STABLE ISOTOPE ANALYSIS: A TOOL FOR STUDYING PAST DIET, DEMOGRAPHY, AND … · 2016-09-28 · direct evidence of past diets, including infant diets. Knowledge gained from bone chemistry

CHAPTER 13

STABLE ISOTOPE ANALYSIS: A TOOLFOR STUDYING PAST DIET,DEMOGRAPHY, AND LIFE HISTORY

M. ANNE KATZENBERG

INTRODUCTION

This chapter is about chemical variation in thevarious components that make up bones andteeth and its application to studies of diet, demo-graphy, and life history. Bones and teeth providedirect evidence of past diets, including infantdiets. Knowledge gained from bone chemistryrelates to other sources of evidence for dietand, in turn, the interaction of diet, nutrition,and disease. Understanding infant diets andthe duration of nursing relates to demographicvariables such as birth spacing and populationgrowth. Chemical variation in bones andteeth can also be linked to chemical variationin the environment and, thus, reveals infor-mation about place of residence and migration.Although stable isotope analysis may beviewed as a fairly technical research specialty,the results of such analyses make a significantcontribution to the reconstruction of pasthuman life.

Uses of Stable Isotope Analysisin Skeletal Biology

The routine use of stable carbon isotopes incurrent skeletal studies is very different from

the excitement generated by the firstapplications of stable carbon isotopes tohuman paleodiet reconstruction. The idea thatone could determine whether prehistoricpeoples of North America consumed corn(maize) by performing chemical tests on theirbones seemed like science fiction to mostarchaeologists and physical anthropologists inthe mid-1970s when it was first attempted(Vogel and van der Merwe, 1977). Today,stable carbon isotope analysis is part of a suiteof technical specialties performed on remainsfrom archaeological sites. In addition to study-ing stable carbon and nitrogen isotopes inpreserved protein, it is now possible to studystable isotopes of carbon, oxygen, and strontiumfrom the mineral portion of bones and teeth.This chapter provides some background to theuse of stable isotopes in bioarchaeologicalstudies. It includes technical information onhow such analyses are performed, with asampling of applications, problems, and finally,promises for the future. This chapter is notintended to be a review of the now vast literatureon methods and applications of stable isotopeanalysis in studies of past peoples. Reviews ofthe earlier literature are available elsewhere(Katzenberg and Harrison, 1997; Pate, 1994;

Biological Anthropology of the Human Skeleton, Second Edition. Edited by M. Anne Katzenberg and Shelley R. SaundersCopyright # 2008 John Wiley & Sons, Inc.

413

Schoeninger and Moore, 1992; Schwarcz andSchoeninger, 1991), while more recent com-prehensive articles focus on particular aspectsof stable isotope applications (e.g., Katzenberget al., 1996 on weaning; Lee-Thorp andSponheimer, 2006 on early hominins). Thefield has grown enormously with many morelaboratories training many more students, apply-ing stable isotope analysis to an ever increasingarray of interesting problems. Hopefully thevarious discussions herein will guide the readerto the relevant literature if more detailed infor-mation is desired.

Developments in Stable IsotopeAnalysis

From the perspective of the archaeologist, thefirst stable isotope studies of past diet werecarried out in the 1970s (DeNiro and Epstein,1978, 1981; van der Merwe and Vogel, 1978;Vogel and van der Merwe, 1977). However,studies of stable isotopes began in the earlyyears of the twentieth century in the laboratoriesof chemists and physicists. After the discoveryof stable isotopes in 1913, improvements ininstrumentation and intensive study resulted inthe identification of most stable isotopes bythe mid-1930s. The first commercial mass spec-trometer was used to analyze petroleum in 1942(Gross, 1979). Throughout the 1950s and 1960smass spectrometers and the applications ofstable isotope studies advanced rapidly in chem-istry, biology, and geochemistry. Efforts weredirected toward understanding variation in therelative abundances of stable isotopes of thevarious elements. For example, geochemistsexplored oxygen isotope variation and its poten-tial for studies of past climate (reviewed by Luzand Kolodny, 1989). Major advances in under-standing stable isotope variation in the biosphereand geosphere occurred during the 1950s and1960s. Botanists and geochemists exploredstable carbon isotope variation in plants (Craig,1954; Smith and Epstein, 1971), and researchersin radiocarbon dating laboratories shared theirinterest (Bender, 1968; Hall, 1967; Lowdon,

1969) since this variation is relevant to radiocar-bon dating methods.

Along with advances in understandingthe processes that cause variation in stableisotope abundance ratios in different substances,major advances have occurred in instrumenta-tion. Although improvements in resolution,detection, and overall design of mass spec-trometers occurred throughout the twentiethcentury, advances in the last 15 years have hadan enormous impact on the use of stableisotope methods because of the ability to runmore samples at a much lower cost. Stableisotope analysis used to be a time-consumingand, therefore, expensive method. Preparationof samples, isolation of gasses containingspecific elements of interest, and the actualanalysis and corrections to standards were verylaborious processes. Only a small number ofsamples per day could be analyzed, and ana-lyses required constant attention by lab person-nel. In the late 1980s new instrumentation wasdeveloped that simplifies sample preparation,automates introduction of samples into thesystem, requires much smaller samples, ismuch faster, and therefore is much less expens-ive. This development has opened up manymore applications of stable isotope analysis,most notably in ecological research (Barrieet al., 1989; Fry, 2006; Griffiths, 1998; Rundelet al., 1989). It has also made it possible formany more samples to be analyzed in archaeolo-gical studies. Instead of selecting a few humanbones for analysis, as was true of many studiescarried out before 1990, researchers now routi-nely analyze nonhuman faunal bone and eitherprehistoric or modern plants to compare poten-tial foods with human stable isotope data.These newer methods have allowed researchersto analyze many more human samples, reveal-ing previously unknown variation withinpopulations, including age differences and indi-viduals who are outliers with respect to theirchemical signatures. Stable isotope methodshave now been applied to studies of demogra-phy, residence patterns, and disease in additionto studies of diet. Although newer instrumentssimplify analyses, it is imperative to have

414 STABLE ISOTOPE ANALYSIS: A TOOL FOR STUDYING PAST DIET, DEMOGRAPHY, AND LIFE HISTORY

trained personnel in reputable laboratoriesrunning the instruments to ensure good accuracyand precision.

History of Applications to Analysisof Past Peoples

The realization that stable isotopes of carboncould be used to investigate past diets can betraced to two different but related fields ofstudy (reviewed by van der Merwe, 1982).Scientists working to determine 14C dates onancient organic remains noted variation indates derived from some human skeletalremains. Also, maize from archaeological sitesgave anomalous dates relative to wood charcoal(Bender, 1968; Hall, 1967). These findingscoincided with research on different biochemi-cal pathways of photosynthesis among plants(Smith and Epstein, 1971). Maize fixes carbonby a different pathway and, as a result, containsmore 13C relative to 12C than most other plantsfrom temperate regions. Since samples forradiocarbon dating were assumed to have thesame stable carbon isotope composition as thestandard, those samples containing relativelymore 13C, such as maize cobs and kernels,gave erroneous dates. In addition to its use indifferentiating human consumption of plantswith different photosynthetic pathways, carbonisotopes have been shown to differentiatemarine from terrestrial-based diets in humans(Chisholm et al., 1982; Tauber, 1981).

Carbon was the first element for which stableisotope variation was used in archaeology,which follows from archaeologists’ familiaritywith radiocarbon. Once the potential of study-ing stable carbon isotopes in preserved proteinwas understood, interest in other elementssuch as nitrogen, oxygen and sulfur flourished.Each of these elements and their stable isotopeshave been studied extensively in geological andecological systems as well. In fact, archaeolo-gists are relative latecomers to the study ofstable isotope variation of the elements, withthe exception of carbon.

The second element to be used in paleodietresearch was nitrogen. DeNiro and Epstein

(1978, 1981) carried out controlled feedingexperiments on several species to study therelationship between the stable carbon andstable nitrogen isotope ratios in diet and inanimal tissues. Shortly after that, DeNiro,working with two postdoctoral researchers,explored trophic level and regional variation innitrogen isotopes (Schoeninger and DeNiro,1984) and trophic level variation and dietarydifferences in east Africa (Ambrose andDeNiro, 1987).

Carbon and nitrogen stable isotopes are theisotopes most commonly studied in humanremains. More recently, stable oxygen isotopesand strontium isotopes have been studied inbone and in tooth enamel. Stable isotopes ofsulphur have been studied in hair and to alesser degree, bone, since hair keratin containsmore sulphur than bone collagen, which con-tains very little sulphur.

Basic Concepts of Stable IsotopeVariation

Isotopes are atoms of the same element with thesame number of protons but different numbersof neutrons. Since the atomic mass is deter-mined by the number of protons and neutrons,isotopes of an element vary in their masses.Table 13.1 shows some of the chemicalelements that have several isotopes and theabundances of those isotopes. In contrast tounstable (radioactive) isotopes, stable isotopesdo not decay over time. For example, 14C in adead organism decays to 14N, whereas theamounts of 12C and 13C in the same organismwill remain constant.

In chemical reactions, such as the conversionof atmospheric CO2 into glucose by plants, therelative amounts of 12C and 13C differ in planttissue relative to that of atmospheric CO2. Thisvariation is due to the fact that isotopes vary inmass and therefore have slightly differentchemical and physical properties. Isotopeswith higher mass (heavier isotopes) such as13C usually react slightly more slowly thanlighter isotopes such as 12C. Physical phenom-ena that occur during chemical reactions as a

INTRODUCTION 415

result of the mass differences in isotopes arereferred to as “isotope effects.” The resultingdifference in the isotope ratio of the carbon inthe plant tissues as compared with the carbonin atmospheric CO2 caused by isotope effectsis termed “fractionation.” Fractionation is thebasis for stable isotope variation in biologicaland geochemical systems, and gaining anunderstanding of the chemical reactions thatresult in stable isotope variation allows thebiological anthropologist, archaeologist, geo-chemist, or ecologist to put stable isotope analy-sis to work to solve a wide range of interestingproblems. More detailed discussions ofisotope effects and fractionation can be foundin textbooks such as that by Hoefs (1997) andFry (2006), with the latter providing an entirechapter on fractionation.

Tissues Used in Stable IsotopeStudies

The first tissue to be used in archaeologicalstable isotope studies of human paleodiet wasthe collagen of bone. Methods for isolating

collagen had already been developed in radio-carbon dating labs since collagen was used indating. Information on isolating collagen frombones and teeth is provided by Ambrose(1990), who critically reviews the variousmethods available.

Bone is composed of an organic matrix of thestructural protein, collagen, which is studdedwith crystals of calcium phosphate, largely inthe form of hydroxyapatite. Dry bone is appro-ximately 70% inorganic and 30% organic byweight. Most of the organic portion (85–90%)is collagen. The remainder includes noncolla-genous proteins, proteoglycans, and lipids(Triffit, 1980). Because of the intimate struc-tural relationship between collagen and hydro-xyapatite, collagen may survive for thousandsof years (Tuross et al., 1980), and protein thatis probably degraded collagen has even beenrecovered from dinosaur fossils (Wyckoff,1980). Collagen contains approximately 35%carbon and 11–16% nitrogen by weight (vanKlinken, 1999), and it is the tissue of choicefor stable carbon and nitrogen isotope analysis.The detection of postmortem degradation ofcollagen has been an active area of research(e.g., Child, 1995; Collins et al., 1995, 2002;DeNiro, 1985; Hedges, 2002; Schoeningeret al., 1989; Tuross, 2002).

Because collagen does degrade over timeand at varying rates depending on the burialenvironment, researchers have sought othersources of carbon that are representative of life-time carbon intake. Another biological sourceof carbon in bones and teeth is in the form ofcarbonate (CO3), which occurs in the mineralportion of bone. Bone mineral is largely com-posed of hydroxyapatite, Ca10(PO4)6OH2.However, several ions can substitute for theconstituent ions of the hydroxyapatite crystals(see Burton, this volume). Well known fromthe trace element literature are the substitutionsof Srþþ (strontium) or Pbþþ (lead) for Caþþ

(calcium). Another common substitution isCO3¼ for PO4

; (phosphate) (LeGeros et al.,1967). Sullivan and Krueger (1981) proposedusing the carbon in bone mineral for stablecarbon isotope studies in fossil bone, when

TABLE 13.1 Average Terrestrial Abundancesof Stable Isotopes of Elements Used in Analysesof Ancient Human Tissues

Element Isotope Abundance (%)

Hydrogen 1H 99.9852H 0.015

Carbon 12C 98.8913C 1.11

Nitrogen 14N 99.6315N 0.37

Oxygen 16O 99.75917O 0.03718O 0.204

Sulfur 32S 95.0033S 0.7634S 4.2236S 0.014

Strontium 84Sr 0.5686Sr 9.8687Sr 7.0288Sr 82.56

Extracted from Table 1.1, Ehleringer and Rundel, 1989.


collagen was too badly degraded to be used.This proposal was challenged by Schoeningerand DeNiro (1982), and a debate in the literaturefollowed (reviewed by Krueger, 1991 and Lee-Thorp and van der Merwe, 1991). The challengecentered on whether bone carbonate was alteredin the postmortem environment by exchangebetween constituents of buried bone and car-bonates in sediments such that the carbonisotope ratios would not reflect lifetime carbondeposition. These concerns were addressed byLee-Thorp (1989), who developed preparationmethods to remove the more soluble carbonates,which are those most likely to be diagenetic inorigin. More recent debate has centered on theuse of carbonate in tooth enamel versus dentinand bone (Koch et al., 1997). There are twocompelling reasons to pursue the use of carbon-ate in biological apatite as a source of carbonisotope ratios. It allows stable isotope studiesto be applied to much older materials whereit is no longer possible to isolate collagen(e.g., Lee-Thorp et al., 1989; Lee-Thorp andSponheimer, 2006; Sponheimer and Lee-Thorp, 1999), and carbon from biologicalapatite records slightly different dietary infor-mation than does collagen.

