Scholarly article on topic 'Dating human skeletal remains using 90Sr and 210Pb: Case studies'

Dating human skeletal remains using 90Sr and 210Pb: Case studies Academic research paper on "History and archaeology"

CC BY-NC-ND
0
0
Share paper
Academic journal
Forensic Science International
OECD Field of science
Keywords
{"Time of death" / "Skeletal remains" / Dating / " 210Pb" / " 90Sr"}

Abstract of research paper on History and archaeology, author of scientific article — Bettina Schrag, Tanya Uldin, Patrice Mangin, François Bochud, Pascal Froidevaux

Abstract In legal medicine, the post mortem interval (PMI) of interest covers the last 50 years. When only human skeletal remains are found, determining the PMI currently relies mostly on the experience of the forensic anthropologist, with few techniques available to help. Recently, several radiometric methods have been proposed to reveal PMI. For instance, 14C and 90Sr bomb pulse dating covers the last 60 years and give reliable PMI when teeth or bones are available. 232Th series dating has also been proposed but requires a large amount of bones. In addition, 210Pb dating is promising but is submitted to diagenesis and individual habits like smoking that must be handled carefully. Here we determine PMI on 29 cases of forensic interest using 90Sr bomb pulse. In 12 cases, 210Pb dating was added to narrow the PMI interval. In addition, anthropological investigations were carried out on 15 cases to confront anthropological expertise to the radiometric method. Results show that 10 of the 29 cases can be discarded as having no forensic interest (PMI>50years) based only on the 90Sr bomb pulse dating. For 10 other cases, the additional 210Pb dating restricts the PMI uncertainty to a few years. In 15 cases, anthropological investigations corroborate the radiometric PMI. This study also shows that diagenesis and inter-individual difference in radionuclide uptake represent the main sources of uncertainty in the PMI determination using radiometric methods.

Academic research paper on topic "Dating human skeletal remains using 90Sr and 210Pb: Case studies"

Contents lists available at ScienceDirect

Forensic Science International

journal homepage www.elsevier.com/locate/forsciint

Forensic Anthropology Population Data

Dating human skeletal remains using 90Sr and 210Pb: Case studies§

Bettina Schraga,b, Tanya Uldinb, Patrice Manginb, Francois Bochudc, Pascal Froidevaux c*

CrossMark

a Hôpital du Valais (RSV) - Institut Central, Av. du Grand-Champsec 86, 1951 Sion, Switzerland

b Centre Universitaire Romand de Médecine Legale, Rue Michel-Servet 1, 1211 Geneve - 21 Rue du Bugnon, 1005 Lausanne, Switzerland c Institute of Radiation Physics, Lausanne University Hospital, Switzerland

ARTICLE INFO

ABSTRACT

Article history:

Received 18 April 2013

Received in revised form 4 September 2013

Accepted 28 October 2013

Available online 12 November 2013

Keywords:

Time of death

Skeletal remains

Dating

In legal medicine, the post mortem interval (PMI) of interest covers the last 50 years. When only human skeletal remains are found, determining the PMI currently relies mostly on the experience of the forensic anthropologist, with few techniques available to help. Recently, several radiometric methods have been proposed to reveal PMI. For instance, 14C and 90Sr bomb pulse dating covers the last 60 years and give reliable PMI when teeth or bones are available. 232Th series dating has also been proposed but requires a large amount of bones. In addition, 210Pb dating is promising but is submitted to diagenesis and individual habits like smoking that must be handled carefully.

Here we determine PMI on 29 cases of forensic interest using 90Sr bomb pulse. In 12 cases, 210Pb dating was added to narrow the PMI interval. In addition, anthropological investigations were carried out on 15 cases to confront anthropological expertise to the radiometric method. Results show that 10 of the 29 cases can be discarded as having no forensic interest (PMI > 50 years) based only on the 90Sr bomb pulse dating. For 10 other cases, the additional 210Pb dating restricts the PMI uncertainty to a few years. In 15 cases, anthropological investigations corroborate the radiometric PMI. This study also shows that diagenesis and inter-individual difference in radionuclide uptake represent the main sources of uncertainty in the PMI determination using radiometric methods.

© 2013 The Authors. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction

To determine the post mortem interval (PMI) when only human skeletal remains are available is a difficult task in forensic sciences. In Switzerland, the statute of limitations in criminal law is 30 years. Nevertheless, for human and social reasons, forensic investigations on bones often cover a PMI of 50 years. Usually, the PMI determination relies on the skill and experience of forensic anthropologists, with few supportive techniques available [1-4]. The state of conservation of the remains depends strongly on the environment in which the body rested for years. Bodies kept for centuries or even a thousand years in glaciers or peat bogs are often excellently preserved due to freeze drying and a reductive environment, while oxidative soils are responsible for a rapid

§ This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial-No Derivative Works License, which permits non-commercial use, distribution, and reproduction in any medium, provided the original author and source are credited.

* Corresponding author at: Institute of Radiation Physics, Grand Pre 1, CH-1007 Lausanne, Switzerland. Tel.: +41 21 314 8185; fax: +41 21 314 8299.

