Electron microprobe analysis of elemental distribution in excavated human femurs.код для вставкиСкачать
AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 62:409-423 (1983) Electron Microprobe Analysis of Elemental Distribution in Excavated Human Femurs JOSEPH B. LAMBERT, SHARON VLASAK SIMPSON, JANE E. BUIKSTRA, AND DOUGLAS HANSON Departments of Chemistry and A nthropobgy, Northwestern University, Euanston, Illinois 60201 KEY WORDS Gibson site, Ledders site, Diagenesis, Ancient diet ABSTRACT Elemental distributions have been determined for femur cross sections of eight individuals from the Gibson and Ledders Woodland sites. The analyses were obtained by x-ray fluorescence with a scanning electron microscope. Movement of a n element from soil to bone should give rise to inhomogeneous distributions within the bone. We found that the distributions of zinc, strontium, and lead are homogeneous throughout the femur. In contrast, iron, aluminum, potassium, and manganese show clear buildup along the outer surface of the femur and sometimes along the inner (endosteal) surface, as the result of postmortem enrichment. The buildup penetrates 10-400 pm into the femur. The major elements calcium and sodium show homogeneous distributions, but considerable material could be lost by leaching (10-15%) without causing a palpable effect on the electron maps. Magnesium shows buildup on the outer edge of some samples. These results suggest that diagenetic contamination may exclude Fe, Al, K, Mn, and probably Mg from use as indicators of ancient data. The homogeneous distributions of Zn, Sr, and Pb suggest that these elements are not altered appreciably and may serve as useful dietary indicators. Chemical analysis of the mineral portion of excavated human bone has come into increased usage as a method for the assessment of the ancient diet (Brown, 1973; Gilbert, 1975; Lambert et al., 1979; Schoeninger, 1979; Sillen, 1981; Price and Kavanagh, 1982). Successful employment of a given chemical element requires that analyzed elemental proportions must accurately reflect those present in the skeleton a t the time of death; i.e., analyzed levels must be equal to or proportional to the original levels. Alteration of elemental levels in buried bone can occur by several mechanisms (Parker and Toots, 1980).Infiltration of elements from the environment can occur either into voids formed by loss of organic components such as collagen or into the inorganic matrix by substitution. Leaching of elements can occur frcm the buried bone into the surrounding soil. Some elements can move in either direction. Because understanding postmortem effects is central to the use of any element for the 0 1983 ALAN R. LISS, INC assessment of ancient diet, we have explored diagenesis by four different approaches. (1) Comparison of observed elemental levels in excavated bones with those in normal, modern samples can suggest diagenetic alteration when differences are substantial (Lambert et al., 1979). (2) Comparison of human ribs and femurs shows differences in elemental levels that may be attributed to diagenetic causes (Lambert et al., 1982). (3) Analysis of soil directly around human burials reveals changes caused both by leaching of elements from bone to soil and by infiltration of elements from soil to bone (Lambert et al., 1983). (4) Finally, we report herein on the distribution of elements in femoral cross sections by electron microprobe analysis. Detailed analysis of the location of buildup or loss of elements can provide information about the processes of infiltration or leaching. Parker and Toots (Parker, 1967; Parker Received June 20, 1983; revision accepted August 16, 1983 410 J.B. LAMBERT, S.V. SIMPSON, J.E. BUIKSTRA, AND D. HANSON and Toots, 1970, 1974) carried out such studies on numerous nonhuman samples by scanning electron microscopy. More recently Badone and Farquhar (1982) analyzed concentration gradients by neutron activation analysis, also on nonhuman samples. In the present study we have examined human samples. We report results on several elements that may be useful dietary indicators. MATERIALS AND METHODS Femurs were chosen from the Gibson (Middle Woodland) population (samples 1-17,2-40, 3-1, and 5-28) and from the Ledders (Late Woodland) population (1-16, 1-28, 1-41,and 1146) (Lambert et al., 1979). Cross-sectional slices were made through the midshaft, and thin sections were prepared for analysis by x-ray fluorescence with a scanning electron microscope (SEM) equipped with a microprobe. Initial cross sections