Isotopic evidence for Maya patterns of deer and dog use at Preclassic Colha

Journal of Archaeological Science (2001) 28, 89–107doi:10.1006/jasc.1999.0560, available online at http://www.idealibrary.com on

Isotopic Evidence for Maya Patterns of Deer and Dog Use atPreclassic Colha

Christine D. White

Department of Anthropology, The University of Western Ontario, London, Ontario N6A 5C2, Canada

Mary E. D. Pohl

Department of Anthropology, Florida State University, Tallahassee, Florida 32306-2023, U.S.A.

Henry P. Schwarcz

Department of Geology, McMaster University, Hamilton, Ontario L8S 4M4, Canada

Fred J. Longstaffe

Department of Earth Sciences, The University of Western Ontario, London, Ontario N6A 5B7, Canada

(Received 1 June 1999, revised manuscript accepted 25 February 2000)

Based on our assessment of human exploitation of animals in the New World, we propose that one criterion fordomestication should be dependence on humans for food, a trait that we test through isotopic analysis of faunalremains. Stable carbon- and nitrogen-isotope compositions of bone collagen have been analysed for 24 dogs and 16 deerfound in well-dated contexts from the prehistoric Maya lithic manufacturing community of Colha, Belize. The samplespans the Early Middle Preclassic period beginning 1000–600 to the Terminal Late Preclassic period ending around 250. The majority of both dogs and deer come from middens, but three dogs come from cache contexts in buildings.The degree to which humans controlled the diets of these animals varies markedly by context and time period. The dietsof midden dogs demonstrate a significant increase in the amount of C4 (maize-based) foods and become moreherbivorous over time. Because the midden dogs were probably dependent scavengers, this phenomenon might reflectthe dynamics of human dietary change as the population at Colha expanded towards the end of the Preclassic period.An increase in the homogeneity of dog diets might also be indicative of either more restrictive human control over theanimals or a reduction in the variability of resources used by humans. Alternatively, because the structure associatedwith the midden in which the dogs were found became more ceremonial in Late Preclassic times, the dogs from thisperiod could be reflecting a general increase in purposeful feeding for ceremonial purposes. Dogs found in special (i.e.non-midden) contexts from both the Late Middle Preclassic and Late Late to Terminal Late Preclassic periods havedistinctive isotopic signatures that strongly suggest a more specific occurrence of purposeful maize feeding. Evidence isprovided from ethnohistory and Maya mythology that may explain their distinctive mortuary and feeding behaviour.We infer that all of the deer in the Preclassic period contexts at Colha were wild and procured by hunting because theyconsumed a herbivorous C3 diet. � 2001 Academic Press

Keywords: STABLE CARBON- AND NITROGEN-ISOTOPE RATIOS, PALAEODIET, MAYA, DEER, DOG,DOMESTICATION, RITUAL, PRECLASSIC, COLHA.

Introduction

A nimals and plants were the grocery store, thepharmacy and the church in early cultures.Animals played a key role in ritual and

medicine, as well as in the diet and economic life ofAmerindians, and were so sought after that they were

890305–4403/01/010089+19 $35.00/0

sometimes traded over long distances (DiPeso, 1974;Gode, 1976; Tyler, 1991; Morales, 1995; Wing & Wing,1995). For the tropical lowland Maya region inMesoamerica (Figure 1), preservation and samplingproblems, combined with traditional paradigms, haveresulted in a limited amount of data on animals andhave inhibited our understanding of the relationship

� 2001 Academic Press

90 C. D. White et al.

Figure 1. Map of the Maya region showing the location of the sites mentioned in the discussion (adapted after Powis et al., 1999).

between the Maya and animals who shared theirenvironment. The practical problems of preservation,sampling, and representativeness in Maya environ-ments were probably also overshadowed by a wide-spread belief that animal domestication was notparticularly important to the Maya. We can see thisEuropean-based outlook in an early ethnohistoricaccount from Colonial Yucatan, in which Friar Diegode Landa (1566, cited in Tozzer, 1941) wrote: ‘‘the

Indians are wanting in the possession of many animals,especially those most necessary for the service of man’’.A second inhibiting belief is the common assumptionthat animals were domesticated for food. As Chaplin(1969) notes, domestication is ‘‘simply a pattern ofexploitation’’. The process of domestication, in fact,may have been a complex interaction of humans withanimals as pets, hunting companions, and scavengers(Downs, 1960; Zeuner, 1963; Tennessen & Hudson,

Isotopic Evidence for Maya Patterns of Deer and Dog Use 91

1981). Domestication may also have involved the needfor animals (or their products) in ritual, medicine,labour, and trade (Cranstone, 1969; Isaac, 1970;Crabtree & Campana, 1989). For the Maya, the use ofanimals in rituals, medicine, and trade appears to havebeen particularly significant (Pohl, 1990). Therefore,we believe that a broader rationale for domestication isappropriate to human–animal relationships in thisregion.

Traditional methodological approaches to providingevidence for animal domestication are dependentprimarily on quantitative and qualitative aspects ofarchaeologial research design. The spatial distributionand relative quantities of species, their size, and demo-graphic and morphological characteristics have beenused to argue domesticated relationships (Chaplin,1969; Davis, 1987). Their specific use as food is deter-mined by evidence of butchery and cooking, as wellas locational context. Additionally, art, writing andethnohistoric accounts provide evidence of domesti-cation, and illustrate animal usage. This study examinesthe potential of stable isotope analysis of bone collagento characterize the relationship between animals (deerand dog) and the Maya at the site of Colha, Belize,during the Preclassic period. Instead of using isotopes todemonstrate the animal component of human diets, as inmost palaeodiet research, here we characterize the foodsconsumed by the animals in order to determine thedegree of animal dependency on humans.

We have focused our study on dog (Canis familiaris)and deer (mostly Odocoileus virginianus) because theyare abundant in faunal remains and because paintingsand ethnohistorical accounts suggest that these animalswere particularly significant to the Maya as symbols ofelemental forces of life: dogs for fire and the hearth(Nicholson, 1971; Thompson, 1972), and deer for sun,rain, and fertility (Pohl, 1981), as well as beliefsconcerning lineage. Pohl (1990) proposed that theextent to which the Maya domesticated deer and dogsdeserves investigation because these animals wereprobably used for ritual sacrifice and as high statusfeasting food. We discovered that, at Preclassic Colha,dogs found in middens were isotopically distinct fromthose found in ritual contexts. We suggest that the dietsof midden dogs may have closely approximated thediets of humans, including changes over time, and, bycontrast, dogs given special mortuary treatment werepurposefully fed maize, probably for ritual reasons. Allof the deer appear to have been hunted.

Background for the Maya Case Study

Stable isotope analysis for the purpose of dietaryreconstruction has been used in archaeology primarilyfor humans, e.g. to document major dietary shiftsassociated with technological change, to track themigration of major crops such as maize, and to recon-

struct past lifeways and social systems (for reviews seeSchwarcz & Schoeninger, 1991; Ambrose, 1993; Pate,1994). For the Maya, this methodology has been usedto further understanding of: the relationship amongthe Classic Maya collapse, subsistence and pathology(White & Schwarcz, 1989; Reed, 1994; Wright, 1994;White, 1997; Wright, 1997; Wright & White, 1996),regional variation in diet (Wright & White, 1996;Gerry & Krueger, 1997; White, 1997), social structure(White et al., 1993, 1996; Reed, 1994; Wright, 1997;Gerry & Krueger, 1997), and the relationship betweenintensive agriculture and population pressure (Whiteet al., 1993; Wright, 1997). Although intent to breedmay be a condition for domestication (Bokonyi, 1969),it is not a condition that we can test using thismethodology. It may also not be a necessary conditionfor New World human–animal relationships. For ourpurposes, we assume that dependency on humans forsurvival is a primary condition for domestication.Notably dietary dependency may not be both a neces-sary and sufficient condition. For example, commensalanimals are dependent scavengers, but would not beconsidered domesticated under the traditional defini-tion. There is a continuum of human control overanimal feeding in human–animal relationships. Stableisotope values in animals that are purposefully fed maynot be easily distinguished from those who are depen-dent scavengers, although animals that are purpose-fully fed could have the most restricted diets. Animalsthat are purposefully fed may have diets which are thesame, or different, from those of humans, dependingon whether they are being fed ‘‘human food’’ or not.Occasional scavenging on human garbage or cropscould be considered semi-domestication, and shouldresult in diets intermediate to those of humans andwild animals.

Several isotopic analyses of archaeological dogremains have been used to provide circumstantialevidence for agriculture or human subsistence changebased on the assumption that dogs were domesti-cated and had diets similar to humans (Burleigh &Brothwell, 1978; Noe-Nygaard, 1988; Clutton-Brock& Noe-Nygaard, 1990; Cannon et al., 1999). Otherstudies, which include analyses of dog remains onlyincidentally as faunal controls (many of which arefrom the Maya region), indicate that dogs have iso-topic signatures much more like those of humans thanmost other animals (Katzenberg, 1989; White &Schwarcz, 1989; White et al., 1993; Reed, 1994; Tykotet al., 1996; Gerry & Krueger, 1997). This similaritycannot simply be explained by the fact that humansand dogs are both omnivores. Dogs are being fed orscavenging the same foods as humans.

Most isotopic studies of deer have focused onecological issues (e.g. Land et al., 1980; Cormie &Schwarcz, 1994). Isotopic analysis of archaeologicaldeer from the Peten, Guatemala, has been used to testthe hypothesis that population pressure in the MayaClassic period caused habitat destruction (Emery,

92 C. D. White et al.

1997; Emery et al., in press). On a smaller scale, deerhave also been analysed for a Late Classic periodceremonial deposit at Lagartero (White et al., 1997),and used as faunal controls in palaeodiet studies(White & Schwarcz, 1989; Gerry, 1993; Reed,1994; White et al., 1993; Tykot et al., 1996). In thesestudies, deer demonstrate a broad range of variationfrom completely wild herbivore diets to diets similarto those of omnivorous humans using maize as astaple.

Figure 2. Late Classic period Maya figurines portraying dogsprobably fattened on maize (Museo de Antropologia, Tabasco).

DogsThe dog was the principal domesticated animalthroughout the Americas (Schwartz, 1997), and theoldest evidence for its domestication in Mesoamericacomes from Late Archaic contexts in the TehuacanValley (Flannery, 1967). Dogs are found in burials asearly as 1400 in Veracruz (Merino Carrion &Garcıa Cook, 1997). In the Maya area, they are foundin deposits from the Early Middle Preclassic period(�1000–600 ) onwards (Andrews & Hammond,1990; Clutton-Brock & Hammond, 1994). Ethno-historic references to dogs are numerous and describetheir use for food, medicine, ritual and hunting (Landa,1566, cited in Tozzer, 1941; Hernandez, 1514–1578,cited in Tozzer & Allen, 1910; Lejon, 1625; Lopez deCogolludo, cited in Allen, 1920; Lopez de Cogolludo,1688; Villagutierre Soto-Major, 1701; Roys, 1931,1943; Duran, 1971; Michelon, 1976) as do archaeologi-cal data from art and osteology (e.g. Pendergast, 1974;Pohl, 1983, 1990; Danien, 1997). The fact that theMaya have at least nine words for dogs (Schwartz,1997) also indicates that they were a significant com-ponent of the Maya world. Yucatec Maya in northernBelize still recount the fact that their ancestors atedogs. According to ethnohistoric data, the fiestaappears to have been the main occasion for commonpeople to eat meat, whereas the nobility are reportedto have consumed meat daily (Landa, 1566, citedin Tozzer, 1941; Lopez de Cogolludo, 1688, cited inTozzer & Allen, 1910; Roys, 1943). Dogs were alsoproffered to elites as tribute at the time of conquest(Pohl, 1985).

Hernandez (1514–1578, cited in Tozzer & Allen,1910) observed dogs being sold in the market atMerida, Yucatan, and noted that dogs were purpose-fully fattened after castration. (Figure 2 illustrates dogsthat were probably fattened on maize.) Markets, how-ever, were rare in the Yucatan at the time of conquest(Fariss, pers. comm.). The production and use of dogsmay have been localized, with households only raisingthe dogs they needed. During early Colonial times,dogs are described as being raised in houses (Ximenez,1967), pits (Lopez de Cogulludo, Book I, cited inTozzer & Allen, 1910), and pens (Herrera, 1934–1947,cited in Hellmuth, 1977). We assume that dogs kept inconfined spaces, such as these, were likely purposefullyfed.