The idea that the carbon in the carbonate ofbones and teeth comes from different dietarycomponents than the carbon in collagen wasfirst proposed by Krueger and Sullivan (1984).Ambrose and Norr (1993) and Tieszen andFagre (1993), in two separate controlled feedingexperiments, demonstrated that Krueger andSullivan were correct in suggesting that collagencarbon comes mainly from ingested protein inthe diet, whereas the carbon in biologicalapatite reflects whole diet. The reason is that col-lagen is composed of a mixture of essential andnonessential amino acids. The essential aminoacids come from ingested protein. The non-essential amino acids may come from ingestedprotein, or they may be formed from otherdietary sources and breakdown products withinthe body. Carbonate in bone is formed from dis-solved bicarbonate in the blood, and this comesfrom dietary carbohydrate, lipid, and protein.Therefore, the carbon in biological apatite

provides a picture of the total diet, whereas col-lagen is more reflective of dietary protein.

The mineral portion of bone is also the sourceof oxygen and strontium used in isotope studies.Oxygen isotopes are most often isolated fromPO4 (Luz and Kolodny, 1989; Stuart-Williamsand Schwarcz, 1997) and have been used inpaleoclimate and, more recently, paleodemo-graphic studies. The oxygen in bone and toothcarbonate has also been analyzed for stableoxygen isotopes (Koch et al., 1997), andalthough it is a technically less demandingprocedure relative to isolating phosphate, car-bonate is more likely to be altered by diageneticprocesses, more so for bone than for toothenamel carbonate (see Garvie-Lok et al., 2004;Nielsen-Marsh and Hedges, 2000a, 2000b; formore detailed discussions of diagenesis andbone carbonate). Strontium is a common traceelement in bone where it substitutes forcalcium. Strontium isotopes have been used inpaleodiet and residence studies (e.g., Bentleyet al., 2005; Ericson, 1989; Ericson et al.,1989; Ezzo et al., 1997; Price et al., 1994a,1994b; Sealy et al., 1991, 1995).

Methods for Isolating SpecificComponents for Stable IsotopeAnalysis

Specific instructions for isolating various com-ponents of bone can be found in the sourcescited throughout this section. It is important tounderstand the chemical principals of eachmethod. A wide range of variation exists inpostmortem environments, duration of inter-ment, and therefore the preservation of hardtissues. It is sometimes necessary to vary themethods for poorly preserved bone samples.For example, by diluting the acid solution, theprocess of dissolving the bone mineral is lessharsh and proceeds more slowly so that partiallydegraded collagen may be recovered. Tropicalenvironments often have poor preservation ofbone with extensive collagen degradation;however, teeth may be better preserved, allowingextraction of collagen from dentine. Sometimes

INTRODUCTION 417

postmortem alteration is so extensive that specificanalyses are not possible.

Collagen. Most researchers use one of threemethods for isolating collagen from bones andtooth dentine. Sealy (1986) describes a simplemethod in which small chunks of bone (1–3gin total) are decalcified in a hydrochloric acidsolution (between 1% and 5%, depending onthe density of the sample and its outwardappearance). An additional soak in sodiumhydroxide (0.1 molar) may follow to removedecayed organic matter from the burial environ-ment. The remaining collagen is freeze-dried.

In another method, (Bocherens et al., 1995;Tuross et al., 1988) small chunks of bone aredemineralized in EDTA (ethylenediaminetetra-cetic acid), a sodium salt, which separates col-lagen from bone mineral. A third method wasoriginally developed by Longin (1971) andlater modified by Schoeninger and DeNiro(1984) and by Brown et al. (1988). In thismethod, powdered bone is demineralized in8% hydrochloric acid for a short period oftime (around 18 minutes). This process isfollowed by a slow hydrolysis in weakly acidichot water (pH 3). This method is preferablefor poorly preserved bone; however, the risk isthat one may obtain other organic matter inaddition to collagen. Thus, an additional soakin sodium hydroxide (0.1 molar) to removedecayed organic matter is usually recommended.

Several researchers have compared methodsand their yields (Chisholm et al., 1983;Schoeninger et al., 1989). Boutton et al.(1984) and Katzenberg (1989; Katzenberget al., 1995) have demonstrated that somecollagen is lost when demineralized bone issoaked in sodium hydroxide, but that theother material removed in the soak containshumic contaminants (decayed organic matter)that may skew d13C values. For example,Katzenberg et al. (1995) demonstrated that theresidue removed during the sodium hydroxidesoak of prehistoric human bones from southernOntario had a much lighter d13C than the col-lagen, indicating that the residue containeddecayed C3 plant remains.

It is essential to demonstrate that the materialbeing analyzed is collagen, and various criteriahave been used for this purpose. DeNiro (1985)proposed that extracted material with a carbon-to-nitrogen (C/N) ratio in the range of 2.9 to3.6 should preserve reasonable stable isotoperatios to those from the lifetime of the organism.Since the ratio of carbon to nitrogen in modern(never buried) bone collagen is 3.2, this rangewas suggested to take into consideration analyti-cal error and some slight alteration over longperiods of time. These figures are based on theatomic ratio of carbon to nitrogen in collagen.Modern light isotope mass spectrometers areusually interfaced with gas analyzers thatprovide data on the content of carbon andnitrogen in samples, and they also calculatecarbon-to-nitrogen ratios. However, these areweight ratios and will be slightly lower (bya factor of 1.16667) than the atomic ratiosgiven by DeNiro (1985, and see Ambrose,1990). Recent publications often include dataon percentage carbon and nitrogen in samplesas well as the C/N ratios. These data are import-ant for evaluating the validity of stable isotoperesults. For example, recent samples thatcontain lipids will have more carbon anderroneous stable carbon isotope ratios sincelipid is less enriched in the heavier isotopethan is collagen. Samples in which collagen isdegraded often have low nitrogen content andmay provide results for stable carbon isotopesbut not stable nitrogen isotopes, since collagencontains much more carbon than nitrogen(thus the expected 3.2 ratio). In all of these pro-cedures, the objective is to isolate collagen frombone mineral and any organic matter introducedin the postmortem environment.

Compound-Specific Analyses. With advan-ces in the field, researchers have focused theirattention on isolating individual amino acids,initially for accelerator radiocarbon dating(Stafford et al., 1991). Stable isotope valuesvary among the different amino acids, so pre-ferential loss of certain amino acids from dia-genesis can alter the overall d13C of a proteinsuch as collagen (Hare and Estep, 1982).


Interest also exists in isolating the indispensable(essential) amino acids (those amino acids thatmust be obtained from the diet) from collagenfor stable isotope analysis for dietary recon-struction (Hare et al., 1991). Since theseamino acids come from dietary protein and areincorporated into human proteins such as col-lagen, they provide a more direct tracer thanbulk collagen, which is made up of both dispen-sable amino acids (those that may be syn-thesized by the organism) and indispensableamino acids. Methods used for such study aremuch more complex than those for simply iso-lating collagen and are described by Staffordet al. (1991) and Hare et al. (1991). There is con-siderable promise in pursuing these methods,particularly with the development of GC/C/IRMS (gas chromatography/combustion/isotope ratio mass spectrometry) (Lichtfouse,2000; Macko et al., 1997). This method allowsthe researcher to isolate specific organic com-pounds and then to introduce them into themass spectrometer. Evershed pioneered thisresearch in paleodiet studies by focusing onlipids and demonstrating the presence of specificsubstances in residues from prehistoric ceramicvessels (Evershed, 1993; Stott and Evershed,1996), including dairy fats (Copley et al.,2005a, 2005b). Corr et al. (2005) have demon-strated that compound-specific stable carbonisotope analysis of collagen amino acids allowsone to differentiate diets with marine proteinand C4 plants in arid regions where d15N valuesare unpredictable. Students interested in graduatestudy in this area are well advised to becomefamiliar with the principles and methods oforganic chemistry and biochemistry.

Biological Apatite. A method for isolatingthe carbonate fraction of bone mineral wasdeveloped by Lee-Thorp (1989; Lee-Thorp andvan der Merwe, 1991; Lee-Thorp et al., 1989).Ground bone is soaked in sodium hypochloriteto remove organic material. Carbonate adsorbedfrom the burial environment is removed withone molar acetic acid. Samples are then reactedwith phosphoric acid to release the structuralcarbonate. CO2 is collected by cryogenic

distillation. Specific steps are described in thereferences cited above and by Tieszen andFagre (1993) and Ambrose and Norr (1993).

More recently, carbon in the carbonatefraction of tooth enamel has been analyzedusing laser ablation stable isotope analysis(Sponheimer et al., 2006). The use of the laserallows sampling of such small quantities thatsampling is nearly imperceptible for purposes oflong-term curation, and seasonal variation canbe explored, as was done by Sponheimer et al.(2006) on teeth from Paranthropus robustus.

Oxygen isotope measurements in boneusually make use of the oxygen in phosphate,which is less affected by diagenetic processesthan carbonate. Stuart-Williams (1996; Stuart-Williams and Schwarcz, 1995, 1997) developeda method of isolating organic phosphate frombone that is simpler and safer than previous pro-cedures. Bryant et al. (1996) compared theoxygen isotope results from phosphate and car-bonate of tooth enamel and demonstrated that itis possible to estimate one from the other.

Hydroxyapatite. Strontium substitutes forcalcium in the hydroxyapatite crystals of bonemineral. Two different methods of isolatingbone mineral for analysis of strontium isotopeshave been described in the recent literature.The method used by Sealy et al. (1995) andby Sillen et al. (1998) makes use of Sillen’s1986 solubility profile method. Bone or toothpowder is washed in acetic acid and sodiumacetate buffer solution repeatedly, saving eachwash. The various washes are analyzed byICP (inductively coupled plasma emissionspectrometry). The first few washes presumablycontain recently deposited contaminants fromthe burial environment and show variation intrace element concentrations. Later washes tendto show less variation in the concentration ofstrontium and other trace elements. It is theselater washes, which are thought to contain the bio-logically deposited strontium, that are used forstrontium isotope analysis by mass spectrometry.

A second method of preparation for stron-tium isotope analysis, described by Price andcolleagues (1994a, 1994b) begins with

INTRODUCTION 419

mechanical cleaning of the outer surface ofbone, followed by an overnight soak in onenormal acetic acid. The acid removes solublecarbonates and the portion of bone most likelyto contain elements from the burial environ-ment. The residue is then wet-ashed in nitricacid in preparation for mass spectrometry.Both methods attempt to isolate bone mineralthat has not been diagenetically altered, andalthough Sillen’s method is more conservative,it is also more labor intensive. Sillen andSealy (1995) have demonstrated that thesolubility profile method does not result in anyrecrystallization of bone mineral, unlikemethods that employ dry ashing (heating tohigh temperature to destroy the organiccomponent).

Lipids. Methods for treating lipids from bonesamples depend on whether one intends toremove the lipids so that only protein or carbon-ate is analyzed, or whether one wants to deter-mine the d13C of lipid. Lipids are less enrichedin the heavier isotope (i.e., d13C is more nega-tive) than bone collagen. Therefore it is necess-ary to remove lipids from bone samples beforeanalysis. This process is particularly importantwhen analyzing bones of recent origin for com-parative purposes. Liden et al. (1995) discussmethods of removing lipids. The most com-monly used method involves soaking the bonein a mixture of chloroform and methanol(Bligh and Dyer, 1959; Folch et al., 1957)after demineralization. The residue must berinsed carefully since these are organic solvents,which may contaminate the sample. Bone mayalso be soaked in diethyl ether before deminer-alization. The normal sodium hydroxide soakthen follows demineralization (Ambrose andNorr, 1993). These methods will effectivelyeliminate lipids from collagen preparations.Bligh and Dyer (1959) describe a methodfor extracting and purifying lipid from biologicalmaterials. Recent work by Evershed and col-leagues (Evershed, 1993; Stott and Evershed,1996) on characterization of lipid extracts fromancient materials, such as ceramic vessels, hasadded another source of evidence for paleodietstudies by allowing d13C determinations of

lipids and, more recently, of specific compoundswithin lipids (see the previous section on com-pound-specific analyses).