E-mail addresses: bettina.schrag@hopitalvs.ch (B. Schrag), tanya.uldin@unige.ch (T. Uldin), patrice.mangin@chuv.ch (P. Mangin), francois.bochud@chuv.ch (F. Bochud), pascal.froidevaux@chuv.ch (P. Froidevaux).

decomposition of tissues and skeletisation [5-7]. Among all the methods proposed to determine reliable PMI, those based on radionuclide determination are very promising. After death, a radionuclide incorporated in the body is only submitted to radionuclide decay, with a well-known half-life [8-10].

Several studies have proposed radionuclide analyses to determine PMI. While the half-life (5730 y) of 14C is much too long to be used in forensic cases, Spalding et al. proposed determining the age at death and the nuclear bomb pulse 14C in teeth to reveal the PMI [11,12]. Marzaioli et al. extended the technique to bone collagen and were able to precisely determine the PMI of an unsolved case investigated by the Rome prosecutor office [13]. However, the forensic determination of pulse bomb 14C requires the use of acceleration mass spectrometry, a technique which is currently not easily available. Nevertheless bomb pulse 14C dating, in combination with the determination of age at death, e.g. by the racemization method, is a promising method yielding PMI with low uncertainties. Another promising technique is based on the isotopic ratio of different radionuclides from the 232Th disintegration series [14-16]. However, this method requires at least 300 g of bone, which can be a strong drawback to the application of the technique in forensic sciences. Moreover, the counting time of a and g spectrometry involved in the measurement is at least 15 days to achieve the desired detection limit. Bomb pulse 90Sr has

0379-0738/$ - see front matter © 2013 The Authors. Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.Org/10.1016/j.forsciint.2013.10.038

also been proposed to determine PMI. The analytical technique is simpler than for bomb pulse 14C but leads to larger uncertainties related to b counting [9,17]. Several authors have also reported on archeological bones having 90Sr above the detection limit [9,17]. The presence of diagenetic 90Sr is thus a serious disadvantage to the application of the technique to the PMI determination. Finally, 210Pb is also a radionuclide of interest for PMI determination. Its physical half-life of 22 years covers the range of elapsed time required in legal investigations. Moreover, it is incorporated by ingestion and inhalation of 222Rn decay products in every individual and it targets bones. In addition, it can be easily measured at very low levels using its granddaughter 210Po, with which it reaches secular equilibrium about one year after death. Nevertheless, 210Po diagenesis has also been observed in most of reported cases of buried skeletons [17,18].

Recently, we reported on 210Po and 90Sr diagenesis if the radiometric method involving both radionuclides is used for the dating of human skeletal remains [17]. We showed that 210Po diagenesis can be handled satisfactorily using solubility profiling. Because of the quantity of bone ash required per analysis (5 g), 90Sr solubility profiling is not recommended. Nevertheless satisfactory results are obtained by using a decontamination method involving Fe(III) reducing agent, metal-binding molecules such as EDTA and DCTA and hydroxyl-apatite abrasion by acetate buffer [17].

Here we report on the use of 90Sr and 210Po dating and anthropological investigations to determine the PMI in 29 cases of forensic interest. We focus on the diagenesis problem that has often been neglected in previous studies on radiometric methods to determine PMI. We compare 90Sr and 210Po dating with anthropological analysis for 15 cases to corroborate PMI. The

limitation of the use of 90Sr and 210Po is also presented, with an attempt to estimate the uncertainty of the method related to the location of the remains, to dietary habits influencing 210Pb and 90Sr intake in the skeleton in a given region and to the analytical aspects of the method.

2. Materials and methods

2.1. Sampling

Most of the human skeletal remains were recovered in different locations in western Switzerland. For the most part they came from 30 to 50 km area close to Lausanne. Two cases were found in Lake Geneva, three in different forests, three in riverbanks, five in construction sites during digging, five in relation to glaciers and one in high alpine territory. The rest of the cases were from different places such as houses and bins. Four cases had a well-established PMI, based on forensic inquiry, before measurements. Glacier cases are particularly important in Switzerland because the discovery of human remains in glaciers increased recently due to climate warming and because the diagenesis of 90Sr in glacier cases is also enhanced (see Section 4).

Vertebrae of individuals who died between 2006 and 2012 have been sampled in the pathology Institutes of Lausanne and Locarno, Switzerland. Ribs were obtained from the Institut Medico-Legal of Lyon (France). These samples were used to determine the 210Po reference value at the time of death for people aged 18-90 years (n = 35). They do not contain diagenetic 210Po because they were not buried before sampling.

To probe the effect of tobacco smoke on 210Po urine excretion and its possible effect on the 210Po content of the bones, we collected urine from non-smoker collaborators of the Institute of Radiation Physics and one sample from a heavy smoker collaborator. 300 ml of urine were spiked with 209Po and 100 ml of HNO3 65% was added. The urine was mineralized in a large volume microwave apparatus (Milestone MLS Ultraclave IV, Germany) under pressure (initial pressure = 60 bars) for 30 min at 180 °C. Afterwards, polonium was co-precipitated on iron hydroxides. The precipitate was dissolved in 80 ml HCl 1 M, ascorbic acid (500 mg) was added and polonium was spontaneously deposited on a silver disk at 50-60 °C for at least 4 h. The alpha spectra were then recorded as described in [17].

