[NPS Arrowhead] U.S. Dept. of Interior National Park Service Archeology Program
Quick Menu Features
* Sitemap * Home
  Back to Kennewick Home Back to C14 Memo  


Radiocarbon Laboratory, Department of Anthropology
Institute of Geophysics and Planetary Physics
University of California, Riverside
Riverside, CA 92521

20 December 1999

TO:  Dr. Frank McManamon

RE:  (1) UCR Kennewick results (2) responses to your inquiries of 12/7/99 and 12/17/99

Dear Frank:

Attached as a table are the results of the UCR 14C analysis of two Kennewick bones compared with our earlier Kennewick results for comparison.

1. Comments on the UCR 14C Results: On the basis of their amino acid carbon contents (AACC) and amino acid profiles, UCR-3806 and 3807 exhibit much lower collagen (protein) preservation than the earlier Kennewick bone my lab previously analyzed (UCR-3476). UCR-3806 has totally lost its collegen-like amino acid pattern. As I reported previously, both UCR-3806 and UCR-3807 exhibited unusual amounts of effervescence in acid which is usually an indication of significant amounts of secondary carbonates and there was unusual difficulty in filtering the hydrolysates.

The AACC that I reported earlier by email has been revised in light of additional analyses. (As I mentioned to you previously, we had just received our new HPLC and were still calibrating with standards when the initial analyses were obtained.) The revised AACC values do not change the fact that both bones are problematical in terms of their suitability to yield accurate bone 14C values due to their degraded biogeochemical condition. Although UCR-3807 turns out to have more protein that I reported earlier (14.3% AACC of our modern bone standard), the amino acid composition is marginal in terms of its collagen- or non-collagen like characteristics. On a routine basis, our criteria for an acceptable bone is at least 5% AACC and where the bone retains a clear collagen-like amino acid profile. On the basis of their amino acid profiles, both UCR-3806 and UCR-3807 are classified as non-collagen.

Because of their biochemically degraded condition, I report the results of the 14C measurements in terms of "fraction modern" with the apparent 14C age cited in footnotes. You will also note that the reported 13C values of these two samples are not typical of collagen amino acids. I would interpret that these values reflect primarily a dietary effect--namely that the individual (assuming that there is only one individual here represented) subsisted largely on a marine diet (e.g., fish). There also could be a fractionation factor involved due to the poor protein preservation. (In the case of UCR-3476, the first Kennewick bone we ran, we also observed a depressed 13C value and, making certain assumptions, we calculated a reservoir corrected age of 7880 (160 BP.)

In summary, UCR-3807 exhibits an younger age offset of about 3% (about 280 14C years) in comparison with UCR-3476 while UCR-3806 is very anomalous with respect to UCR-3476. One interpretation is that the age offsets reflect varying percentages of more recent and/or modern contamination in both UCR-3806 and UCR-3607, with the percentage contribution of contamination increasing as a function of the decreasing residual collagen protein content. For UCR-3807, there is enough residual collagen so that the offset is limited to a few percent, while for UCR-3806, the very low AACC is reflected in the much more recent anomalous age.

2. Responses to Questions:

A. Questions of December 7, 1999

(1) First set

  1. Did any of you observe any structure or other characteristics of the extracted carbon that indicates it is deteriorated collagen rather than an intrusive element?

    Without sequencing data, it would be difficult to establish definitively that the amino acids came only from collagen peptides. The observation that the age offset increases in inverse relationship to the collagen content in both UCR-3806 and UCR-3607 strongly suggests that there are exogenous amino acids in these samples. As you know, in bone, it is usually assumed that the older the inferred 14C age the more likely that this is closer to the actual age since typically non-carbonate contamination that has not been sufficiently removed generally renders samples "too young."

  2. Did any of you observe any structure or other characteristics of the extracted carbon that indicates that it is from a source external to the bone sample?

    The SEM images did reveal some microstructures that we could not identify and thus it is not possible to determine if they were organic in nature. It was difficult to filter the hydrolysate of both UCR-3806 and UCR-3807 which is rarely a problem with high collagen yield bone such as UCR-3476.
  3. In your experience, is it invariable/common/rare/impossible for "old" intrusive carbon to contaminate a bone sample from a riverine, floodplain, or lower river terrace geomorphologic context?

    It entirely depends on the characteristics of the humic and other soil organic compounds contained in the soil together with the nature of the ground water conditions over the time period that the bone has been exposed to the environment. Also, can it be assumed that the bone was always buried in the same soil profile? May it have been exposed and then reburied as some unknown period in the past?
  4. Are there other structural, physical, chemical, or visual characteristics of the sample and extracted carbon that suggest to you that it is uncontaminated?

    On the contrary, the chemical state of the amino acid extract from UCR-3807, and especially that from UCR-3806, in my view, points strongly to the possibility that it may be contaminated with exogenous carbon compounds.
  5. Are there other structural, physical, chemical, or visual characteristics of the sample and extracted carbon that suggest to you that it is contaminated? If so, what do you believe the contaminate is?