Mass Spectrometry

Stable isotope abundance ratios are measured inisotope ratio mass spectrometers (IRMS), whichshould not be confused with organic mass spec-trometers that are used to characterize complexorganic molecules. Isotope ratio mass spec-trometers have four components: an inletsystem, an ion source, a mass analyzer, and aseries of ion detectors (Fig. 13.1). For mostelements of interest (H, O, N, C), the sampleis introduced to the mass spectrometer as a gas(H2, CO2, N2, and CO2, respectively). Untilrecently, most stable isotope work performedwith collagen or carbonate required combustionof the sample in sealed tubes. After combustion,the resultant CO2 and H2O were separated off-line before the CO2 was let into the mass spec-trometer. Modern instruments now interfacecombustion furnaces and gas analyzers withmass spectrometers to simplify and ease the con-version of the sample into the requisite gaseousform. In such a setup, collagen is weighed intotin sample holders, which are then placed intoan automated sample tray. The revolving traydrops samples into the furnace where N2, CO2,and H2O are produced. These gasses, carriedby helium carrier gas, are separated beforebeing swept into the mass spectrometer. (For adetailed presentation of continuous-flow stableisotope analysis, see Barrie et al., 1989 andBarrie and Prosser, 1996.)

Once in the mass spectrometer, the gas ofinterest is let into the second component of themass spectrometer, the ion source. In the ionsource, some gas molecules are ionized by elec-tron bombardment, which allows them to becontrolled and focused into a beam. The ionbeam is then directed, via a flight tube, into themass analyzer zone of the mass spectrometer.As the name suggests, the mass analyzer separ-ates the ion beam into several smaller beams bypassing it between the poles of a magnet. Thisprocess is directly analogous to the separation


of white light into its constituent wavelengthsthrough a prism. The separation of one ionbeam into several beams according to massresults in the desired “mass spectrum.” Thebeam intensities of the respective ion beamscan then be measured in the ion collectorsection of the instrument. The relative intensi-ties of the individual isotope ion beams arethen reported as isotope ratios, for example,13CO2:12CO2. To report a meaningful value,the mass spectrometer alternately analyzes ali-quots of the unknown sample and a known stan-dard gas, thereby providing a ratio of the stableisotopes in the sample relative to that same ratioin the standard.

Because the element of interest cannotalways be converted into an easily handledgas, it is sometimes necessary to introduce thesample in solid form. For strontium isotopeanalysis, the sample is deposited directly on afilament near the ion source where it is heatedto evaporation and ionized under vacuum.This process is referred to as thermal ionizationmass spectrometry (TIMS) and requires a differ-ent ion source (solid source) for analysis. Manylabs will have separate instruments for analyz-ing gasses and solids. More recently, laser abla-tion ICP-MS (inductively coupled plasma–mass spectrometery) has been used for pinpointsampling to analyze isotopes of heavierelements such as strontium (Latkoczy et al.,2001; Prohaska et al., 2002).

For compound-specific analyses, a gas chro-matograph (GC) can be interfaced with an

IRMS resulting in GC-C-IRMS (gas chromato-graphy combustion isotope ratio mass spec-trometry). This configuration allows one todetermine the isotope ratio of a specific com-pound, such as cholesterol, that has been iso-lated using gas chromatography (Stott et al.,1999; Tripp and Hedges, 2004).

It is helpful to know enough about massspectrometry to be able to discuss one’s needswith various laboratory personnel at the begin-ning of a research project. It is important tobe able to understand problems that maydevelop when the data have been collected.Considerations include the general compositionof the sample and the range of the expectedresults. For example, departures from theexpected range of carbon to nitrogen in bonesand teeth may signal poor preservation or con-tamination. It is necessary to have some ideaof the range of isotope compositions expectedsince the standards used should bracket theexpected values.

Mass spectrometers are analytical instru-ments that have many uses in chemistry, biochem-istry, geochemistry, and ecology. Instrumentsvary in their setup depending on the needsof the researcher. A laboratory may be set upto perform analyses on certain types ofsamples and may specialize in certain elements.Laboratories carrying out ecological researchare most likely able to accommodate mostanalyses of interest to the archaeologist andphysical anthropologist interested in light iso-topes (hydrogen, carbon, nitrogen, oxygen,

Figure 13.1 Diagram showing three of the four components of the mass spectrometer: Gas is introduced into the sourcewhere molecules are ionized and then accelerated; the resulting ion beam is directed into the mass analyzer where ions of differ-ent masses are separated; ion collectors measure intensities of the separated ion beams. (Courtesy of H.R. Krouse.)

INTRODUCTION 421

and sulphur), whereas geochemists are morelikely to have TIMS and laser ablation setups.Organic chemists and biochemists may haveGC-C-IRMS capabilities, and enterprising bio-logical anthropologists and archaeologists mayhave their own instruments. Given the highcost of mass spectrometers and the need forfull-time technical support, most institutionshave shared facilities that serve the needs ofmultiple users.

Standards, Precision, and Accuracy

Stable isotope abundance ratios are determinedrelative to the ratios of those same isotopes instandard materials. The mass spectrometer com-pares the stable isotope abundance ratio in thesample with the stable isotope ratio in a stan-dard. Thus, the reported value uses the follow-ing notation:

d in ‰ ¼ R(sample) � R(standard)

R(standard)� 1000

where R ¼ the ratio of the number of heavierto lighter isotopes, so that for carbon isotopes,the equation is

d13C‰ PDB¼13C=12Csample�13C=12Cstandard� �

b 13C=12Cstandard c

�1000

The ‰ (permil sign) means “per thousand”since the ratio is multiplied times 1000.1 Thiscalculation is done to amplify the differencebetween the ratio of the stable isotopes in thesample and the ratio of stable isotopes in thestandard, which is usually a very small number.

International standards are available throughthe National Bureau of Standards (NBS) and theInternational Atomic Energy Agency (IAEA),Vienna. The circulation of these standardsamong laboratories allows comparison ofresults from different researchers working indifferent laboratories. However, because of thehigh cost, individual laboratories normallyalso have internal standards. These standardsare substances whose isotopic ratio is wellcharacterized relative to an international stan-dard and that are run routinely with batches ofunknowns to check for consistency in theinstrument. Internal and reference standardsare reported relative to a primary reference stan-dard for a particular element, which by defi-nition has a d value of 0. Absolute isotopeabundances of some primary reference stan-dards are available in textbooks such as that ofHoefs (1997) and Fry (2006). Primary andother reference standards for elements discussedin this chapter are listed in Table 13.2.

The sensitivity of mass spectrometers varies,and it is important to know the precision of theinstrument used before making interpretationsfrom the data. Most light isotope mass spec-trometers can measure d13C values with aprecision of +0.1‰ and d15N values with aprecision of +0.2‰. Newer models haveimproved sensitivity, although some newer con-tinuous flow systems that simultaneouslymeasure more than one element in a samplesacrifice precision for speed and economy ofsample size. Precision should be determinedfor individual instruments using multiple ana-lyses of samples with similar composition tothose of interest. Data should be reported in amanner that is consistent with the precision ofthe instrument. If the precision is +0.2‰,then it is incorrect to report results to 0.01‰.

APPLICATION OF STABLE ISOTOPEANALYSIS TO SELECTED PROBLEMSIN SKELETAL BIOLOGY

During the 1990s, with the growing use of stableisotopes to address problems in biological

1The term “permil,” meaning per thousand, is similar to theterm “percent,” meaning per hundred. In the isotope litera-ture, using the delta notation, values are reported as somenumber permil (‰). In the trace element literature, elementsmay be measured in parts per million (ppm). This measure isvery different in that it is not a ratio and it refers to a muchsmaller proportion (out of one million). It is best to avoidusing the phrase “parts permil,” which is sometimes heardfrom anthropologists. Although not incorrect, is itawkward and incongruous, since people do not say partsper cent. Standard terminology is “permil.”


anthropology and archaeology, several reviewarticles were published. Katzenberg andHarrison (1997) discuss developments andreview the literature since 1989. Otherreviews, which also cover basic concepts instable isotope studies, include Schwarcz andSchoeninger (1991), Schoeninger and Moore(1992), and Pate (1994). A series of conferencesentitled “Advanced Seminars in Paleodiet” hasresulted in several publications (edited booksas well as guest-edited issues of journals) ofpapers presented at those conferences (Price,1989; Sillen and Armelagos, 1991; Lambertand Grupe, 1993; Bocherens et al., 1999;Ambrose and Katzenberg, 2001; Koch andBurton, 2003;). A symposium in honor of thework of Dr. Harold W. Krueger, held at the2001 meetings of the Society for AmericanArchaeology, was also published as a specialissue of a journal (Ambrose and Krigbaum,2003). Collectively, these volumes provide adetailed look at the history and developmentsin the use of stable isotopes for studying archae-ological human remains. For a similar review ofdevelopments in research on diagenesis, a seriesof conferences have been held on bone diagen-esis, and these are also published as specialissues of journals (Schwarcz et al., 1989;Hedges and van Klinken, 1995; Bocherens andDenys, 1997; Fernandez-Jalvo et al., 2002).

Initially, most applications of stable isotopeanalysis to human remains were concernedwith reconstructing diet. Subsequently otherresearch questions have been addressed withstable isotope methods. These questions includedetermining the duration of breastfeeding,effects of disease processes, and determination

of residence and migration patterns. The follow-ing sections highlight some of these applicationsin addition to more traditional approaches topaleodiet studies.

Paleodiet

C3 and C4 plants. Maize is one of severaltropical grasses that fixes carbon by a differentphotosynthetic pathway (referred to as theHatch–Slack or C4 pathway) than most plantsfound in temperate regions. C4 plants, whichalso include sorghum, millet, and sugar cane,adapt to heat and aridity by minimizing theamount of time that the leaf pores (stomata)are open, thereby minimizing water loss.These plants discriminate less against theheavier isotope, 13C, than do temperate plantspecies, which use the C3 (Calvin) photosyn-thetic pathway. Atmospheric CO2 has a d13Cvalue of 28‰ today, but before the widespreadburning of fossil fuels, the value was around27‰.2 C4 plants range from 29 to 214‰,whereas C3 plants range from 220 to 235‰(Deines, 1980). The non-overlapping rangesof C3 and C4 plants provide the basis for usingstable isotopes of carbon in preserved humantissue for revealing diet.

Several studies have been carried out todetermine the difference between the d13C ofthe diet and the d13C value of various body

2Fossil fuels are depleted in 13C, and their use has resulted ina decrease in atmospheric d13C of approximately 1‰. Thevalue of 27‰ is used in archaeological research to reflectthe conditions present during the lifetime of the individualsbeing studied.

TABLE 13.2 Primary (International) and Reference Standards for Selected Elements

Element Primary Standard Other Reference Standards

Hydrogen Vienna Standard Mean Ocean Water (VSMOW) V-GISP, V-SLAP, NBS-30Oxygen Standard Mean Ocean Water (VSMOW) NBS 19, 20, 18, 28, 30, V-GISP, V-SLAPCarbon PeeDee Belemnite (VPDB) NBS 18, 19, 20, 21Nitrogen Atmospheric Nitrogen (Air)Sulfur Canyon Diablo meteorite troilite (VCDT)

Adapted From Hoefs, 1997 and Coplen, 1994.

APPLICATION OF STABLE ISOTOPE ANALYSIS 423

tissues (Ambrose and Norr, 1993; Lyon andBaxter, 1978; Tieszen and Fagre, 1993; Vogel,1978). Bone collagen d13C is approximately5‰ greater than d13C of the diet, which is thebasis for the expression, “you are what you eatþ5‰.” Interestingly, this number was firstsuggested by van der Merwe and Vogel (1978)based on measurements of free-ranging largemammals and their diets and then was con-firmed recently by Ambrose and Norr (1993).In controlled feeding studies of rats, the diet tocollagen spacing was only 5‰ when thedietary protein, carbohydrate, and fats were allfrom similar sources (C3 or C4). In situationswhere the protein source differed from that ofthe carbohydrates and fats, the spacing valuevaried (Ambrose and Butler, 1997). Such a situ-ation in humans might occur if people were con-suming the meat of C3 browsers such as deer andC4 plants such as maize. In the controlledfeeding experiments on rats, such a situationresults in a diet to collagen spacing of lessthan 5‰. Thus, although the spacing factor of5‰ may be used as a guide, it should not betaken as an absolute value.

After the demonstration that stable carbonisotopes in bone collagen could be used todocument the consumption of C4 plants suchas maize against a background of C3 plants(Vogel and van der Merwe, 1977; van derMerwe and Vogel, 1978), several other research-ers applied these same principles to other regionswhere maize was the major introduced cultigen(e.g., Buikstra and Milner, 1991; Katzenberget al., 1995; Larsen et al; 1992; Schurr andRedmond, 1991; Schwarcz et al., 1985). Themethod works very nicely in eastern NorthAmerica where maize is the predominant and,in some places, the only C4 plant consumed inany quantity. Follow-up studies took into con-sideration the fact that if the animals exploitedby human groups consumed C4 plants, thentheir tissues would be enriched in the heavierisotope and this would show up in humanbone collagen carbon (Katzenberg, 1989).Subsequent work on the differential routing ofprotein and nonprotein nutrients to the synthesis

of collagen reinforces the importance of under-standing the diets of the animals consumed bypeople (Ambrose and Norr, 1993; Tieszen andFagre, 1993).