Table 1

90Sr and 210Po activities (Bq/g Ca) of human skeletal remains found between 2001 and 2012 in different locations in Switzerland.

Case number Discovery location 90Sr (Bq/g Ca) 210Po (Bq/g Ca) Bone type

GRE-12-439 Alpine forest 0.020 ± 0.006 0.025 ± 0.005b Trabecular bone (vertebrae)

GRE-12-415 High alpine glacier 0.019 ± 0.002 0.005 ± 0.001c Trabecular bone (calcaneum)

GRE-12-414 High alpine, glacier 0.032 ± 0.003 0.005 ± 0.001 c Trabecular bone (calcaneum)

GRE-12-413 High alpine, glacier 0.012 ± 0.001 0.002 ± 0.0005 c Trabecular bone (calcaneum)

GRE-12-413/414/415 High alpine glacier, mixture of the three samples <0.005a Trabecular bone (calcaneum)

GRE-12-417 Lowland lake <0.005 0.003 ± 0.001b Trabecular bone (femur head)

GRE-11-363 Lowland, bed of a river 0.039 ± 0.007 0.009 ± 0.001c Trabecular bone (femur head

GRE-11-320 Lowland, construction site <0.008 Trabecular bone (vertebrae)

GRE-11-163 Alpine deep valley, rocky slope 0.019 ± 0.004 0.005 ± 0.002b Cortical bone (femur)

GRE-11-162 Alpine deep valley, rocky slope 0.019 ± 0.003 0.007 ± 0.003b Trabecular bone (vertebrae)

GRE-11-064 Inside, garbage bag <0.005 Trabecular bone (vertebrae)

GRE-10-285 Sub-alpine, gravel pit 0.032 ± 0.002 Trabecular bone (jaw)

GRE-10-278 Lowland, construction site 0.032 ± 0.004 Trabecular bone (vertebrae)

GRE-10-019 Lowland, lake (fishing net) <0.008 <0.015c Cortical bone (femur)

GRE-10-204 Lowland, under 20 cm of soil <0.005 Trabecular bone (vertebrae)

GRE-09-307 Lowland, close to ancient cemetery 0.014 ± 0.004 0.015 ± 0.003c Cortical bone (skull)

GRE-09-306 Lowland, forest 0.067 ± 0.005 0.021 ± 0.003b Trabecular bone (vertebrae)

GRE-08-199 Lowland, construction site 0.032 ± 0.004 0.020 ± 0.003a Trabecular bone (iliac crest)

0.028 ± 0.004 Trabecular bone (vertebrae)

GRE-08-018 High alpine, rocky slope 0.036 ± 0.005 0.008 ± 0.004 Cortical (femur)

0.055 ± 0.005 0.050 ± 0.010 Trabecular bone (iliac crest)

GRE-07-162 Lowland, construction site <0.005 Trabecular bone (internal ear)

GRE-07-063 Lowland, construction site 0.014 ± 0.003 Trabecular bone (vertebrae)

<0.005 Cortical bone (skull)

GRE-07-001 Lowland, in concrete behind a kitchen wall <0.005 Cortical bone (skull)

GRE-06-111 Lowland, bank of a river 0.008 ± 0.003 Trabecualr bone (skull)

0.005 ± 0.003 Cortical bone (skull)

GRE-05-325 High alpine glacier 0.098 ± 0.006 Trabecular bone (iliac crest)

GRE-05-191 Alpine, inside in a box <0.005 Trabecular bone (skull)

GRE-05-072 Lowland, on top of accumulated leaves 0.012 ± 0.004 Cortical bone (skull)

GRE-04-203 Lowland, lake, in a fish trap <0.005 Cortical bone (skull)

GRE-04-085 Lowland, small river 0.008 ± 0.002 Cortical bone (skull)

GRE-01-256 High alpine glacier 0.107 ± 0.003 Cortical bone (femur)

GRE-01-155 Lowland, forest 0.010 ±0.002 Cortical bone (skull)

a After additional decontamination steps involving iron(III)-reduction and EDTA washing, followed by acetate buffer at pH 4.5 washings. b Target value after solubility profile.

c Decontamination by solubility profiling, only fraction corresponding to fraction 5 + 6 analyzed for 210Po.

to remove adhering soil particles. Afterwards, EDTA and DCTA washings were introduced to decontaminate bones from metal species. Furthermore, reduction of iron hydroxide particles to soluble iron(ll) with hydroxylamine 1 M was introduced to remove many interfering ions adsorbed on adhering iron hydroxides particles. At the end of the 2000s, we introduced acetate buffer washings (90Sr) and acetate buffer solubility profiling (210Po) to suppress diagenesis. Currently, a sequence including all of these decontamination steps is used to avoid diagenetic 90Sr and 210Po.

