- Email: [email protected]
Metal content in medieval skeletal remains from Southern Croatia
Journal of Archaeological Science 46 (2014) 393e400
Contents lists available at ScienceDirect
Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas
Metal content in medieval skeletal remains from Southern Croatia Angela Stipisic a, *, Maja Versic-Bratincevic b, Zlatka Knezovic a, Davorka Sutlovic b, c a
Public Health Institute of Split e Dalmatia County, Vukovarska 46, 21 000 Split, Croatia Department of Forensic Medicine, University of Split, School of Medicine, 21 000 Split, Croatia c Department of Pathology and Forensic Medicine, University Hospital Centre Split, 21 000 Split, Croatia b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 January 2013 Received in revised form 11 March 2014 Accepted 28 March 2014 Available online 8 April 2014
This work is a contribution to the existing knowledge of lifestyle and diet of the South Croatian population who lived in Early Medieval Period. The one hundred samples dating from 9th century were discovered and collected at the burial sites Ostrovica and Naklice. Concentrations of metals and their mutual relationships were examined in regards to gender and age of skeletal remains. Differences were observed in diet between men and women as well as among age groups. For a correct interpretation of the results it is necessary to determine the metal content in the soil. Namely, metals in archaeological bones are influenced by changes in soil e diagenesis, which is confirmed by our results. We concluded that there were no influences of diagenesis on lead, calcium, strontium and zinc content while cadmium, iron, manganese and copper are most exposed. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Archaeological bones Early Medieval Period Heavy metals Diet reconstruction Diagenesis
1. Introduction Environmental conditions, dietary habits and uptake of some elements from the surrounding soil may reflect on heavy metal content in human bones. This statement is supported in several papers with results of metals in archaeological human bones from different locations and different periods (Martinez-Garcia et al., 2005; Dobrovolskaya, 2005; Shafer et al., 2008; Donno et al., 2010; Nakashima et al., 2010; Yamada et al., 1997). Martinez-Garcia et al. (2005) determined the content of lead, copper, zinc, cadmium and iron in archaeological bones from the area of Cartagena in Spain from different historical periods. According to their study, the lowest concentrations of lead were in the Neolithic period (median Pb 45 mg/g), the values from the Bronze Age were higher (median Pb 80 mg/g), and maximum were recorded during the Roman (median Pb 742 mg/g) and the Byzantine era (median Pb 898 mg/g). The concentrations of copper were highest in Byzantine era bones (median Cu 111 mg/g), and the concentrations of iron during the Ottoman Empire (median Fe 15 383 mg/g) (Martinez-Garcia et al., 2005). Schutkowski and Herrmann (1999) analyzed metals in archaeological human bones from 6th to 8th century, from Northwestern
* Corresponding author. Tel.: þ385 021401175; fax: þ385 021539825. E-mail address: [email protected] (A. Stipisic). http://dx.doi.org/10.1016/j.jas.2014.03.032 0305-4403/Ó 2014 Elsevier Ltd. All rights reserved.
Germany (Schutkowski and Herrmann, 1999). They found low zinc content, increased content of copper and strontium as well as enlarged ratio of Sr/Ca. Strontium is not an essential element. It is found in plant origin food. Ratio of Sr/Ca in bones is very important. Higher ratio indicates that foods of plant origin (cereals and legumes) were predominated in the diet, while a smaller ratio suggests higher incidence of animal origin food (milk and dairy products) (Schutkowski and Herrmann, 1999). According to Mays (2003), strontium is mostly located in the skeleton, with a strong relationship between Sr/Ca ratio in diet and in the bones of the consumer (Mays, 2003). The analyses of Sr/Ca ratio in human skeletal remains from medieval times from England were carried out, particularly in archaeological bones of breastfeeding mothers. Mays found association not only between Sr/Ca ratio and diet from that period, but also with metabolic changes during lactation. During lactation, transfer of calcium is faster than strontium, as well as through the placenta during pregnancy. Potentially, according to Mays, bone Sr/Ca ratios is a sensitive technique for detecting the administration of lowprotein or animal-milk supplements into infant diets. This is important as historical sources indicate that such foods were frequently used as infant feeding supplements in the past (Mays, 2003). In archaeological studies analysis of metals in bones reflect diet and lifestyle habits of archeological populations. Contamination and diagenesis of archaeological bones can compromise the credibility of this information (Martinez-Garcia et al., 2005; Nielsen-
394
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
Marsh and Hedges, 2000). Diagenesis is a series of changes in the soil, including: dissolution, precipitation, crystallization and exchange of minerals between the soil and bones. According to Nielsen-Marsh et al. (2006), there are three important alterations to bone apatite that can be used as indicators of diagenetic change: porosity increase, crystallinity increase and inclusion of exogenous ione (Nielsen-Marsh et al., 2006). These processes can significantly affect the archaeological bones. Preservation of archaeological bone depends on the composition and acidity of the soil (Hedges and Millard, 1995; Hedges, 2002). Archaeological sites that are located on gentle hills and not being exposed to groundwater are less exposed to the influence of diagenesis (Hedges and Millard, 1995; Nielsen-Marsh and Hedges, 2000). The composition of calcareous soil with slightly alkaline pH values is appropriate for preservation of osteological material over thousands of years. The alkaline soil can preserve bone far better than the sour one, which easily destroys hydroxyapatite, an integral part of the bone (Hedges, 2002). Previous studies in Croatia indicated that the osteological material is best preserved at the archaeological sites along the Adriatic coast, under the condition that their location is not close to the sea (Slaus, 2006c; Novak and Slaus, 2010). These sites have soil with neutral to slightly acidic pH values and high content of sand, whereas in Central and Continental Croatia, level of preservation varies depending on the acidity of the soil and the amount of groundwater (Slaus, 2006c). Selection of archaeological bone is also very important for reducing the impact of diagenesis (Vuorinen et al., 1990a; Price et al., 1992). Long parts of femur are less exposed to changes in the soil than the short bones. Trabecular bone is more exposed to the diagenesis than compact cortical bone (Carvalaho et al., 2004; Ambrose and Krigbaum, 2003). Contamination of archaeological bones can occur with the long laying in the soil, but the sampling and preparation techniques may also cause it (Price et al., 1992). The part of the skeletal remains must be selected very carefully for chemical analysis before metal determination. During the preparation of samples, metal accessories such as saws, forceps, etc. should not be used. Contemporary archaeological research in Croatia became more intensive at the beginning of nineties. Analysis included anthropological and archaeological research (determination of gender, age, trauma, dental analysis, as well as osteological changes indicating different diseases) both for the continental and coastal Croatian (Slaus, 1997a, 2000b, 2006c, 2008d; Novak and Slaus, 2010). During the 90-ies of the 20th century the Museum of Croatian Archaeological Monuments in the excavations carried out a number of sites from the early Middle Ages, or the 9e10th century (Delonga and Buric, 1998). However, in the aforementioned studies in Croatia so far no one has dealt with the analysis of metals in archaeological bones. In our study the content of metals was analyzed in 100 samples of archaeological bones in compact part of the bone e cortical femoral. Archaeological bones were collected from two archaeological sites of Southern Croatia: Ostrovica and Naklice (Fig. 1). The influence of diagenesis on the distribution of metals in archaeological samples was investigated by determining metals and pH in soil samples which were collected from the burial site during the archaeological excavations. The aim of our study was to determine the difference of metal distribution in regards to gender and age, as well as potential differences in social status between men and women, and children and adults. We also tried to determine whether there are significant differences referring to metal distribution in the bones of the early Croatian population for two different archaeological sites from the same historical period.
2. Methods 2.1. Instrumentation and operating conditions Lead and cadmium measurements were carried out by using a Model AAS vario 6 GFAAS atomic absorption spectrometer (Analytik Jena AG, 2001) equipped with a transversely heated graphite atomizer with autosampler (Model MPE 50), a deuterium background correction system and a hollow cathode lamp for lead operated at 3 mA (wavelength 283.3 nm) and for cadmium operated at 3 mA (wavelength 228.8 nm). Pyrolytic coated graphite tubes with PIN-platform (Analytik Jena, Part No. 407-A81.025) were used during the anlytical determination (Analytik Jena AG, 2001). The injection volume was 20 mL and integrated absorbance (peak area) was used for signal evaluation. Calcium, strontium, zinc, copper, iron and manganese determinations were carried out by using a Model AAS vario 6 FAAS atomic absorption spectrometer equipped with deuterium background correction system and a hollow cathode lamp for calcium, strontium, zinc, copper, iron and manganese (wavelength 422.7 nm for Ca; 460.7 nm for Sr; 213.9 nm for Zn; 324.8 for Cu nm; 248.3 nm for Fe and 279.5 nm for Mn). Concentrations of Zn, Cu, Fe and Mn were determinated by Flame AAS with C2H2/air burner, while Ca and Sr were determinated with C2H2/N2O2 burner (Analytik Jena AG, 2001). Samples were weighted on the Mettler Toledo balance e Model AX 205DR (Mettler, Germany) with resolution of 0.01 mg. A closed microwave system, CEM Model Mars 5 was used for wet digestion (CEM Corporation, 2006). pH values of the soil samples were determined with pH-meter Model inoLab e pH720; measuring range and resolution e 2.000. þ 19.999 with accuracy of 0.005 (Wissenchaftlich Technishe Werkstatten WTW, 2004). 2.2. Reagents Working standards of Pb, Cd, Ca, Sr, Zn, Cu, Fe and Mn for measurements were prepared from Merck (Darmstadt, Germany) stock solutions (1000 2) mg/L, suprapur. Standard solutions were prepared in range of expected concentration values. We applied the same technique for the metal analysis in two different matrices: soil and human bones. To confirm the success of the method performed, we used the available Certified Standard Reference soil material: SRM e 2710a Montana I Soil, from the National Institute of Standards and Technology (NIST, 2009). Pyrolysis and atomization curves were established in the presence of chemical modifier e 0.1% Pd(NO3)2 þ 0.05% Mg(NO3)2 6H2O. Modifier was prepared from Merck stock solution: Art.1.07289 palladium matrix modifier for graphite furnace AAS and Art.1.05855 magnesium nitrate hexahydrate. Volumen of added modifier was 5 mL. 2.3. Samples preparation After archaeological excavations, bone samples were classified according to the anthropological parameters: gender, age (0e15, 16e25, 26e39 and above 40 years) as well as geographical location. For chemical analysis, bone samples were prepared with special attention to avoid the post-mortem contamination. A part of a dense cortical bone was sawed for analysis. Bone pieces were crushed into small fragments using razor blades and stored in sterile polypropylene tubes at 20 C until analyzed. After drying to a constant weight samples were washed in 6 ml 65% nitric acid (HNO3) over night, subsequently washed in distilled water and finally dried at room temperature.