Several regions of archaeological interestexist where stable isotope analysis was notattempted because it did not seem that suchanalysis could provide any additional infor-mation. If all potential food sources havesimilar stable isotope ratios, then little can belearned, or if the central research questionfocuses on a C4 plant, such as maize, but thereare other C4 plants in the region, it will not bepossible to distinguish when maize was firstintroduced. The application of stable isotopemethods came somewhat later to the AmericanSouthwest, where maize made an early appear-ance, but there are also several indigenousC4 and CAM plants (a third photosyntheticpathway with values intermediate to C3 and C4

plants) in addition to human exploitation ofanimals that consume C4 plants. Nevertheless,researchers did tackle this more complex region(Benson et al., 2006; Coltrain et al., 2006;Decker and Tieszen, 1989; Katzenberg andKelley, 1991; Matson and Chisholm, 1991;Spielmann et al., 1990) with useful results interms of documenting the intensity of maize use.

Many recent studies include stable carbonisotope data from both bone collagen and bonecarbonate or, in cases where collagen is notpreserved, just bone carbonate. Since bonecarbonate is thought to better reflect the wholediet (Ambrose and Norr, 1993; Tieszen andFagre, 1993) and collagen preferentially reflectsdietary protein, analyses of both componentsprovide additional information. For example,Harrison and Katzenberg (2003) used stablecarbon isotope data from bone apatite to detectsmall amounts of maize in the diet that do notshow up in collagen, when it was first intro-duced into southern Ontario. Ambrose et al.(2003) could better differentiate the diets ofhigh and low status burials from Cahokia (inIllinois) using stable carbon isotope ratiosfrom both collagen and carbonate and stablenitrogen isotope ratios from collagen.


Marine Versus Terrestrial Based Diets.Stable carbon isotopes are also useful instudies of people inhabiting or exploitingcoastal areas where it is possible to test hypo-theses about the relative importance of marineand terrestrial foods in the diet (Blake et al.,1992; Chisholm et al., 1982; Hayden et al.,1987; Keegan and DeNiro, 1988; Lubell et al.,1994; Norr, 1991; Walker and DeNiro, 1986;Tauber, 1981). Such studies now span theglobe and include research in the Arctic(Coltrain et al., 2004), Tierra del Fuego(Yesner et al., 2003), and the Pacific (Ambroseand Butler, 1997; Pate et al., 2001). The mainsource of carbon for marine organisms is dis-solved carbonate, which has a d13C value of0‰, whereas the main source of carbon for ter-restrial organisms is atmospheric CO2, whichhad a d13C value of 27‰ in pre-industrialtimes. Tauber (1981) and Chisholm and col-leagues (1982, 1983) demonstrated that this7‰ difference is reflected in mammals, includ-ing humans, feeding from these two differentecosystems. Potential dietary change in north-ern Europe from the Mesolithic Period to theNeolithic Period has been investigated usingstable carbon and nitrogen isotopes. Richardset al. (2003) have suggested that use of marinefoods stopped when domesticated plants andanimals were introduced into Britain, based ond13C in bone collagen. Others working in north-ern Europe have argued that such an abrupt tran-sition did not occur and that the exploitation ofmarine foods continued into the NeolithicPeriod (Liden et al., 2004; Milner et al., 2004).This debate provides a good illustration ofboth the promises and the limitations of stableisotope analysis (Hedges, 2004), and it empha-sizes the need to use the results of stable isotopedata in the larger context of other archaeologicalevidence (Milner et al., 2004).

Nitrogen Isotopes and Diet

Trophic Level Distinctions. At the sametime that DeNiro and Epstein (1981) demon-strated that carbon isotope ratios of diet are

reflected in the tissues of an animal, theycarried out a study of the relationship of dietand tissues for stable isotopes of nitrogen.Nitrogen isotopes vary depending on trophiclevel. Atmospheric nitrogen (N2) is the primarystandard, and its value is set at 0‰. Someplants (legumes) have a symbiotic relationshipwith bacteria of the genus Rhizobium. The bac-teria live in the roots and can fix nitrogen(combine it with other elements such as hydro-gen or oxygen), thereby making it available tothe plant (Brill, 1977). Other plants must gettheir nitrogen from decomposed organicmatter, which breaks down to compounds suchas ammonia (NH3) or nitrate (NO3). Legumeshave d15N values closer to that of atmosphericnitrogen, whereas nonleguminous plants aremore enriched in 15N and therefore have higherd15N values. Herbivore d15N values are approxi-mately 3‰ higher than the d15N of their diet;thus, herbivores consuming legumes will havelower d15N values than those consuming non-leguminous plants. Carnivore tissues are againenriched in the heavier isotope resulting ind15N values approximately 3‰ higher thantheir diet. This principal of enrichment throughsuccessively higher trophic levels, which wasfirst pointed out by Minagawa and Wada(1984) and Schoeninger and DeNiro (1984),provides the basis for using stable nitrogen iso-topes to infer trophic level. Variation occurs inthe magnitude of the trophic-level effect instable nitrogen isotopes both among differenttissues within the same organism and amongdifferent taxa (Vanderklift and Ponsard, 2003).It is important to understand this variation whenworking with tissues other than bone collagenand with fauna other than mammals.

Ideally, a range of animals and plants fromthe environment under study is analyzed andhumans are viewed relative to the other organ-isms in their environment. An example is pro-vided in Fig. 13.2, which shows carbon andnitrogen stable isotope ratios for humans fromseveral prehistoric sites in New Mexico relativeto other mammals from the region (fromKatzenberg and Kelley, 1991). Humans have


the highest d15N and d13C values. In this region,humans consumed maize as well as animals thatfed on C4 plants. The figure indicates thatboth jackrabbits and bison consumed some C4

plants. Humans are approximately 3‰ higherthan deer, antelope, and bison for d15N. Theelevated d15N for deer and antelope relative tocottontail may reflect consumption of somelegumes by cottontail, but it may also be relatedto habitat. Cottontail and deer prefer forested,moister habitats, whereas antelope and jackrabbitare adapted to open range habitat. Heaton et al.(1986) and Ambrose (1991) have demonstratedthat d15N is sensitive to climate and is elevatedin arid regions. For this reason, Ambrose (1991)suggests that species from different ecosystemscannot be compared directly without consider-ing the isotopic composition of the local foodweb. This caution has been verified in sub-sequent studies.

Freshwater Resources

Initially, it was assumed by archaeologiststhat freshwater fish had d13C values similar tothose of terrestrial C3-consuming organisms.Little was known about stable nitrogen

isotope ratios in freshwater systems. In 1989,Katzenberg explored this question by analyzingbones of freshwater fish from archaeologicalsites around the Great Lakes. It was discoveredthat freshwater fish exhibit a trophic leveleffect, which results in higher d15N and slightlyincreased d13C in carnivorous fish. Thus, it ispossible to estimate reliance on fish in regionslike the Great Lakes where they are an abundantresource. This reliance is particularly significantsince fish are frequently underrepresented in thezooarchaeological record becasue of culturalpractices and, in earlier excavations, becauseof methods of recovery. Cultural practices,such as filleting and drying fish where they arecaught and then transporting the dried filletsback to the village, will result in underrepresen-tation of fish bones relative to their dietaryimportance. Excavations carried out withoutfine screening will miss bones of fish as wellas other small skeletal elements.

Freshwater fish also exhibit more variation ind13C than had been assumed. Several studies offreshwater ecosystems by ecologists have pro-vided the background to understanding thesources of this variation (France, 1995; Heckyand Hesslein, 1995; Kiyashko et al., 1991;

Figure 13.2 d13C and d15N values (mean+one standard deviation) for collagen from human and faunal bone samples fromsites in the Sierra Blanca region of New Mexico, dating from A.D. 800 to A.D. 1400. (Reproduced from Katzenberg and Kelley(1991) with permission of the publisher.)


Zohary et al., 1994). Freshwater plants havenumerous sources of carbon unlike terrestrialplants whose source is atmospheric CO2. Infreshwater ecosystems, carbon comes fromatmospheric CO2, CO2 in the water, bicarbonateand carbonate from rocks and soils, and organiccarbon as waste and decomposition productsfrom plants and animals living in the water(Zohary et al., 1994). The result is that fishliving in different habitats within freshwaterlakes display widely varying d13C. This is illus-trated in Fig. 13.3, which shows a reconstructionof the stable isotope ecology of the regionaround Lake Baikal, Siberia (from Katzenbergand Weber, 1999). The d13C of fish bonesranges from 214.2 to 224.6‰ with higherd13C in species inhabiting the shallow watersand lower d13C for fish inhabiting the deeper,open waters of the lake (Katzenberg andWeber, 1999). The higher d13C for some fishexplains the observed variation in human bonecollagen d13C from the region. This finding isimportant since there are no C4 plants in thearea and yet some human samples analyzed

showed evidence of food that was enriched inthe heavier isotope. In other words, stableisotope evidence from humans and theirpresumed food sources can indicate when some-thing has been missed.

Figure 13.3 also illustrates the trophic-leveleffect in d15N in both terrestrial and freshwaterorganisms. Large terrestrial herbivores haved15N around 4‰ to 5‰. Humans from a largenumber of sites in the region range from10.1‰ to 14.4‰. Fish and the freshwaterseals of Lake Baikal vary according to theirtrophic position. Carp (Caras a.) are bottom-feeders and have the lightest d15N. Lenok(Branchimystax l.) are highest in the littoralfood web and have higher d15N. Seals occupythe highest position in the freshwater food webwith d15N around 14‰.

Nitrogen Isotopes and Water Stress

Environmental variation in d15N of plants hasbeen demonstrated in coastal versus inlandregions and in arid versus wetter regions

Figure 13.3 d13C and d15N values (means) for collagen from human and faunal bone samples from sites from the westernshore of Lake Baikal, the Angara River Valley, and the Upper Lena River Valley, Siberia, dating from the Early Neolithic(5800–4900 B.C.) to the Early Bronze Age (3400–1700 B.C.). Fish bones are from modern specimens. Mean values forhuman bones are plotted by site and location (e.g., Baikal E refers to Early Neolithic Lake Baikal site; Lena M refers toMiddle Neolithic Lena Valley site, and Angara L refers to Early Bronze Age Angara Valley site). (Reprinted fromKatzenberg and Weber (1999) with permission.)


(Heaton, 1987; Shearer et al., 1983; Virginiaand Delwiche, 1982). Heaton et al. (1986) andSealy et al. (1987) have also shown that d15Nvaries in animals of the same species from aridversus wetter regions. Sealy et al. (1987) pointout the importance of recognizing reasons forelevated d15N in arid regions when thoseregions are also in close proximity to coastalresources, as is the case on the southern capeof Africa. Incorrect dietary interpretations forhuman samples can result if higher d15N isattributed to the use of marine resourceswithout recognizing that the same d15N mightresult from consumption of water-stressed ter-restrial animals. Ambrose (1991) has exploredthe physiologic basis for d15N variation inmammals living in arid regions. He andDeNiro (1986) proposed a model based onvarying nitrogen loss in urea, which is excretedin urine. Urea is depleted in 15N relative to diet.Under conditions of water stress, more urea isexcreted relative to the total volume of urine,and therefore, more of the lighter isotope, 14N,is lost. Therefore, more 15N is retained in thebody where it is available for tissue synthesis.The result is that tissue d15N will increaseunder prolonged water stress conditions.

Nitrogen Isotopes and Protein Stress

Another cause of elevated d15N, relative toexpectations from diet, is protein stress. Thisstress is related to the model above in that insuf-ficient protein intake results in the breakdownand reutilization of existing tissues in the bodythat are already enriched in 15N because ofpreferential excretion of 14N. Hobson andcolleagues, in two different studies on birds(Hobson and Clark, 1992; Hobson et al.,1993), have found that under conditions ofnutritional stress, new protein is synthesizedfrom the products of catabolism of existingprotein. Katzenberg and Lovell (1999) have pre-sented evidence that suggests this may bedetected in humans. They compared d13C andd15N in normal versus pathological segmentsof bones from individuals of known medical

history. One individual, who died of AIDSand experienced some new bone depositionbecause of osteomyelitis, showed elevatedd15N in the diseased segment relative to thetwo unaffected segments of bone. This casesuggests that the recently deposited bone col-lagen was synthesized, at least in part, fromamino acids liberated from the catabolism ofexisting proteins in the body. White andArmelagos (1997) reported elevated d15N inbones from individuals with osteoporosis froman arid region of Sudan. They discuss waterstress as a possible factor to explain the nitrogenisotope data, but it is possible that protein stresswas also a factor. The use of stable isotopesto address questions in paleopathology is a rela-tively new area of inquiry that promises to addto our understanding of disease processes aswell as to our understanding of stable isotopevariation in metabolism.