3. Results

The results of 90Sr bomb pulse dating are presented for 29 cases in Table 1. The uncertainty on the 90Sr measurement was between 10% and 20% (k = 2). Eleven cases showed activity below the detection limit. These cases are therefore not relevant to any forensic investigation because the time of death is prior to 1955, when no 90Sr was present in the environment and food. When 90Sr activity in bones was above the detection limit, the result was compared to the calibration curve determined previously and presented in Fig. 1. One can see that 90Sr activities in vertebrae have rapidly followed the changes appearing in environmental 90Sr [17,19], reflecting a high bone turnover in the trabecualar fraction. Note that the 90Sr activities are presented as a function of the year of death, as an average of at least 12 measurements of individuals who died in Switzerland. Thus these activities must be recalculated for every year of measurement of the skeletal remains using 28.8 years as the physical half-life of 90Sr before comparison. Results of Table 1 show that 3 cases were compatible with a time of death corresponding to the highest 90Sr activity in bones (0.067-1.07 Bq/ gCa), yielding a period of death close to 1963. The results of 210Pb

Table 2

Estimation of the time of death based on the 90Sr and 210Po dating method for 29 cases of human skeletal remains found between 2001 and 2012 in different location in Switzerland.

Case number Discovery location 90Sr (Bq/g Ca) 210Po (Bq/g Ca) Estimated time of death

GRE-12-439 Alpine forest 0.020 ± 0.006 0.025 ± 0.005a After 1990

GRE-12-415 High alpine glacier 0.019 ± 0.002 0.005 ±0.001b Before 1940

GRE-12-414 High alpine, glacier 0.032 ± 0.003 0.005 ± 0.001" Before 1940

GRE-12-413 High alpine, glacier 0.012 ± 0.001 0.002 ± 0.0005b Before 1940

GRE-12-413/414/415 High alpine glacier, mixture of the three samples <0.005c Before 1940

GRE-12-417 Lowland lake <0.005 0.003 ± 0.001a Before 1940

GRE-11-363 Lowland, bed of a river 0.039 ± 0.007 0.009 ± 0.001b After 1970

GRE-11-320 Lowland, construction site <0.008 Before 1955

GRE-11-163 Alpine deep valley, rocky slope 0.019 ± 0.004 0.005 ± 0.002a d After 1955, before 1970

GRE-11-162 Alpine deep valley, rocky slope 0.019 ± 0.003 0.007 ± 0.003a d After 1955, before 1970

GRE-11-064 Inside, garbage bag <0.005 Before 1955

GRE-10-285 Sub-alpine, gravel pit 0.032 ± 0.002

GRE-10-278 Lowland, construction site 0.032 ± 0.004

GRE-10-019 Lowland, lake (fishing net <0.008 <0.015b Before 1955

GRE-10-204 Lowland, under 20 cm of soil <0.005 Before 1955

GRE-09-307 Lowland, closed to ancient cemetery 0.014 ± 0.004 0.015 ± 0.003b After 1955, before 1980

GRE-09-306 Lowland, forest 0.067 ± 0.005 0.021 ± 0.003b Close to 1965

GRE-08-199 Lowland, construction site 0.032 ± 0.004 0.020 ±0.003a After 1955, before 1963

0.028 ± 0.004

GRE-08-018 High alpine, rocky slope 0.036 ± 0.005 0.008 ± 0.004b After 1955, before 1975

0.055 ± 0.005 0.050 ± 0.010b

GRE-07-162 Lowland, construction site <0.005 Before 1955

GRE-07-063 Lowland, construction site 0.014 ± 0.003 After 1950, before 1963

<0.005

GRE-07-001 Lowland, in concrete behind a kitchen wall <0.005 Before 1955

GRE-06-111 Lowland, bank of a river 0.008 ± 0.003 After 1950, before 1963

0.005 ± 0.003

GRE-05-325 High alpine glacier 0.098 ± 0.006 Close to 1965

GRE-05-191 Alpine, inside in a box <0.005 Before 1955

GRE-05-072 Lowland, on top of accumulated leaves 0.012 ± 0.004 1990-2000

GRE-04-203 Lowland, lake, in a fish trap <0.005 Before 1955

GRE-04-085 Lowland, small river 0.008 ± 0.002 After 1955, before 1965

GRE-01-256 High alpine glacier 0.107 ± 0.003 Close to 1965

GRE-01-155 Lowland, forest 0.010 ± 0.002 After 1955

a Target value after solubility profile.

b Decontamination by solubility profiling, only fraction 5 + 6 analyzed for 210Po. c After additional decontamination steps. d Not buried, submitted to weathering.

0.05 -_

i-1-1-1-1-1-1-1-1-1-1

1955 1350 1965 1970 1975 1980 1985 1990 1995 2000 2005

year of death

Fig. 1. Calibration curve for activity of 90Sr in vertebrae based on measurements carried out on bones sampled at autopsy of individuals who died in Switzerland between 1960 and 2001. Uncertainty given for k = 2.

The calibration curve of bomb pulse 90Sr for Switzerland has been determined previously as described in [17,19]. The samples are from individuals aged between 20 and 92 years old at the time of death. Based on 5 g of bone ash introduced in the analysis, all samples have a 90Sr activity above the detection limit of 5 mBq/g Ca.

2.2. 90Sr and 2mPo determination

The method of 210Po and 90Sr determination has been described in detail in [17,19]. Because our measurements from forensic cases cover the last ten years of human remains findings, different protocols of decontamination have been used, according to the state of the art and our own results. Briefly, at the beginning of the 2000s, only basic decontamination steps such as ultrasonication in water were used

Table 3

Anthropological estimates (taphonomy of time since death and 90Sr and 210Po activity (Bq/g Ca).