The archaeological evidence for the use of dogsamong the Maya as food versus ritual objects isdualistic. The Preclassic site of Cuello providesindirect osteological evidence of dog consumption(Clutton-Brock & Hammond, 1994), as do Tikal andSeibal during the Late Classic period (Pohl, 1990). Bycontrast the evidence for the nature of dog use atPostclassic Cozumel is less clear. The burial contexts,treatment of the bones, and the unusually highrepresentation of canine teeth in the sample mightindicate ritual use (Hamblin, 1984).

The ideological significance of dogs is evident inboth artistic and ethnohistoric records. In art, dogsappear to be integral to the most significant politicalrituals, such as inaugurations and founding of newcivic religious centres as lineages fissioned (Danien,1997). The association of dogs with death and thejourney to the underworld is well known from both artand traditional lore (Trik, 1963; Coe, 1982; Schele &Miller, 1986; Freidel et al., 1993; Grube & Nahm,1994). The significance of the dog as a travellingcompanion and protector of the hearth may haveextended beyond the ultimate trip to the afterlife.

In the written record, Landa (1566, in Tozzer, 1941;in Coe & Coe, 1996) describes two contexts for dogsacrifice. In one, he records a New Year’s rite wheredogs are sacrificed in place of humans, a practice inkeeping with the common theme of new beginningsand the association of dogs with yearly renewalceremonies (Pohl, 1983). In another, he documentssacrifices of dogs ‘‘spotted with the color of cacao’’ aspart of cacao rituals, which would also have hadsymbolic economic importance.

In terms of the archaeological evidence, there aremany examples of special deposits of dog remainsconsistent with sacrificial offering. Dog burials arefound in ceremonial and administrative contexts atCozumel (Wing, 1978), and in natural features associ-ated with fertility and ancestor worship (Pohl, 1983)such as cave caches from the Terminal Classic (along

Isotopic Evidence for Maya Patterns of Deer and Dog Use 93

with deer) (Pendergast, 1969, 1974), and cenotedeposits from the Postclassic (Pollock & Ray, 1957).(The fact that such esoteric deposits contained so manyindividuals may indicate that the numbers of dogbones at archaeological habitation sites are deceptivelylow.) Dogs also accompany humans in graves (Kidderet al., 1946; Pollock & Ray, 1957; Hamblin, 1984;Tourtellot, 1990) including burial in royal residences(e.g. at Copan, E.W. Andrews V, pers. comm.).

Temporal distinctions in the ritual use of dogs havenot been well studied. Shaw (1995) notes that at Colha,evidence for ceremonial use of dog bones seems toincrease toward the end of the Preclassic period. Otherspecial deposits of Late Preclassic dogs, i.e. burials andcaches, are also found at Cuello (Clutton-Brock &Hammond, 1994), Cerros (Carr, 1985) and Rıo Azul(Valdez & Buttles, 1991; Valdez, 1992).

The degree to which dogs are generally representedin the faunal record varies by region and time period.Dogs appear to have been unusually important tocoastal populations, not only throughout Mesoamericabut also in South America (Burleigh & Brothwell,1978; Wing, 1978; Hamblin, 1984; Carr, 1985).Presumably, domesticated dogs would have provided arenewable, easily accessible and convenient terrestrialprotein resource which, as Carr (1985) notes, couldhave been more reliable than either fished or huntedresources. The proportion of dogs to other fauna tendsto change from one period to another within sites,making it difficult to generalize patterns of usagecommon to all Maya. Preclassic sites provide evidencethat dog remains were major components of middens(Wing, 1981; Carr, 1985; Pohl, 1990; Wing & Scudder,1991; Shaw, 1995). Late Classic and Postclassic dogdeposits tend not to be in middens. It is currentlyunknown whether the dietary versus symbolic functionof dogs changed over time, or whether this pattern isa reflection of archaeological research design whereexcavation is selective.

DeerThe Maya dog, with its ginger coat, upright ears, andsleek body (Olsen, 1978), bears a striking similarity tothe white-tailed deer. The role of deer in Maya lifeis also intriguingly similar to dogs, but their statusas domesticates is more questionable. Althoughindigenous bovids, caprines, and camelids were avail-able for domestication in the New World, with theexception of some of the camelids (Wing, 1978; Wing& Steadman, 1980), Amerindians generally preferredto hunt or selectively manage herds of native animalsinstead of domesticating them. The deer hunt amongthe Colonial Maya is documented ethno-historically (Landa, 1566, cited in Tozzer, 1941) andin high-status Late Classic period painted pottery(Pohl, 1981; Hellmuth, 1996). Techniques used in theprocurement of deer appear to have been variable.Late Classic ceramics illustrate deer hunting using

several techniques including atlatls (Pendergast, 1969);deadfall traps and possibly snares (Pohl, 1990); nets(Pohl & Feldman, 1982); blowguns, dogs, and whistlesused as lures (Carr, 1985). Deer were hunted com-munally in round-ups, which may have been a meta-phoric dramatization of harvesting successful maizecrops, and/or capturing rivals in warfare (Pohl, 1985,1994). An optimal location for hunting today is themilpa field and its surrounding zones (Reina, 1967;Pohl, 1977). Deer are attracted to the salty ashes of theburned fields, and, later, to a lesser extent, to thesupply of immature grain in the fields. They primarilyfeed or browse in the forest edge environment aroundthe fields. It would be surprising if they were nothunted in the same location in ancient times as well.

White-tailed deer (Odocoileus virginianus), whichthrive in conditions of forest disturbance, are com-monly found among the faunal remains at ceremonialcentres. The smaller brocket deer (Mazama sp.), moreat home in undisturbed areas, also occur althoughmore rarely. Both occur early in the archaeologicalrecord (Early Preclassic, c. 2400–1300 , Pohl et al.,1996). Deer are found in high proportions at Preclassicsites (Pohl, 1985; Shaw, 1995), and deer was alsothe most popular food at Classic and Postclassicceremonial and political centres (Olsen, 1978; Wing &Steadman, 1980; Hamblin, 1984; Pohl, 1985; etc). Thisfood preference continued until Colonial times andinto the present (Pohl, 1977). How often the Mayaconsumed deer, and whether they were eaten onlyceremonially or on a daily basis, is not known.

Like dogs, deer embody natural forces such as thesun and rain, and are associated with ceremoniesmarking yearly renewal as well as political inaugur-ation (the cuch rite) (Pohl, 1981). They frequentlyappear in art, marking the investiture of authority(e.g. iconography on stelae). Maya rulers associatedthemselves with the predators of the natural world,especially the jaguar, and they show defeated sacrificialcaptives from neighbouring polities as deer (Pohl,1994). The predator symbolism with its message ofcontrol over the Maya world provides a link with theyearly-renewal agricultural fertility theme. As top‘‘predators’’, naturally the elite at Late Classic Seibaland Late Preclassic Cerros had favoured access tovenison according to an analysis of the distribution ofbones by economic contexts (Pohl, 1985; Carr, 1986).Like dog bones, deer bones occur in cave (Pendergast,1969, 1974) and cenote deposits (Pollock & Ray, 1957),which were focal locations of cults marking politicalauthority and fertility. The ritual importance of deer isstarkly illustrated at Cozumel, an island where deerwould not have been indigenous (Hamblin, 1984).Although deer constitute only a tiny portion (1·4%) ofthe fauna there, the majority of deer remains arecomprised of head, foot and ankle bone elements andsome are burned. This pattern strongly suggests thatdeer, or only parts of deer, were imported for ritualpurposes.

94 C. D. White et al.

Today, Maya sometimes raise captive deer, usuallycaught as fawns on hunting trips. The actualhusbandry of deer has been suggested by the ethno-historic record, but has never been establishedarchaeologically. Deer were so passive to the SpanishConquistadors that they were easily killed, accord-ing to the accounts of Cortes (Means, 1917). TheSpaniards were told that deer were protected and‘‘treated like gods’’ for religious reasons. The sacrificeof a captive deer is documented ethnohistorically byPalacio (1860, cited in Tozzer & Allan, 1910). Further-more, Landa refers to the husbandry of deer by womenin early Colonial times:

deer suck their breast, by which means they raise them andmake them so tame that they never will go into the woods,although they take them and carry them through thewoods and raise them there (Landa, 1566, cited in Tozzer,1941: 127).

As Pohl (1990) has noted, this behaviour suggestsmanagement of animal herds, a practice found else-where (Taber & Dasmann, 1957; Wilkinson, 1972). Ithas been hypothesized that deer might have beenpenned at political–ceremonial centres and purpose-fully fed for sacrifice in ancient times (Carr, 1996; Pohl& Feldman, 1982; Pohl, 1983). Deer, as well as dogs,may have been substituted for humans in sacrificialrituals (Pohl, 1985). One possible pen structure wasidentified at Terminal Classic Seibal and was associ-ated with remains of antler (Pohl, 1990). Although itis difficult to maintain a penned herd of deer, par-ticularly in the tropics (Crandall, 1964; Severinghaus& Cheatum, 1956), deer penning has been donesuccessfully in modern Guatemala (Lewis, 1970).

We might expect deer diets to vary considerably,because they can be completely wild forest feeders,semi-domesticates feeding from agricultural fields atthe forest edge, herds managed in the wild, or enclosedcaptives being fed by humans. Deer, however, are lesslikely to have scavenged on household garbage ormature maize crops and, therefore, are less likely toreflect human diets as dogs might.

Theoretical Background for Stable IsotopeAnalysisThe isotopic method of reconstructing palaeodiet inthis study uses the variation in carbon- and nitrogen-isotope ratios as it exists in the foodweb, and thesystematic ‘‘fractionation’’ of those isotopes as theymove up the food chain. Because the isotopic com-position of foods becomes incorporated into the tissuesof consumers, we can determine their general dietaryregime.

Figure 3 is a composite graph showing the isotopicranges of the types of foods available to the Maya. Themost basic level of isotopic variation is created byplants, which selectively incorporate or exclude atmos-pheric 13C during photosynthesis. The most common

isotopic plant types discriminate the most against 13Cand therefore have the most negative �13C values(modern average of �26·5‰, Smith & Epstein, 1971;O’Leary, 1988). These are often called C3 plants be-cause their metabolic processes result in a 3-carbonmolecule. In the Maya area, most wild plants, trees,nuts, fruits and vegetable cultigens fall into thiscategory (Table 1). Plants that incorporate more 13Cduring photosynthesis have the least negative �13Cvalues (modern average of �12·5‰, O’Leary, 1988).These are called C4 plants and are usually comprised oftropical grasses. In the Maya area, maize falls into thiscategory and is the main agricultural staple, based onevidence from ethnohistory (Landa, 1566, in Tozzer,1941; Hellmuth, 1977; Marcus, 1982), macrobotanicaland palynological evidence (e.g. Mikisicek et al., 1981;Crane, 1986; Rue, 1987; Pohl et al., 1996), and stablecarbon-isotope analysis of humans (White & Schwarcz,1989; White et al., 1993; Reed, 1994; Wright, 1994;Tykot et al., 1996; Gerry & Krueger, 1997). Amaranth(Amaranthus spp.) and epazote (Chenopodium ambro-soides) are also C4 plants that grow in the area, andthey may have made some contribution to the diet.Nevertheless, there is no archaeological evidence ofheavy consumption of these species, unlike the case formaize. The distribution of ranges in �13C values for C3and C4 plants is bimodal (O’Leary, 1988). Thus, therelative dependency on these plant types for foodshould be clear in �13C values of tissues. There is oneother photosynthetic category of plants (CAM orCrassulacean acid metabolism) that has a range of�13C values spanning both those of C3 and C4 plants(�27 to �12‰). CAM plants are mainly succulentsthat have more flexible photosynthetic processes. TheMaya might have consumed a few of these, specificallythe nopal cactus (Opuntia), the pinuela (Bromeliakaratas), and possibly the pineapple (Ananas cosmosus).