Infant Feeding and Weaning Studies

One of the big questions that carries throughmany approaches to studying past peoplesfrom their skeletal and cultural remains is thetiming and rate of population growth. Buikstraet al. (1986) presented an interesting hypothesisthat draws together the adoption of agricul-ture, the development of thin-walled ceramicvessels, and therefore the ability to preparecereal pap as a weaning food. The resultwould be earlier weaning (defined here as theintroduction of non-milk foods rather thanthe complete cessation of breastfeeding) andthus decreased inter-birth intervals, increasedpopulation growth, and larger population size.This result is based on the fact that nursingsuppresses ovulation so that supplementingmother’s milk with cereal gruel will result in ashorter period of infertility after childbirth.The hypothesis that weaning occurs sooner inagricultural societies was tested by Fogel andcolleagues (1989) in a study that includedstable nitrogen isotopes from fingernails ofnursing mothers and their infants. They founda trophic-level increase in d15N (þ2.4‰ on


average) of infant protein beginning shortlyafter birth (around three months) and decreasingwhen supplemental foods were introduced. Bythree to five months after nursing ceased, fin-gernail d15N for mothers and babies was thesame. Fogel and colleagues (1989, 1997) thenanalyzed bone collagen from preagriculturaland agricultural sites to see whether the latterchildren were weaned at an earlier age. In bothgroups, d15N decreased at 18–20 months ofage. This study demonstrated that the trophic-level effect of increased d15N is reflected inprotein, including bone collagen, and that thepotential exists to apply this method to skeletalsamples. In a more recent study, Schurr andPowell (2005) analyzed stable carbon and nitro-gen isotopes from individuals from four archae-ological sites in eastern North America, twopreagricultural and two agricultural, and foundno difference in the duration of nursing amongthe four samples.

Others have applied this method and haveattempted to refine estimates of the duration ofnursing in the past (reviewed by Katzenberget al., 1996; Schurr, 1998). For example,Herring et al. (1998) studied the process ofweaning (introduction of non-breast milkfoods and the gradual cessation of breastfeed-ing) in a nineteenth-century skeletal samplefrom Ontario by using stable nitrogen isotopes,parish records, census data, and skeletal evi-dence. The sample is from an AnglicanChurch cemetery (see Saunders, this volume,for additional background information on thecemetery). Many skeletons of infants andyoung children were included in the sample,and parish records as well as the regional cen-suses could be consulted to obtain multiplesources of mortality data. Careful age determi-nation of the individuals allowed the construc-tion of a mortality profile, which could becompared with the parish records. Since mor-tality increases as infants are weaned, becasueof exposure to infectious agents and loss ofpassive immunity, it was possible to comparemortality with the nitrogen isotope evidencefor weaning. The combined data show that

non-breast milk foods were introduced aroundthe age of 5 months, as indicated by increasingmortality, and that milk continued to be themajor source of protein until around 14months, as indicated by decreasing d15N.

Fuller et al. (2006) carried out a longitudinalstudy of d13C and d15N in fingernails and hair ofeight modern mother–infant pairs. Five infantswere exclusively breastfed, whereas two werebreastfed initially and then formula-fed andone was fed only formula. The results of thisstudy provide confirmation for the use ofstable nitrogen isotopes to document the dur-ation of nursing and show the smaller trophic-level effect in stable carbon isotopes. Such care-fully controlled studies are very important forproviding a more precise understanding of theinformation conveyed through stable isotopeanalysis.

A potential problem with using stable iso-topes of nitrogen to study past breastfeedingpatterns is that the subjects under study died ininfancy and early childhood of unknowncauses. Since d15N may be increased in situ-ations of nutritional stress, it may not be possibleto separate the trophic effect of breastfeedingfrom cause of death in some cases. This possi-bility has been examined, but it is not yet clearthat it is a serious problem (Katzenberg, 1999).

One way to avoid the problem of analyzingbones from individuals who died in infancy orearly childhood is to analyze tissues that wereformed early in postnatal development that canbe isolated from older children and adults. Thepermanent tooth crowns are formed betweenthe ages of three months (incisors) and sevenyears (second premolars and second molars)(Hillson, 1996). The third molar crown formslater but is highly variable. Analysis of nitrogenisotopes in dentine collagen and oxygen andcarbon isotopes in enamel apatite can provideinformation about nursing and weaning for indi-viduals who survived into at least late child-hood. This approach has been followed byWright and Schwarcz (1998). They analyzedstable isotopes of carbon and oxygen from thecarbonate in tooth enamel. Oxygen isotope


measurements are normally performed on thephosphate of bones and teeth since phosphateis less likely to undergo diagenetic changethan is carbonate. Wright and Schwarcz(1998) point out that the procedure for isolatingphosphate is more complex than that forisolating carbonate, and that tooth enamel isless likely to be affected by diagenetic processesthan is bone mineral. The principle of usingstable isotopes of carbon and oxygen is asfollows. Carbon isotopes reflect the introductionof weaning foods, which, in the Americas, areusually maize-based gruel. Therefore, d13C isexpected to increase as this C4-based weaningfood is introduced into the diet. Oxygen iso-topes reflect water source. Body water has ahigher d18O than ingested water because more16O relative to 18O is lost in expired watervapor. Since breast milk incorporates bodywater, it is enriched in the heavier isotope incomparison to other water sources for theinfant. Therefore as the infant is weaned, d18Oshould decrease and breastfed infants shouldhave higher d18O than infants who are notbreastfed. Wright and Schwarcz (1998) com-pared d13C and d18O in enamel carbonatefrom tooth crowns formed at different agesand showed that, as expected, d13C increaseswith age, whereas d18O decreases with age.

RESIDENCE AND MIGRATIONSTUDIES

From the preceding information, it is obviousthat some principles regarding stable isotopevariation may be used to indicate place of resi-dence. Stable carbon isotope abundance ratiosvary based on diet, whereas stable nitrogenisotope abundance ratios vary based on dietand habitat. Oxygen isotope abundance ratiosvary based on climate and water source. Ingeneral, d18O decreases with increasing lati-tude, increasing distance from the coast, andincreasing altitude, because more of the heavyisotope, 18O, falls in precipitation (Dansgaard,1964). Other variables that affect the d18O of

human bone phosphate include humidity andthe plants and animals consumed (Luz andKolodny, 1989). Comparisons of stable isotoperatios in tissues laid down early in life andthose that turn over throughout life may beused to determine whether individuals havemoved. An elegant example of this type ofstudy is that of Sealy et al. (1995), who com-pared enamel, dentine, and bone from five indi-viduals from the southern Cape of Africa. Twoindividuals are prehistoric Khoisan hunter-gatherers, and three others date to the historicperiod. Of the historic burials, two were thoughtto be males of European ancestry and one wasthought to be a female who was brought to theCape as a slave. Enamel, dentine, and bonewere analyzed for stable isotopes of carbon,nitrogen, and strontium. The prehistoric individ-uals have similar isotope values for all tissuesanalyzed. However, two historic individualsshow significant variation in stable isotopevalues for the different tissues, indicating thatthey moved between their childhood andseveral years before their death as adults.

The use of stable oxygen isotopes to identifythe geographical origin of prehistoric peopleshas been carried out on prehistoric humanbones from Mexico (White et al., 1998, 2004)and on historic soldiers from northeasternNorth America (Schwarcz and Schoeninger,1991). The method becomes problematic inmore recent peoples because our food andwater comes from many sources (e.g., Perrierwater from France is popular in NorthAmerica, as is New Zealand lamb). However,in people who were unlikely to move frequently,and who had fairly monotonous diets, oxygenisotope analysis is a useful indicator of residence.

Strontium isotopes are also useful in resi-dence and migration studies since their variationis tied to local geology. Strontium isotopes varydepending on the underlying bedrock that givesrise to soils, and they vary between peoplefeeding on marine versus on terrestrial resources(Sealy et al., 1991). In addition to the studydescribed above by Sealy et al. (1995), stron-tium isotopes have been used in studies of


residence at the Grasshopper Pueblo in Arizona(Ezzo et al., 1997; Price et al., 1994b) and ina Bell Beaker sample from Bavaria (Priceet al., 1994a). They have also been used toidentify immigrants to large ceremonial sitessuch as the Mayan site of Tikal (Wright,2005), and to investigate residential mobilityin the Andes (Knudson et al., 2004). The poten-tial of using strontium isotopes for residencestudies was first pointed out by Ericson(1989). Because of the strontium isotope differ-ences between marine and terrestrial foods, theycan also be used in paleodiet studies (Sealyet al., 1991).

A DAY WITHOUT STABLE ISOTOPES:WHAT HAS THEIR USE ADDED TOOUR KNOWLEDGE?

The use of stable isotope methods to addressarchaeological questions now dates back over30 years. Do we really need these sophisticatedanalyses to tell us what people ate, where theylived, and whether they migrated? Othersources of evidence exist in the archaeologicalrecord for all of these questions. The presenceof a food item in an archaeological assemblagedoes not necessarily imply that it was locallyproduced or that it was eaten. Differential pres-ervation of food remains may confuse interpret-ations of the relative importance of particularfoods. Charred maize cobs and kernels maybe well preserved in comparison with otherplants, and this has resulted in situations wherethe importance of maize was overestimated, asshown by stable isotope analyses. The use ofnitrogen isotopes has allowed researchers tohave a clearer picture of the importance of fresh-water fish in the diets of inland populations.Collectively the use of stable isotopes fordietary reconstruction refines estimates of therelative importance of various foods and there-fore leads to more accurate interpretations ofthe effects of changing diet on health and demo-graphy. It is also the only means currently avail-able of detecting sex and age differences in diet

within past human groups. The use of nitrogenas well as carbon and oxygen isotopes toestimate the duration of breastfeeding also tiesinto demographic reconstruction. Residenceand migration studies help explain populationinteraction and movement in the past.

With advances in instrumentation and theincreasing use of stable isotopes in ecologicalstudies, there has been a very fruitful exchangeof information between bioarchaeologistsand ecologists. This exchange is illustrated inthe example of the Lake Baikal study byKatzenberg and Weber (1999), where much ofthe information on stable carbon isotopevalues for freshwater fish was obtained fromecological studies. The exchange has alsobeen beneficial in the other direction in thatecological studies have benefited from theextensive stable isotope ecology reconstruc-tions carried out by archaeologists such asAmbrose, working in East Africa (1986, 1991;Ambrose and DeNiro, 1986) and Sealy,working in Southern Africa (1986; Sealy andvan der Merwe, 1988; Sealy et al., 1987).

There are still problems to solve, but many ofthe problems identified in 1989, in an importantpaper by Sillen et al. entitled “Chemistry andPaleodietary Research: No More Easy Answers”have seen significant progress. Controlledfeeding studies such as those carried out byAmbrose and Norr (1993; Ambrose, 2001)and Tieszen and Fagre (1993) have addressedquestions about d13C differences in collagenand bone carbonate. Others (Sponheimeret al., 2003) have focused on nitrogen isotopesin the diet and tissues of mammals. Researchhas become more sophisticated in terms ofunderstanding the underlying biochemical pro-cesses that affect stable isotopes. Examplesinclude Ambrose’s (2001) work on nitrogen iso-topes and water stress, Fogel et al.’s (1997)insights into protein stress and amino acidmetabolism, and Evershed’s (1993, Stott andEvershed 1996) research on stable isotopes inlipids. Research carried out on living subjects,using hair and fingernails as sources of proteinfor analysis, allows much more control over

A DAY WITHOUT STABLE ISOTOPES: WHAT HAS THEIR USE ADDED TO OUR KNOWLEDGE? 431

diet, dietary change, and variation (Fuller et al.,2006; O’Connell and Hedges, 1999).

Ethical issues have an impact on all skeletalstudies of past peoples (see Walker, thisvolume). In some situations, visual study is per-mitted but no destructive analyses are allowed.Even though analytical methods have increas-ingly been improved and refined so as torequire milligram quantities of material, somelegislation completely prohibits any destructiveanalyses of human remains. Even when destruc-tive analyses are permitted on skeletal samples,one must be careful to consider the long-termintegrity of collections. Are there other waysto obtain the information? Is the method likelyto be successful and informative? Biologicalanthropologists must continue to work withconcerned groups and to explain the value ofsuch studies to everyone.

It is also important to be aware of limitationson the information that may be provided. Thisawareness is particularly important with paleo-diet studies where consumption of severaldifferent combinations of foods with varyingemphasis on each one may result in the samestable isotope ratios. Use of multiple elements

and familiarity with other archaeologicalsources of dietary information can go a longway toward unraveling this problem. Recentdevelopments using GC/IRMS in whichspecific amino acids, or other specific biomole-cules, are isolated and analyzed can also help tosolve this problem.

The future of stable isotope analysis willinclude many applications that have becomeroutine over the last 30 years, but the fieldis also moving forward very quickly. It isincreasingly important for individuals wishingto pursue this area of study to have trainingin chemistry and biochemistry, since the levelof sophistication has increased substantially. Itis true of both the methods for isolating specificcomponents in bone samples as well as theinterpretation of the results. Indeed manyrecent advances have been made by individualswith backgrounds in biochemistry, geochemis-try, and ecology who have become interestedin archaeological applications. At the sametime, some of the most fruitful collaborationshave been between individuals trained in thephysical sciences and those trained in archaeol-ogy and physical anthropology.

BOX 13.1

When working with prehistoric skeletal remains, actual age at death is not known nor are actualfamilial relationships. However, when studying past diet in individuals buried in historiccemeteries, analyses may be carried out on individuals of known sex, age at death, and familialrelationship since grave markers or coffin plates may be present. Such was the case with asmall historic cemetery in southern Ontario (Katzenberg, 1991). Stable carbon and nitrogen3 iso-topes were analyzed in collagen from 15 individuals, of whom 9 were adults and 6 were childrenranging in age from around the time of birth to six years. The oldest adults probably spent theirearlier years in Great Britain before migrating to Canada. In some cases, this migration is alsoknown since year of death is known (Saunders and Lazenby, 1991). Thus, it is possible in thissmall sample to view stable isotope data from migrants and those born in Canada, to assessdiet, and to look for evidence of nursing and weaning in the bones of infants. In one particularcase, a mother and child were buried together. Two presumed mother/infant pairs were alsoburied in the small cemetery.