Case number 90Sr (Bq/g Ca) 210Po (Bq/g Ca) Taphonomy

GRE-12-439 0.020 ± 0.006 0.025 ± 0.005 Adipocere, preserved hairs and textiles, partly bleached by weathering, not covered by algae or moss

GRE-12-415 0.019 ± 0.002 By texture and color: ancient

GRE-12-414 0.032 ± 0.003 By texture and color: ancient

GRE-12-413 0.012 ± 0.001 By texture and color: ancient

GRE-12-417 <0.005 0.003 ± 0.001 Signs of a certain period of contact with water, by texture and color: ancient

GRE-11-363 0.039 ± 0.007 0.009 ± 0.001 Signs of a certain period of contact with water, grease saturated, adipocere, cartilage preservation

GRE-11-320 <0.008 By texture and color: ancient

GRE-11-064 <0.005 By texture and color: ancient

GRE-10-285 0.032 ± 0.002a By texture and color: ancient

GRE-10-278 0.032 ± 0.004a By texture and color: ancient

GRE-10-019 <0.008 <0.015 Signs of a certain period of contact with water; by texture and color: ancient

GRE-10-204 <0.005 By texture and color: ancient

GER-09-307 0.014 ± 0.004 0.015 ± 0.003 Covered by plants, exposed to weathering

GRE-09-306 0.067 ± 0.005 0.021 ± 0.003 Covered by plants, exposed to weathering

GRE-08-18 0.036 ± 0.005 0.008 ± 0.004 Adipocere, grease saturated, soft tissue preservation/mummification, partly exposed to weathering,

not covered by algae or moss

0.055 ± 0.005 0.050 ± 0.010

a Strong suspicion of diagenesis, based on a very low loss of mass on ignition, presence of high level of insoluble (HCl 1M) material and general aspect of fossilization.

dating for 12 cases are presented in Table 1. The uncertainty on the 210Po measurement was between 10% and 20% (k = 2). Half of the cases had 210Po activity below 15 mBq/g Ca. These values were compatible with a PMI longer than 30 years, with limitations presented in Section 4. Table 2 presents an estimation of the period of death based on 90Sr and 210Po dating. The results of anthropological investigations are presented in Table 3. All these results aimed to determine the PMI with a high level of confidence. In all 29 cases, a reliable PMI or a PMI limit (e.g. before 1955) was attributed (Table 2), either by 90Sr bomb pulse dating alone, or in combination with 210Po dating, or still with the addition of anthropological investigation.

The results of the determination of 210Po at the time of death are presented in Fig. 2. The activities cover a wide range of values, from 15 to 97 mBq/gCa. The activities of the 14 ribs do not differ significantly from the activities of the 21 vertebrae and average value is 31 ± 22 mBq/g Ca (k = 1). Finally, we measured 210Po in urine sampled from non smoker and one heavy smoker collaborators of the Institut of Radiation physic to probe the effect of smoking on the incorporation of 210Pb/210Po. Results showed that the non smokers have 210Po activity in urine below 5 mBq/l while the heavy smoker has 210Po activity in urine of 15 mBq/l (Fig. 4). This increase

Po urine excretion is a result of an increase in incorporation of 210Pb/210Po that may also be reflected in 210Pb/210Po activity in bones.

Fig. 2. Activity of 210Po (Bq/g Ca) based on measurements carried out on bones sampled at autopsy of individuals who died in Switzerland (vertebrae) or were autopsied in Lyon (ribs) between 2006 and 2011 (reference date 2011).

4. Discussion

The results of our study suggest that bomb pulse 90Sr dating in combination with 210Pb dating (through 210Po measurement) can be used to reveal PMI as long as severe precautions are taken to avoid 90Sr and 210Po diagenesis. This was confirmed by the anthropological investigations, which gave a similar PMI to the radiometric method in all studied cases. Being able to reject all cases with 90Sr activity below the detection limit of 0.005 Bq/g Ca as being of no forensic interest because of a time of death before the nuclear era represents the major advantage of using the radiometric method in forensic science. This rather simple measure will prevent a tremendous amount of forensic work, e.g. extraction of DNA. In our study, one third of all cases were archeological and not legal matters and were then rejected on the basis of 90Sr dating alone. Nevertheless, it is important to note that finding 90Sr above the detection limit may mislead the final PMI attribution because of two drawbacks: (i) 90Sr can come from diagenesis and (ii) 90Sr bomb pulse, while not symmetric (Fig. 1), has an upslope and downslope to the curve with similar activity for very different years of death.