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
395
Fig. 1. Skeletal remains of adults from archaeological burial site Ostrovica. Samples are arranged for anthropological process and sampling for chemical analysis.
Approximately 0.5 g of the sample was weighed in a teflon-TFM vessel and with the addition of 65% nitric acid and hydrogen peroxide wet e ashed in automated (temperature regulated) microwave digestion unit (CEM, USA Model Mars 5 with 1600 W power). Digested samples were diluted with deionized water and metal content was determined quantitatively. We conducted analysis of soil samples to determine the extent of the archaeological bone contamination with the soil. We analyzed the soil from two archaeological sites; Ostrovica and Naklice. Each of them included two soil layers: the upper sample (0.5 m from the surface) and the lower sample (1.0 m from the surface). Lower layer of the sampled soil is at an approximate distance of 30 cm from the skeleton. Each of ground samples was analyzed four times. Preservation of osteological material depends on the acidity of the soil, thus pH value in the 5% water solution of soil was determined. For quantitative analysis of metal samples were prepared by microwave digestion in the same way as bone samples (CEM Corporation, 2006).
2.4. Method validation The linear calibration of Pb was obtained using five standard solutions prepared in 0.5% HNO3 in the range (10e50) mg/L Pb. The linear calibration of Cd was obtained using five standard solutions prepared in 0.5% HNO3 in the range (0.2e1) mg/L Cd. Relative standard deviations of five replicate samples for each standard solution were lower than 10%. The correlation coefficient (R) obtained by linear fitting to the calibration points was better than 0.995. The detection limit for Pb and Cd analyzed by GFAAS technique is 1 mg/L. Matrix effects and interferences to other metals were minimized with graphite platform and modifier 0.01% Mg(NO3)2 þ 0.5% Pd(NO3)2. The detection limits for other metals analysis technique FAAS are: Ca 100 mg/L; Sr 200 mg/L; Zn 10 mg/L; Cu 50 mg/L; Fe 50 mg/L; Mn 10 mg/L. The standard reference material NIST e SRM 2710a (Montana I Soil) was used throughout this study to validate the analytical
396
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
technique (National Institute of Standard and Technology, 2009). The repeatability of sample preparation and measurements are expressed as relative standard deviation. All the determined values of relative standard deviation (coefficient of variation) were less than 10%, indicating a good precision of measurement. Effect of matrix and chemical interference has been reduced to a minimum of the methods which confirm the values obtained percentage recovery (91.3e110%) (Table 1).
Table 2 Statistical analysis of metals in archeological bones (N ¼ 100). The results of metals are presented in mg/g, except the results of Ca in %.
Mean Median SD SE Minimum Maximum
Pb
Cd
Ca*
Sr
Zn
Cu
Fe
Mn
1.266 0.610 2.420 0.242 0.002 16.400
0.074 0.047 0.008 0.047 0.001 0.459
33.97 32.87 4.20 0.42 20.90 49.31
395 383 104 10 202 659
112 111 53 5 14 311
5.83 3.01 9.45 0.95 0.05 72.90
1135 244 3315 331 11 29787
58.08 13.28 106 10.6 0.40 684
3. Results and discussion The results were statistically analyzed by determining the average value (mean and median), standard deviation (SD), standard error (SE), measurement range (minimum and maximum value) for each metal in the test groups (Table 2). Results of the analysis of metals are expressed as mean and median. For comparing and expressing the results in our paper, we used the median. Namely, it is useful to express the results in this way, when the distribution of results has very large extreme values (minimum and maximum values). Determination of lead concentration in recent human bones indicates exposure of “modern” humans to harmful effects of environmental pollution (Baranowska et al., 1995; Jurkiewicz et al., 2004). Lead is a toxic metal, which in the form of lead phosphate has a significant ability to accumulate in the bones, particularly in the fast growing bones (femur and tibia) (Baranowska et al., 1995). It is very slowly excreted from the body. In various studies of toxic metals and their harmful effects on man, lead is the most important indicator of exposure to toxic metals. However, in an archaeological study lead can be used for determining the way of life of the population (eg. use of pottery with lead glaze). Determination of lead in archaeological studies is extremely important for the Roman period and the Bronze Age, particularly in archaeological population at that time when pottery with lead glaze was used (Martinez-Garcia et al., 2005; Donno et al., 2010; Shafer et al., 2008; Pyatt et al., 2005; Nakashima et al., 2010; Sutlovic et al., in press). Pyatt et al. study (2005) results showed significantly higher concentrations of lead than results of our study (median Pb 0.610 ug/g). They analyzed skeletal remains from the Bronze and Roman Age. The reason for extremely high concentrations of lead in bones (human femur 170e196 mg/g) from that period was due to using pottery with lead glaze (Pyatt et al., 2005). Nakashima et al. (2010) also showed high values of lead (Pb median 313 ug/g) in archaeological bones of children samurai from the Edo period 1600the1867th in Japan, who also used pottery with lead glazes (Nakashima et al., 2010). Lead concentrations in archaeological bones in our study are lower than in both mentioned studies. Results of the concentration of lead (range 0.02e14.76 mg/g) in the study of archaeological bones by Gonzalez-Reiemers et al. (2005) from Al Hair, Canary Islands area, are similar to the results of our study (range 0.002e16.4 mg/g). On the other hand the values of cadmium in the above study are higher (range 0.003e2.56 mg/g) Table 1 Results of metals in certified material SRM e 2710 Montana Soil were determined by AAS. The results of metals are presented in %, except result of Cd* and Sr* in mg/g. Metal
Certified
Measured (n ¼ 6)
SD
RSD (%)
Recovery (%)
Pb Cd Ca Sr Zn Cu Fe Mn
0.552 12.3* 0.964 255* 0.418 0.342 4.32 0.214
0.547 11.2* 0.905 237* 0.446 0.349 4.08 0.212
0.024 0.186* 0.070 20.9* 0.008 0.008 0.055 0.004
4.33 1.66 7.76 8.83 1.91 2.35 1.35 1.98
99.1 91.3 93.9 92.9 106.7 102.0 94.4 99.1
than in our study (range 0.001e0.459 mg/g) (Gonzalez-Reiemers et al., 2005). The available archaeological sources of Early Medieval Croatian population showed their hard living conditions i.e. the extreme poverty (Goldstein, 1995). Archaeological objects discovered along the graves indicate that they did not use pottery with lead glazes (Sutlovic et al., 2010; Delonga and Buric, 1998). Early Medieval man in Croatia had to struggle with nature continuously. Poverty and hunger led to various diseases (leprosy and tuberculosis) and death (Goldstein, 1995; Novak and Slaus, 2010). Anthropological research on the skeletons indicates the frequent occurrence of hypertrophied muscle insertions and ossification tendons, which indicates a high level of physical activity e heavy physical work. According to anthropological studies, life expectancy generally did not exceed 45 years, with a high mortality rate of infants and children (Slaus, 2006c). A common osteological indicator of iron deficiency anemia due to inadequate diet, unhygienic living conditions and chronic gastrointestinal disease in archaeological bone is cribira orbitalia (Slaus, 2006c, 2008d). More often it was observed on the archaeological bones of children. Content of calcium, strontium, zinc and their ratios in archaeological bones are indicator of the diet type (Schutkowski and Herrmann, 1999; Lambert et al., 1985). The characteristics of these metals, their representation in a certain food type and other archaeological characteristics in osteological remains (analysis of teeth, etc.) can be related to diet type and lifestyle of population (Schutkowski and Herrmann, 1999). Calcium is an important metal in the human body, and it is mostly represented in bones. It is responsible for many metabolic processes as well as for teeth and bones growth. Dairy food and legumes contain calcium in high concentrations. Although strontium is not an essential element, high concentrations were found in the bones. It is present only in foods of plant origin, particularly in legumes (Schutkowski and Herrmann, 1999). Its determination in archaeological bones and determining the ratio of Sr/Ca in bioarchaeological studies is very significant (Schutkowski and Herrmann, 1999; Mays, 2003). The value of this ratio indicates the origin of calcium. An increased ratio suggests that calcium is from plant origin food (Runia Lex, 1987). Zinc is an important element in the human organism and is responsible for activity of many enzymes. It is usually found in red meat, and legumes. However, if the source of zinc is vegetable food, the human body cannot absorb it. Vegetables contain high concentrations of phytic acid, which creates a form of insoluble zinc phytate. Therefore, the introduction of zinc through the food of animal origin is better absorbed in the human body. Low zinc content in bones indicates a diet with reduced amount of animal origin food (Schutkowski and Herrmann, 1999).