3Reanalysis of stable isotopes of nitrogen was carried out after the 1991 publication to obtain a complete set of data. The morerecent results are provided herein.


The data are presented in the table below:

Burial Number Sex Age d13C d15N

1 Female 25 years 218.3 12.01a Female 1 year, 6 months 218.9 14.12 Male 1 year 218.7 13.13 Male 31 years 218.6 12.04 Male 71 years 218.9 11.25 neonate 216.5 12.06 Male 71 years 219.5 11.57 Female 98 years 217.8 12.28 Male 71 years 220.9 10.79 1 year 218.9 12.711 Female 34 years 218.7 11.112 Female 6 years 219.5 10.313 Female 31 years 218.4 11.814 Male 57 years 217.7 11.715 neonate 218.9 12.0

The known mother and child are burials 1 and 1a, respectively. Notice that the d15N of the child,aged one year and six months at death, is 2.1‰ greater than that of the mother, indicating that thechild nursed and may have still been nursing at the time of death. There is some time lag betweenthe cessation of nursing and the deposition of new collagen that no longer reflects the trophic-leveleffect of nursing. In other studies, a small trophic-level shift in d13C has been observed, but it is notpresent here. The d13C of the child is 0.6‰ less enriched than that of the mother.

One of the presumed mother–infant pairs is burial 13 and burial 15. In this case, the infant diedwithin a month of birth and the difference in d15N between mother and infant is only 0.2‰. Giventhat the precision of the analysis is +0.2‰, this difference is not significant, and repeat analysesmay result in the numbers being the same or in the mother having a 0.1% or 0.2‰ difference fromthe infant in the other direction. The third presumed mother–infant pair (and the least certain,based on other sources of evidence) includes burials 5 and 11. Once again, the infant diedwithin the first month of life. Here the d15N of mother and infant differ by 0.9‰. If wecompare the d13C for these mother and infant pairs, we see that they are similar for burials 13and 15 (difference of 0.5‰) but that they differ by 2.2‰ for burials 5 and 11. One wouldexpect a newborn baby to have similar isotope ratios to its mother, so we might suggest thatthis is not a mother–infant pair, based on the stable isotope evidence. (In a matched study of fin-gernails from living mother–infant pairs, Fuller et al. (2006) found that d13C did not differ bymore than 1‰ between mothers and their infants.)

The diet of people from Great Britain is based exclusively on C3 plants such as wheat, oats, andbarley and on domesticated animals that also consumed C3 plants. The average d13C is around220% to 221‰ (van Klinken et al., 2001). Settlers to southern Ontario brought that diet withthem and characteristically have d13C values around 219% to 218‰, as seen here(Katzenberg et al., 2001). The slight enrichment is undoubtedly from the incorporation of theNew World domesticate, maize, which is a C4 plant, that was used in bread. Another C4 plantproduct, cane sugar, was used in baked goods (Katzenberg et al., 2001).

The people buried in the cemetery came from Scotland and moved first to New York in1810 and then to Ontario in 1817. Is it possible to see any dietary differences between theoldest adults, who would have spent some time in Scotland, and the younger individuals? At

A DAY WITHOUT STABLE ISOTOPES: WHAT HAS THEIR USE ADDED TO OUR KNOWLEDGE? 433

first glance, only one individual stands out and that is burial 8, a 71-year-old man. His d13C valueis 220.9‰, the most depleted in the heavier isotope of carbon. Burial 8 died in 1825, so he wouldhave spent the first 56 years of his life in Scotland and only the last 15 years in North America.Collagen turnover is estimated to be about 10–20 years so it is possible that his bones still con-tained some collagen that formed during his many years in Scotland. Burial 7 is the oldest indi-vidual in the cemetery, and her d13C value is 217.8‰, which is different from burial 8 and more inline with a North American diet. However, she died in 1894 so she spent almost her entire life (thelast 84 years) in North America and only the first 14 in Scotland. Therefore, her bone collagenreflects her North American diet.

This example illustrates some of the diverse information that may be obtained from stableisotope data. Obviously, some information in a historic sample will not be available in othersituations; however, the certainty that is gained through such analyses is important in strengthen-ing interpretations in less well-documented samples.

ACKNOWLEDGMENTS

For the first edition of this chapter, whichappeared in 2000, I wish to express my appreci-ation to the Killam Foundation for granting mea Killam Resident Fellowship at the Universityof Calgary during the fall term 1998 for thepurpose of preparing this chapter. I am also grate-ful to Steve Taylor and Roy Krouse, Departmentof Physics, University of Calgary, for providinguseful comments on some sections of thechapter, and to my colleagues Shelley Saundersand Susan Pfeiffer for reading and commentingon the complete chapter. Thanks also to mygraduate students, Sandra Garvie-Lok, TamaraVarney, and Roman Harrison who read and com-mented on the chapter. For the second edition,special thanks to Andrea Waters-Rist for researchassistance and editorial comments, to KristineHepple for assistance with references, and toKate Faccia and Shelley Saunders for editorialsuggestions.

REFERENCES

Ambrose SH. 2001. Controlled diet and climateexperiments on nitrogen isotope ratios of rats.In: Ambrose SH and Katzenberg MA, editors.Biogeochemical Approaches to PaleodietaryAnalyses. New York: Plenum Press. pp. 243–259.

Ambrose SH, Katzenberg MA. 2001.Biogeochemical Approaches to PaleodietaryAnalyses. New York: Plenum Press.

Ambrose SH. 1986. Stable carbon and nitrogenisotope analysis of human and animal diet inAfrica. J Hum Evol 15:707–731.

Ambrose SH. 1990. Preparation and characterizationof bone and tooth collagen for isotopic analysis.J Archaeol Sci 17:431–451.

Ambrose SH. 1991. Effects of diet, climate andphysiology on nitrogen isotope abundances interrestrial foodwebs. J Archaeol Sci 18:293–317.

Ambrose SH, Buikstra J, Krueger HW. 2003. Statusand gender differences in diet at Mound 72,Cahokia, revealed by isotopic analysis of bone.J Anthropol Archaeol 22:217–226.

Ambrose SH, Butler BM. 1997. Stable isotopicanalysis of human diet in the MarianasArchipelago, Western Pacific. Am J PhysAnthropol 104:343–361.

Ambrose SH, DeNiro MJ. 1986. Reconstruction ofAfrican human diet using bone collagen carbonand nitrogen isotope ratios. Nature 319:321–324.

Ambrose SH, DeNiro MJ. 1987. Bone nitrogenisotope composition and climate. Nature325: 201.

Ambrose SH, Krigbaum J. 2003. Bone chemistryand bioarchaeology. J Anthropol Archaeol 22:191–192.

Ambrose SH, Norr L. 1993. Experimental evidencefor the relationship of the carbon isotope ratiosof whole diet and dietary protein to those of


bone collagen and carbonate. In: Lambert JB,Grupe G, editors. Prehistoric Human Bone:Archaeology at the Molecular Level. Berlin:Springer-Verlag. pp. 1–37.

Barrie A, Davies JE, Park AJ, Workman CT. 1989.Continuous-flow stable isotope analysis for biol-ogists. Spectroscopy 4:42–52.

Barrie A, Prosser SJ. 1996. Automated analysis oflight-element stable isotopes by isotope ratiomass spectrometry. In: Boutton TW, Yamasaki S,editors. Mass Spectrometry of Soils. New York:Marcel Dekker. pp. 1–46.

Bender MM. 1968. Mass spectrometric studies ofcarbon 13 variation in corn and other grasses.Radiocarbon 10:468–472.

Benson S, Lennard C, Maynard P, Roux C. 2006.Forensic applications of isotope ratio massspectrometry—A review. Forensic Sci Internat157:1–22.

Bentley RA, Pietrusewsky M, Douglas MT,Atkinson TC. 2005. Matrilocality during the pre-historic transition to agriculture in Thailand?Antiquity 79:865–881.

Blake M, Chisholm BS, Clark JE, Voorhies B,Love MW. 1992. Prehistoric subsistence in theSoconusco region. Curr Anthropol 33:83–94.

Bligh EG, Dyer WJ. 1959. A rapid method of totallipid extraction and purification. Canadian JBiochem Physiol 37:911–917.

Bocherens H, Denys C. 1997. Third InternationalConference on Bone Diagenesis. Bulletin de laSociete Geologique de France, 168 and 169.

Bocherens H, Fogel ML, Tuross N, Zeder M. 1995.Trophic structure and climatic information fromisotopic signatures in Pleistocene cave fauna ofSouthern England. J Archaeol Sci 22:327–340.

Bocherens H, van Klinken GJ, Pollard AM.1999. Proceedings of the 5th AdvancedSeminar on Paleodiet, Centre de RecherchesArcheologiques, Valbonne, 1–5, September1997 J Archaeol Sci 26.

Boutton TW, Klein PD, Lynott MJ, Price JE,Tieszen LL. 1984. Stable carbon isotope ratiosas indicators of prehistoric human diet. In:Turnlund J-R, Johnson PE, editors. StableIsotopes in Nutrition. ACS Symposium Series258, 191–204. Washington, D.C.: AmericanChemical Society.

Brill WJ. 1977. Biological nitrogen fixation. In:Kretchmer N, Robertson W, editors. Human

Nutrition: Readings from Scientific American.San Francisco, Calif.: W.H. Freeman andCompany. pp. 18–27.

Brown TA, Nelson DE, Vogel JS, Southon JR. 1988.Improved collagen extraction by modified Longinmethod. Radiocarbon 30:171–177.

Bryant JD, Koch PL, Froelich PN, Shoewers WJ,Genna BJ. 1996. Oxygen isotope partitioningbetween phosphate and carbonate in mammalianapatite. Geochim Cosmochim Acta 60:5145–5148.

Buikstra JE, Konigsberg LW, Bullington J. 1986.Fertility and development of agriculture in theprehistoric midwest. Am Antiq 51:528–546.

Buikstra JE, Milner GR. 1991. Isotopic and archaeo-logical interpretations of diet in the centralMississippi valley. J Archaeol Sci 18:319–329.

Child AM. 1995. Towards an understanding of themicrobial decomposition of archaeological bonein the burial environment. J Archaeol Sci 22:165–174.

Chisholm BS, Nelson DE, Hobson KA, SchwarczHP, Knyf M. 1983. Carbon isotope measurementtechniques for bone collagen: notes for thearchaeologist. J Archaeol Sci 10:355–360.

Chisholm BS, Nelson DE, Schwarcz HP. 1982.Stable carbon isotope ratios as a measure ofmarine versus terrestrial protein in ancient diets.Science 216:1131–1132.

Collins MJ, Nielsen-Marsh CM, Hiller J, Smith CI,Roberts JP, Prigodich RV, Wess TJ, Csapo J,Millard AR, Turner-Walker G. 2002. The survi-val of organic matter in bone: a review.Archaeometry 44:383–394.

Collins MJ, Riley MS, Child AM, Turner-Walker G.1995. A basic mathematical simulation of thechemical degradation of ancient collagen.J Archaeol Sci 22:175–183.

Coltrain JB, Hayes MG, O’Rourke DH. 2004.Sealing, whaling and caribou: the skeletalisotope chemistry of Eastern Arctic foragers.J Archaeol Sci 31:39–57.

Coltrain JB, Hayes MG, O’Rourke DH. 2006.Hrdlicka’s Aleutian population-replacementhypothesis: a radiometric evaluation. CurrAnthropol 47:537–548.

Coplen TB. 1994. Reporting of stable hydrogen,carbon, and oxygen isotopic abundances. Pure& Applied Chemistry, 66(2):273–276.

REFERENCES 435

Copley MS, Berstan R, Dudd SN, Straker V,Payne S, Evershed RP. 2005a. Dairying in anti-quity. I. Evidence from absorbed lipid residuesdating to the British Iron Age. J Archaeol Sci32:485–503.

Copley MS, Berstan R, Straker V, Payne S,Evershed RP. 2005b. Dairying in antiquity. II.Evidence from absorbed lipid residues dating tothe British Bronze Age. J Archaeol Sci 32:505–521.

Corr LT, Sealy JC, Horton MC, Evershed RP. 2005.A novel marine dietary indicator utilisingcompound-specific bone collagen amino acidd13C values of ancient humans. J Archaeol Sci32:321–330.

Craig H. 1954. Carbon 13 in plants and the relation-ships between carbon 13 and carbon 14 variationsin nature. J Geol 62:115–149.

Dansgaard W. 1964. Stable isotopes in precipitation.Tellus 16:436–468.

Decker KW, Tieszen LL. 1989. Isotopic reconstruc-tion of Mesa Verde diet from Basketmaer III toPueblo III. Kiva 55:33–47.

DeNiro MH. 1985. Postmortem preservation andalteration of in vivo bone collagen isotope ratiosin relation to palaeodietary reconstruction.Nature 317:806–809.