    As noted in 4, the chemical state of the collagen in UCR-3807 and especially UCR-3806 raises the strong possibility that both may be contaminated. Soil humics of various types are the most obvious candidates.
  6. In your experience, what magnitude of time span would be required for the characteristics you observed in the extracted carbon from these samples to have deteriorated from normal bone collagen?

    This is very difficult to determine since there are many environmental variables that can influence rates of biogeochemical diagenesis processes in bone structures.
  7. Before we took the samples from the Kennewick remains in September, we consulted with experts, including each of you about the kind of bone to select. Dense bone in weight bearing areas and mid-shaft were the main suggestions we got and followed. If we were to take additional samples, is there a way to determine visually which bones would be rich in collagen? If not visually, what other means would be needed to detect collagen levels?

    Except with highly degraded bone where there is a "chalk-like" appearance, it is usually difficult to determine which bones have retained more unaltered collagen on the basis of gross visual appearance. Some have used responses to ultraviolet to gauge collagen content but there are a number of variables that interfere with good responses. (I believe that I suggested previously to you that it would be very helpful to take very small amounts of bone from 20 different Kennewick bones and determine their amino acid composition. This would give you an objective basis on which to gauge differential preservation.)

(2) Second Set

  1. In your experience is it common or rare for samples from the same skeleton to display such a range in collagen structure and content?

    Few specific experiments have addressed this directly. The Haverty skeletons exhibited significant variability in protein content but, in this case, the analyses were done on different skeletons that were assumed to have been buried in close spacial and temporal proximity. (Brooks, S., R. H. Brooks, J. Austin, G. Kennedy, J. R. Firby, L. A. Payen, C. A. Prior, P. J. Slota, Jr., and R. E. Taylor. 1991. The Haverty Human Skeletons: Morphologial, Depositional and Geochronological Characteristics. Journal of California and Great Basin Anthropology 12:60-83.). In cases where different parts of a skeleton have been subjected to different alternating ground water/moisture cycle (wet/dry/wet) regimes, there can be significant differences among the bones. This can occur if different parts of a skeleton are not being exposed to the same ground water conditions or has been exposed to different soil types by redeposition.
  2. Do you have any suggestions that could explain this difference reasonably?

    As noted above, differential ground water cycle (wet/dry/wet) regimes could explain the difference in the same skeleton. Conditions would depend on the relationship between the position of different bones in the skeleton with reference to the soil profile/ground water regime, i.e., if different bones were exposed to varying soil/ground water conditions.

B. 12/17/99 Question Set

  1. Have you or some other expert ever summarized the characteristics of skeletal remains earlier than 7000 years BP that have been dated? We are checking articles and books on the subject, such as articles by Powell and Steele that review early skeletal evidence; "Brule Woman" article; "Arlington Springs Woman" info; Windover site burial population; Pyramid Lake and Spirit Cave mummies; other?

    There is an extensive literature on the 14C dating of bone and the problems of dealing with collagen degraded bone extending back for several decades. For example, Taylor 1987: 53-61 reviews the research as of the mid-1980s and cites the earlier literature. Hedges and Law 1989 and Hedges and Van Klinken 1992 are excellent overviews and present the experiences of the Oxford Laboratory. Stafford et al. 1988 and 1991 reports extensive and excellent studies carried out by him at the Carnegie Geophysical Laboratory and at the University of Arizona. Taylor 1982, 1987b, 1992, 1994 reports some of the work of my lab. Burkey et al. 1998 reports our work in attempting to deal with collagen-degraded bone.

    All of these studies highlight the significant variability in the degree to which endogenous carbon-containing fractions in bone are retained and are, or are not, protected from contamination by a wide variety of physical and chemical diagenetic mechanisms. It is widely acknowledged that obtaining accurate 14C age estimates on bone requires attention to detail in sample preparation and an appreciation that each bone may present an unique chemical challenge if the isolation of a fraction that contains only autochthonous carbon atoms is to be consistently achieved.
    It should be reiterated that the biochemical condition of bone reflects more directly the diagenetic conditions to which it is exposed--which can be highly variable--so that, in one environment, 7,000, 10,000, or 40,000 year old bones can retain close to 100% of their in vivo collagen, while in another environment, a 1,000 year old bone may have lost most of its collagen content.


Burky, R. R., D. L. Kirner, R. E. Taylor, P.E. Hare, and J. R. Southon.
(1998)        Radiocarbon Dating of Bone Using Gamma-Carboxyglutamic Acid and Alpha-Carboxyglycine (Aminomalonate). Radiocarbon 40:11-20.

Hedges, R. E. M. and I. A. Law.
(1989)        The radiocarbon dating of bone. Applied Geochemistry 4:249-233.

Hedges, R. E. M. and Van Klinken, G. J.
(1992)        A review of current approaches in the pretreatment of bone for radiocarbon dating by AMS. Radiocarbon 34:279-291.