The remains of three skeletons (cases # GRE-12-415; GRE-12-414 and GRE-12-413) found in 2012 on a Swiss glacier illustrate the first problem. Formally identified, they have been missing before 1940. Their bones should thus contain no 90Sr (death before nuclear era) and should have a low 210Po activity. The bones (calcaneum) received in the laboratory for radiometric dating were well preserved, with an organic matter lost on ignition of 50%. After decontamination (see [17] for details) 90Sr was nevertheless found above the detection limit. This 90Sr was obviously of a diagenetic origin. On the other hand, 210Po activity was very low in the three bones, indicating a long PMI. This finding showed that skeletal remains from glaciers could be mixed with the ice and snow layers formed during the beginning of the sixties, when 90Sr atmospheric fallout was at its highest. We measured 90Sr again, adding several steps of hydroxylamine and acetate buffer washings; 90Sr activity dropped below the detection limit. The case of skeletal remains found on another Swiss glacier (case # GRE-01-256) posed a similar problem; the 90Sr activity was 1.07 Bq/g Ca, obtained after a very basic decontamination using ultrasonication in water that was in use in 2001, when the measurement was carried out. Based on our calibration curve of Fig. 1, the time of death should be close to 1962. However, several objects found in proximity of the remains were from the end of the 19th century, including coins from the 1880s which were no longer in use in the sixties. In this

respect the relics were of archeological interest. Here again, the remains have probably been mixed in with the ice layer from the sixties due to melting, leading to 90Sr diagenesis. The five cases of glacier remains showed the highest activities (e.g. case # GRE-01-256), with possibilities of diagenesis. This leads us to determine that when a positive value is observed, it should always be confirmed by a second measurement, after additional decontamination steps. When possible, 210Pb dating should be added, preferably as solubility profiling, as in case # GRE-08-018. The solubility profiling method is based on the difference in solubility of the apatite of biogenic and diagenetic origin. The first washings with acetate buffer at pH 4.5 will dissolve less crystallized apatite from diagenetic origin [20]. Further washings will contain biogenic 210Pb/210Po [17]. In the cases of the three skeletons found on a glacier, biogenic 210Po confirmed a long PMI, with 210Po activities less than 5 mBq/g Ca.

Cases # 11-162/163 illustrate the limitation of 210Po dating when the human remains are not buried and therefore submitted to weathering. It also illustrates the interest of conducting 90Sr and 210Po measurements in trabecular and cortical bone, if both are recovered. The remains from this case were found not buried, at about 1100 m above sea level, in April 2011. Formally identified, the forensic investigations established that the individual had been missing since 2008. The analyzed bones were completely bleached, with a very low content in organic matter (loss on ignition < 10%). Results show that vertebrae and femur have the same 90Sr activity of 0.019 Bq/g Ca. This activity in vertebrae is expected for a time of death close to 1960 or after 2000. Thus we carried out a 210Po solubility profile to get a better estimate of the PMI. The solubility profile (Fig. 3) reveals no diagenetic 210Po (stable values) and 210Po activity below 10 mBq/g Ca. This low value of 210Po alone yields a long PMI, not compatible with the forensic investigation. It is then possible that weathering, with heavy rains and snow melting, would have leached the 210Pb out of the bone structure. As a matter of fact, the Ca content of the bone ash in this case was significantly higher (0.40 g/g ash) than for bone ash in our database (0.33 ± 0.04 g Ca/g ash, n = 310). This might express a loss of biogenic lead phosphate where lead replaces Ca in the apatite structure. In addition, weathering might have leached 210Po preferentially because of the local destruction of the crystalline lattice due to nucleus recoil when 210Pb and 210Bi, grandmother and mother of 210Po, respectively, disintegrate, as observed in the radioactive disequilibrium between 238U and 234U in water [21].

Often, human skeletal remains consist of only the skull and/or long bone diaphyses. On the other hand, trabecular bones found in vertebrae, ribs, etc. are less often preserved. In this case, our problem was to extend results established for trabecular bone to compact bone. Compact bone takes up 90Sr and 210Pb in food at the time of bone formation. In contrast, trabecular bone takes up 90Sr and 210Pb in food at the time of death, because of a higher bone remodeling rate. In our study, seven cases consisted only of the skull. Among them, three showed 90Sr activities below the detection limit and were discarded as archeological cases. The presence of 90Sr in the 4 remaining cases demonstrates that the time of death is posterior to 1955. It was nevertheless impossible to give a more precise time of death because we do not have an established calibration curve for this kind of bone. Higher 90Sr activity in the skull compared to vertebrae is expected for a recent time of death only if the skull was forming during the sixties, where food contamination was at its highest. Thus it is necessary to turn to the anthropologist evaluation of the age at death to improve the determination of the time of death. For instance, human skeletal remains found in 2010 from an individual older than 60 years and who died recently should contain only traces of 90Sr because the skull was already formed during the middle of the sixties, when food contamination was high. 210Po/210Pb dating

O vertebrae

□ femur ^

-s- 10-1

1+2 3+4 5+6

washing number

Fig. 3. Solubility profile of 210Po in vertebrae (circle) and femur (square) from the #GRE-11-162 and GRE-11-163 cases.

should indicate a possible time of death but was carried out here only for the GRE-09-307 case. The activity of 210Po was 0.015 Bq/g Ca. This activity is significantly lower than expected for a short PMI. Consequently we attributed a period of death after 1955 (presence of 90Sr) and before 1980 (intermediate 210Po value).