DeNiro MJ, Epstein S. 1978. Influence of diet on thedistribution of carbon isotopes in animals.Geochim Cosmochim Acta 42:495–506.

DeNiro MJ, Epstein S. 1981. Influence of diet on thedistribution of nitrogen isotopes in animals.Geochim Cosmochim Acta 45:341–351.

Ehleringer JR, Rundel PW. 1989. Stable Isotopes:History, Units, and Instrumentation. In: RundelPW, Ehleringer JR, Nagy KA, editors. StableIsotopes in Ecological Research. New York:Springer-Verlag. pp. 1–15.

Ericson JE. 1985. Strontium isotope characterizationin the study of prehistoric human ecology. J HumEvol 14:503–514.

Ericson JE, West M, Sullivan CH, Krueger HW.1989. The development of maize agriculture inthe Viru valley, Peru. In: Price TD, editor. TheChemistry of Prehistoric Bone. Cambridge:Cambridge University Press. pp. 68–104.

Evershed RP. 1993. Biomolecular archaeology andlipids. World Archaeol 25:74–93.

Ezzo JA, Johnson CM, Price TD. 1997. Analyticalperspectives on prehistoric migration: A case

study from east-central Arizona. J Archaeol Sci24:447–466.

Fernandez-Jalvo Y, Sanchez-Chillon B, Alcala L.2002. The fourth international meeting on bonediagenesis. Archaeometry 44:315–318.

Fogel ML, Tuross N, Johnson BJ, Miller GH. 1997.Biogeochemical record of ancient humans.Organic Geochem 27:275–287.

Fogel M, Tuross N, Owsley DW. 1989. Nitrogenisotope tracers of human lactation in modernand archaeological populations. Annual reportof the Director; Geophysical Laboratory.Washington, D.C.: Carnegie Institution. pp.111–117.

Folch J, Lees M, Stanley GHS. 1957. A simplemethod for the isolation and purification of totallipids from animal tissues. J Biol Chem 226:497–509.

France RL. 1995. Differentiation between littoral andpelagic food webs in lakes using stable carbonisotopes. Limnol Oceanog 40:1310–1313.

Fry B. 2006. Stable Isotope Ecology. New York:Springer.

Fuller BT, Fuller JL, Harris DA, Hedges REM.2006. Detection of breastfeeding and weaningin modern human infants with carbon and nitro-gen stable isotope ratios. Am J Phys Anthropol129:279–293.

Garvie-Lok SJ, Varney TL, Katzenberg MA. 2004.Preparation of bone carbonate for stable isotopeanalysis: the effects of treatment time and acidconcentration. J Archaeol Sci 31:763–776.

Griffiths H. 1998. Stable Isotopes: Integration ofBiological, Ecological and GeochemicalProcesses. Environmental Plant Biology Series.Oxford: BIOS Scientific Publishers, Ltd.

Gross ML. 1979. Mass spectrometry. In: Bauer HH,Christian GD, O’Reilly JE, editors. InstrumentalAnalysis. Boston: Allyn and Bacon. pp. 443–486.

Hall RL. 1967. Those late corn dates: isotopic frac-tionation as a source of error in Carbon-14dates. Michigan Archaeol 13:171–180.

Hare PE, Estep M. 1982. Carbon and nitrogen isoto-pic composition of amino acids in modern andfossil collagens. Carnegie Inst Wash Yrbk 82:410–414.

Hare PE, Fogel ML, Stafford TW Jr, Mitchell AD,Hoering TC. 1991. The isotopic composition ofcarbon and nitrogen in individual amino acids


isolated from modern and fossil proteins.J Archaeol Sci 18:227–292.

Harrison RG, Katzenberg MA. 2003. Paleodietstudies using stable carbon isotopes from boneapatite and collagen: examples from SouthernOntario and San Nicolas Island, California.J Anthropol Archaeol 22:227–244.

Hayden B, Chisholm B, Schwarcz HP. 1987. Fishingand foraging: marine resources in the upperpaleolithic of France. In: Soffer O, editor. ThePleistocene Old World: Regional Perspectives.New York: Plenum. pp. 279–291.

Heaton THE. 1987. The 15N/14N ratios of plantsin South Africa and Namibia: relationship toclimate and coastal/saline environments.Oecologia 74:236–246.

Heaton THE, Vogel JC, von la Chevallerie G,Collett G. 1986. Climatic influence on the isoto-pic composition of bone nitrogen. Nature 322:822–823.

Hecky RE, Hesslein RH. 1995. Contributions ofbenthic algae to lake food webs as revealed bystable isotope analysis. J N Am Benthol Soc 14:631–653.

Hedges REM. 2004. Isotopes and red herrings: com-ments on Milner et al. and Liden et al. Antiquity78:34–38.

Hedges REM, van Klinken GJ. 1995. Special issueon bone diagenesis. J Archaeol Sci 22:145–341.

Hedges REM. 2002. Bone diagenesis: an overviewof processes. Archaeometry 44:319–328.

Herring DA, Saunders SR, Katzenberg MA. 1998.Investigating the weaning process in past popu-lations. Am J Phys Anthropol 105:425–439.

Hillson A. 1996. Dental Anthropology. Cambridge:Cambridge University Press.

Hobson KA, Alisauskas RT, Clark RG. 1993. Stablenitrogen isotope enrichment in avian tissuesdue to fasting and nutritional stress: implicationsfor isotopic analyses of diet. Condor 95:388–394.

Hobson KA, Clark RG. 1992. Assessing avian dietsusing stable isotopes II: factors influencing diet-tissue fractionation. Condor 94:187–195.

Hoefs J. 1997. Stable Isotope Geochemistry. Berlin:Springer-Verlag.

Katzenberg MA. 1989. Stable isotope analysis ofarchaeological faunal remains from southernOntario. J Archaeol Sci 16:319–329.

Katzenberg MA. 1999. A re-examination of factorscontributing to elevated stable nitrogen isotopevalues in infants and young children (abstract).Am J Phys Anthropol 28(suppl):165.

Katzenberg MA, Weber A. 1999. Stable isotopeecology and paleodiet in the Lake Baikal regionof Siberia. J Archaeol Sci 26:651–665.

Katzenberg MA. 1991. Stable isotope analysis ofremains from the Harvie family. In: SaundersSR, Lazenby R, editors. The Links that Bind:The Harvie Family Nineteenth Century BuryingGround. Occasional Papers in NortheasternArchaeology, No. 5. Dundas, Ontario:Copetown Press. pp. 65–69.

Katzenberg MA, Harrison RG. 1997. What’s in abone? Recent advances in archaeological bonechemistry. J Archaeol Res 5:265–293.

Katzenberg MA, Herring DA, Saunders SR. 1996.Weaning and infant mortality: evaluating theskeletal evidence. Yrbk Phys Anthropol 39:177–199.

Katzenberg MA, Kelley JH. 1991. Stable isotopeanalysis of prehistoric bone from the SierraBlanca region of New Mexico. In: Beckett PH,editor. Mogollon V: Proceedings of the 1988Mogollon Conference. Las Cruces, N.M.:COAS Publishing and Research. pp. 207–209.

Katzenberg MA, Lovell NC. 1999. Stable isotopevariation in pathological bone. Internat JOsteoarchaeol 9:316–324.

Katzenberg MA, Saunders SR, Abonyi S. 2001.Bone chemistry, food and history: a case studyfrom 19th century upper Canada. In: AmbroseSH, Katzenberg MA, editors. BiogeochemicalApproaches to Paleodietary Analysis.New York: Plenum Press. pp. 1–22.

Katzenberg MA, Schwarcz HP, Knyf M, Melbye FJ.1995. Stable isotope evidence for maizehorticulture and paleodiet in southern Ontario,Canada. Am Antiquity 60:335–350.

Keegan WF, DeNiro MJ. 1988. Stable carbon- andnitrogen-isotope ratios of bone collagen used tostudy coral-reef and terrestrial components ofprehistoric Bahamian diet. Am Antiquity 53:320–336.

Kiyashko SI, Mamontov AM, Chernyaev M Zh.1991. Food web analysis of Lake Baikal fish byratios of stable carbon isotopes. Doklady.Biological Sciences, 318(1–6):274.

REFERENCES 437

Knudson KJ, Price TD, Buikstra JE, Blom DE. 2004.The use of strontium isotope analysis to investi-gate Tiwanaku migration and mortuary ritual inBolivia and Peru. Archaeometry 46:5–18.

Koch PL, Tuross N, Fogel ML. 1997. The effects ofsample treatment and diagenesis on the isotopicintegrity of carbonate in biogenic hydroxylapa-tite. J Archaeol Sci 24:417–429.

Koch PL, Burton J. 2003. Bone chemistry. Internat JOsteoarchaeol 13:1–113.

Krueger HW. 1991. Exchange of carbon with bio-logical apatite. J Archaeol Sci 18:355–361.

Krueger HW, Sullivan CH. 1984. Models for carbonisotope fractionation between diet and bone.In: Turnlund JE, Johnson PE, editors. StableIsotopes in Nutrition. American ChemicalSociety Symposium Series 258. Washington,D.C.: American Chemical Society. pp. 205–222.

Lambert JB, Grupe G. 1993. Prehistoric humanbone: archaeology at the molecular level.Berlin: Springer-Verlag.

Larsen CS, Schoeninger MJ, van der Merwe NJ,Moore KM, Lee-Thorp JA. 1992. Carbon andnitrogen stable isotopic signatures of humandietary change in the Georgia Bight. Am J PhysAnthropol 89:197–214.

Latkoczy C, Prohaska T, Watkins M, Teschler-Nicola M, Stingeder G. 2001. Strontium isotoperatio determination in soil and bone samplesafter on-line matrix separation by coupling ionchromatography (HPIC) to an inductivelycoupled plasma sector field mass spectrometer(ICP-SFMS). J Anal Atomic Spectrom 16:806–811.

Lee-Thorp J, Sponheimer M. 2006. Contributions ofbiogeochemistry to understanding hominindietary ecology. Yrbk Phys Anthropol 49:131–148.

Lee-Thorp JA. 1989. Stable Carbon Isotopes in DeepTime: the Diets of Fossil Fauna and Hominids.Ph.D. Dissertation, University of Cape Town,Rondebosch, South Africa.

Lee-Thorp JA, Sealy JC, van der Merwe MJ. 1989.Stable carbon isotope ratio differences betweenbone collagen and bone apatite, and their relation-ship to diet. J Archaeol Sci 16:585–599.

Lee-Thorp JA, van der Merwe NJ. 1991. Aspects ofthe chemistry of modern and fossil biologicalapatites. J Archaeol Sci 18:343–354.

LeGeros RZ, Trautz OR, LeGeros JP, Klein E. 1967.Apatite crystallites: effects of carbonate on mor-phology. Science 155:1409–1411.

Lichtfouse E. 2000. Compound-specific isotopeanalysis. Application to archaeology, biomedicalsciences, biosynthesis, environment, extraterres-trial chemistry, food science, forensic science,humic substances, microbiology, organic geo-chemistry, soil science and sport. RapidCommun Mass Spectrom 14:1337–1344.

Liden K, Takahashi C, Nelson DE. 1995. The effectsof lipids in stable carbon isotope analysis and theeffects of NaOH treatment on the composition ofextracted bone collagen. J Archaeol Sci 22:321–326.

Liden K, Eriksson G, Nordqvist B, Gotehrstrom A,Bendixen E. 2004. The wet and the wild followedby the dry and the tame—or did they occur at thesame time? Diet in Mesolithic—Neolithicsouthern Sweden. Antiquity 78:23–33.

Longin R. 1971. New method of collagen extractionfor radiocarbon dating. Nature 230:241–242.

Lowdon JA. 1969. Isotopic fractionation in corn.Radiocarbon 11:391–393.

Lubell D, Jackes M, Schwarcz H, Knyf M,Meiklejohn C. 1994. The Mesolithic-Neolithictransition in Portugal: Isotopic and dentalevidence of diet. J Archaeol Sci 21:201–216.

Luz B, Kolodny Y. 1989. Oxygen isotope variationin bone phosphate. Appl Geochem 4:317–323.

Lyon TDB, Baxter MS. 1978. Stable carbon isotopesin human tissues. Nature 48:187–191.

Macko SA, Uhle ME, Engel MH, Andrusevich V.1997. Stable nitrogen isotope analysis of aminoacid enantiomers by gas chromatography/com-bustion/isotope ratio mass spectrometry. AnalyChem 69:926–929.

Matson RG, Chisholm B. 1991. Basketmaker IIsubsistence: Carbon isotopes and other dietaryindicators from Cedar Mesa, Utah. AmAntiquity 56:444–459.

Milner N, Craig OE, Bailey GN, Pederson K,Anderson SH. 2004. Something fishy in theNeolithic? A re-evaluation of stable isotopeanalysis of Mesolithic and Neolithic coastalpopulations. Antiquity 78:9–22.

Minagawa M, Wada E. 1984. Stepwise enrichmentof 15N along food chains: further evidence andthe relation between d15N and animal age.Geochim Cosmochim Acta 48:1135–1140.