of 1-inch width were embedded in a methyl methacrylate polymer resin. From these, thin sections were cut transversely with a high-speed diamond blade bandsaw at the Argonne National Laboratory. Thin sections were soaked in ethanol to remove surface debris. In order to identify the outer and inner edges of the femur, the sections were cut with an Exacto knife into triangular wedges, the longer edge of which corresponded to the outer surface (the exterior) of the bone. The wedges were mounted with tweezers onto a 0.5-inch-diameter carbon stub by means of liquid graphite. The dried carbon stub was mounted with epoxy onto a 0.5-inch-diameter aluminum stub, which was used as a mount in the holder of the SEM vacuum chamber. Graphite was applied to all edges of the carbon stub to ensure good conductive contact through the samples. Since bone is nonconductive, the entire mount was carbopcoated to a depth of approximately 300 A in a JEOL Model JEE 4C evaporator. Analysis of Na, Mg, and A1 was carried out at Northern Illinois University on a JEOL JSM-50A scanning electron microscope with a wavelength dispersive crystal x-ray spectrometer attachment. The SEM was equipped with a preamplified pulse-height analyzer, pX system 7000 Naisel control computer, and Canberra gas flow proportional counter. The gas flow detector contained a mixture of 10% methane and 90% argon. Polypropylene windows on the detector were used to allow penetration of light x-rays from Na, Mg, and A1 to the counter. The analyzing crystal was rubidium acid phthalate (RAP, ortho-RbOOCC6H4-COOH). The electron microprobe analysis of all other elements was carried out in the Northwestern University Materials Research Center SEM Facility, with a n identical instrument that used a xenon-filled proportional counter. The windows of the detector were covered with a thin layer of beryllium to allow penetration of heavy metal x-rays to the counter. The linear focusing x-ray spectrometer contained pentaerythritol (PET, C(CHzOH),) or lithium fluoride (LiF) crystal. The RAP, PET, or LiF crystals were used to calibrate the spectrometer wavelength, which was optimized with appropriate standards: KaKanoi Horneblende (A1and Mg), AN 30%NdCa Anarthite (Na), KN03, Sr(NO3)2, ZnO, CaC03, Pb(N03)2, Mn metal (99.0%), and Fe metal (99.9%). During analysis, the samples were irradiated by a finely focused electron beam. The samples were placed in the vacuum chamber of the instrument (5 x lop5 torr), and the distribution of elements within the section was determined by x-ray mapping. Each element was analyzed separately. Experimental conditions are given in Table 1. Each section was analyzed from the outer edge through to the inner edge. X-ray maps were photographed with Polaroid positivehegative Type 55 film (ASA 50). Specific regions of interest were analyzed three times for the number of x-ray counts per 20-second period. The volume of these regions was approximately 1 pm3. The sample holder was normally level. The stage occasionally was tilted k6.0" to test for aberrations. RESULTS AND DISCUSSION As a visual representation of elemental distributions we have used a dot map in which the density of dots signifies relative concentration. The density of dots depends on the actual level of a given element, on background noise, and on the sensitivity of the particular wavelength. Samples were analyzed throughout the entire cross section, but we illustrate here only the elemental densities on the outer or inner surfaces of the femur. Zron Figure 1shows the distribution of iron from Gibson 3-1. All figures are 400 pm wide. A buildup of approximately 20-fold occurs along the outer surface, and a buildup of twofold along the inner surface. The surfaces of the Fig. 1. Iron distribution in femur from Gibson 3-1. The width of the scan is 400 pm. The figures on the right are the average number of x-ray counts in 20 seconds for a volume of approximately 1 pm’ located at the left end of the white line. 412 J.B. LAMBERT, S.V. SIMPSON, J.E. BUIKSTRA, AND D. HANSON TABLE 1. Experimental conditons of the electron microprobe analysis Element Crystal' Fe A1 K Mn Zn Sr Pb Na Mg Ca LiF RAP PET LiF LiF PET LiF RAP RAP PET x-ray' K,, K,, K," K<, K, L, L', K" K," K,. Wavelength (nm) Absorbed3 current (PA x lo8) Voltage (KeV) 134.60 89.39 119.84 146.12 99.78 219.80 83.94 127.68 106.01 107.56 1.0- 3.0 6.0-10.0 2.5- 6.0 1.4- 4.7 1.6- 2.1 1.6- 8.6 2.4- 3.0 6.0-10.0 6.0-10.0 1.0- 2.4 30 15 30 30 30 30 30 15 15 25 Magnification x 300 x 300 x 300 x300 x 300 x 300 x 300 x 300 x 300 x 300 'LiF indicates lithium fluoride: RAP indicates rubidium acid phthalate; PET indicates pentaerythritol. 