3.1. The difference between archaeological sites The material found in the graves of both archaeological sites (Ostrovica and Naklice) determines the cultural and temporal
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400 Table 3 Statistical analysis the ratio of metals in archaeological bones (N ¼ 100).
Mean Median SD SE Minimum Maximum
397
Table 4 Sperman’s coefficient correlation between iron and manganese.
Ostrovica
Naklice
Bones
Sperman’s coefficient (rs)
Statistical correlation (p)
Sr/Ca 103 Zn/ Fe/Mn Ca 103
Sr/ Zn/ Fe/Mn Ca 103 Ca 103
0.855
0.000
1.07 1.04 0.04 0.29 0.53 1.85
1.32 1.23 0.41 0.07 0.68 2.39
Archeological (N ¼ 100) Archeological site e Ostrovica (N ¼ 64) Archeological site e Naklice (N ¼ 36)
0.744
0.000
0.820
0.000
0.36 0.35 0.01 0.10 0.15 0.78
39.49 33.92 3.20 25.57 4.88 149
0.31 0.19 0.30 0.05 0.05 1.19
11.63 11.32 6.02 1.06 1.77 25.42
affiliation to the burial site of the Early Medieval period. Regional studies examine differences between geographically separated populations within a relatively small and well-defined space. Comparing the results from two archaeological sites, Naklice and Ostrovica, distinctions in metal concentrations were analyzed demonstrating differences in their diet and lifestyle. Correlations among the metals were investigated in an attempt to reconstruct the dietary habits of the population. Values of the ratio of metals are specified in Table 3. Bones samples from location Naklice contain increased value of Sr and high ratio of Sr/Ca (median 1.23 103) while bone samples from Ostrovica burial site contain higher concentrations of Zn and Zn/Ca ratio (median 0.35 103). Our results indicate that Naklice population probably ate mainly legumes and cereals (Fig. 2). The archaeological bones from the burial site Naklice have higher concentrations of iron (median Fe 799 mg/g) and manganese (median Mn 78 mg/g) than sample from the site Ostrovica (median Fe 130 mg/g and median Mn 4.8 mg/g) (Table 3.). The high concentrations of Fe and Mn were analyzed in archaeological bones with lower quality from location Naklice. ManneWhitney nonparametric statistical test confirmed statistically significant difference for three ratios metals between the two archaeological sites: Sr/Ca (z ¼ 2.940; P ¼ 0.003); Zn/Ca (z ¼ 2.764; P ¼ 0.006) and Fe/Mn (z ¼ 6.206; P ¼ 0.000). Archaeological burial site Ostrovice was apparently on the crossroads, which allowed better traffic and availability of different foods and therefore better living conditions of this population. At the same time archaeological population from Naklice burial site was more isolated, and the food that was available to them was reduced in quantity and quality. Increased values of the ratio of strontium and Sr/Ca indicate that the Naklice populations mostly ate foods of plant origin. Previous archaeological and anthropological studies on these locations confirm our assumptions about the diet habits. Very high incidence of dental caries was established
indicating the greater intake of food rich in carbohydrates (starch and sugar), particularly barley and wheat (Slaus, 2008d; Sutlovic et al., 2010a). Iron and manganese are characterized by high mobility in soil. Manganese has a large number of oxidation states. For these reasons they cannot be considered relevant in the reconstruction of diet. Their values in archaeological bones are always higher than the values in recent bones. For these two metals significant correlation was shown. Spearman’s correlation coefficient statistically proved very high positive correlation between iron and manganese for all tested groups (Table 4). The concentrations of iron and manganese are important indicators of the preservation of archaeological bones. Bones that are not well preserved, have a very high content of both metals. Similar conclusions were presented by Vuorinen et al. (1990a). He analyzed metal content in 141 bone samples from periods 1580 to 1650 from Turku, Finland. Particular group of well-preserved and group of damaged femur were analyzed. Bones with more defects indicate a high content of both metals (Fe 9.16 mg/g, Mn 1.80 mg/g) (Vuorinen et al., 1996b).
3.2. The relationship of biological variables and metal content Anthropometric methods were used to determine gender and age (Table 5). The gender was determined in 71 adult samples by means of DNA analysis, while age was determined in 87 samples. The diet of adults (men and women) was associated with the distribution of metals in archaeological bones. We showed a statistically significant difference in the distribution of metals by sex. ManneWhitney test confirmed the statistical significance of strontium (z ¼ 2.587; P ¼ 0.010) and zinc (z ¼ 2.285; P ¼ 0.022) (Fig. 3). The difference in diet may explain the social conditions at Early Medieval Period, where women are placed in a subordinate position. Quality and abundant food was provided for men. However, the difference in the content of metals in archaeological bone between men and women are not solely due to the different diet, but may be the result of various metabolic processes in the body of a woman (pregnancy and lactation).
Table 5 Number of samples (N) analyzed by gender and age (1625; 2639; >40 age), according to the archaeological site.
Fig. 2. Metal concentration e median (strontium and zinc) in archaeological bones. Comparison results between archaeological sites: Ostrovica and Naklice.
Archeological bones e femur
Ostrovica
Naklice
Male Female Children (0e15) Adult (16e25) Adult (26e39) Adult (>40) Unknown sex Unknown age All samples (N)
36 15 11 9 17 22 2 5 64
10 7 8 4 15 1 11 8 36
398
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
Fig. 3. Metal concentration e median (strontium and zinc) in archaeological bones. Comparison results between male and female.
Distribution of calcium, strontium and zinc in men’s bones did not differ depending on age (Fig. 4). However, between women the distribution of these metals varies with age. Values for calcium and zinc in all examined groups were similar (Fig. 5). Strontium content was significantly higher in the second group (26e39 years). KruscaleWallis test confirmed statistically significant differences in the distribution of strontium in women, depending on age (c2 ¼ 6.803, P ¼ 0.033). In younger women’s archaeological bones zinc and strontium concentrations were lower than the concentrations in the bones of men from the same age group. Older women had higher levels of both metals than men of the same age. These differences in the concentration of metals (Ca, Sr and Zn) between women of different ages are not only the consequence of different diet, but also result of the metabolic process in the woman’s body during pregnancy and lactation. Comparing the results of the bones from children group (N ¼ 19) with the results of adults bones, it was observed that values of zinc (median 85 mg/g) as well as ratio of Zn/Ca (0.26 103) were decreased. Extremely low values of Zn (35.06 mg/g) and Zn/Ca ratio (0.11 103 mg/g) were measured in the bones of children from archaeological site Naklice indicating insufficient nutrition (Fig. 6). The results of these metals in children’s bones were in good correlation with archaeological and anthropological knowledge;
Fig. 4. Metal concentration e median (strontium and zinc) in male archaeological bones per three age groups.
Fig. 5. Metal concentration e median (strontium and zinc) in female archaeological bones per three age groups.
therefore it is easier to explain the impact of diet. One indicator of anemia and malnutrition of archaeological population was c. orbitalia. Anthropologists have observed the frequency of c. orbitalia in the osteological remains of old Croatian archaeological populations (Slaus, 2008d). The results of his study showed that the frequency was higher in children’s bones than in adults’. C. orbitalia is more pronounced in children from 1 to 5 years and lowest in children under one year. Diet was the cause of it. Infants had enough iron, vitamins and minerals during the breastfeeding period, in contrast to older children who were gradually losing these sources. On the other hand, after the end of breastfeeding and introduction of hard foods, children are becoming more exposed to a variety of gastrointestinal disorders. The exact cause of death cannot be determined in archaeological populations. However, the mortality rate among children in archaeological populations was often caused by inadequate nutrition and anemia. On the other hand, their immune system was weakening and various diseases could cause death. 3.3. Diagenesis The mineral composition of bones depends on conditions during lifetime as well as the processes in the burial soil. Chemical and physical interactions between the bones and soil may cause an enrichment of metal content in archeological bones (Ambrose and
Fig. 6. Metal concentration (median strontium and zinc) in archaeological bones from two archaeological sites per adults and children.