The �13C values mentioned above are all based onvalues for modern plants. Because plants derive theircarbon from the atmosphere, and because the burningof fossil fuels, which began in the modern industrialera, has decreased the �13C of plants by about 1·5‰since then (Freidli et al., 1986; Marino & McElroy,1991), we must assume that Preclassic Maya plants hadcorrespondingly more positive values of �13C.

Although plants form the base of the food chain andtheir isotopic signatures are passed from one trophiclevel to the next, the �13C values of animals will reflectthose of both the plants and the animals they con-sumed. The isotopic composition of collagen preferen-tially represents the protein component of the diet(Krueger & Sullivan, 1984; Lee-Thorp et al., 1989;Ambrose & Norr, 1993; Tieszen & Fagre, 1993). Thus,collagen �13C values represent both plant and animalproteins. Wild terrestrial animals consumed by theMaya were most likely to have been herbivoresthat would have lived in a C3 plant environment.Fauna that had a domesticated, or semi-domesticated,association with the Maya would have lived in an

Isotopic Evidence for Maya Patterns of Deer and Dog Use 95

00

25

–35

δ13C (‰, VPDB)

δ15N

(‰, A

IR)

20

15

10

5

–30 –25 –20 –15 –10 –5

Freshwatersnail meat

C3 N2-fixing plants

Terrestrialherbivore collagen

Colhadeer

Colhadogs

Reefshellfishmeat

Reef fishmeat

Marine mammals

C3 plants

Marine fish

Freshwater fishmeat

CAM

C4 plants

Figure 3. Theoretical model of major Maya food resources (ranges for most plants are shaded) and positioning of the range of �13C and �15Nvalues for dogs and deer from Colha within the model. (Food values from White & Schwarcz, 1989; Wright, 1994; Tykot et al., 1996.) Notethat the �13C values are not adjusted for a diet-collagen offset (�5‰) and the �15N values are not adjusted for trophic level (�3‰).

environment characterized by more C4 plants, i.e.mostly maize. Although deer are herbivores, dogs areomnivores and may have consumed the meat (andfaeces) of other animals.

The proportion of protein in the diet, relative to theother macronutrients, i.e. carbohydrates and lipids,can affect the carbon source in collagen and hence thedifference in �13C values between diet and collagen(Kennedy, 1988; Chisholm, 1989; Ambrose & Norr,1993; Schwarcz, 1999). In low-protein maize-dependent diets, the carbon in collagen is synthesizedfrom carbohydrates. Although the Maya were maize-dependent, the relative contributions of proteins,carbohydrates, and fats to their diets is still poorlyunderstood (Coyston, 1994; Wright, 1994; Tykot et al.,1996). Previous studies of Maya diet indicate regionaland temporal variability in the protein sources con-sumed (for a review see Wright & White, 1996).Accordingly, it is quite possible that the �13C valuesmeasured from Maya bone collagen represent carbonfrom both protein and carbohydrate sources, and thatthe relative proportions of these may shift by timeperiod, and differ by location and social status. Furtherstudies that include carbon isotope analysis of boneapatite may help to resolve these issues, but only if thebone is well-preserved (Wright & Schwarcz, 1996).

Although the �13C values of bone collagen provideindirect evidence of plant and animal protein sourcesbecause dietary protein is routed to collagen during itssynthesis (see above), it is not possible using �13Cvalues alone to differentiate between protein sources.Freshwater fish and snails in this region have flesh thatresembles C3 plants (Wright, 1994; Tykot et al., 1996)and were consumed in quantity at some sites (Moholy-Nagy, 1978; Healy et al., 1990; Stanchly, 1995). Bycontrast, marine fish and shellfish from the CaribbeanBarrier Reef are enriched in 13C and have �13C valuessimilar to maize (Keegan & DeNiro, 1988; Tykot et al.,1996) (Figure 3). The consumption of marine andreef resources was significant at Cahal Pech, anotherPreclassic site in Belize (Powis et al., 1999). BecauseColha was even closer to the reef than Cahal Pech, andits faunal assemblage contains marine species, it ispossible that reef resources may have been a significantpart of the diets of both humans and animals. Tonarrow down the source of protein and its level in thefood chain (e.g. freshwater fish (C3) versus C3 plants,freshwater snails and legumes, or reef fish (C3) versusC4 plants or marine shellfish), nitrogen-isotope ratiosare used. With the exception of legumes (e.g. beans,Phaseolus vulgaris, which were part of Maya diets)and blue-green algae (which is abundant along the

96 C. D. White et al.

Caribbean Barrier Reef), nitrogen-isotope ratios donot vary among plant types. Legumes and algae fixnitrogen differently and have �15N values close to 0‰.Other plants have �15N values ranging from 2 to 6‰.As nitrogen passes from one level in the food chain tothe next, it becomes enriched by about 3 to 4‰ in 15Nin the consumers, making it a useful indicator oftrophic level (DeNiro & Epstein, 1981; Schoeninger,1985). In the Maya area, fish tend to have higher �15Nvalues than terrestrial herbivores, and freshwater fishhave higher values than reef fish (Keegan & DeNiro,1988; Norr, 1991; van der Merwe et al., in press)(Figure 3). Terrestrial herbivores have �15N values thatoverlap with reef fish, shellfish, freshwater snail meat,and most C3 plants, but are intermediate to freshwaterfish and legumes.

Table 1. �13C values for modern plants and animals from Belize

Food source�13C(‰)

Maize �11·2

Wild and domesticated C3 foodsRamon �29·2Camote-cacas �29·4Camote �27·2Palm �24·7Cassava �27·0Taciste �28·4Macal �26·8Salsa (beb) �28·2Oregano �28·8Trumpet wood �27·2Mangle �29·3Golondrina �28·7Jicaco �27·3Mean �27·9.. 1·3

Plants consumed by deerHabin �28·6Papaya �29·3Xkintal �29·8Chichibe �30·6Pixoy �27·0Jobo �29·3Mean �28·8.. 1·4FaunaDeer (Odocoileus viginianus) �24·3Armadillo �27·2Peccarya �23·3Tapira �24·2Kinkajoua �22·4Gibnuta �22·0Armadillo earth worms �22·0Ants �19·8Termites �26·9

avan der Merwe et al. (1999).All other materials collected by M. Pohl.

Site and Sample DescriptionColha was a significant Late Preclassic LowlandMaya centre located in northern Belize about half way

between the New River and the Belize River, andabout 22 km away from the Caribbean Sea and itsbarrier reef. According to the ceramic-based chron-ology (which is yet to be calibrated with 14C dates), thetime span we are examining in this paper extends fromthe Early Middle Preclassic period (1000 to 600 ) tothe Late Late and Terminal Late Preclassic periods(100 to 250) (Adams & Valdez, 1980; Valdez,1987). Situated near a large outcrop of chert, andalong an established trade route, Colha was a majorindustrial centre of lithic tool production unique innorthern Belize (Wilk, 1975; Nash, 1982; Potter, 1982).It came to dominate the production and export of lithiccommodities by the Late Preclassic period (Shafer,1982). In fact, the degree and patterning of movementof lithics from Colha has been used as an exampleof craft specialization in a ‘‘tightly integratedregional economy’’ (McAnany, 1986). Colha waslinked by both lithics and ceramics to virtually all otherPreclassic Maya centres in northern Belize, includingCuello, Nohmul, Santa Rita Corazol, El Pozito,San Estevan, Altun Ha, Cahal Pech, and Kichpanha(Hammond, 1991). Possibly as a result of its economicsuccess, and probably as part of political consoli-dation, in which smaller centres were subsumed, Colhaexperienced a population explosion at the end of thePreclassic period that levelled out in the followingClassic period (Hammond, 1991).

The faunal remains reflect the proximity of the siteto the coast. There is a variety of freshwater turtles,as well as an abundance of crocodiles and sea turtles,but a large proportion (42%) of the assemblage ismade up of marine and reef fishes (Wing & Scudder,cited in Hammond, 1991). Nevertheless, dog and deeralso occur in large quantities and in a variety ofcontexts (Shaw, 1991). In this study, samples of bonefrom 24 dogs and 16 deer found in well-dated con-texts have been analysed for their stable carbon andnitrogen isotopes. The samples come from two neigh-bouring areas of residential activity (Operations 2031and 2012) that saw 1250 years of occupation begin-ning in the Early Middle Preclassic period, and whichwere used for ceremonial purposes at the end of thePreclassic period (Potter, 1982; Anthony & Black,1994). Shaw (1991) notes that although dog is com-mon throughout the Preclassic sequence at Colha,the frequency of dog remains increases over time.Three temporal periods are represented by the dogs,i.e. the Early Middle Preclassic, the Late MiddlePreclassic, and the Late Late to Terminal LatePreclassic (Table 2). Most of the dogs come fromprimary middens and are represented by randomelements, i.e. they are assumed to have been humanfood remains. Three dogs were, however, found inspecial locations, two of which were caches. Two ofthese belong to the Late Middle Preclassic period(one was located on a platform associated withStructure II; the other, a cache, occurred in a sub-floor of Structure III). One other dog cache from the

Isotopic Evidence for Maya Patterns of Deer and Dog Use 97

Late Late Preclassic period was found in a collapsedwall associated with Structure V (Table 2).

The deer analysed here have been dated to only twotemporal periods: Early Middle Preclassic and LateLate to Terminal Late Preclcassic (Table 3). Theyoccurred in primary or secondary middens, with twoexceptions where the deer were found on the groundsurface. None were found in special structural con-texts, unlike the dogs.

We analysed a selection of modern plants and ani-mals from northern Belize for their �13C values inorder to provide a picture of the local food web fromwhich deer and dog might have derived their meals.These include maize, wild plants, plants known to beconsumed by deer in the wild, and a variety of smallanimals and insects (Table 1).

Table 2. Isotopic data for Colha dogs by time period

Sample no.time period Context

Boneelement

�13C(‰)

�15N(‰)

Collagenyield (%) C/N

Early Middle Preclassic (1000–600 BC)2031-38A-854 PM Axis �16·8 9·9 3·3 3·52031-38A-4501 PM Radius �13·8 9·9 5·3 3·42012-15-44 OGS Mandible �17·5 9·9 6·1 3·52012-15-51 OGS Metacarpal �20·0 9·1 4·9 3·2Mean �17·0 9·7 4·9 3·4.. 2·6 0·4 1·2 0·1

Late Middle to Early Late Preclassic (600–100 BC)2031-37-3365 Platform Ulna �12·1* 9·3* 1·7 3·42031-45-3995 PM Femur �13·8 8·3 5·5 3·42031-45-3997 PM Radius �13·4 9·7 — 3·32031-51-3023 Subfloor Radius �8·0* 7·5* 4·6 3·42031-34-3660 PM Radius �12·6 9·3 1·9 3·12031-34-3666 PM Metacarpal �14·0 9·3 2·7 3·32031-34-3671 PM Femur �8·1? 7·2? 1·8 3·2?2031-39-2170 PM Humerus �14·6 8·9 2·2 3·32031-39-3208 PM Femur �22·5 6·6 2·3 3·52031-4B-1074 P/SM Humerus �13·2 8·5 3·3 3·72031-4B-1097 P/SM Femur �10·9 8·7 1·9 3·22031-4B-1098 P/SM Tarsal �11·1 9·9 1·3 —2031-4B-2181 P/SM Tibia �17·5 10·1 2·9 3·5Mean �13·8 8·8 2·7 3·4.. 3·7 1·1 1·3 0·2

Late Late–Terminal Late Preclassic (100 BC–AD 250)2031-25-2335 PM Mandible �15·0 9·3 2·9 3·22031-25-3446 PM Pelvis �12·5 7·8 2·0 3·12031-25-3455 PM Ulna �11·5 7·5 2·8 3·22031-25-3467 PM Calcaneus �11·8 9·0 2·6 3·12031-13-3603 Wall Tibia �9·0* 6·7* 2·0 3·32031-13-3610 PM Radius �13·3 8·6 1·0 3·4Mean �12·8 8·4 2·2 3·2.. 1·4 0·8 0·6 0·1

*�13C and �15N values were not included in the means because of the special context.PM, primary midden; SM, secondary midden; OGS, on ground surface; ?, may have been purposefully fed maize.