The uncertainties in the 90Sr bomb pulse and 210Pb dating are mostly the consequence of a large uncertainty on the radionuclide measurement, to the possibility of diagenesis and to the differences in individual 90Sr and 210Pb uptake. For both 90Sr and 210Po measurements at the encountered activities in human, the uncertainty (k = 2) is above 10% and can reach 20%. The contamination of food with 90Sr can differ depending on the location and dietary habits; for instance, 90Sr activity in dairy food (milk and cheese) is proportional to the altitude of cows grazing, with higher activities measured in Alpine milk than in milk from lowland regions of Europe [22]. Thus in Switzerland, where people are living between 300 and 2000 m above sea level, the contamination of food, reflected in the vertebrae, may significantly differ, depending on the living area. This will add uncertainty in estimating PMI based only on the 90Sr activity found in vertebrae. In fact, our 90Sr calibration curve of Fig. 1 has been established by measuring vertebrae of people having lived in the Swiss lowland (mostly Lausanne, Zurich, Basel and Lugano regions). In addition, to compare the 90Sr activity found in human remains to the ones of Fig. 1 requires the activity equilibrium between vertebrae and local food, which will take about two years to establish [19]. In a country with a high immigration rate such as Switzerland, this condition may not be totally fulfilled.

To estimate the PMI based on 210Pb dating requires the knowledge of the 210Pb level at the time of death. The results of 35 measurements show a large dispersion of values (min = 6.4 mBq/ gCa; max = 97 mBq/gCa). The average value is 31 mBq/g Ca. The standard deviation is 22 mBq/g Ca. Thus, the 210Pb content in human ribs and vertebrae varies greatly at the time of death between individuals. In a previous study, we found a target value of 45 ± 11 mBq/g Ca in vertebrae of individuals who died in 1990, excavated and measured in 2008. Recalculated to the reference date 2008, the mean activity at the time of death would be 70 mBq/g Ca, about twice the activity found in this work. So it seems that the target value at death determined in [17] is over-evaluated, possibly because the samples came from buried bodies. On the other hand, the samples in [17] were excavated from a Milanese cemetery and might illustrate differences in the place they lived, medications, age at death, and smoking and dietary habits. Fig. 2 also shows that 7 values are below 20 mBq/g Ca at the time of death. This will cause a large over-estimation of PMI based on 210Pb dating for these individuals. On the other hand, two cases have values close to 90 mBq/g Ca, causing an under-estimation of PMI, should we encounter these cases in forensic

Fig. 4. 209Po tracer (left peak) and 210Po (right peak) spectra of urine sampled from non-smoker collaborators of the Institute of Radiation Physics (dotted line) and a heavy smoker collaborator of the Institute of Radiation Physics (black line).

evaluation. The main cause of the large inter-individual difference in the 210Po activity at the time of death also probably reflects a smoking habit. Tobacco fumes are known to contain both 210Pb and 210Po and to significantly increase 210Po excretion (Fig. 4). Thus heavy smokers are expected to have more 210Pb-210Po in their bones than non-smokers. In this respect, the two cases with 210Po activity close to 90 mBq/g Ca at the time of the death might well have been heavy smokers.

5. Conclusions

Here we show that a third of the cases can be discarded as ''of archeological interest'' using 90Sr determination only. This fact alone saves a lot of forensic work. In addition, the radiometric measurements yield hard data that do not rely on the investigator. In spite of large uncertainties on the model due to differences between individuals in 210Pb and 90Sr uptake, these hard data provide estimates which can be used to refine anthropological hints. We conclude that 90Sr and 210Po dating, if carried out with the necessary precautions discussed above, can be used to determine PMI. Nevertheless, anthropological investigations should always be carried out to confirm the radiometric estimate. Eventually, our study demonstrates that 90Sr and 210Pb dating should be carried out together to increase the power of the radiometric estimate, particularly because these two radionuclides are not submitted to the same degree of diagenesis, depending on the burial site.

Nevertheless, the method suffers drawbacks that the investigators must keep in mind; once the diagenetic problem is resolved, weathering can leach 210Po from bones, yielding to unrealistic long PMI. In addition, heavy smoking habits will yield too short PMI, if based on 210Po dating only. In this respect, cases of forensic interest presenting the above-described limitations could be submitted to the more precise but less available 14C-AMS bomb pulse dating.

As a final remark, we propose that all the activities reported for radionuclides in human bone are normalized to the calcium content. We strongly think that a normalization to the bone wet or dry weight, as often reported in the literature, is not the right way to report activities in bones because the level of skeletisation (lost of organic matter) is not known. This is particularly true for human remains. In addition, calcium diagenesis is not susceptible to significantly change the calcium content in bone apatite.

Role of the funding source

This work has been funded by the Swiss Federal Office of Public Health and by the University of Lausanne.

Acknowledgements

The authors would like to thank: J.-J. Geering for the sampling and the determination of 90Sr in vertebrae from 1960 to 2001; F. Barraud for her help in the 90Sr measurements; A. Savary, for her help in the 210Po solubility profile and 210Po determination; Dr. L. Fenton for the gift of the ribs set; the Pathology Institutes of Lausanne and Lugano for the yearly sampling and gift of the vertebra set since 1962.

References

M.A. Granrud, G.R. Dabbs, A preliminary study of incisor exfoliation as an estimator of the postmortem interval using accumulated degree days, Forensic Sci. Int. 222 (2012) E29-E32.