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
Krigbaum, 2003). Changes in soil next to archaeological sites can affect the growth of metals in archaeological bones (Nielsen-Marsh and Hedges, 2000). Therefore the bones were collected, prepared and cleaned with particular care. Three preparation methods were used to reduce the possibility of contamination: mechanical cleaning, chemical cleaning with acid and finally washing and rinsing with a reducing agent. Preservation of archaeological bones depends on the composition and soil acidity. Archaeological sites from our research are located on the gentle hills and are not exposed to the groundwater. Samples of soil were calcareous with slightly alkaline pH values (8.17e8.30). That composition of the soil is suitable for preservation of osteological material for more than a thousand years. Alkaline soil structure protects the bone better than acid soil, which easily destroys hydroxyapatite. Metals are differently exposed to diagenesis. According to different studies iron, manganese, cooper and cadmium are most exposed to changes, while lead, calcium, strontium and zinc are significantly less exposed to diagenesis (Lambert et al., 1985; Nielsen-Marsh and Hedges, 2000; Carvalaho et al., 2004). Soluble and exchange ions will be most useful in the development of geochemical models to address the composition of secondary minerals and ionic substitution phases in archaeological bone (Pate and Hutton, 1988). Under the alkaline conditions of most arid-land environments, iron and manganese will occur as relatively insoluble oxides and hydroxides, and phosphorus as sparingly soluble calcium and magnesium phosphates. In addition, phosphorus and heavy metals (Cu, Zn, Pb and Ni) are mainly insoluble due to adsorption of ionic species in silicate clays. The majority of soil solution ions will be derived from water soluble carbonate, sulfate and chloride salts of Ca, Mg, K and Na. These salts accumulate in arid land environments because annual precipitation is insufficient to leach the soil beyond depths of about 0.7e1 m (Pate and Hutton, 1988). Comparison of metal concentrations in soil and bone samples from two archaeological sites is shown in Table 6. Lead content in bones is higher than in soil samples suggesting that archaeological bones were not influenced by changes in the soil. However, cadmium is significantly exposed to diagenesis. Its content in the upper and lower soil layers is higher than in the bones, indicating that elevated concentrations of cadmium in the bones are the consequence of diagenesis. Results of analysis for both archaeological
Table 6 Comparison analyzed metals in archaeological bones and in soil from two archaeological sites. The results of metals are presented in mg/g dry matter, except Ca* in %. Lower layer of the soil
Archaeological bones
Top layer of the soil
Median SD
Median SD
Median SD
Ostrovica Pb 0.075 0.010 Cd 0.502 0.021 Ca* 15.30 0.72 Sr 89.60 14.62 Zn 46.30 6.96 Cu 39.90 1.02 Fe 52 564 2459 Mn 138 11.02 Naklice Pb 0.009 0.008 Cd 0.803 0.020 Ca* 2.61 0.76 Sr 22 35.4 Zn 196 10.0 Cu 39.8 1.52 Fe 23 184 1547 Mn 637 13.9
0.412 0.053 34.49 371 121 2.38 130 4.81
0.622 0.068 3.51 89 35.2 8.98 865 9.72
0.098 0.005 0.356 0.033 27.80 0.73 597 52.42 38.10 3.03 10.10 1.27 10 745 1039 156 10.97
0.918 0.029 32.60 418 66 5.64 799 77.5
3.797 0.111 4.53 111 74 9.63 1215 94.2
0.020 0.009 0.758 0.012 5.23 0.28 28 4.16 178 12.7 19.8 1.31 10 990 1205 441 11.3
399
sites confirm that iron, manganese and copper are metals significantly exposed to diagenesis. Determined values of these metals in archaeological bones were extremely lower than the values in the soil.
4. Conclusions This study showed that analysis of individual metals in archaeological bones had the potential to strengthen archaeological interpretations in terms of its contribution to the existing knowledge on population who had lived in South Croatia during the Early Medieval Period. Though we dealt with two sites from similar periods (9th century), spatial differences notably affected the differences in lifestyle and diet of these people. Both archaeological sites belong to the Early Medieval Period, famous for natural economy. At the archaeological site Ostrovica single byzantine currency of that period was found, witnessing the relationship of Early Medieval Croatia and its Counties with the Empire on Bospor. The site was at intersection of roads, which enabled better exchange of goods, and consequently a food exchange. However, the other archaeological site Naklice is located at the mild hill, and this archaeological population was more isolated. The available food was poor in quality and quantity. Considering the above mentioned archaeological facts, the difference between these two archaeological groups is evident. The concentration of some metals analyzed in archaeological bones may be useful to draw conclusions about lifestyle and diet of archaeological populations. With regard to these two archaeological sites, differences have been found between values of the individual metals and their ratios (Sr/Zn; Sr/Ca; Zn/Ca). The increased values of Sr and the ratio Sr/Ca in archaeological population from Naklice location indicate that this population mostly consumed vegetable origin food. Extremely low values of Zn and the Zn/Ca ratio were determined for a group of children bones, from Naklice burial site, because they were living in exceptionally difficult circumstances. The comparison between adults and children within the same archaeological site shows great similarity in diet. In regards to the aim of our study, we found statistically significant differences between gender which is confirmed with different strontium and zinc concentrations among men and women probably due to differences in diet and activity of metabolic processes (pregnancy and lactation) in women’s bodies. There were no differences in the distribution of metals between men in regards to age, while differences in regards to age between women, were present. They were incurred as a result of metabolic processes (Sr is higher in the age group 25e39 years). It can be concluded that the analysis of metals in archaeological bones, particularly Sr, Zn, Ca and their ratios may be useful as a supplement to other archaeological and anthropological knowledge although geological changes in soil may affect the credibility of the results of metals in archaeological bones. Certain metals as Cd, Fe, Mn are subject to significant influence of diagenesis. However, the determination of the concentration of these metals is also important. It shows the impact of changes in the soil. The analysis of metals in archaeological bones cannot be only used as a way of reconstruction of population’s habits, but can be useful complement to other archaeological and anthropological studies. In order to make these results more authentic, it is imperative to analyze the composition of the soil, and to observe the extent of diagenesis in archaeological bone. Many factors determine the exposure of archaeological bones to diagenesis (length of the historical period, type and part of the bone, groundwater, soil acidity, etc.). Taking into consideration not only the above-mentioned, but also the fact that different metals behave
400
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
differently in the soil, the relationship between metal in bones and diet of archaeological populations was shown. In future studies, it would be interesting to make a comparative analysis of metals in the bones and teeth of a certain archaeological population. Namely, teeth contain more inorganic substances than archaeological bones and are more resistant to changes in the soil, and thus expected to be less exposed to diagenesis than bones. Acknowledgments The authors would like to thank dr. Vedrana Delonga and dr. Tonci Buric from the Archaeological Museum in Split for enabling the access to skeletal material from the Naklice burial site. This study was financially supported by a grant from the Ministry of Science, Education and Sports of the Republic Croatia (Grant number 216-2160800-0655). References Ambrose, S., Krigbaum, J., 2003. Bone chemistry and bioarchaeology. J. Anthropol. Archaeol. 22, 193e199. Analytik Jena AG, Jena, Germany, 2001. Operating Manual AAS Vario 6. Baranowska, I., Czernicki, K., Aleksandrowicz, R., 1995. The analysis of lead, cadmium, zinc, cooper and nickel content in human bones from the upper Silesian industrial district. Sci. Total Environ. 159, 155e162. Carvalaho, M.L., Marques, A.F., Lima, M.T., Reus, U., 2004. Trace elements distribution and post-mortem intake in human bones from middle Age by total reflection X-ray fluorescence. Spectrochim. Acta B 59, 1251e1257. CEM Corporation, Illinos, USA, 2006. Operating Manual CEM Mars 5. Delonga, V., Buric, T., 1998. Ostrovica near the Bribir. In: Archaeological and Historial Sketch. The Museum of Croatian Archaeological Monuments, Split. Dobrovolskaya, M.V., 2005. Upper palaeolithic and late stone age human diet. J. Physiol. Anthropol. Appl. Hum. Sci. 24, 433e438. Donno, A., Santoro, V., Fazio, A., Corrado, S., Urso, D., Baldassarra, S.L., 2010. Analysis of neolithic human remains discovered in southern Italy. J. Archaeol. Sci. 37, 3482e3487. Goldstein, I., 1995. Hrvatski srednji vijek. Novi Liber, Zagreb, Croatia, pp. 84e116. Gonzalez-Reiemers, E., Arnay-de-la-Rosa, M., Velasco-Vazquez, J., GalindoMartin, L., Satolaria-Fernandez, F., 2005. Bone cadmium and lead in the ancient population from El Hierro, Canary Islands. Biol. Trace Elem. Res. 105, 37e51. Hedges, R.E.M., 2002. Bone diagenesis: an overview of processes. Archacometry 44, 319e328. Hedges, R.E.M., Millard, A.R., 1995. Bones and groundwater, towards the modeling of diagenetic processes. J. Archaeol. Sci. 22, 155e164. Jurkiewicz, A., Wiechula, D., Nowak, R., Gazdzik, T., Loska, K., 2004. Metal content in femoral head spongious bone of people living in regions of different degrees of environmental pollution in Southern and Middle Poland. Ecotoxicol. Environ. Saf. 59, 95e101. Lambert, J.B., Vlasak, S., Szpunar, C.B., Bukistra, J.E., 1985. Bone diagenesis and dietary analysis. J. Hum. Evol. 14, 477e482. Martinez-Garcia, J.M., Moreno, J.M., Moreno-Clavel, J., Vergara, N., GarciaSanchez, A., Guillamon, A., 2005. Heavy metals in human bones in different historical epochs. Sci. Total Environ. 348, 51e72.