Method

Collagen was extracted from bone using the proceduredescribed by Longin (1971). Bone collagen from low-land Maya sites, particularly early ones, is not expected

to be very well preserved because of the tropicalenvironment (White & Schwarcz, 1989; White et al.,1993, 1996; Wright, 1994; Tykot et al., 1996). In orderto reduce further collagen loss due to preparation, amore dilute solution (0·25 N) of HCL (see Chisholmet al., 1983) was used, followed by refluxing thecollagen-based ‘‘gelatin’’ at a lower temperature (58�C)(see Brown et al., 1988). Samples of gelatin werecombusted with CuO at 555�C, and the N2 and CO2gases were analysed for 13C/12C and 15N/14N ratios onthe SIRA mass spectrometer at McMaster University.The analytical precision for �13C is �0·1‰ and�0·2‰ for �15N. The isotope ratios of carbon andnitrogen are expressed in per mil (‰) as � values:

�=[(Rsample/Rstandard)�1]�1000

where R=13C/12C for �13C values, and R=15N/14N for�15N values. The carbon standard is Vienna PeeDeebelemnite (VPDB) (Coplen, 1996) and the nitrogenstandard is atmospheric nitrogen (Coplen et al., 1992),for which an inter-laboratory standard sample ofglycine was used.

98 C. D. White et al.

The C/N ratios of collagen have been established asa common means of detecting the presence of post-mortem alteration (DeNiro et al., 1985). As a check onthe validity of the data, both C/N ratios and thepercentage yield of collagen were measured and testedfor associations with the �13C and �15N values. TheC/N ratios were determined on a Carlo-Erba Analyserat the University of Western Ontario. Reproducibilityof the measured value for both carbon and nitrogenaveraged �2%.

Results and Discussion

Collagen yield is a simple measure of preservationin archaeological bone (DeNiro & Weiner, 1988;Ambrose, 1990) and can be used to predict the validityof isotopic ratios (White, 1991). Maya human materialis generally poorly preserved because of the tropicalenvironment, particularly exposure to water, and com-monly yields less than 5% collagen (White & Schwarcz,1989; White et al., 1993). Previous analysis of Mayamaterial has indicated that where collagen yield islower than 1%, original collagen �-values are no longerpreserved (White et al., 1993). The animal bones at thissite have the gross appearance of relatively goodpreservation. The average collagen yield in the Colhasample is low (2·7%, range 0·8 to 6·1%) (Tables 2 & 3)relative to modern bone (DeNiro & Weiner, 1988;Ambrose, 1990). The good physical appearance of thebone is, however, supported by statistical analysis ofthe isotopic and yield data. There is no correlation

between �13C values and yield (Pearsons r= �0·052,df=34). For �15N values versus collagen yield, theonly deer sample (2031-38A-837) with a yield of lessthan 1% creates a significant correlation at P>0·05(Pearsons r=0·364, df=32). However, no correlationexists (Pearson’s r=0·227, df=31) when this outlier isremoved.

As a further check on the chemical integrity ofthe samples, atomic C/N ratios were analysed. Bonecollagen that has a C/N ratio outside the range of 2·9to 3·7 can give anomalous �15N values and somewhatanomalous �13C values (DeNiro, 1985). At Colha, themean of the C/N ratios is 3·35�1·5. There is only onesample (deer 2031-25-3445) that falls slightly outsidethe acceptable range at 3·8 (Tables 2 & 3). Neither its�13C nor �15N values are anomalous, however, and wehave retained it in the data set.

Table 3. Isotopic data for Colha deer by time period

Sample no.time Period Context

Boneelement

�13C(‰)

�15N(‰)

Collagenyield (%) C/N

Early Middle Preclassic (1000–600 BC)2031-38A-837 PM Tibia �21·8 3·4 0·8* 3·42031-38A-857 PM Tibia �19·9 8·8 2·2 3·32031-38A-4225 PM Metatarsal �20·1 4·9 1·6 3·22031-38A-4238 PM Tibia �21·0 4·5 1·8 3·32012-14-2467 OGS Ulna �20·6 4·4 — 3·52012-14-2481 OGS Metatarsal �22·4 4·2 2·0 3·5Mean �21·0 5·0 1·7 3·4.. 1·0 1·9 0·5 0·1

Late Late and Terminal Late Preclassic (100 BC–AD 250)2031-4B-1078 PM Phalanx �21·8 4·2 1·7 3·42031-4B-4393 PM Tibia �19·9 7·8 4·5 3·32031-4C-3836 PM Metacarpal �21·5 3·6 1·4 3·42031-4D-3853 PM Phalanx �21·1 3·9 2·9 3·42031-25-3421 PM Tibia �21·5 5·8 2·1 3·32031-25-3445 PM Vertebra �20·7 5·9 5·3 3·82031-25-3450 PM Radius �20·5 4·9 2·4 3·42031-25-3457 PM Phalanx �21·7 4·4 2·9 3·42031-17-399 P/SM Calcaneus �21·0 4·7 2·2 3·42031-17-404 P/SM Metatarsal �22·6 4·3 1·7 3·4Mean �21·2 5·0 2·7 3·4.. 0·8 1·2 1·3 0·1

PM, primary midden; SM, secondary midden; OGS, on ground surface. *Outlier.

Flora and faunaThe bimodal distribution between C3 and C4 plants inmodern Belizean flora is apparent from Table 1. Thewild plants, which include nuts, fruits, trees, groundplants and roots, all fall into the category of C3 plantswith a mean �13C value of �27·9�1·3‰. Plantscommonly consumed by deer in the wild are also C3,with a mean of �28·8�1·4‰. The somewhat lowerthan average �13C values in this region may be due toa forest canopy effect (van der Merwe & Medina,1991). The C4 plant, maize, has a contrasting �13Cvalue of �11·2‰, which is near the low end of therange for modern maize (�12·0 to �9·6‰) (Tieszen

Isotopic Evidence for Maya Patterns of Deer and Dog Use 99

& Fagre, 1993). With the adjustment of 1·5‰ fordepleted atmospheric carbon since the Industrial Era,it also falls within the range for archaeological maize(�11·0 to �7·9‰) (Tieszen & Fagre, 1993).

The modern wild deer (Odocoileus virginianus) has a�13C value of �24·3‰ (Table 1), which is what onewould expect of exclusive C3 feeding, given an esti-mated 5‰ fractionation for herbivores between dietspecific to deer (�28·8‰) and bone collagen. Otherforest animals (peccary, tapir, kinkajou, and gibnut),and termites who feed on wood, have �13C valuesconsistent with a C3 environment. Ants are herbi-vorous insects that, in this sample, appear to havesome C4 component to their diet. There are no �15Nvalues available for these samples.

Lamanai

–5δ13C (‰, VPDB)

–25

Cuello

Colha

–7–9–11–13–15

n = 3 n = 1

–21 –19 –17

Pacbitun

–23

n = 16

n = 5 n = 1n = 20

n = 6 n = 11 n = 28

n = 16 n = 23

Figure 4. Inter-site comparison of �13C values for dogs (�), deer(�) and humans (�). Time period is not controlled. Bars representone standard deviation. (Data from White & Schwarcz, 1989; Whiteet al., 1993; Tykot et al., 1996.)

3–10

12

–24

δ13C (‰, VPDB)

Odocoileusvirgimianus

Canis familiaris

Canisfamiliaris

Homosapiens

δ15N

(‰, A

IR)

11

10

9

8

7

6

5

4

–22 –20 –18 –16 –14 –12

Figure 5. Isotopic variability in Preclassic dogs and deer from Cuello(�) and Colha (�) compared with isotopic variability in Preclassichumans from Cuello. Special context dogs are included. Bars repre-sent one standard deviation. (Data for Cuello from Tykot et al.,1996.)

Early MiddlePreclassic

12δ15N (‰, AIR)

2

Last MiddlePreclassic

Late Late toTerminalLatePreclassic

1110987

n = 6 n = 4

n = 11

(b)

3 4 5 6

n = 5n = 10

Early MiddlePreclassic

12δ13C (‰, VPDB)

2

Last MiddlePreclassic

Late Late toTerminalLatePreclassic

1110987

n = 6 n = 4

n = 11

(a)

3 4 5 6

n = 5n = 10

Figure 6. Isotopic values of midden dogs (�) and deer (�) forthe Preclassic sequence at Colha. (), Special context dogs. Barsrepresent one standard deviation. (a) �13C values; (b) �15N values.

DogsThe most striking aspect of the �13C data for dogs istheir extreme variability, which indicates diets thatrange from dominantly C3 (sample 2031-39-3208= �22·5‰) to dominantly C4 (samples 2031-51-3023, 2031-34-3671, 2031-13-3603= �9·0 to �8·0‰)(Table 2, Figure 4). This degree of variability alsoappears at the contemporaneous site of Cuellodescribed by Tykot et al. (1996) (Figures 4 & 5), whohave suggested that two of the Cuello dogs may havebeen fattened on maize for sacrifice. Although there isa great deal of variability between sites representingdifferent time periods and geographical locations(White & Schwarcz, 1989; Gerry, 1993; White et al.,1993; Tykot et al., 1996) (Figure 4), the dog samplesin Figure 4 are not sufficiently large or well-dated toestablish sound temporal or spatial patterning. Bycontrast, the Colha dog data are patterned in anarchaeologically meaningful way. Among the dogsfrom the middens, which are taken to represent reposi-tories of ancient dinners, there is a steady enrich-ment of 13C from the Early Middle Preclassic to the

Terminal Late Preclassic periods (Figure 6(a)). In theearly Middle Preclassic period, dogs were consumingsignificant quantities of C3 foods. By the Late MiddlePreclassic period, more C4 foods were present, and bythe end of the Preclassic period dogs were mainlyconsuming C4 foods. The total shift in average isotopicvalues amounts to 4·2‰ and is highly significant

100 C. D. White et al.

statistically (ANOVA F(2,17)=45·40, P<0·001). Fur-thermore, the variability in the �13C values dropsconsiderably (from 3·7‰ in the Late Middle Preclassicto 1·4‰ in the Late Late to Terminal Late Preclassicperiod).

The �15N values of dogs in this sequence com-plement the interpretation of the �13C values. If weassume a 3‰ offset between diet and collagen, therange of �15N values falls within that of terrestrialherbivores (Figure 3), and the �13C and �15N values inthe latter two time periods, especially, could includereef foods such as shellfish (Figure 6). The faunalassemblage contains a strong representation of reefmarine fauna (which humans might also have con-sumed). There is a decrease in �15N values from thebeginning of the sequence (Early Middle Preclassicperiod, 9·7‰) to the end (Late Late to Terminal LatePreclassic period, 8·4‰) (Table 2, Figure 6(b)), thatis statistically significant (ANOVA, F(2,17)=4·94,P<0·05). Dog diets became increasingly herbivorousover time, but they were still more omnivorous and ata higher trophic level than deer which had a mean �15Nvalue of 5·0‰ (see below). Although the sample issmall, the variability in �15N values also appearsto increase over time from 0·4‰ to 1·1 and 0·8‰.This fact suggests that, relative to the Early MiddlePreclassic period when dog diets were consistentlyomnivorous, not all dogs in the following two periodsconsumed plants to the same degree. It is also interest-ing to note that, although the range of �13C values fordogs at Cuello is almost the same as that for Colha, theColha dogs have a much narrower range of �15Nvalues and are at a higher trophic level, perhaps due togreater inclusion of marine and reef resources in theirdiets.

There are three possible explanations for thisphenomenon of increasing mean �13C values (withdecreasing variability) accompanied by decreasingmean �15N values (with increasing variability). Thefirst possibility is that towards the end of the LatePreclassic period, the dogs reflect more purposefulfeeding for consumption in ceremonial feasts, giventhat the structure they come from assumed a moreceremonial function at this time (Anthony & Black,1994). The second explanation is that there may havebeen a change in the nature of human control, wherebydogs were fed or allowed to scavenge increasingamounts of maize over time as humans became moreinvolved in feeding and breeding. Recall that thenumber of dogs in the faunal record at Colha increasestoward the end of the Preclassic period (Shaw, 1995).The final possibility is that the �13C values of dogsrepresent human dietary behaviour during theseperiods, i.e. as dependent scavengers, dogs were con-suming the leftovers of human meals that includedincreasing amounts of maize. Wright (pers. comm.)plans to use an analogous sequence of human data totest this hypothesis. The change in variability at Colhacould indicate that, toward the end of the Preclassic

period, dogs were either being purposefully fed andhad a more controlled diet, or (using the dependentscavenger model) that human diets became moremaize-dependent.