'X-ray line in which K, is from the n = 1 shell and L, is from the n = 2 shell. ?rhe absorbed current on the bone from the beam. femur are easily detected by the abrupt dropoff in iron concentration. The numerals to the right of the dot pictures are the average x-ray counts over a 20-second period a t the indicated spot. The distribution of Fe is homogeneous throughout the interior of the femur. The large buildup along the outer surface penetrates only 10-20 pm, is not present continuously along the surface, and has a variable depth of penetration. Samples from Ledders 1-16, 1-41, and 1-146 and Gibson 5-28 also were examined in this fashion. In all cases a strong buildup (fivefold or more) is observed along the outer surface and a lesser buildup along the inner surface. The buildup penetrates 5-20 pm from the surfaces. The interior contains a homogeneous distribution of Fe. In Ledders 1-41, however, isolated areas of excess Fe (up to fivefold) were observed further from the surface, but never more than 150 pm in. The pattern of Fe distribution corroborates our earlier conclusion that Fe is contaminative, based on overall concentration (Lambert et al., 1979),on higher concentrations in rib than in femur (Lambert et al., 1982), and on Fe distribution in the soil (Lambert et al., 1983).These x-ray pictures, however, provide the first direct evidence in human samples. After deterioration of organic matter, the outer surface of the bone is in direct contact with the soil and experiences a n influx of Fe. The medullary cavity normaliy is more protected, so that there is less buildup of Fe on the inner surface. Because these SEM observations are not quantitative, it is not possible to calculate the absolute level of contamination. Moreover, it is not possible to generalize from these results, because other conditions of soil pH, duration of burial, temperature, flooding, and so on might lead to different results. In the present case, the soil was neutral to slightly alkaline, and the bones had been buried since c. A.D. 175 (Gibson) or c. A.D. 1000 (Ledders) (Lambert et al., 1979). Aluminum Distribution of A1 was studied in Ledders 1-16, 1-41, and 1-146 and Gibson 3-1. Figure 2 shows the dot map for A1 in Ledders 1-16. In this sample and the others, there is a buildup of twofold to 26-fold along the outer surface, with a penetration of 10-20 pm. The distribution of A1 is homogeneous throughout the interior and along the inner surface of the femurs. The only exception to this pattern is Ledders 1-41, in which there is no buildup of A1 in any portion of the femur. The present observation of penetration of A1 into the surface of the femur gives direct evidence for contamination and corroborates our earlier conclusions (Lambert et al., 1979, 1982, 1983). Although relative levels of buildup are similar for Fe and Al, the absence of buildup along the inner surface of all femurs and along all surfaces of one femur suggests that A1 is slightly less prone than Fe to infiltration into buried bone. Potassium X-ray data were obtained for K in Ledders 1-16, 1-41, and 1-146 and Gibson 1-17, 2-40, and 3-1. Figure 3 shows the dot map for Ledders 1-146.For this sample, there is a buildup of fivefold to 13-fold along the outer surface, extending 10-20 pm into the femur. The concentration of K is uniform throughout the MICROPROBE ANALYSIS OF FEMURS Fig. 2. Aluminum distribution in femur from Ledders 1-16.See Figure 1 for details. 413 414 J.B. LAMBERT, S.V. SIMPSON, J.E. BUIKSTRA, AND D. HANSON Fig. 3. Potassium distribution in femur from Ledders 1-146.See Figure 1 for details MICROPROBE ANALYSIS OF FEMURS rest of the femur, including along the inner surface. Similar patterns are observed for Gibson 1-17 and 3-1. For Ledders 1-16, a twoto threefold buildup is also observed along the inner surface. In contrast, Ledders 1-41 and Gibson 2-40 exhibit a homogeneous distribution of K through all portions of the femur. Like A1 and Fe, K was previously classified as a contaminative element (Lambert et al., 1979, 1982, 19831, and these x-ray results confirm this conclusion. The variable degree of influx of K into the bone, however, s u g gests a sensitivity of the process to specific environmental characteristics of the burial. Manganese This element shows the highest levels of contamination of any we examined. Buildup along the outer surfaces of Ledders 1-16 and 1-41 is 30- to 60-fold greater than in areas of Mn