[2] R. Alaeddini, M. Ahmadi, S.J. Walsh, A. Abbas, Semi-quantitative PCR analysis of DNA degradation, Aust. J. Forensic Sci. 43 (2011) 53-64.

[3] R.C. Griffin, H. Moody, K.E.H. Penkman, M.J. Collins, The application of amino acid racemization in the acid soluble fraction of enamel to the estimation of the age of human teeth, Forensic Sci. Int. 175 (2008) 11-16.

[4] F. Ramsthaler, K. Kreutz, K. Zipp, M.A Verhoff, Dating skeletal remains with luminol-chemiluminescence: validity, intra- and interobserver error, Forensic Sci. Int. 187 (2009) 47-50.

[5] A.F. Schilling, T. Kummer, R.R. Marshall, A. Bauerochse, E. Jopp, K. Pueschel, M. Amling, Brief communication: two and three-dimensional analysis of bone mass and microstructure in a bog body from the Iron Age, Am. J. Phys. Anthropol. 135 (2008) 479-483.

M.V. Monsalve, E. Humphrey, D.C. Walker, C. Cheung, W. Vogl, M. Nimmo, Brief communication: state of preservation of tissues from ancient human remains found in a glacier in Canada, Am. J. Phys. Anthropol. 137 (2008) 348-355.

[7] W.A. Murphy, D. zur Nedden, P. Gostner, R. Knapp, W. Recheis, H. Seidler, The Iceman: discovery and imaging, Radiology 226 (2003) 614-629.

[8] B. Swift, I. Lauder, S. Black, J. Norris, An estimation of the post-mortem interval in human skeletal remains: a radionuclide and trace element approach, Forensic Sci. Int. 117 (2001) 73-87.

P. Neis, R. Hille, M. Paschke, G. Pilwat, A. Schnabel, C. Niess, H. Bratzke, Stron-tium90 for determination of time since death, Forensic Sci. Int. 99 (1999) 47-51.

10] B. Swift, Dating human skeletal remains: investigating the viability of measuring the equilibrium between Po-210 and Pb-210 as a means of estimating the postmortem interval, Forensic Sci. Int. 98 (1998) 119-126.

11] C.F. Speller, K.L. Spalding, B.A. Buchholz, D. Hildebrand, J. Moore, R. Mathewes, M.F. Skinner, D.X. Yang, Personal identification of cold case remains through combined contribution from anthropological, mtDNA, and bomb-pulse dating analyses, J. Forensic Sci. 57 (2012) 1354-1360.

12] K.L. Spalding, B.A. Buchholz, L.E. Bergman, H. Druid, J. Frisen, Age written in teeth by nuclear tests, Nature 437 (2005) 333-334.

13] F. Marzaioli, V. Fiumano, M. Capano, I. Passariello, N. De Cesare, F. Terrasi, Forensic applications of C-14 at CIRCE, Nucl. Instrum. B 269 (2011) 3171-3175.

14] B. Zinka, R. Kandlbinder, R. Schupfner, G. Haas, O.S. Wolfbeis, M. Graw, The activity ratio of Th-228 to Ra-228 in bone tissue of recently deceased humans: a new dating method in forensic examinations, Anthropol. Anzeiger 69 (2012) 147-157.

15] B. Zinka, R. Kandlbinder, G. Haas, R. Schupfner, O. Wolfbeis, M. Graw, Radionuclide analysis of Th-228 and Ra-228, Rechtsmedizin 21 (2011) 124-130.

16] R. Kandlbinder, V. Geissler, R. Schupfner, O. Wolfbeis, B. Zinka, Analysing of Th-228, Th-232, Ra-228 in human bone tissues for the purpose of determining the post mortal interval, J. Radioanal. Nucl. Chem. 280 (2009) 113-119.

17] B. Schrag, T. Uldin, P. Mangin, P. Froidevaux, Dating human skeletal remains using a radiometric method: biogenic versus diagenetic Sr-90 and Pb-210 in vertebrae, Forensic Sci. Int. 271 (2012) 271-278.

18] N. Ziad, R. Zarki, M. Benmansour, T. Sayerh, A. Laissaoui, Determination of Pb-210 in human skeletal remains from Morocco: implications for time since death assessment, J. Radioanal. Nucl. Chem. 292 (2012) 315-319.

19] P. Froidevaux, F. Bochud, M. Haldimann, Retention half times in the skeleton of plutonium and 90Sr from above-ground nuclear tests: a retrospective study of the Swiss population, Chemosphere 80 (2010) 519-524.

20] A. Sillen, J.C. Sealy, Diagenesis of strontium in fossil bone: a reconsideration of Nelson et al. (1986), J. Archaeol. Sci. 22 (1995) 313-320.

21] R.L. Fleischer, O.G. Raabe, Recoiling alpha-emitting nuclei - mechanisms for uranium-series disequilibrium, Geochim. Cosmochim. Acta 42 (1978) 973-978.

22] L. Pourcelot, P. Steinmann, P. Froidevaux, Lower variability of radionuclide activities in upland dairy products compared to soils and vegetation: implication for environmental survey, Chemosphere 66 (2007) 1571-1579.