The reason for the patterning in the data can beaddressed, in part, by examining the � values of thedogs found in special contexts in both the Late Middleand Late Late to Terminal Late Preclassic periods.Special contexts are defined here as associations withstructures, in this case on a platform, under a subfloor,and in a collapsed wall. Dogs excavated from thesenon-midden locations are more enriched in 13C thanthe average of their contemporary midden dogs(Figure 6(a)). Of these, the dogs from subfloor and wallcaches were dramatically different and suggest con-sumption of markedly more C4 foods. The theoretical�13C value for a pure maize diet should be �6·0‰ ifwe use an end value of �11·0‰ for archaeologicalcorn and a diet-collagen offset of 5‰ for animals thatare not protein deficient. Although the dogs with �13Cvalues between �8 and �9‰ would not have had apure maize diet, its contribution was significant.

The carbon turnover rate in bone is debated (Libbyet al., 1964; Stenhouse & Baxter, 1979; Tieszen et al.,1983; Kennedy, 1988; Parkington, 1991), but in theabsence of extremely high protein diets (Trilock &Draper, 1989a, 1989b) or a sudden shift to a highprotein diet (Kennedy, 1988), the replacement rate ofcollagen in adult humans is between 10 and 30 years.The nitrogen-isotope ratios of these dogs do not indi-cate extremely high protein intake (see below), butdogs are relatively small animals, and some in thissample may have been young. Because age and size arerelated to metabolism (Klepinger, 1984), their carbon-turnover rates were probably faster than those of adulthumans. The lifespan of most modern dogs is, how-ever, usually under 10 years, the lowest estimatedperiod for carbon turnover (Cannon et al., 1999).Therefore, it is reasonable to assume that their �13Cvalues are reflecting a long-term C4 diet. These cachedogs were probably fed maize for most of their lives,and they might have been kept in captivity to preventthem from consuming other foods. The same twocache dogs exhibited much lower than average �15Nvalues (Figure 6(b)), indicating long-term consumptionof a restricted herbivorous diet. Compared to the dog(2031-37-3365) found on top of the platform, the cachedogs were distinctive in both mortuary and dietarytreatments for their respective time periods. Only onemidden dog (2031-34-3671, from the Late MiddlePreclassic period) has almost identical �13C and �15Nvalues to the cache dogs. We suspect that this dog wasalso purposefully fed.

The higher average �13C and �15N values found inthe midden dogs of the Late Late to Terminal LatePreclassic period could be a result of structure 2031becoming part of the ceremonial precinct if the dogswere scavenging from humans who had more maize intheir diets because of preferred status. Nevertheless, it

Isotopic Evidence for Maya Patterns of Deer and Dog Use 101

is unlikely that the midden dogs as a group were beingpurposefully fed maize to suit the ceremonial purposeof the structure because the only dog (2031-13-3606)found in an unambiguous ritual location has distinc-tively higher �13C and �15N values than its middencontemporaries. This difference between midden andspecial-context dogs is consistent with that found in theprevious time period, and suggests that the selection ofdogs for purposeful feeding and ritual use was fairlyspecific.

Maya mythology may provide an explanationfor such distinctive feeding and mortuary behaviour.The Popol Vuh (‘‘Council Book’’), the Quiche Mayacreation story that probably has its roots in the pre-Columbian past, describes three unsuccessful attemptsat the creation of humans. One of these failures isattributed to an unacceptable feeding relationshipbetween humans and dogs after humans had beencreated out of wood.

‘‘Why is it you can’t seem to give us our food? We justwatch and you just keep us down, and you throw usaround. You keep a stick ready when you eat, just so youcan hit us. We don’t talk, so we’ve received nothing fromyou. How could you not have known? You did know wewere wasting away there, behind you.

So this very day you will taste the teeth in our mouths.We shall eat you’’, their dogs told them, and their faceswere crushed (Tedlock, 1985).

Humans survived as a species only after being createdfrom maize according to the Popul Vuh. The �13Cvalues of the dogs in this study might indicate that the‘‘civilized’’ Maya ensured that their dogs were fed ona respectable human regime as well. Dogs kept bymodern Maya consume an average of six to eighttortillas per day (Benedict & Steggerda, 1936). The factthat some dogs were purposefully and exclusively fedmaize for ritual sacrifice may have a dual significance.This feeding behaviour might have not only denotedrespect for the dog, based on the ancient mythology,but it may also have created a symbolic representationof humans (who were also made of maize). Thesespecially treated dogs may well have been humansubstitutes in sacrifice, as was documented in theethnohistoric record.

The sacrifice of dogs probably relates to anothermyth from the Popul Vuh. As part of the preparationfor the human presence on earth, the Hero twins visitXibalba, the Maya Underworld, and face its rulers, theLords of Xibalba (Lords of Death) (Tedlock, 1985).After successfully meeting challenges presented tothem, the Hero twins are asked to sacrifice a dogbelonging to one of the Lords, and bring him back tolife again. After the dog is brought back to life, theLords of Xibalba ask that they too be sacrificed andbrought back to life. Needless to say, the Hero twinsvanquish the Lords of Death by carrying out onlythe first part of the Lords’ request, and thus theymade way for humans to live on earth. Dogs are,therefore, not only associated with regeneration, but

also with the very existence of human life and its placein nature.

The most likely explanation for the temporalpatterning in �13C values of midden dogs is that,as dependent scavengers, they reflect human dietarybehaviour. Although we do not have human data fromColha to compare to the dog data, humans can becompared to dogs at a few other sites (Figure 4). Notethat the dogs at Cuello, on average, did not have asmany C4 foods in their diets as the humans, and theyexhibited more than twice as much variability in both�13C and �15N values (Figures 4 & 5). This patternwould be consistent with scavenging behaviour, but italso brings into question the validity of using dogs asproxies for humans for dietary reconstruction. At otherlater period Maya sites, dogs have more C4 foods intheir diets than humans, e.g. Lamanai (White &Schwarcz, 1989), Pacbitun (White et al., 1993), and 19dogs from three Classic Maya sites (Gerry & Krueger,1997). The difference in the isotopic compositionbetween dogs and humans is either a result of sampling(where the archaeological context of either the dogs orthe humans is dominantly either ceremonial or lower-status midden), or a change in feeding behaviour. Inthe first instance, the archaeological relationship ofdogs to the rest of the site is not always well-understood, and the number of dog samples analysedis usually dramatically smaller than that for humans.In the second instance, a change in feeding behaviourimplies a shift in the relationship between humans anddogs. This could involve more purposeful feeding withmaize or, as suggested by Tykot et al. (1996), dogsmight also increase the C4 component of their diet byscavenging human faecal material in combination withmaize. It is unlikely that dogs would alter the propor-tions of faeces to other foods unless there was moreavailable of one than the other, e.g. sanitation con-ditions became worse, or there was a scarcity of foodscraps.

There are few carbon or nitrogen isotopic data forhumans available from other Preclassic Maya sites(White & Schwarcz, 1989; Wright, 1994; Tykot et al.,1996; White et al., 1996). In the data that are available,however, there is a considerable amount of vari-ability, which may reflect either regional differences inresources (as is seen in later time periods; Wright &White, 1996), unknown chronological assignment forhumans within the Preclassic period, or both. OnlyCuello provides a Preclassic sequence contempor-aneous with Colha (Tykot et al., 1996). The �13Cvalues for humans at Cuello indicate a dramaticallygreater enrichment in 13C, i.e. they reflect a greatermaize or C4 plant consumption than found for dogsat Colha in the Early Middle Preclassic period(Figure 7(a)). Because Cuello was probably based onan agricultural economy (Hammond, 1991), asopposed to Colha, whose economy was mainly basedon lithic manufacture (Shafer, 1982), this finding is notsurprising. Nevertheless, the C4 component of the

102 C. D. White et al.

human diet at Cuello diminishes slightly towards theend of the Preclassic sequence in contrast to Colhawhere dogs are consuming more maize as time goes by.The pattern of increasing C4-food consumption atColha might be an indirect reflection of intensi-fied maize production resulting from a regionalpopulational explosion at the end of the Preclassicperiod.

There is a corresponding contrast in �15Nvalues between Cuello humans and Colha dogs as well(Figure 7(b)). At Cuello, the mean �15N values ofhumans also decline from 8·8‰ in the Early MiddlePreclassic period to 8·0‰ in the Middle Preclassicperiod, but then they rise in the Late Preclassic periodto 9·3‰. Tykot et al. (1996) have suggested frommultiple lines of evidence (i.e. �15N values, �13C ofbone apatite (which indicates whole diet), and thepattern of age and trauma in the faunal record) thatdogs may have constituted the main source of animalprotein in human diets. The �15N values of dogs atColha are similar to those of humans at Cuello, butthey show a steady decline from the Early Middle toLate Preclassic periods (9·7 to 8·4‰) (Figure 7(b)),indicative of a shift to a more herbivorous diet. Colhahumans may also have consumed dogs. If we assumethat the midden dogs in this sample were dependent

scavengers, then the inverse relationship between �15Nvalues and �13C values in these dogs suggests that theColha humans may have had less meat in their dietstowards the end of the Preclassic period.

Early MiddlePreclassic

12δ15N (‰, AIR)

6

MiddlePreclassic

Last MiddlePreclassic

Late Late toTerminalLatePreclassic

1110987

n = 5 n = 4

n = 3

n = 11

n = 11

n = 5(b)

Early MiddlePreclassic

–8δ13N (‰,VPDB)

–22

MiddlePreclassic

Last MiddlePreclassic

Late Late toTerminalLatePreclassic

–10–14–16–20

n = 5n = 4

n = 4

n = 11

n = 14

n = 5(a)

–12–18

Figure 7. Comparison of isotopic values in the Preclassic sequencefor humans at Cuello (�) with those for dogs at Colha (�). Barsrepresent one standard deviation. (Data for Cuello from Tykot et al.,1996. No data are available for the Late Middle Preclassic at Cuello.)(a) �13C values; (b) �15N values.

DeerBoth �13C and �15N data for deer provide a strikingcontrast with those for dogs (Figures 2, 3 & 5).The �13C values for archaeological deer (Odocoileusvirginianus, or white-tailed deer) are remarkablyuniform (mean= �21·1�1·0‰; Table 3). Althoughthey are more enriched in 13C by about 3·0‰ than theirmodern counterparts, much of this difference can beaccounted for by the Industrial Era correction factor(1·5‰) (Marino & McElroy, 1991). The Colha deermay have consumed very small quantities of maize,possibly immature plants grazed from fields located onforest margins, but their diets are dominantly C3. Theuniformity of the Colha deer carbon-isotope data isconsistent between the two time periods examined aswell. The average �13C values do not change from theEarly Middle Preclassic to the Late Late to TerminalLate Preclassic periods and neither does the variability(Table 3; Figure 7(a), (b)).

These data are consistent with those for deer fromother Lowland Maya sites (White & Schwarcz, 1989;White et al., 1993; Tykot et al., 1996; Gerry & Krueger,1997), where deer have also been interpreted to havebeen wild forest animals. Notably, at Cuello, thereare two species of deer: Odocoileus virginianus andMazama americana. The latter (brocket deer) hasslightly less enriched �13C values. This differencebetween species has also been found at Classic periodPacbitun (White et al., 1993), and Tikal (White et al.,1997), and reflects the previously mentioned differencein feeding behaviour between the two species.