homogeneity and extends 200-400 pm into the interior (Fig. 4). Furthermore, Mn concentrates within the femur as patterns that correspond to osteonal and other canals within the framework of the bone matrix. These structures were filled with collagen and other proteins during the lifetime of the individual (Goffer, 1980).Deterioration of the organics after death permits minerals to fill the voids. We see that Mn is particularly well suited to fill this role. These results with Mn are similar to those of Parker and Toots (1970). Beyond about 400 pm, the distribution of Mn is homogeneous. There is no apparent buildup along the inner surface. Similar results were obtained with Ledders 1-146 and Gibson 1-17 and 3-1. We observe stronger buildup of Mn than of Fe, Al, or K, particularly within structures in the interior of the bone. The observation of severe contamination is in agreement with our earlier conclusions CLambert et al., 1979,1982,1983). Zinc The distribution of Zn in Ledders 1-16(Fig. 5) is seen to be entirely homogeneous. There is no evidence for buildup of Zn a t any point in the bone. Analyses of Ledders 1-28, 1-41, and 1-146 and Gibson 1-17 show similar homogeneous distributions. Unlike Fe, Al, K, and Mn, there is no apparent influx of Zn into the femur. This conclusion agrees with what we previously observed (Lambert et al., 1979,1982,1983). All indications are that Zn concentrations in excavated bone are the same as those at the time of death and should 415 be a valid indicator of the ancient bone condition and, by inference, diet. Strontium Analysis for Sr was carried out on Ledders 1-16and 1-41and Gibson 2-40 and 3-1. Figure 6 shows the results for Ledders 1-16. The xray counts are low because of the inherently low sensitivity of Sr even a t the best wavelength. Background counts were two to five times lower than the signal from the bone. In all samples, the Sr concentration is homogeneous throughout the femur, in agreement with conclusions from other methods (Lambert et al., 1979, 1982, 1983). This result confirms the validity of using Sr as an indicator of ancient diet. Lead The distribution of Pb in Ledders 1-16 and 1-41 and Gibson 2-40 was found to be entirely homogeneous along both surfaces and throughout the femurs (Fig. 7). In our comparison of Pb levels in ribs and femurs (Lambert et al., 1982), we found a lower concentration in the rib. In most cases, such a n observation was associated with leaching of the element from bone to soil. Mackie et al. (19751, however, suggested that Pb is not distributed evenly throughout the skeleton during lifetime, so that the lower levels in rib may be physiologically inherent rather than caused by diagenesis. Thus Pb becomes useful in the analysis of dietary levels and health (Aufderheide et al., 1981). Our observation of a homogeneous distribution of Pb in femurs is in agreement with these conclusions. Magnesium We have obtained contradictory indicators concerning Mg from the other techniques we have used (Lambert et al., 1979,1982, 1983). Comparison of ribs and femurs showed no difference in amounts, but examination of soil around burials suggested that Mg had leached from the bone to the soil. We examined Ledders 1-16, 1-41,and 1-146and Gibson 3-1 by electron microprobe and found the distribution of Mg to be inhomogeneous (Fig. 8). For Ledders 1-16, there is a buildup of twoto 20-fold along the outer surface. Excess Mg penetrates into the bone 10-30 pm. The distribution of Mg is homogeneous in the interior and along the inner edge of the femurs. Thus in contradiction to the riblfemur and soil studies, we observe Mg distributions that 416 J.B. LAMBERT, S.V. SIMPSON, J.E. BUIKSTRA, AND D. HANSON Fig. 4. Manganese distribution in femur from Ledders 1-16(upper) and 1-41 (lower). See Figure 1for details. MICROPROBE ANALYSIS OF FEMURS Fig. 5. Zinc distribution in femur from Ledders 1-16. See Figure 1for details. 417 Fig. 6 . Strontium distribution in femur from Ledders 1-16,See Figure 1 for details. Fig. 7. Lead distribution in femur from Ledders 1-16. See Figure 1 for details 420 J.B. LAMBERT, S.V. SIMPSON, J.R. BUIKSTRA, AND D. HANSON Fig. 8. Magnesium distribution in femur from Ledders 1-16.See Figure 1 for details. 