Nielsen-Marsh CM, Hedges REM. 2000a. Patternsof diagenesis in bone I: the effects of site environ-ments. J Archaeol Sci 27:1139–1150.

Nielsen-Marsh CM, Hedges REM. 2000b. Patternsof diagenesis in bone II: effects of acetic acidtreatment and the removal of diagenetic CO3

22.J Archaeol Sci 27:1151–1159.

Norr LC. 1991. Nutritional Consequences ofPrehistoric Subsistence Strategies in LowerCentral America. Unpublished Doctoral Thesis,University of Illinois at Urbana-Champaign.

O’Connell TC, Hedges REM. 1999. Investigationsinto the effect of diet on modern human hairisotopic values. Am J Phys Anthropol 108:409–425.

Pate FD. 1994. Bone chemistry and paleodiet.J Archaeol Method Theory 1:161–209.

Pate FD, Craib JL, Heathcote GM. 2001. Stable iso-topic analysis of prehistoric human diet in theMariana Islands, western Pacific. AustralianArchaeol 52:1–4.

Price TD. 1989. The Chemistry of Prehistoric Bone.Cambridge: Cambridge University Press.

Price TD, Grupe G, Schroter P. 1994a.Reconstruction of migration patterns in the BellBeaker period by stable strontium isotopeanalysis. Appl Geochem 9:413–417.

Price TD, Johnson CM, Ezzo JM, Ericson J, BurtonJH. 1994b. Residential mobility in the prehistoricsouthwest United States: a preliminary studyusing strontium isotope analysis. J Archaeol Sci21:315–330.

Prohaska T, Latkoczy C, Schultheis G, Teschler-Nicola M, Stingeder G. 2002. Investigation ofSr isotope ratios in prehistoric human bones andteeth using laser ablation ICP-MS and ICP-MSafter Rb/Sr separation. J Analy Atom Spectrom17:887–891.

Richards MP, Schulting RJ, Hedges REM. 2003.Sharp shift in diet at onset of Neolithic. Nature425:366.

Rundel PW, Ehleringer JR, Nagy KA. 1989. StableIsotopes in Ecological Research. EcologicalStudies: Analysis and Synthesis No. 68.New York: Springer-Verlag.

Saunders SR, Lazenby R. 1991. The Links thatBind: The Harvie Family Nineteenth CenturyBurying Ground. Dundas, Ontario: CopetownPress.

Schoeninger MJ, DeNiro MJ. 1982. Carbon isotoperatios of apatite from fossil bone cannot be usedto reconstruct diets of animals. Nature 297:577–578.

Schoeninger MJ, DeNiro MJ. 1984. Nitrogen andcarbon isotopic composition of bone collagenfrom marine and terrestrial animals. GeochimCosmochim Acta 48:625–639.

Schoeninger MJ, Moore K. 1992. Bone stableisotope studies in archaeology. J WorldPrehistory 6:247–296.

Schoeninger MJ, Moore KM, Murray ML, KingstonJD. 1989. Detection of bone preservation inarchaeological and fossil samples. AppliedGeochem 4:281–292.

Schurr MR, Powell ML. 2005. The role of changingchildhood diets in the prehistoric evolution offood production: an isotopic assessment. Am JPhys Anthropol 126:278–294.

Schurr MR. 1998. Using stable nitrogen-isotopes tostudy weaning behavior in past populations.World Archaeol 30:327–342.

Schurr MR, Redmond BG. 1991. Stable isotopeanalysis of incipient maize horticulturists fromthe Gard Island 2 Site. Midcontinent J Archaeol16:69–84.

Schwarcz HP, Hedges REM, Ivanovich M. 1989.Editorial comments on the first internationalworkshop on fossil bone. Applied Geochem 4:211–213.

Schwarcz HP, Melbye FJ, Katzenberg MA, Knyf M.1985. Stable isotopes in human skeletons ofsouthern Ontario: reconstructing paleodiet.J Archaeol Sci 12:187–206.

Schwarcz HP, Schoeninger MJ. 1991. Stable isotopeanalyses in human nutritional ecology. Yrbk PhysAnthropol 34:283–322.

Sealy JC, Armstrong R, Schrire C. 1995. Beyondlifetime averages: tracing life histories throughisotopic analysis of different calcified tissuesfrom archaeological human skeletons. Antiquity69:290–300.

Sealy JC, van der Merwe NJ. 1988. Social, spatialand chronological patterning in marine food useas determined by 13C measurements ofHolocene human skeletons from the south-western Cape, South Africa. World Archaeol20:87–102.

Sealy JC, van der Merwe NJ, Sillen AKFJ, KruegerHW. 1991. 87Sr/86Sr as a dietary indicator in

REFERENCES 439

modern and archaeological bone. J Archaeol Sci18:399–416.

Sealy JC, van der Merwe NJ, Thorp JA, Lanham JL.1987. Nitrogen isotopic ecology in southernAfrica: implications for environmental anddietary tracing. Geochim Cosmochim Acta 51:2707–2717.

Sealy JC. 1986. Stable carbon isotopes and prehisto-ric diets in the southwestern Cape Province,South Africa. Oxford: BAR International Series293.

Shearer G, Kohl DH, Virginia RA, Bryan BA,Skeeters JL, Nilsen ET, Sharifi MR, RundelPW. 1983. Estimates of N2-fixation from vari-ation in the natural abundance of 15N inSonoran desert ecosystems. Oecologia 56:365–373.

Sillen A, Hall G, Richardson S, Armstrong R. 1998.87Sr/86Sr ratios in modern and fossil food-webs of the Sterkfontein Valley: Implicationsfor early hominid habitat preference. GeochimCosmochim Acta 62:2463–2473.

Sillen A, Sealy JC. 1995. Diagenesis of strontium infossil bone: a reconsideration of Nelson et al.J Archaeol Sci 22:313–320.

Sillen A. 1986. Biogenic and diagenetic Sr/Ca inPlio-Pleistocene fossils of the Omo ShanguraFormation. Paleobiology 12:311–323.

Sillen A, Armelagos G. 1991. Proceedings of the 2ndAdvanced Seminar on Paleodietary Research:Chemistry and Paleodiet. J Archaeol Sci 18(3):225–416.

Smith BN, Epstein S. 1971. Two categories of13C/12C ratios for higher plants. Plant Physiol47:380–384.

Spielmann KA, Schoeninger MJ, Moore K. 1990.Plains-Pueblo interdependence and human dietat Pecos Pueblo, New Mexico. Am Antiquity55:745–765.

Sponheimer M, Lee-Thorp JA. 1999. Isotopic evi-dence for the diet of an early hominid,Australopithecus africanus. Science 283:368–370.

Sponheimer M, Robinson T, Ayliffe L, Roeder B,Hammer J, Passey B, West A, Cerling T,Dearing D, Ehleringer J. 2003. Nitrogen isotopesin mammalian herbivores: hair d15N valuesfrom a controlled feeding study. Internat JOsteoarchaeol 13:80–87.

Sponheimer M, Passney BH, de Ruiter DJ, Guatelli-Steinberg D, Cerling TE, Lee-Thorp JA. 2006.Isotopic evidence for dietary variability in theearly hominin Paranthropus robustus. Science314:980–982.

Stafford TW Jr, Hare PE, Currie L, Jull AJT,Donahue DJ. 1991. Accelerator radiocarbondating at the molecular level. J Archaeol Sci 18:35–72.

Stott AW, Evershed RP. 1996. d13C analysis ofcholesterol preserved in archaeological bonesand teeth. Analy Chem 68:4402–4408.

Stott AW, Evershed RP, Jim S, Jones V, Rogers JM,Tuross N, Ambrose S. 1999. Cholesterol as a newsource of palaeodietary information: experimen-tal approaches and archaeological applications.J Archaeol Sci 26:705–716.

Stuart-Williams HLQ. 1996. Analysis of theVariation of Oxygen Isotopic Composition ofMammalian Bone Phosphate. UnpublishedDoctoral Thesis, McMaster University,Hamilton, Ontario.

Stuart-Williams HLQ, Schwarcz HP. 1995. Oxygenisotopic analysis of silver orthophosphate usinga reaction with bromine. Geochim CosmochimActa 59:3837–3841.

Stuart-Williams HLQ, Schwarcz HP. 1997. Oxygenisotopic determination of climatic variationusing phosphate from beaver bone, toothenamel, and dentine. Geochim CosmochimActa 61:2539–2550.

Sullivan CH, Krueger HW. 1981. Carbon isotopeanalysis in separate chemical phases in modernand fossil bone. Nature 292:333–335.

Tauber H. 1981. 13C evidence for dietary habits ofprehistoric man in Denmark. Nature 292:332–333.

Tieszen LL, Fagre T. 1993. Effect of diet quality andcomposition on the isotopic composition ofrespiratory CO2, bone collagen, bioapatite andsoft tissues. In: Lambert JB, Grupe G, editors.Prehistoric Human Bone: Archaeology at theMolecular Level. Berlin: Springer-Verlag. pp.121–155.

Triffit JT. 1980. The organic matrix of bone tissue.In: Urist MR, editor. Fundamental and ClinicalBone Physiology. Philadelphia, Pa.: J.B.Lippincoot Co. pp. 45–82.


Tripp JA, Hedges REM. 2004. Single-compoundisotopic analysis of organic materials in archaeol-ogy. LC GC North America 22:1098–1106.

Tuross N. 2002. Alterations in fossil collagen.Archaeometry 44:427–434.

Tuross N, Eyre DR, Holtrop ME, Glimcher MJ,Hare PE. 1980. Collagen in Fossil Bones.New York: John Wiley & Sons.

Tuross N, Fogel ML, Hare PE. 1988. Variability inthe preservation of the isotopic composition ofcollagen from fossil bone. GeochimCosmochim Acta 52:929–935.

van der Merwe NJ, Vogel JC. 1978. 13C content ofhuman collagen as a measure of prehistoric diet inWoodland North America. Nature 276:815–816.

van der Merwe NJ. 1982. Carbon isotopes, photo-synthesis, and archaeology. Am Scient 70:596–606.

van Klinken GJ. 1999. Bone collagen quality indi-cators for palaeodietary and radiocarbonmeasurements . J Archaeol Sci 26:687–695.

van Klinken GJ, Richards MP, Hedges REM. 2001.An overview of causes for stable isotopic vari-ations in past European human populations:environmental, ecophysiological and culturaleffects. In: Ambrose SH, Katzenberg MA,editors. Biogeochemical Approaches toPaleodietary Analysis. New York: KluwerAcademic/Plenum Publishers. pp. 39–63.

Vanderklift MA, Ponsard S. 2003. Sources of vari-ation in consumer-diet d15N enrichment: ameta-analysis. Oecologia 136:169–182.

Virginia RA, Delwiche CC. 1982. Natural 15N abun-dance of presumed N2-fixing and non-N2-fixingplants from selected ecosystems. Oecologia 54:317–325.

Vogel JC. 1978. Isotopic assessment of the dietaryhabits of ungulates. S African J Sci 74:298–301.

Vogel JC, van der Merwe NJ. 1977. Isotopic evi-dence for early maize cultivation in New YorkState. Am Antiquity 42:238–242.

Walker PL, DeNiro MJ. 1986. Stable nitrogen andcarbon isotope ratios in bone collagen asindices of prehistoric dietary dependence onmarine and terrestrial resources in southernCalifornia. Am J Phys Anthropol 71:51–61.

White CD, Armelagos GJ. 1997. Osteopenia andstable isotope ratios in bone collagen of Nubianfemale mummies. Am J Phys Anthropol 103:185–199.

White CD, Spence MW, Stuart-Williams HLQ,Schwarcz HP. 1998. Oxygen isotopes and theidentification of geographical origins: the Valleyof Oaxaca versus the Valley of Mexico.J Archaeol Sci 25:643–655.

White C, Longstaffe FJ, Law KR. 2004. Exploringthe effects of environment, physiology anddiet on oxygen isotope ratios in ancientNubian bones and teeth. J Archaeol Sci 31:233–250.

Wright LE, Schwarcz HP. 1998. Stable carbon andoxygen isotopes in human tooth enamel: identify-ing breastfeeding and weaning in prehistory. AmJ Phys Anthropol 106:1–18.

Wright LE. 2005. Identifying immigrants to Tikal,Guatemala: defining local variability in strontiumisotope ratios of human tooth enamel. J ArchaeolSci 32:555–566.

Wyckoff RWG. 1980. Collagen in fossil bones. In:Hare PE, Hoering TKCKJ, editors.Biogeochemistry of Amino Acids. New York:John Wiley & Sons. pp. 17–22.

Yesner DR, Torres MJF, Guichon RA, Borrero LA.2003. Stable isotope analysis of human boneand ethnohistoric subsistence patterns in Tierradel Fuego. J Anthropol Archaeol 22:279–291.

Zohary T, Erez J, Gophen M, Berman-FrankIlana, Stiller M. 1994. Seasonality of stablecarbon isotopes within the pelagic food webof Lake Kinneret. Limnol Oceanog 39:1030–1043.

REFERENCES 441

Documents

STABLE ISOTOPE ANALYSIS: A TOOL FOR STUDYING PAST DIET, DEMOGRAPHY, AND … · 2016-09-28 · direct evidence of past diets, including infant diets. Knowledge gained from bone chemistry