The �15N values of the Colha deer (mean=5·0�1·5‰) are consistent with herbivory and similarto those obtained for a collection of 46 archaeologicaldeer of the same species analysed by Gerry & Krueger(1997) from a variety of Maya lowland sites. Againthere is no apparent difference between time period inthe source of protein consumed by deer, as indicatedby �15N values, but the variability decreases slightlyfrom a standard deviation of 1·9‰ to 1·2‰ in thelater time period (Table 3; Figure 6(b)). Althoughthis difference is not statistically significant usinga Student’s t-test, it hints at the possibility that anarrower range of C3 sources were available to the deerduring the Late Middle Preclassic period. The humanpopulation explosion during this time might havecaused encroachment on deer lands. Emery et al. (inpress) have tested this hypothesis at a number of Petensites using �13C values of deer bone-collagenbut do not find it supported there. Alternately, thephenomenon of reduced variability could be explainedby keeping herds in a restricted forest range. Althoughthe deer from Colha were forest feeders, 13C-enriched

Isotopic Evidence for Maya Patterns of Deer and Dog Use 103

deer have been found at other Maya sites in laterClassic period contexts (White et al., 1993, 1997, inpress). This fact gives us reason to recommend that therestricted forest range hypothesis be tested further,especially given Landa’s (1566) mention of the practiceof training deer to remain in certain places in theforest.

ConclusionDietary dependency can be considered one conditionof domestication. Although domestication involvesmany other factors, reconstructing the diets of animalsis a useful means of understanding relationshipsbetween humans and animals. At Colha, the degree towhich humans controlled the diets of their animalsappears to have varied according to the purposefor maintaining the relationship. A full range of re-lationships involving different degrees of dependency isrepresented in deer and dog samples in this study.Strictly controlled diets consisting of food grownand valued by humans, i.e. maize, are found in thecached dogs used for ritual. There is a basis in Mayamythology for feeding maize to dogs and their sacri-fice. According to the Quiche Maya creation text, theMaya have traditionally believed that: (1) dogs neededto be fed well in order for humans to survive, (2) thesacrifice of a dog allowed humans a place on earth, (3)sacrificed dogs could be brought back to life symbol-izing regeneration, and (4) humans were created frommaize. By feeding dogs maize, the Maya may not onlyhave given dogs the respect they demanded, but alsocreated good symbolic replacements for humans insacrificial rites. Thus, this ritual, mortuary, and feedingbehaviour maintains, justifies, and renews the existenceof humans on earth. The combined archaeologicalcontext and isotopic data substantiate the ethno-historic descriptions of dogs being fattened on maizefor sacrifice and playing a significant role in religiousbelief.

At the other end of the spectrum, there are the dietsof Colha deer, which consist almost entirely of C3plants, such as would be obtained in a wild context.Deer that are 13C-enriched have been found at other,Classic period Maya sites, which could indicate eithergrazing from maize fields or purposeful feeding (Whiteet al., 1993, 1997, in press). The deer that are foundin the Preclassic middens at Colha must have beenhunted, however. The decreased variation found atthe end of the sequence in the Late Late to TerminalLate Preclassic period hints at the possibility ofeither reduced feeding territory caused by humanexpansion or forest husbandry, which is described byan ethnohistoric account from the Colonial period. Weurge further isotopic testing of the deer husbandryhypothesis.

The diets of midden dogs at Preclassic Colha changeover time to include increasing amounts of maize and

different protein sources. This pattern may be a reflec-tion of human diets, particularly the greater consump-tion of maize. At other Maya sites where data for bothdogs and humans are available, dogs appear to bedependent scavengers. Nevertheless, because of therelationship between mortuary treatment and diet ofdogs at Colha, we suggest caution in the use of dogsfrom non-midden contexts to infer human diets.

Dogs become more herbivorous from the EarlyMiddle Preclassic period, to the Late Late to TerminalLate Preclassic period, but the amount of C4 foods intheir diets also increases dramatically. Dogs alsoexhibit the highest degree of homogeneity towards theend of the Preclassic sequence. These data could indi-cate either greater human control of dog diets, or areduction in the diversity of resources used by humans.As such, the dog diets could be providing indirectevidence for a change in the use of the structure inwhich they were found (i.e. to more ritual use ofmaize), or a general subsistence shift related to the needto feed an expanding population at the end of thePreclassic period (i.e. a rise in intensity of agricultureor increased importation of maize). Although wefavour the latter hypothesis because of the clear iso-topic distinction between the cache dog and the rest ofmidden dogs from the Late Late to Terminal LatePreclassic period, a controlled analysis of humans willbe needed to test these hypotheses.

AcknowledgementsWe would like to thank Leslie Shaw, Bowdoin College,for providing the bones. We also thank John Pohlfor information on dog symbolism in Mesoamerica.Martin Knyf, McMaster University, and Kim Law, theUniversity of Western Ontario, provided us with thetechnical assistance including graphics. This researchwas funded by an SSHRC grant to Henry Schwarczand Christine White, and an SSHRC grant to ChristineWhite, Fred Longstaffe and Michael Spence.

ReferencesAdams, R. E. W. & Valdez, F. Jr (1980). The ceramic sequence of

Colha, Belize: 1979 and 1980 seasons. In (T. R. Hester, J. D. Eaton& H. J. Shafer, Eds) The Colha Project. Second Season, 1980,Interim Report. San Antonio and Venice: Center for Archaeologi-cal Research, University of Texas and Centro Studie RichercheLigabue, pp. 15–40.

Allen, G. M. (1920). Dogs of the American Aborigines. Bulletin ofthe Museum of Comparative Zoology, Harvard University 63,431–517.

Ambrose, S. H. (1990). Preparation and characterization of boneand tooth collagen for isotopic analysis. Journal of ArchaeologicalScience 17, 431–451.

Ambrose, S. H. (1993). Isotopic analysis of paleodiets: methodologi-cal and interpretive considerations. In (M.K. Sandford, Ed.)Investigations of Ancient Human Tissue. Langhorne, Pennsylvania:Gordon and Breach Science Publishers.

Ambrose, S. H. & Norr, L. (1993). Experimental evidence for therelationship of the carbon isotope ratios of whole diet and dietary

104 C. D. White et al.

protein to those of bone collagen and carbonate. In (J. Lambert &G. Grupe, Eds) Prehistoric Human Bone: Archaeology at theMolecular Level. Berlin: Springer-Verlag, pp. 1–37.

Andrews, E. W. V. & Hammond, N. (1990). Redefinition of theSwasey Phase at Cuello, Belize. American Antiquity 54, 570–584.

Anthony, D. & Black, S. (1994). Excavations at Operation 2031. In(T. R. Hester, H. J. Shafer & J. D. Eaton, Eds) Continu-ing Archaeology at Colha, Belize. Studies in Archaeology,No. 16. University of Texas at Austin: Archaeological ResearchLaboratory.

Benedict, F. G. & Steggerda, M. (1936). The Food of the Present-dayMaya Indians Yucatan. Washington, DC: Carnegie InstituteContributions to American Archaeology, No. 18.

Bokonyi, S. (1969). Archaeological problems and methods ofrecognizing animal domestication. In (P. J. Ucko & G. W.Dimbleby, Eds) The Domestication and Exploitation of Plants andAnimals. Chicago: Aldine, pp. 219–229.

Brown, T. A., Nelson, D. E., Vogel, J. S. & Southern, J. R. (1988).Improved collagen extraction by modified Longin method.Radiocarbon 30, 171–177.

Burleigh, R. & Brothwell, D. (1978). Studies on Amerindian Dogs, 1:Carbon isotopes in relation to maize in the diet of domestic dogsfrom early Peru and Equador. Journal of Archaeological Science 5,335–362.

Cannon, A., Schwarcz, H. P. & Knyf, M. (1999). Marine-basedsubsistence trends and the stable isotope analysis of dog bonesfrom Namu, British Columbia. Journal of Archaeological Science26, 399–407.

Carr, H. S. (1985). Subsistence and ceremony: faunal utilization in aLate Preclassic community at Cerros, Belize. In (M. Pohl, Ed.)Prehistoric Lowland Maya Environment and Subsistence Economy:The Data Base and Its Problems. Papers of the Peabody Museum,Vol. 77. Cambridge, MA: Harvard University Press, pp. 115–132.

Carr, H. S. (1986). Faunal utilization in a Late Preclassic communityat Cerros, Belize. Ph.D. Thesis. Tulane University.

Carr, H. S. (1996). Precolumbian Maya exploitation and manage-ment of deer populations. In (S. L. Fedick, Ed.) The ManagedMosaic: Ancient Maya Agriculture and Resource Populations. SaltLake City: University of Utah Press, pp. 251–261.

Chaplin, R. E. (1969). The use of non-metric criteria in the study ofanimal domestication from bones found on archaeological sites. In(P. J. Ucko & G. W. Dimbleby, Eds) The Domestication andExploitation of Plants and Animals. Chicago: Aldine, pp. 231–245.

Chisholm, B. S. (1989). Variation in diet reconstructions based onstable isotope evidence. In (T. D. Price, Ed.) The Chemistryof Prehistoric Bone. Cambridge: Cambridge University Press,pp. 10–37.

Chisholm, B. S., Nelson, D. E., Hobson, K. A., Schwarcz, H. P.& Knyf, M. (1983). Carbon isotope measurement techniquesfor bone collagen: notes for the archaeologist. Journal ofArchaeological Science 10, 355–360.

Clutton-Brock, J. & Hammond, N. (1994). Hot dogs: comestiblecanids in Preclassic Maya culture at Cuello, Belize. Journal ofArchaeological Science 21, 819–826.

Clutton-Brock, J. & Noe-Nygaard, N. (1990). New osteological andC-isotope evidence on Mesolithic dogs: companions to huntersand fishers at Star Carr, Seamer Carr and Kongemose. Journal ofArchaeological Science 17, 643–653.

Coe, M. D. (1982). Old God and Young Heroes: The PearlmanCollection of Maya Ceramics. Jerusalem: Israel Museum.

Coe, S. D. & Coe, M. D. (1996). The True History of Chocolate.New York: Thames and Hudson.

Coplen, T. B. (1996). New guidelines for reporting stable hydro-gen, carbon, and oxygen isotope-ratio data. Geochimica etCosmochimica Acta 60, 3359–3360.

Coplen, T. B., Krouse, H. R. & Bohlke, J. K. (1992). Reporting ofnitrogen-isotope abundances. Pure and Applied Chemistry 64,907–908.

Cormie, A. B. & Schwarcz, H. P. (1994). Stable isotopes of nitrogenand carbon of North American white-tailed deer and implicationsfor paleodietary and other food web studies. Palaeogeography,Palaeolimatology, Palaeoecology 107, 227–241.

Coyston, S. (1994). An application of carbon isotopic analysis of boneapatite to the study of Maya diets and subsistence at Pacbitun andLamanai, Belize. M.A. Thesis. Trent University.

Crabtree, P. J. & Campana, D. (Eds) (1989). Early AnimalDomestication and its Cultural Context. MASCA Research Papersin Science and Archaeology 6, The Museum of Archaeology andAnthropology. Philadelphia: University of Pennsylvania.

Crandall, L. S. (1964). The Management of Wild Animals inCaptivity. Chicago: University of Chicago Press.

Crane, C. (1986). Late Preclassic Maya agriculture, wild plantutilization and land-use practices. In (D. A. Freidel, Ed.)Archaeology at Cerros, Belize, Central America. Austin: Universityof Texas Press, pp. 51–62.

Cranstone, B. A. L. (1969). Animal husbandry: the evidence fromethnography. In (P. J. Ucko & G. W. Dimbleby, Eds) TheDomestication and Exploitation of Plants and Animals. Chicago:Aldine, pp. 247–264.

Danien, E. (1997). The ritual on the Ratinlixul Vase: pots andpolitics in Highland Guatemala. Expedition 39, 37–48.

Davis, S. J. (1987). The Archaeology of Animals. New Haven: YaleUniversity Press.

DeNiro, M. J. (1985). Post-mortem preservation and alteration of‘‘in vivo’’ bone collagen ratios: implications for paleodietaryanalysis. Nature 317, 806–809.

DeNiro, M. J. & Epstein, S. (1981). Influence of diet on thedistribution of nitrogen isotopes in animals. Geochimica etCosmochimica Acta 45, 341–351.

DeNiro, M. J. & Weiner, S. (1988). Chemical, enzymatic andspectroscopic characterization of ‘‘collagen’’ and other organicfractions in prehistoric bone. Geochimica et Cosmochimica Acta52, 2197–2206.

DeNiro, M. J., Schoeninger, M. J. & Hastorf, C. (1985). Effect ofheating on the stable carbon and nitrogen isotope ratios of bonecollagen. Journal of Archaeological Science 10, 199–203.

