. )*a$emi* Press: +ew ,ork & -on$on. Pri*e, 2.D. & e!a0er, ).B. %e$s.&. 199(. *ast Hunters! First Farmers. 3*hool o4 )meri*an "esear*h Press: 3anta 5e. "eit6, 7.J. & 8in1, 7.3. 1999. 4ooarchaeology. .am!ri$1e Universit/ Press: .am!ri$1e. 3mith, B.D. 199(. he Emergence of Agriculture. 3*ienti4i* American *ibrary: 5e6 7or3. 4ohary! D. , Hopf! 8. %&&&. 5omestication of 'lants in the 3ld 6orld. The 3rigin and Spread of )ulti%ated 'lants in 6est Asia, Europe and the 0ile 7alley. #9rd ed.$. )larendon Press: /+ford.