Mays, S., 2003. Bone strontium:calcium ratios and duration of breastfeeding in a Mediaeval skeletal population. J. Archaeol. Sci. 30, 731e741. Nakashima, T., Matsuno, K., Matsushita, M., Matsushita, T., 2010. Severe lead contamination among children of samurai families in Edo period Japan. J. Archaeol. Sci. 38, 23e28. National Institute of Standard and Technology, Gaithersburg, Maryland, USA, 2009. Certificate of Analysis Standard Reference Material Montana I Soil 2710a. Nielsen-Marsh, C.M., Hedges, R.E.M., 2000. Patterns of diagenesis in bone I: the effects of site environments. J. Archaeol. Sci. 27, 1139e1150. Nielsen-Marsh, C., Gernaey, A., Turner-Walker, G., Hedges, R., Pike, A., Collins, M., 2006. The chemical degradation of bone. In: Cox, M., Mays, S. (Eds.), Human e Osteology: In Archaeology and Forensic Science. Cambridge University Press, New York, pp. 439e453. Novak, M., Slaus, M., 2010. Health and disease in a Roman walled city: an example of Colonia Iulia Iader. J. Anthropol. Sci. 88, 189e206. Pate, F.D., Hutton, J.T., 1988. The use of soil chemistry data to address post-mortem diagenesis in bone mineral. J. Archaeol. Sci. 15, 729e739. Price, T.D., Blitz, J., Burton, J., Ezzo, J.A., 1992. Diagenesis in prethistoric bone: problems and solutions. J. Archaeol. Sci. 19, 513e529. Pyatt, F.B., Pyatt, A.J., Walker, C., Sheen, T., Grattan, J.P., 2005. The heavy metal content of skeletons from an ancient metalliferous polluted area in southern Jordan with particular reference to bioaccumulation and human health. Ecotoxicol. Environ. Saf. 60, 295e300. Runia Lex, T., 1987. Strontium and calcium distribution in plants effect on Palaeodietary studies. J. Archaeol. Sci. 14, 599e608. Schutkowski, H., Herrmann, B., 1999. Diet, status and decomposition at Weingarten: trace element and isotope analyses on early mediaeval skeletal material. J. Archaeol. Sci. 26, 675e685. Shafer, M.M., Siker, M., Overdier, J.T., Ramsl, P.C., Teschler-Nicola, M., Farrell, P.M., 2008. Enhanced methods for assessment of the trace element composition of Iron Age bone. Sci. Total Environ. 401, 144e161. Slaus, M., 1997a. Discriminant function sexing of fragmentary and complete femora from medieval cites in continental Croatia. Opvscvla. Archeol. 21, 167e175. Slaus, M., 2000b. Biocultural analysis of sex differences in mortality profiles and stress levels in the late medieval population from Nova Raca, Croatia. Am. J. Phys. Anthropol. 111, 193e209. Slaus, M., 2006c. In: Bioarcheology-demography, Health, Nutrition and Trauma Ancient Croatian population. Skolska knjiga, Zagreb, Croatia, pp. 61e98. Slaus, M., 2008d. Osteological and dental markers of health in the transition from the Late Antique to the Early Medieval period in Croatia. Am. J. Phys. Anthropol. 136, 455e469. Sutlovic, D., Stipisic, A., Marusic, J., Pavlov, J., Versic, M., Definis-Gojanovic, M., 2010a. Elemental status in individual from Naklice Burial site (Southern Croatia): Mediaeval Diet reconstruction. Croat. Chem. Acta 83, 217e221. Sutlovic, D., Ljubkovic, J., Stipisic, A., Knezovic, Z., Drmic-Hofman, I., Andjelinovic, S., 2014b. Skeletal remains from late Roman period: “As old as Diocletian’s Palace”. Coll. Antropol. (in press). Vuorinen, H., Tapper, U., Mussalo-Rauhamaa, H., 1990a. Trace and heavy metals in infants, analysis of long bones from Ficana, Italy, 8e6th Century BC. J. Archaeol. Sci. 17, 237e254. Vuorinen, H., Pihlman, S., Mussalo-Rauhamaa, H., 1996b. Trace and heavy metal analyses of a skeletal population representing the town people in Turku (Abo), Finland in the 16e17th centuries: with special reference to gender, age and social background. Sci. Total Environ. 177, 145e160. Wissenchaftlich Technishe Werkstatten-WTW, Weilheim, Germany, 2004. Operating Manual inoLab-pH720. Yamada, M., Minami, T., Yamada, G., Tohno, Y., Tohno, S., Ikeda, Y., 1997. Different element ratios of red cosmetics excavated from ancient burials of Japan. Sci. Total Environ. 199, 293e298.
Contents lists available at ScienceDirect
Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas
Metal content in medieval skeletal remains from Southern Croatia Angela Stipisic a, *, Maja Versic-Bratincevic b, Zlatka Knezovic a, Davorka Sutlovic b, c a
Public Health Institute of Split e Dalmatia County, Vukovarska 46, 21 000 Split, Croatia Department of Forensic Medicine, University of Split, School of Medicine, 21 000 Split, Croatia c Department of Pathology and Forensic Medicine, University Hospital Centre Split, 21 000 Split, Croatia b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 22 January 2013 Received in revised form 11 March 2014 Accepted 28 March 2014 Available online 8 April 2014
This work is a contribution to the existing knowledge of lifestyle and diet of the South Croatian population who lived in Early Medieval Period. The one hundred samples dating from 9th century were discovered and collected at the burial sites Ostrovica and Naklice. Concentrations of metals and their mutual relationships were examined in regards to gender and age of skeletal remains. Differences were observed in diet between men and women as well as among age groups. For a correct interpretation of the results it is necessary to determine the metal content in the soil. Namely, metals in archaeological bones are influenced by changes in soil e diagenesis, which is confirmed by our results. We concluded that there were no influences of diagenesis on lead, calcium, strontium and zinc content while cadmium, iron, manganese and copper are most exposed. Ó 2014 Elsevier Ltd. All rights reserved.
Keywords: Archaeological bones Early Medieval Period Heavy metals Diet reconstruction Diagenesis
1. Introduction Environmental conditions, dietary habits and uptake of some elements from the surrounding soil may reflect on heavy metal content in human bones. This statement is supported in several papers with results of metals in archaeological human bones from different locations and different periods (Martinez-Garcia et al., 2005; Dobrovolskaya, 2005; Shafer et al., 2008; Donno et al., 2010; Nakashima et al., 2010; Yamada et al., 1997). Martinez-Garcia et al. (2005) determined the content of lead, copper, zinc, cadmium and iron in archaeological bones from the area of Cartagena in Spain from different historical periods. According to their study, the lowest concentrations of lead were in the Neolithic period (median Pb 45 mg/g), the values from the Bronze Age were higher (median Pb 80 mg/g), and maximum were recorded during the Roman (median Pb 742 mg/g) and the Byzantine era (median Pb 898 mg/g). The concentrations of copper were highest in Byzantine era bones (median Cu 111 mg/g), and the concentrations of iron during the Ottoman Empire (median Fe 15 383 mg/g) (Martinez-Garcia et al., 2005). Schutkowski and Herrmann (1999) analyzed metals in archaeological human bones from 6th to 8th century, from Northwestern
* Corresponding author. Tel.: þ385 021401175; fax: þ385 021539825. E-mail address: [email protected] (A. Stipisic). http://dx.doi.org/10.1016/j.jas.2014.03.032 0305-4403/Ó 2014 Elsevier Ltd. All rights reserved.