DiPeso, C. C. (1974). Casas Grandes: A Fallen Trading Center of theGran Chichimeca, Vol. 2. Dragoon, New Mexico: The AmerindFoundation Inc.

Downs, J. F. (1960). Domestication: an examination of thechanging social relationship between man and animals. KroeberAnthropological Society Papers 22, 18–67.

Duran, D. (1971). Book of the Gods and Rites and the AncientCalendar (F. Horcasitas & D. Heyden, translators). Norman:University of Oklahoma Press.

Emery, K. F. (1997). The Maya collapse: a zooarchaeological inquiry.Ph.D. Thesis. Cornell University.

Emery, K. F., Schwarcz, H. P. & Wright, L. E. (in press). Isotopicanalysis of ancient deer bone: biotic stability in Collapse PeriodMaya land-use. Journal of Archaeological Science.

Flannery, K. V. (1967). Vertebrate fauna and hunting patterns. In(D. S. Byers, Ed.) The Prehistory of the Tehuacan Valley, Vol. 1.Austin: University of Texas Press, pp. 132–177.

Freidel, D., Schele, L. & Parker, J. (1993). Maya Cosmos: ThreeThousand Years on the Shaman’s Path. New York: William andMorrow.

Friedli, H., Lotscher, H., Oeschger, H., Siegenthaler, U. & Stauffer,B. (1986). Ice core record of the 13C/12C ratio of atmospheric CO2

in the past two centuries. Nature 324, 237–238.Gerry, J. P. (1993). Diet and status among the Classic Maya: an

isotopic perspective. Ph.D. Thesis. Harvard University.Gerry, J. P. & Krueger, H. W. (1997). Regional diversity in Classic

Maya diets. In (S. Whittington & D. Reed, Eds) Bones ofthe Maya: Studies of Ancient Skeletons. Washington, DC:Smithsonian Institution Press, pp. 196–207.

Gode, D. (1976). The guinea pig in Andean folk culture.Geographical Review 57, 213–224.

Grube, N. & Nahm, W. (1994). A census of Xibalba: a completeinventory of way characters on Maya ceramics. In (J. Kerr, Ed.)The Maya Vase Book. New York: Kerr Associates, pp. 686–712.

Hamblin, N. (1984). Animal Use by the Cozumel Maya. Tucson: TheUniversity of Arizona Press.

Hammond, N. (Ed.) (1991). Cuello: An Early Maya Community inBelize. Cambridge: Cambridge University Press.

Isotopic Evidence for Maya Patterns of Deer and Dog Use 105

Healy, P. F., Emery, K. & Wright, L. E. (1990). Ancient and modernexploitation of the Jute snail (Pachychilus). Latin AmericanAntiquity 1, 170–183.

Hellmuth, N. M. (1977). Cholti-Lacandon (Chiapas) and Peten-Itzaagriculture. In (N. Hammond, Ed.) Social Process in MayaPrehistory. New York: Academic Press, pp. 421–428.

Hellmuth, N. M. (1996). Headdresses and Skirts Shared by DeerHunters and Ballplayers. Osaka, Japan: National Museum ofEthnology.

Hodell, D. A., Curtis, J. H. & Brenner, M. (1995). Possible role ofclimate in the collapse of Classic Maya civilization. Nature 375,391–394.

Isaac, E. (1970). Geography of Domestication. Englewood Cliffs:Prentice-Hall.

Katzenberg, M. A. (1989). Stable isotope analysis of archaeologicalfaunal remains from Southern Ontario. Journal of ArchaeologicalScience 16, 319–329.

Keegan, W. F. & DeNiro, M. J. (1988). Stable carbon- and nitrogen-isotope ratios of bone collagen used to study coral-reef andterrestrial components of prehistoric Bahamian diet. AmericanAntiquity 53, 320–336.

Kennedy, B. V. E. (1988). Variation in �13C values of post-Medievaleuropeans. Ph.D. Thesis. University of Calgary.

Kidder, A. V., Jennings, J. D. & Shook, E. M. (1946). Excavations atKaminaljuju, Guatemala. Carnegie Institute Publication No. 561.Washington: Carnegie Institute.

Klepinger, L. L. (1984). Nutritional assessment from bone. AnnualReview of Anthropology 13, 75–96.

Krueger, H. W. & Sullivan, C. H. (1984). Models for carbonisotope fractionation between diet and bone. In (J. E. Turnland &P. E. Johnson, Eds) Stable Isotopes in Nutrition. AmericanChemical Society Symposium Series No. 258. Washington, D.C.,pp. 205–220.

Land, L. S., Lundelius, E. L. Jr & Valastro, S. Jr (1980). Isotopicecology of deer bones. Palaeogeography, Palaeoclimatology,Palaeoecology 32, 143–151.

Lee-Thorp, J. A., Sealy, J. & van der Merwe, N. J. (1989). Stablecarbon isotope ratio differences between bone collagen and boneapatite, and their relationship to diet. Journal of ArchaeologicalScience 16, 585–599.

Lejon, D. (1625). Bocabulario de Mayathan por su Abeseario (ViennaDictionary). Manuscript in the National Library in Vienna,Austria. From microfilm in the Library of Pomona College,Claremont, California.

Lewis, R. D. (1970). Animal domestication in Guatemala. M.A.Thesis. University of Oregon.

Libby, W. F., Berger, R., Mead, J. F., Alexander, G. V. & Ross,J. F. (1964). Replacement rates for human tissue from atmosphericcarbon. Science 146, 1170–1172.

Longin, R. (1971). A new method of collagen extraction forradiocarbon dating. Science 230, 241–242.

Lopez de Cogolludo, D. (1688). Los Tres Siglos de la DominacionEspanola en Yucatan o sea Historia de esta Provincia, 2 Vols. Graz,Austria: Akademische Druk und Verlaganstalt (1971).

Marcus, J. (1982). The plant world of the 16th and 17th centuryMaya. In (K. Flannery, Ed.) Maya Subsistence. New York:Academic Press, pp. 239–273.

Marino, B. D. & McElroy, M. B. (1991). Isotopic composition ofatmospheric CO2 inferred from carbon in C4 plant cellulose.Nature 349, 127–131.

McAnany, P. A. (1986). Lithic technology and exchange amongwetland farmers of the Eastern Maya lowlands. Ph.D. Thesis.University of New Mexico.

Means, P. A. (1917). History of the Spanish Conquest of Yucatan andof the Itzas. Papers of the Peabody Museum, Vol. 7. Cambridge,MA: Harvard University.

Merino Carrion, B. L. & Garcıa Cook, A. (1997). Enterramientos deFormative en el noreste de Mexico. In (A. Garcıa Cook, A. G.Mastache, L. Merino Carrion & S. Rivero Torres, Eds) Homenajeal Profesor Cesar A Saenz. Mexico: Instituto Nacional deAntropologıa e Historia.

Michelon, O. (Ed.) (1976). Dictionario de San Francisco. BibliotecaLinguistica Americana, Vol. II. Graz: Akademische Druck-uVerlagenstalt.

Miksicek, C. H., Bird, R. M., Pickersgill, B., Donaghey, S.,Cartwright, J. & Hammond, N. (1981). Preclassic lowland maizefrom Cuello, Belize. Nature 289, 56–59.

Moholy-Nagy, H. (1978). The utilization of pomacea snails at Tikal,Guatemala. American Antiquity 43, 65–73.

Morales, E. (1995). The Guinea Pig: Healing, Food and Ritual in theAndes. Tucson: University of Arizona Press.

Nash, M. A. (1982). An analysis of a debitage collection from Colha,Belize. In (T. R. Hester, J. D. Eaton & H. J. Shafer, Eds) TheColha Project: Second Season, 1980 Interim Report. Center forArchaeological Research: University of Texas at San Antonio,pp. 333–352.

Nicholson, H. B. (1971). Religion in Prehispanic Central Mexico. In(G. Ekholm & F. Bernal, Eds) Handbook of Middle AmericanIndians, Vol. 10. Archaeology of Northern Mesoamerica, Part 1.Austin: University of Texas Press, pp. 395–446.

Noe-Nygaard, N. (1988). �13C values of dog bones reveal the natureof changes in man’s food resources at the Mesolithic-Neolithistransition, Denmark. Chemical Geology 73, 87–96.

Norr, L. C. (1991). Nutritional consequences of prehistoric subsistencestrategies in Lower Central America. Ph.D. Thesis. University ofUrbana-Champaign.

Olsen, S. J. (1978). Vertebrate faunal remains. In (G. R. Willey, Ed.)Excavations at Seibal, Department of Peten, Guatemala. Papers ofthe Peabody Museum, Vol. 64, No. 1. Cambridge, MA: HarvardUniversity, pp. 172–176.

O’Leary, M. H. (1988). Carbon isotopes in photosynthesis.Bioscience 38, 328–336.

Parkington, J. (1991). Approaches to dietary construction in theWestern Cape: are you what you have eaten? Journal ofArchaeological Science 18, 331–342.

Pate, F. D. (1994). Bone chemistry and paleodiet. Journal ofArchaeological Method and Theory 1, 161–209.

Pendergast, D. R. (1969). The Prehistory of Actun Balam, BritishHonduras. Occasional Paper 16. Toronto: Royal Ontario Museum.

Pendergast, D. R. (1974). Excavations at Actun Polbiche. Toronto:Royal Ontario Museum.

Pohl, M. D. (1977). Hunting in the Maya village of San Antonio, RioHondo, Orange Walk District, Belize. Journal of Belizean Affairs5, 52–63.

Pohl, M. D. (1981). Ritual continuity and transformation inMesoamerica: reconstructing the ancient Maya cuch ritual.American Antiquity 46, 513–529.

Pohl, M. D. (1983). Maya ritual faunas: vertebrate remains fromburial, caches, caves and cenotes in the Maya lowlands. In (R. M.Leventhal & A. L. Kolata, Eds) Civilization in the AncientAmericas. Albuquerque: University of New Mexico Press,pp. 55–103.

Pohl, M. D. (1985). The privileges of Maya elites: Late Classicvertebrate fauna from Seibal, Guatemala. In (M. Pohl, Ed.)Prehistoric Lowland Maya Environment and Subsistence Economy:The Data Base and its Problems. Papers of the Peabody Museum,Vol. 77. Cambridge, MA: Harvard University, pp. 133–145.

Pohl, M. D. (1990). The ethnozoology of the Maya: faunal remainsfrom five sites in Peten, Guatemala. In (G. R. Willey, Ed.)Excavations at Seibal. Peabody Museum Memoirs, Vol. 18, No. 3.Cambridge, MA: Harvard University, pp. 144–174.

Pohl, M. D. (1994). Mexican codices: maps and liengas as socialcontrasts. In (E. Hill Boone & W. D. Mignolo, Eds) AlternativeLiteracies in Mesoamerica and the Andes. Durham: DukeUniversity Press, pp. 137–160.

Pohl, M. D. & Feldman, L. H. (1982). The traditional role of womenand animals in Lowland Maya economy. In (K. Flanner, Ed.)Maya Subsistence. New York: Academic Press, pp. 295–312.

Pohl, M. D., Pope, K. O., Jones, J. G., Jacob, J. S., Piperno, D. R.,deFrance, S. D., Lentz, D. L., Gifford, J. A., Danforth, M. E. &Josserand, J. K. (1996). Early agriculture in the Maya Lowlands.Latin American Antiquity 7, 355–372.

106 C. D. White et al.

Pollock, H. E. D. & Ray, C. E. (1957). Notes on Vertebrate Remainsfrom Mayapan. Carnegie Institute Report No. 41. Washington:Carnegie Institute.

Potter, D. R. (1982). Some results of the second year of excavationsat operation 2012. In (T. R. Hester, H. J. Shafer & J. D. Eaton,Eds) Archaeology at Colha, Belize: The 1981 Interim Report.Center for Archaeological Research: University of Texas at SanAntonio, pp. 98–122.

Powis, T., Stanchly, N., Lentz, D., White, C., Awe, J., Healy, P. F.& Longstaffe, F. (1999). A reconstruction of Middle Formativediet and subsistence from the periphery of Cahal Pech, Belize.Antiquity 73, 364–376.

Reed, D. M. (1994). Ancient Maya diet at Copan, Honduras,as determined through the analysis of stable carbon andnitrogen isotopes. In (K. D. Sobolik, Ed.) Paleonutrition: The Dietand Health of Prehistoric Americans. Carbondale: Center forArchaeological Investigations, Southern Illinois University,pp. 210–221.