421 MICROPROBE ANALYSIS OF FEMURS appear to result from infiltration of Mg from the soil into the bone. It is also possible that surface buildup could result from outward migration from within the bone as well as from external contamination. Since we cannot distinguish the two processes, magnesium may in fact be different from, e.g., iron. A possible explanation of these results has been provided by Parker et al. (1974) and by Parker and Toots (1976,19801, who suggested that Mg moves in both directions; i.e., it leaches from the bone after death but is reintroduced into the pore spaces of the bone. Thus any use of Mg levels to assess ancient diet should prove to be very difficult. Sodium We now come to elements that are major constituents of bone and are found in percentage rather than ppm levels. Thus extremely large amounts of mineral would have to move in either direction to be detected by the x-ray technique. Visual examination of dot maps for Ledders 1-16, 1-28, and 1-41and Gibson 3-1 shows completely homogeneous distributions of Na, with the possible exception of small pockets of buildup along the outer edge. These areas are just barely significant and do not extend along the entire edge. We observed Na depletion from the bone into the surrounding soil (Lambert et al., 1983), so it is possible that Na, like Mg, moves in both directions. Calcium Analysis for Ca in Ledders 1-16 and 1-41 and Gibson 3-1 shows a homogeneous distribution throughout the bone, with two qualifications. First, the high level of Ca skews the x-ray counts toward the center of the map. This distortion is a n experimental arti- fact rather than a reflection of the Ca distribution. Second, certain regions appear to be almost devoid of Ca. By comparison with the electron micrograph of the same portion of the femur, these empty regions were found to correspond to osteonal canals or other features (Fig. 9). These regions of the bone structure are not part of the inorganic matrix but were filled with other materials during the lifetime of the individual. Thus the absence of Ca in these regions reflects the actual levels before death. Results from the rib/femur and soil studies (Lambert et al., 1982, 1983) suggested that Ca is leached from the bone into the soil. Because of high levels of Ca in the bone, leaching can occur to a significant extent without influencing the dot drawing representation. SUMMARY AND CONCLUSIONS Influx of elements from soil to human bone during burial is successfully and graphically demonstrated by x-ray analysis with the electron microprobe. Results for the analysis of ten elements in eight human femurs from Woodland sites are summarized in Table 2. The concentrations of Sr and Zn are homogeneous throughout the cross sections observed. By this technique, these elements appear to be free from diagenetic effects and should be useful in the assessment of ancient diet. Postmortem enrichment by Fe, K, Al, and Mn is documented by observation of buildup along the outer surface and sometimes along the inner surface of the femur. The areas of buildup usually penetrate 1030 pm into the surface for Fe, Al, and K, and up to 400 pm for Mn. These elements previously were classified as contaminative on the basis of comparison of ribs and femurs and on analysis of the adjacent soil (Lambert et al., 1982, 1983). TABLE 2. Elemental distribution in femurs' Sample Fe A1 K Mn Zn Sr Pb Mg Na Ca Ledders 1-16 Ledders 1-28 Ledders 1-41 Ledders 1-146 Gibson 1-17 Gibson 2-40 Gibson 3-1 Gibson 5-28 C C C C N N N C N 2 N N - - - - C C N C N - - C C C C C N N N N N - - C - C - - - C C C - - N N N N N - - N N C - N - - - N C - N - - - N - 2 N - 'C denotes contamination on the inner or outer surface; N denotes homogeneous distribution throughout the femur. 'Low level of contamination on the outer surface. - Fig, 9. (Upper) Calcium distribution in femur from Gibson 3-1.See Figure 1 for details. (Lower) Photomicrograph of bone structure for same region ( ~ 3 0 0 ) . MICROPROBE ANALYSIS OF FEMURS The concentrations of Pb, Na, and Ca generally are homogeneous within the femurs. In the case of Pb, lower levels in ribs compared with femurs were attributed to antemortem physiological differences between the bones rather than to leaching of Pb out of the bone (Lambert et al., 1982). For Na and Ca, other results suggested that these elements leach into the soil (Lambert et al., 1982, 1983). Leaching of significant amounts of these major elements would not show up in the x-ray maps but could still cause the rib/ femur differences and observable buildup in the soil. For example, a decrease in Ca from 37% to 33% would not be visible in the maps but would still represent a large mass if moved to the soil and a significant difference between rib and femur. The greater resistance of femurs to diagenesis also may make the effects of leaching from