Germany (Schutkowski and Herrmann, 1999). They found low zinc content, increased content of copper and strontium as well as enlarged ratio of Sr/Ca. Strontium is not an essential element. It is found in plant origin food. Ratio of Sr/Ca in bones is very important. Higher ratio indicates that foods of plant origin (cereals and legumes) were predominated in the diet, while a smaller ratio suggests higher incidence of animal origin food (milk and dairy products) (Schutkowski and Herrmann, 1999). According to Mays (2003), strontium is mostly located in the skeleton, with a strong relationship between Sr/Ca ratio in diet and in the bones of the consumer (Mays, 2003). The analyses of Sr/Ca ratio in human skeletal remains from medieval times from England were carried out, particularly in archaeological bones of breastfeeding mothers. Mays found association not only between Sr/Ca ratio and diet from that period, but also with metabolic changes during lactation. During lactation, transfer of calcium is faster than strontium, as well as through the placenta during pregnancy. Potentially, according to Mays, bone Sr/Ca ratios is a sensitive technique for detecting the administration of lowprotein or animal-milk supplements into infant diets. This is important as historical sources indicate that such foods were frequently used as infant feeding supplements in the past (Mays, 2003). In archaeological studies analysis of metals in bones reflect diet and lifestyle habits of archeological populations. Contamination and diagenesis of archaeological bones can compromise the credibility of this information (Martinez-Garcia et al., 2005; Nielsen-
394
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
Marsh and Hedges, 2000). Diagenesis is a series of changes in the soil, including: dissolution, precipitation, crystallization and exchange of minerals between the soil and bones. According to Nielsen-Marsh et al. (2006), there are three important alterations to bone apatite that can be used as indicators of diagenetic change: porosity increase, crystallinity increase and inclusion of exogenous ione (Nielsen-Marsh et al., 2006). These processes can significantly affect the archaeological bones. Preservation of archaeological bone depends on the composition and acidity of the soil (Hedges and Millard, 1995; Hedges, 2002). Archaeological sites that are located on gentle hills and not being exposed to groundwater are less exposed to the influence of diagenesis (Hedges and Millard, 1995; Nielsen-Marsh and Hedges, 2000). The composition of calcareous soil with slightly alkaline pH values is appropriate for preservation of osteological material over thousands of years. The alkaline soil can preserve bone far better than the sour one, which easily destroys hydroxyapatite, an integral part of the bone (Hedges, 2002). Previous studies in Croatia indicated that the osteological material is best preserved at the archaeological sites along the Adriatic coast, under the condition that their location is not close to the sea (Slaus, 2006c; Novak and Slaus, 2010). These sites have soil with neutral to slightly acidic pH values and high content of sand, whereas in Central and Continental Croatia, level of preservation varies depending on the acidity of the soil and the amount of groundwater (Slaus, 2006c). Selection of archaeological bone is also very important for reducing the impact of diagenesis (Vuorinen et al., 1990a; Price et al., 1992). Long parts of femur are less exposed to changes in the soil than the short bones. Trabecular bone is more exposed to the diagenesis than compact cortical bone (Carvalaho et al., 2004; Ambrose and Krigbaum, 2003). Contamination of archaeological bones can occur with the long laying in the soil, but the sampling and preparation techniques may also cause it (Price et al., 1992). The part of the skeletal remains must be selected very carefully for chemical analysis before metal determination. During the preparation of samples, metal accessories such as saws, forceps, etc. should not be used. Contemporary archaeological research in Croatia became more intensive at the beginning of nineties. Analysis included anthropological and archaeological research (determination of gender, age, trauma, dental analysis, as well as osteological changes indicating different diseases) both for the continental and coastal Croatian (Slaus, 1997a, 2000b, 2006c, 2008d; Novak and Slaus, 2010). During the 90-ies of the 20th century the Museum of Croatian Archaeological Monuments in the excavations carried out a number of sites from the early Middle Ages, or the 9e10th century (Delonga and Buric, 1998). However, in the aforementioned studies in Croatia so far no one has dealt with the analysis of metals in archaeological bones. In our study the content of metals was analyzed in 100 samples of archaeological bones in compact part of the bone e cortical femoral. Archaeological bones were collected from two archaeological sites of Southern Croatia: Ostrovica and Naklice (Fig. 1). The influence of diagenesis on the distribution of metals in archaeological samples was investigated by determining metals and pH in soil samples which were collected from the burial site during the archaeological excavations. The aim of our study was to determine the difference of metal distribution in regards to gender and age, as well as potential differences in social status between men and women, and children and adults. We also tried to determine whether there are significant differences referring to metal distribution in the bones of the early Croatian population for two different archaeological sites from the same historical period.
2. Methods 2.1. Instrumentation and operating conditions Lead and cadmium measurements were carried out by using a Model AAS vario 6 GFAAS atomic absorption spectrometer (Analytik Jena AG, 2001) equipped with a transversely heated graphite atomizer with autosampler (Model MPE 50), a deuterium background correction system and a hollow cathode lamp for lead operated at 3 mA (wavelength 283.3 nm) and for cadmium operated at 3 mA (wavelength 228.8 nm). Pyrolytic coated graphite tubes with PIN-platform (Analytik Jena, Part No. 407-A81.025) were used during the anlytical determination (Analytik Jena AG, 2001). The injection volume was 20 mL and integrated absorbance (peak area) was used for signal evaluation. Calcium, strontium, zinc, copper, iron and manganese determinations were carried out by using a Model AAS vario 6 FAAS atomic absorption spectrometer equipped with deuterium background correction system and a hollow cathode lamp for calcium, strontium, zinc, copper, iron and manganese (wavelength 422.7 nm for Ca; 460.7 nm for Sr; 213.9 nm for Zn; 324.8 for Cu nm; 248.3 nm for Fe and 279.5 nm for Mn). Concentrations of Zn, Cu, Fe and Mn were determinated by Flame AAS with C2H2/air burner, while Ca and Sr were determinated with C2H2/N2O2 burner (Analytik Jena AG, 2001). Samples were weighted on the Mettler Toledo balance e Model AX 205DR (Mettler, Germany) with resolution of 0.01 mg. A closed microwave system, CEM Model Mars 5 was used for wet digestion (CEM Corporation, 2006). pH values of the soil samples were determined with pH-meter Model inoLab e pH720; measuring range and resolution e 2.000. þ 19.999 with accuracy of 0.005 (Wissenchaftlich Technishe Werkstatten WTW, 2004). 2.2. Reagents Working standards of Pb, Cd, Ca, Sr, Zn, Cu, Fe and Mn for measurements were prepared from Merck (Darmstadt, Germany) stock solutions (1000 2) mg/L, suprapur. Standard solutions were prepared in range of expected concentration values. We applied the same technique for the metal analysis in two different matrices: soil and human bones. To confirm the success of the method performed, we used the available Certified Standard Reference soil material: SRM e 2710a Montana I Soil, from the National Institute of Standards and Technology (NIST, 2009). Pyrolysis and atomization curves were established in the presence of chemical modifier e 0.1% Pd(NO3)2 þ 0.05% Mg(NO3)2 6H2O. Modifier was prepared from Merck stock solution: Art.1.07289 palladium matrix modifier for graphite furnace AAS and Art.1.05855 magnesium nitrate hexahydrate. Volumen of added modifier was 5 mL. 2.3. Samples preparation After archaeological excavations, bone samples were classified according to the anthropological parameters: gender, age (0e15, 16e25, 26e39 and above 40 years) as well as geographical location. For chemical analysis, bone samples were prepared with special attention to avoid the post-mortem contamination. A part of a dense cortical bone was sawed for analysis. Bone pieces were crushed into small fragments using razor blades and stored in sterile polypropylene tubes at 20 C until analyzed. After drying to a constant weight samples were washed in 6 ml 65% nitric acid (HNO3) over night, subsequently washed in distilled water and finally dried at room temperature.
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
395
Fig. 1. Skeletal remains of adults from archaeological burial site Ostrovica. Samples are arranged for anthropological process and sampling for chemical analysis.
Approximately 0.5 g of the sample was weighed in a teflon-TFM vessel and with the addition of 65% nitric acid and hydrogen peroxide wet e ashed in automated (temperature regulated) microwave digestion unit (CEM, USA Model Mars 5 with 1600 W power). Digested samples were diluted with deionized water and metal content was determined quantitatively. We conducted analysis of soil samples to determine the extent of the archaeological bone contamination with the soil. We analyzed the soil from two archaeological sites; Ostrovica and Naklice. Each of them included two soil layers: the upper sample (0.5 m from the surface) and the lower sample (1.0 m from the surface). Lower layer of the sampled soil is at an approximate distance of 30 cm from the skeleton. Each of ground samples was analyzed four times. Preservation of osteological material depends on the acidity of the soil, thus pH value in the 5% water solution of soil was determined. For quantitative analysis of metal samples were prepared by microwave digestion in the same way as bone samples (CEM Corporation, 2006).
2.4. Method validation The linear calibration of Pb was obtained using five standard solutions prepared in 0.5% HNO3 in the range (10e50) mg/L Pb. The linear calibration of Cd was obtained using five standard solutions prepared in 0.5% HNO3 in the range (0.2e1) mg/L Cd. Relative standard deviations of five replicate samples for each standard solution were lower than 10%. The correlation coefficient (R) obtained by linear fitting to the calibration points was better than 0.995. The detection limit for Pb and Cd analyzed by GFAAS technique is 1 mg/L. Matrix effects and interferences to other metals were minimized with graphite platform and modifier 0.01% Mg(NO3)2 þ 0.5% Pd(NO3)2. The detection limits for other metals analysis technique FAAS are: Ca 100 mg/L; Sr 200 mg/L; Zn 10 mg/L; Cu 50 mg/L; Fe 50 mg/L; Mn 10 mg/L. The standard reference material NIST e SRM 2710a (Montana I Soil) was used throughout this study to validate the analytical
396
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
technique (National Institute of Standard and Technology, 2009). The repeatability of sample preparation and measurements are expressed as relative standard deviation. All the determined values of relative standard deviation (coefficient of variation) were less than 10%, indicating a good precision of measurement. Effect of matrix and chemical interference has been reduced to a minimum of the methods which confirm the values obtained percentage recovery (91.3e110%) (Table 1).
Table 2 Statistical analysis of metals in archeological bones (N ¼ 100). The results of metals are presented in mg/g, except the results of Ca in %.