Reina, R. E. (1967). Milpas and milperos: implications forprehistoric times. American Anthropologist 69, 1–20.

Roys, R. L. (1931). The Ethnobotany of the Maya. Middle AmericanResearch Institute Publication 2, New Orleans.

Roys, R. L. (1943). The Indian Background of Colonial Yucatan.Carnegie Institution of Washington, Publication 548. Washington,DC.

Rue, D. J. (1987). Early agriculture and early Postclassic Mayaoccupation in western Honduras. Nature 326, 285–286.

Schele, L. & Miller, M. E. (1986). The Blood of Kings: Dynasty andRitual in Maya Art. Fort Worth: Kimbell Art Museum.

Schoeninger, M. J. (1985). Trophic level effects on 15N/14N and13C/12C ratios in bone collagen and strontium levels in bonemineral. Journal of Human Evolution 14, 515–525.

Schwarcz, H. P. (1999). Some biochemical aspects of carbon isotopicpaleodiet studies. In (S. Ambrose & M. A. Katzenberg, Eds)Isotopic Aspects of Paleodiet. New York: Kluwer (in press).

Schwarcz, H. P. & Schoeninger, M. J. (1991). Stable isotope analysesin human nutritional ecology. Yearbook of Physical Anthropology34, 283–321.

Schwartz, M. (1997). A History of Dogs in the Early Americas.New Haven: Yale University Press.

Severinghaus, C. W. & Cheatum, E. L. (1956). Life and times of thewhite-tailed deer. In (W. P. Taylor, Ed.) The Deer of NorthAmerica. Harrisburg, Pennsylvania: The Stackpole Co.

Shafer, H. J. (1982). A preliminary report on the lithic technology atKichpanha, Northern Belize. In (T. R. Hester, H. J. Shafer & J. D.Eaton, Eds) Archaeology at Colha, Belize: The 1981 InterimReport. Center for Archaeological Research: University of Texasat San Antonio, pp. 167–181.

Shaw, L. (1991). The articulation of social inequality and faunalresource use in the Preclassic community of Colha, Northern Belize.Ph.D. Thesis. University of Massachusetts-Amherst.

Shaw, L. (1995). The importance of dog in ritual feasting in the MayaPreclassic. Paper presented at the 60th Annual Meeting of theSociety for American Archaeology, Minneapolis.

Smith, B. N. & Epstein, S. (1971). Two categories of 13C/12C ratiosfor higher plants. Plant Physiology 47, 380–384.

Stanchly, N. (1995). Formative period Maya faunal utilization atCahal Pech, Belize. In (P. F. Healy & J. J. Awe, Eds) Belize ValleyPreclassic Maya Project: Report on the 1994 Field Season.Occasional Papers in Anthropology, No. 10, Trent University,pp. 124–149.

Stenhouse, M. J. & Baxter, M. S. (1979). The uptake of bomb 14C inhumans. In (R. Berger & H. Suess, Eds) Radiocarbon Dating.Berkeley: The University of California Press, pp. 324–341.

Taber, R. D. & Dasmann, R. F. (1957). The dynamics of threenatural populations of deer Odocoileus hemionus columbianus.Ecology 38, 233–246.

Tedlock, D. (1985). Popol Vuh. The Definitive Edition of the MayanBook of the Dawn of Life and the Glories of Gods and Kings.New York: Simon and Schuster.

Tennessen, T. & Hudson, R. J. (1981). Traits relevant to thedomestication of herbivores. Applied Animal Ethology 7, 87–102.

Thompson, J. E. S. (1972). A Commentary on the Dresden Codex.Philadelphia: American Philosophical Society.

Tieszen, L. L. & Fagre, T. (1993). Carbon isotopic variability inmodern and archaeological maize. Journal of ArchaeologicalScience 20, 25–40.

Tieszen, L. L., Boutton, T. W., Tesdahl, K. G. & Slade, N. A. (1983).Fractionation and turnover of stable carbon isotopes in animaltissues: implications for the �13C analysis of diet. Oecologia 57,32–37.

Tourtellot, G. (1990). Excavations at Seibal: Burials. A CulturalAnalysis. Memoirs of the Peabody Museum of Archaeology andEthnology, Vol. 17, No. 2. Cambridge, MA: Harvard UniversityPress.

Tozzer, A. M. (Ed.) (1941). Landa’s Relacion de las Cosas deYucatan: A Translation. Papers of the Peabody Museum, Vol. 18.Cambridge, MA: Harvard University.

Tozzer, A. M. & Allen, G. M. (1910). Animal Figures in theMaya Codices. Papers of the Peabody Museum, Vol. 4, No. 3.Cambridge, MA: Harvard University.

Trik, A. S. (1963). The splendid tomb of Temple I at Tikal,Guatemala. Expedition 6, 2–18.

Trilok, G. & Draper, H. H. (1989a). Effect of a high protein dieton acid-base balance in adult rats. Calcified Tissue ResearchInternational 44, 339–342.

Trilok, G. & Draper, H. H. (1989b). Sources of a protein-inducedendogenous acid production and excretion by human adults.Calcified Tissue Research International 44, 335–338.

Turner, B. L. II & Harrison, P. D. (1983). Pulltrouser Swamp:Ancient Maya Habitat, Agriculture and Settlement in NorthernBelize. Austin: University of Texas Press.

Turner, G. L., Bergerson, F. J. & Tantala, H. (1983). Naturalenrichment of 15N during decomposition of plant material in soils.Soil Biology and Biochemistry 15, 495–497.

Tykot, R. H., van der Merwe, N. J. & Hammond, N. (1996). Stableisotope analysis of bone collagen, bone apatite, and tooth enamelin the reconstruction of human diet: a case study from Cuello,Belize. In (M. V. Orna, Ed.) Archaeological Chemistry V.Washington, DC: American Chemical Society, pp. 355–365.

Tyler, H. A. (1991). Pueblo Birds and Myths. Norman: University ofOklahoma Press.

Ucko, P. J. & Dimbleby, G. W. (Eds) (1969). The Domestication andExploitation of Plants and Animals. London: Duckworth.

Valdez, F. Jr (1987). The Prehistoric ceramics of Colha, NorthernBelize. Ph.D. Thesis. Harvard University.

Valdez, F. Jr (1992). Late Preclassic architecture and ritual at RioAzul, Guatemala. Paper presented at the 57th Annual Meeting ofthe Society for American Archaeology, New Orleans.

Valdez, F. Jr & Buttles, P. (1991). G-103: a late Preclassic monu-mental structure at Rıo Azul, Guatemala. Paper presented at the56th Annual Meeting of the Society for American Archaeology,Pittsburg.

van der Merwe, N. J. & Medina, E. (1991). The canopy effect,carbon isotope ratios and foodwebs in Amazonia. Journal ofArchaeological Science 18, 249–259.

van der Merwe, N. J., Tykot, R. H. & Hammond, N. (1999). Dietand animal husbandry of the Preclassic Maya at Cuello, Belize. In(S. Ambrose & M. A. Katzenberg, Eds) Isotopic Aspects ofPaleodiet. New York: Kluwer, in press.

Villagutierre Soto-Major, J. de. (1701). Historia de las conquista de laprovincia de el Itza. Guatemala City: Biblioteca ‘‘Goathemala’’ dela Sociedad de Geografia et Historia, Vol. 9 (1933).

Virginia, R. A. & Delwiche, C. C. (1982). Natural 15N abundance ofpresumed N2-fixing and non-N2-fixing plants from selected ecosys-tems. Oecologia 54, 317–325.

White, C. D. (1991). Isotopic analysis of multiple human tissues fromthree ancient Nubian populations. Ph.D. Thesis. University ofToronto.

White, C. D. (1997). Ancient diet at Lamanai and Pacbitun: impli-cations for the ecological model of collapse. In (S. Whittington &D. Reed, Eds) Bones of the Ancestors: Recent Studies of AncientMaya Skeletons. Washington, DC: Smithsonian Institution Press,pp. 171–180.

Isotopic Evidence for Maya Patterns of Deer and Dog Use 107

White, C. D. & Schwarcz, H. P. (1989). Ancient Maya diet: asinferred from isotopic and chemical analysis of human bone.Journal of Archaeological Science 16, 451–474.

White, C. D., Healy, P. F. & Schwarcz, H. P. (1993). Intensiveagriculture, social status and Maya diet at Pacbitun, Belize.Journal of Anthropological Research 49, 347–375.

White, C. D., Longstaffe, F. & Song, R. (1996). Preclassic Maya dietat Cahal Pech: the isotopic evidence. Paper presented at the61st Annual Meeting of the Society for American Archaeology,New Orleans.

White, C. D., Pohl, M., Schwarcz, H. P. & Longstaffe, F. J. (1997).Maya husbandry of deer and dog. Paper presented at the 62ndAnnual Meeting of the Society for American Archaeology,St Louis.

White, C. D., Pohl, M., Schwarcz, H. P. & Longstaffe, F. J. (inpress). Feast, field and forest: deer and dog diets at Lagartero andCopan. In (K. Emery & N. Stanchly, Eds) Maya Zooarchaeology:New Directions in Method and Theory. Los Angeles: UCLAInstitute of Archaeology.

Wilk, R. R. (1975). Superficial examination of Structure 100, Colha.In (N. Hammond, Ed.) Archaeology of Northern Belize: BritishMuseum–Cambridge University Corozal Project 1974–1975 InterimReport. Cambridge: Cambridge University Centre of LatinAmerican Studies, pp. 152–173.

Wilkinson, P. F. (1972). Current experimental domestication and itsrelevance to prehistory. In (E. S. Higgs, Ed.) Papers in EconomicPrehistory. Cambridge: Cambridge University Press.

Wing, E. S. (1978). Use of dogs for food: an adaptation to the coastalenvironment. In (B. L. Stark & B. Voorhies, Eds) PrehistoricCoastal Adaptation: The Economy and Ecology of Maritime MiddleAmerica. New York: Academic Press, pp. 29–41.

Wing, E. S. (1981). A comparison of Olmec and Maya foodways. In(M. D. Coe, D. Grove & E. P. Benson, Eds) The Olmec and Their

Neighbors: Essays in Memory of Mathew W. Stirling. Washington,DC: Dumbarton Oaks, pp. 21–28.

Wing, E. S. & Scudder, S. J. (1991). The exploitation of animals. In(N. Hammond, Ed.) Cuello: An Early Maya Community in Belize.Cambridge: Cambridge University Press, pp. 84–97.

Wing, E. S. & Steadman, D. (1980). Vertebrate faunal remainsfrom Dzibilchaltun, Yucatan. In (E. W. Andrews IV & E. W.Andrews V, Eds) Excavations at Dzibilchaltun, Yucatan. MiddleAmerican Research Institute Publication 48. New Orleans: TulaneUniversity, pp. 326–331.

Wing, E. S. & Wing, S. R. (1995). Prehistoric ceramic age adaptationto varying diversity of animal resources along the West Indianarchipelago. Journal of Ethnobiology 15, 119–148.

Wright, L. E. (1994). The sacrifice of the Earth? Diet, health, andinequality in the Pasion Maya Lowlands. Ph.D. Thesis. Universityof Chicago.

Wright, L. E. (1997). Ecology or society? Paleodiet and the collapseof the Pasion Maya Lowlands. In (S. Whittington & D. Reed, Eds)Bones of the Ancestors: Recent Studies of Ancient Maya Skeletons.Washington, DC: Smithsonian Institution Press, pp. 181–195.

Wright, L. E. & Schwarcz, H. P. (1996). Infrared and isotopicevidence for diagenesis of bone apatite at Dos Pilas, Guatemala:paleodietary implications. Journal of Archaeological Science 23,933–944.

Wright, L. E. & White, C. D. (1996). Human biology in the ClassicMaya Collapse: evidence from paleopathology and paleodiet.Journal of World Prehistory 10, 147–198.

Ximenez, F. (1967). Historia natural del reino de Guatemala.Sociedad de Geografia e Historia de Guatemala, Special Publication14.

Zeuner, F. E. (1963). A History of Domesticated Animals. New York:Harper and Row.