the bone more difficult to observe in the present study. It would be of interest to examine cross sections of ribs for this group of elements, since more extensive effects of diagenesis should be observable. The role of Mg remains ambiguous. Although proportions of Mg were identical in ribs and femurs, we have observed leaching of Mg into the soil and buildup on the outer edge of femurs. The best explanation for these results is that Mg undergoes flux in both directions, leaching out of the bone and redepositing on the surface. As the first study of elemental distributions in excavated human bone, these results cannot be considered to be general. Differences in soil and climate conditions, duration of burial, and choice of bone might lead to different results. More acid soil, older burials, or analysis of more porous bone could result in greater contamination. Since most of the contamination of femurs reaches only a few tens of microns into the surface, it would be of interest in the assessment of ancient diet to analyze bones whose surface has been removed down to several hundred microns. Such a study would have to be done in conjunction with the x-ray analysis of cross sections. Some workers have removed surface material, but without using SEM a s a monitor. The procedure would offer the opportunity to augment the list of dietarily important elements from Sr, Zn, and Pb possibly to include Fe, Al, K, Na, Mg, and Ca, but probably not Mn. The inclusion of Fe, Mg, and Na would be particularly useful because of their physiological importance. ' ACKNOWLEDGMENTS This work was supported by the Research 423 Committee of Northwestern University and by funds from the local and national organizations of Sigma Xi, the Scientific Research Society. We are grateful to Dr. Robert Bailey of Northern Illinois University for the use of the SEM. LITERATURE CITED Aufderheide, AC, Neiman, FD, Wittmers, LE, Jr, and Rapp, G (1981) Lead in bone 11: Skeletal-lead content as an indicator of lifetime lead ingestion and social correlates in a n archaeological population. Am. J. Phys. Anthropol. 55:285-291. Badone, E, and Farquhar, RM (1982)Application of neutron activation analysis to the study of element concentration and exchange in fossil bones. J. Radioanal. Chem. 69:291-311. Brown, AB (1973)Bone Strontium as a Dietary Indicator in Human Skeletal Populations. PhD dissertation, University of Michigan, Ann Arbor. Gilbert, RI, Jr (1975)Trace Elemental Analyses of Three Skeletal Amerindian Populations at Dickson Mounds. PhD dissertation, University of Massachusetts, Amherst. Goffer, 2 (1980) Archaeological Chemistry. New York: Wiley and Sons. Lambert, JB, Szpunar, CB, and Buikstra, J E (1979) Chemical analysis of excavated human bone from Middle and Late Woodland sites. Archaeometry 21: 115-129. Lambert, JB, Simpson, SV, Buikstra, JE, and Charles, DK (1983) Analysis of soil associated with Woodland Burials. In JB Lambert (ed):Archaeological Chemistry 111. Washington, DC: American Chemical Society, pp. 000-000. Lambert, JB,Vlasak, SM, Thometz, AC, and Buikstra, J E (1982) A comparative study of the chemical analysis of ribs and femurs in Woodland populations. Am. J. Phys. Anthropol. 59:289-294. Mackie A, Townshend, A, and Waldron, HA (1975)Lead concentrations in bones from Roman York. J. Arch. Sci. 2:235-237. Parker, RB (1967)Electron microprobe analyses of fossil bones and teeth. Geol. SOC.Am. Spec. Paper 101:415. Parker, RB, and Toots, H (1970)Minor elements in fossil bones. Bull. Geol. Soc. Am. 815325-932. Parker, RB, and Toots, H (1974) Minor elements in fossil bone-application to Quaternary samples. Geol. Survey Wyoming, Rept. Invest. 10:74-77. Parker, RB, and Toots, H (1976)Minor element distribution among vertebrate calcified tissue. Arizona Acad. Sci. J. Proc. 11:158. Parker, RB, and Toots, H (1980)Trace elements in bones as paleobiological indicators. In AK Behrens-Mayer and AP Hill (eds): Fossils in the Making. Chicago: University of Chicago Press, pp. 197-207. Parker, RB, Toots, H, and Murphy, JW (19741 Leaching of sodium from skeletal parts during fossilization. Geochim. Cosmochim. Acta 38:1317-1321. Price, TD, and Kavanagh, JT (1982) Bone composition and the reconstruction of diet: Examples from the Midwestern United States. Mid-Cont. J. Arch. 7:61-79. Schoeninger, MJ (1979) Diet and status at Chalcatzingo: Some empirical and technical aspects of strontium analysis. Am. J. Phys. Anthropol. 51:295-309. Sillen, A (1981) Strontium and diet at Hayonim Cave, Israel. Am. J. Phys. Anthropol. 56:131-137.