Mean Median SD SE Minimum Maximum
Pb
Cd
Ca*
Sr
Zn
Cu
Fe
Mn
1.266 0.610 2.420 0.242 0.002 16.400
0.074 0.047 0.008 0.047 0.001 0.459
33.97 32.87 4.20 0.42 20.90 49.31
395 383 104 10 202 659
112 111 53 5 14 311
5.83 3.01 9.45 0.95 0.05 72.90
1135 244 3315 331 11 29787
58.08 13.28 106 10.6 0.40 684
3. Results and discussion The results were statistically analyzed by determining the average value (mean and median), standard deviation (SD), standard error (SE), measurement range (minimum and maximum value) for each metal in the test groups (Table 2). Results of the analysis of metals are expressed as mean and median. For comparing and expressing the results in our paper, we used the median. Namely, it is useful to express the results in this way, when the distribution of results has very large extreme values (minimum and maximum values). Determination of lead concentration in recent human bones indicates exposure of “modern” humans to harmful effects of environmental pollution (Baranowska et al., 1995; Jurkiewicz et al., 2004). Lead is a toxic metal, which in the form of lead phosphate has a significant ability to accumulate in the bones, particularly in the fast growing bones (femur and tibia) (Baranowska et al., 1995). It is very slowly excreted from the body. In various studies of toxic metals and their harmful effects on man, lead is the most important indicator of exposure to toxic metals. However, in an archaeological study lead can be used for determining the way of life of the population (eg. use of pottery with lead glaze). Determination of lead in archaeological studies is extremely important for the Roman period and the Bronze Age, particularly in archaeological population at that time when pottery with lead glaze was used (Martinez-Garcia et al., 2005; Donno et al., 2010; Shafer et al., 2008; Pyatt et al., 2005; Nakashima et al., 2010; Sutlovic et al., in press). Pyatt et al. study (2005) results showed significantly higher concentrations of lead than results of our study (median Pb 0.610 ug/g). They analyzed skeletal remains from the Bronze and Roman Age. The reason for extremely high concentrations of lead in bones (human femur 170e196 mg/g) from that period was due to using pottery with lead glaze (Pyatt et al., 2005). Nakashima et al. (2010) also showed high values of lead (Pb median 313 ug/g) in archaeological bones of children samurai from the Edo period 1600the1867th in Japan, who also used pottery with lead glazes (Nakashima et al., 2010). Lead concentrations in archaeological bones in our study are lower than in both mentioned studies. Results of the concentration of lead (range 0.02e14.76 mg/g) in the study of archaeological bones by Gonzalez-Reiemers et al. (2005) from Al Hair, Canary Islands area, are similar to the results of our study (range 0.002e16.4 mg/g). On the other hand the values of cadmium in the above study are higher (range 0.003e2.56 mg/g) Table 1 Results of metals in certified material SRM e 2710 Montana Soil were determined by AAS. The results of metals are presented in %, except result of Cd* and Sr* in mg/g. Metal
Certified
Measured (n ¼ 6)
SD
RSD (%)
Recovery (%)
Pb Cd Ca Sr Zn Cu Fe Mn
0.552 12.3* 0.964 255* 0.418 0.342 4.32 0.214
0.547 11.2* 0.905 237* 0.446 0.349 4.08 0.212
0.024 0.186* 0.070 20.9* 0.008 0.008 0.055 0.004
4.33 1.66 7.76 8.83 1.91 2.35 1.35 1.98
99.1 91.3 93.9 92.9 106.7 102.0 94.4 99.1
than in our study (range 0.001e0.459 mg/g) (Gonzalez-Reiemers et al., 2005). The available archaeological sources of Early Medieval Croatian population showed their hard living conditions i.e. the extreme poverty (Goldstein, 1995). Archaeological objects discovered along the graves indicate that they did not use pottery with lead glazes (Sutlovic et al., 2010; Delonga and Buric, 1998). Early Medieval man in Croatia had to struggle with nature continuously. Poverty and hunger led to various diseases (leprosy and tuberculosis) and death (Goldstein, 1995; Novak and Slaus, 2010). Anthropological research on the skeletons indicates the frequent occurrence of hypertrophied muscle insertions and ossification tendons, which indicates a high level of physical activity e heavy physical work. According to anthropological studies, life expectancy generally did not exceed 45 years, with a high mortality rate of infants and children (Slaus, 2006c). A common osteological indicator of iron deficiency anemia due to inadequate diet, unhygienic living conditions and chronic gastrointestinal disease in archaeological bone is cribira orbitalia (Slaus, 2006c, 2008d). More often it was observed on the archaeological bones of children. Content of calcium, strontium, zinc and their ratios in archaeological bones are indicator of the diet type (Schutkowski and Herrmann, 1999; Lambert et al., 1985). The characteristics of these metals, their representation in a certain food type and other archaeological characteristics in osteological remains (analysis of teeth, etc.) can be related to diet type and lifestyle of population (Schutkowski and Herrmann, 1999). Calcium is an important metal in the human body, and it is mostly represented in bones. It is responsible for many metabolic processes as well as for teeth and bones growth. Dairy food and legumes contain calcium in high concentrations. Although strontium is not an essential element, high concentrations were found in the bones. It is present only in foods of plant origin, particularly in legumes (Schutkowski and Herrmann, 1999). Its determination in archaeological bones and determining the ratio of Sr/Ca in bioarchaeological studies is very significant (Schutkowski and Herrmann, 1999; Mays, 2003). The value of this ratio indicates the origin of calcium. An increased ratio suggests that calcium is from plant origin food (Runia Lex, 1987). Zinc is an important element in the human organism and is responsible for activity of many enzymes. It is usually found in red meat, and legumes. However, if the source of zinc is vegetable food, the human body cannot absorb it. Vegetables contain high concentrations of phytic acid, which creates a form of insoluble zinc phytate. Therefore, the introduction of zinc through the food of animal origin is better absorbed in the human body. Low zinc content in bones indicates a diet with reduced amount of animal origin food (Schutkowski and Herrmann, 1999).
3.1. The difference between archaeological sites The material found in the graves of both archaeological sites (Ostrovica and Naklice) determines the cultural and temporal
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400 Table 3 Statistical analysis the ratio of metals in archaeological bones (N ¼ 100).
Mean Median SD SE Minimum Maximum
397
Table 4 Sperman’s coefficient correlation between iron and manganese.
Ostrovica
Naklice
Bones
Sperman’s coefficient (rs)
Statistical correlation (p)
Sr/Ca 103 Zn/ Fe/Mn Ca 103
Sr/ Zn/ Fe/Mn Ca 103 Ca 103
0.855
0.000
1.07 1.04 0.04 0.29 0.53 1.85
1.32 1.23 0.41 0.07 0.68 2.39
Archeological (N ¼ 100) Archeological site e Ostrovica (N ¼ 64) Archeological site e Naklice (N ¼ 36)
0.744
0.000
0.820
0.000
0.36 0.35 0.01 0.10 0.15 0.78
39.49 33.92 3.20 25.57 4.88 149
0.31 0.19 0.30 0.05 0.05 1.19
11.63 11.32 6.02 1.06 1.77 25.42
affiliation to the burial site of the Early Medieval period. Regional studies examine differences between geographically separated populations within a relatively small and well-defined space. Comparing the results from two archaeological sites, Naklice and Ostrovica, distinctions in metal concentrations were analyzed demonstrating differences in their diet and lifestyle. Correlations among the metals were investigated in an attempt to reconstruct the dietary habits of the population. Values of the ratio of metals are specified in Table 3. Bones samples from location Naklice contain increased value of Sr and high ratio of Sr/Ca (median 1.23 103) while bone samples from Ostrovica burial site contain higher concentrations of Zn and Zn/Ca ratio (median 0.35 103). Our results indicate that Naklice population probably ate mainly legumes and cereals (Fig. 2). The archaeological bones from the burial site Naklice have higher concentrations of iron (median Fe 799 mg/g) and manganese (median Mn 78 mg/g) than sample from the site Ostrovica (median Fe 130 mg/g and median Mn 4.8 mg/g) (Table 3.). The high concentrations of Fe and Mn were analyzed in archaeological bones with lower quality from location Naklice. ManneWhitney nonparametric statistical test confirmed statistically significant difference for three ratios metals between the two archaeological sites: Sr/Ca (z ¼ 2.940; P ¼ 0.003); Zn/Ca (z ¼ 2.764; P ¼ 0.006) and Fe/Mn (z ¼ 6.206; P ¼ 0.000). Archaeological burial site Ostrovice was apparently on the crossroads, which allowed better traffic and availability of different foods and therefore better living conditions of this population. At the same time archaeological population from Naklice burial site was more isolated, and the food that was available to them was reduced in quantity and quality. Increased values of the ratio of strontium and Sr/Ca indicate that the Naklice populations mostly ate foods of plant origin. Previous archaeological and anthropological studies on these locations confirm our assumptions about the diet habits. Very high incidence of dental caries was established
indicating the greater intake of food rich in carbohydrates (starch and sugar), particularly barley and wheat (Slaus, 2008d; Sutlovic et al., 2010a). Iron and manganese are characterized by high mobility in soil. Manganese has a large number of oxidation states. For these reasons they cannot be considered relevant in the reconstruction of diet. Their values in archaeological bones are always higher than the values in recent bones. For these two metals significant correlation was shown. Spearman’s correlation coefficient statistically proved very high positive correlation between iron and manganese for all tested groups (Table 4). The concentrations of iron and manganese are important indicators of the preservation of archaeological bones. Bones that are not well preserved, have a very high content of both metals. Similar conclusions were presented by Vuorinen et al. (1990a). He analyzed metal content in 141 bone samples from periods 1580 to 1650 from Turku, Finland. Particular group of well-preserved and group of damaged femur were analyzed. Bones with more defects indicate a high content of both metals (Fe 9.16 mg/g, Mn 1.80 mg/g) (Vuorinen et al., 1996b).
3.2. The relationship of biological variables and metal content Anthropometric methods were used to determine gender and age (Table 5). The gender was determined in 71 adult samples by means of DNA analysis, while age was determined in 87 samples. The diet of adults (men and women) was associated with the distribution of metals in archaeological bones. We showed a statistically significant difference in the distribution of metals by sex. ManneWhitney test confirmed the statistical significance of strontium (z ¼ 2.587; P ¼ 0.010) and zinc (z ¼ 2.285; P ¼ 0.022) (Fig. 3). The difference in diet may explain the social conditions at Early Medieval Period, where women are placed in a subordinate position. Quality and abundant food was provided for men. However, the difference in the content of metals in archaeological bone between men and women are not solely due to the different diet, but may be the result of various metabolic processes in the body of a woman (pregnancy and lactation).