Stafford, T. W., Jr., K. Brendel, and R. C. Duhamel.
(1988)        Radiocarbon, 13C and 15N analysis of fossil bone: Removal of humates with XAD-2 resin. Geochimica et Cosmochimica Acta 52: 2197-2206.

Stafford, T. W., Jr. P. E. Hare, L. Currie, A. J. T. Jull and D. J. Donahue.
(1991)        Accelerator radiocarbon dating at the molecular level. Journal of Archaeological Sciences 18:35-72.

Taylor, R. E.
(1982)        Problems in the radiocarbon dating of bone. In Nuclear and Chemical Dating Techniques. L. A. Currie, ed., pp. 453-473. Washington, D.C.: American Chemical Society.

Taylor, R. E.
(1987a)        Radiocarbon Dating: An Archaeological Perspective. New York: Academic Press.

Taylor, R. E.
(1987b)        AMS 14C Dating of critical bone samples: Proposed protocol and criteria for evaluation. Nuclear Instruments and Methods in Physics Research B29:159-163.

Taylor, R. E.
(1992)        Radiocarbon Dating of Bone: Beyond Collagen. In R. E. Taylor, A. Long, and R. Kra, eds. Radiocarbon After Four Decades: An Interdisciplinary Perspective, pp. 375-402. New York: Springer-Verlag.

  1. For these relatively ancient remains (post 7000) is the collagen and its structure typically deteriorated? Is the amount of carbon in the bones that is available for
    14C dating consistently low, if not consistently low, what seems to be cause of the variation?

    As noted previously, there are many environmental variables that can influence rates of biogeochemical diagenesis. In most cases, the most critical variables are probably effective mean annual temperature and effective moisture. Typically, bone in tropical contexts is rapidly biochemically and physically degraded. Bone from cold environments, e.g. arctic or high altitudes and bone from special environments that excludes water (e.g., La Brea Tar Pits or in desiccated desert caves or rock shelters) can retain their collagen content for extended periods of time measured, in some cases, in excess of several tens of thousands of years.
  2. Can you point me to any general or summary statements in your articles or radiocarbon texts and general articles about bone carbon deterioration over time, any graphs or tables on this?

    Please see the comments on question 1 above.
  3. In the processing of the bone samples has your lab needed to use all the bone? If so, is this because of the deterioration of the collagen carbon, if not what factor has required use of most of the bone?

    We used about 20% of the UCR-3807 bone we received and about 30% of UCR-3806 to obtain our dates. (We will need most of the remaining bone to undertake the additional studies to determine the source of the contamination. Please see answer to the next question.)
  4. Can you explain to me in writing the dating of additional fractions that you and I have discussed, what do we hope to learn from this, will it be done with both samples or only the most deteriorated? How long do you estimate it will take?

    As we discussed, I would like to determine, if possible, where the contamination is coming from. The most likely candidate is the humic fraction. We wish to do an XAD-extraction and also look directly at a total humic fraction. It may be necessary to request additional bone to do these tests, but we will start on the remaining bone currently in the lab. This may take up to another month to 6 weeks, depending on the problems we encounter.
  5. What description is available of the first Kennewick sample from the Benton Co. coroner? What portion of the bone remained after the sample extraction at UCR?

    All we have by way of a description of the first Kennewick sample is the paperwork that we received from the submitter. Our results were published in Science. [Taylor, R. E. et al. (1998) Science 280:1171-1172].

I trust these responses and suggestions have been responsive and helpful. If and when this data is released to the popular press, I know that you will find some way to get them to report it appropriately.


R. E. Taylor
Director, Radiocarbon Laboratory

UCR/CAMS Radiocarbon Analyses of Kennewick Human Bone

Bone Preservationa Fraction measured 13C (permil)   Radiocarbon analysis
          Fmb 14C age (BP)

5th left metacarpal APS-CPS-01 68.8%(C) total amino acids -15.4 ---- 841060c
14.3%(NC)d total amino acids -10.8 0.36330.0014 ---e
2.3%(NC)f total amino acids -10.3 0.42160.0015 ---g

aExpressed as % of amino acid carbon content (AACC) of modern bone standard. C = collagen-like amino acid composition. NC = non-collagen amino acid composition.
bFm = fraction modern where 1.0 = "modern." pM (percent modern) = Fm x 100.
cConventional radiocarbon age in 14C years BP. Reservoir corrected age = 7880160 [Taylor et al. (1998) Science 280:1171-1172]
dRevised AACC after duplicate analysis and recalibration of HPLC. Initial analysis = 3.2% AACC of modern bone standard. Gly/Glu ratio and other indices of collagen-like amino acid profile indicates significant biogeochemical diagenesis has occurred and on this basis the profile is characterized as non-collagen.
eApparent 14C age = 813040 BP
fRevised AACC after duplicate analysis and recalibration of HPLC. Initial analysis = 5.3% AACC of modern bone standard.
gApparent 14C age = 694030 BP

Back to:Kennewick Man | C14 Memo