Table 5 Number of samples (N) analyzed by gender and age (1625; 2639; >40 age), according to the archaeological site.
Fig. 2. Metal concentration e median (strontium and zinc) in archaeological bones. Comparison results between archaeological sites: Ostrovica and Naklice.
Archeological bones e femur
Ostrovica
Naklice
Male Female Children (0e15) Adult (16e25) Adult (26e39) Adult (>40) Unknown sex Unknown age All samples (N)
36 15 11 9 17 22 2 5 64
10 7 8 4 15 1 11 8 36
398
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
Fig. 3. Metal concentration e median (strontium and zinc) in archaeological bones. Comparison results between male and female.
Distribution of calcium, strontium and zinc in men’s bones did not differ depending on age (Fig. 4). However, between women the distribution of these metals varies with age. Values for calcium and zinc in all examined groups were similar (Fig. 5). Strontium content was significantly higher in the second group (26e39 years). KruscaleWallis test confirmed statistically significant differences in the distribution of strontium in women, depending on age (c2 ¼ 6.803, P ¼ 0.033). In younger women’s archaeological bones zinc and strontium concentrations were lower than the concentrations in the bones of men from the same age group. Older women had higher levels of both metals than men of the same age. These differences in the concentration of metals (Ca, Sr and Zn) between women of different ages are not only the consequence of different diet, but also result of the metabolic process in the woman’s body during pregnancy and lactation. Comparing the results of the bones from children group (N ¼ 19) with the results of adults bones, it was observed that values of zinc (median 85 mg/g) as well as ratio of Zn/Ca (0.26 103) were decreased. Extremely low values of Zn (35.06 mg/g) and Zn/Ca ratio (0.11 103 mg/g) were measured in the bones of children from archaeological site Naklice indicating insufficient nutrition (Fig. 6). The results of these metals in children’s bones were in good correlation with archaeological and anthropological knowledge;
Fig. 4. Metal concentration e median (strontium and zinc) in male archaeological bones per three age groups.
Fig. 5. Metal concentration e median (strontium and zinc) in female archaeological bones per three age groups.
therefore it is easier to explain the impact of diet. One indicator of anemia and malnutrition of archaeological population was c. orbitalia. Anthropologists have observed the frequency of c. orbitalia in the osteological remains of old Croatian archaeological populations (Slaus, 2008d). The results of his study showed that the frequency was higher in children’s bones than in adults’. C. orbitalia is more pronounced in children from 1 to 5 years and lowest in children under one year. Diet was the cause of it. Infants had enough iron, vitamins and minerals during the breastfeeding period, in contrast to older children who were gradually losing these sources. On the other hand, after the end of breastfeeding and introduction of hard foods, children are becoming more exposed to a variety of gastrointestinal disorders. The exact cause of death cannot be determined in archaeological populations. However, the mortality rate among children in archaeological populations was often caused by inadequate nutrition and anemia. On the other hand, their immune system was weakening and various diseases could cause death. 3.3. Diagenesis The mineral composition of bones depends on conditions during lifetime as well as the processes in the burial soil. Chemical and physical interactions between the bones and soil may cause an enrichment of metal content in archeological bones (Ambrose and
Fig. 6. Metal concentration (median strontium and zinc) in archaeological bones from two archaeological sites per adults and children.
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
Krigbaum, 2003). Changes in soil next to archaeological sites can affect the growth of metals in archaeological bones (Nielsen-Marsh and Hedges, 2000). Therefore the bones were collected, prepared and cleaned with particular care. Three preparation methods were used to reduce the possibility of contamination: mechanical cleaning, chemical cleaning with acid and finally washing and rinsing with a reducing agent. Preservation of archaeological bones depends on the composition and soil acidity. Archaeological sites from our research are located on the gentle hills and are not exposed to the groundwater. Samples of soil were calcareous with slightly alkaline pH values (8.17e8.30). That composition of the soil is suitable for preservation of osteological material for more than a thousand years. Alkaline soil structure protects the bone better than acid soil, which easily destroys hydroxyapatite. Metals are differently exposed to diagenesis. According to different studies iron, manganese, cooper and cadmium are most exposed to changes, while lead, calcium, strontium and zinc are significantly less exposed to diagenesis (Lambert et al., 1985; Nielsen-Marsh and Hedges, 2000; Carvalaho et al., 2004). Soluble and exchange ions will be most useful in the development of geochemical models to address the composition of secondary minerals and ionic substitution phases in archaeological bone (Pate and Hutton, 1988). Under the alkaline conditions of most arid-land environments, iron and manganese will occur as relatively insoluble oxides and hydroxides, and phosphorus as sparingly soluble calcium and magnesium phosphates. In addition, phosphorus and heavy metals (Cu, Zn, Pb and Ni) are mainly insoluble due to adsorption of ionic species in silicate clays. The majority of soil solution ions will be derived from water soluble carbonate, sulfate and chloride salts of Ca, Mg, K and Na. These salts accumulate in arid land environments because annual precipitation is insufficient to leach the soil beyond depths of about 0.7e1 m (Pate and Hutton, 1988). Comparison of metal concentrations in soil and bone samples from two archaeological sites is shown in Table 6. Lead content in bones is higher than in soil samples suggesting that archaeological bones were not influenced by changes in the soil. However, cadmium is significantly exposed to diagenesis. Its content in the upper and lower soil layers is higher than in the bones, indicating that elevated concentrations of cadmium in the bones are the consequence of diagenesis. Results of analysis for both archaeological
Table 6 Comparison analyzed metals in archaeological bones and in soil from two archaeological sites. The results of metals are presented in mg/g dry matter, except Ca* in %. Lower layer of the soil
Archaeological bones
Top layer of the soil
Median SD
Median SD
Median SD
Ostrovica Pb 0.075 0.010 Cd 0.502 0.021 Ca* 15.30 0.72 Sr 89.60 14.62 Zn 46.30 6.96 Cu 39.90 1.02 Fe 52 564 2459 Mn 138 11.02 Naklice Pb 0.009 0.008 Cd 0.803 0.020 Ca* 2.61 0.76 Sr 22 35.4 Zn 196 10.0 Cu 39.8 1.52 Fe 23 184 1547 Mn 637 13.9
0.412 0.053 34.49 371 121 2.38 130 4.81
0.622 0.068 3.51 89 35.2 8.98 865 9.72
0.098 0.005 0.356 0.033 27.80 0.73 597 52.42 38.10 3.03 10.10 1.27 10 745 1039 156 10.97
0.918 0.029 32.60 418 66 5.64 799 77.5
3.797 0.111 4.53 111 74 9.63 1215 94.2
0.020 0.009 0.758 0.012 5.23 0.28 28 4.16 178 12.7 19.8 1.31 10 990 1205 441 11.3
399
sites confirm that iron, manganese and copper are metals significantly exposed to diagenesis. Determined values of these metals in archaeological bones were extremely lower than the values in the soil.
4. Conclusions This study showed that analysis of individual metals in archaeological bones had the potential to strengthen archaeological interpretations in terms of its contribution to the existing knowledge on population who had lived in South Croatia during the Early Medieval Period. Though we dealt with two sites from similar periods (9th century), spatial differences notably affected the differences in lifestyle and diet of these people. Both archaeological sites belong to the Early Medieval Period, famous for natural economy. At the archaeological site Ostrovica single byzantine currency of that period was found, witnessing the relationship of Early Medieval Croatia and its Counties with the Empire on Bospor. The site was at intersection of roads, which enabled better exchange of goods, and consequently a food exchange. However, the other archaeological site Naklice is located at the mild hill, and this archaeological population was more isolated. The available food was poor in quality and quantity. Considering the above mentioned archaeological facts, the difference between these two archaeological groups is evident. The concentration of some metals analyzed in archaeological bones may be useful to draw conclusions about lifestyle and diet of archaeological populations. With regard to these two archaeological sites, differences have been found between values of the individual metals and their ratios (Sr/Zn; Sr/Ca; Zn/Ca). The increased values of Sr and the ratio Sr/Ca in archaeological population from Naklice location indicate that this population mostly consumed vegetable origin food. Extremely low values of Zn and the Zn/Ca ratio were determined for a group of children bones, from Naklice burial site, because they were living in exceptionally difficult circumstances. The comparison between adults and children within the same archaeological site shows great similarity in diet. In regards to the aim of our study, we found statistically significant differences between gender which is confirmed with different strontium and zinc concentrations among men and women probably due to differences in diet and activity of metabolic processes (pregnancy and lactation) in women’s bodies. There were no differences in the distribution of metals between men in regards to age, while differences in regards to age between women, were present. They were incurred as a result of metabolic processes (Sr is higher in the age group 25e39 years). It can be concluded that the analysis of metals in archaeological bones, particularly Sr, Zn, Ca and their ratios may be useful as a supplement to other archaeological and anthropological knowledge although geological changes in soil may affect the credibility of the results of metals in archaeological bones. Certain metals as Cd, Fe, Mn are subject to significant influence of diagenesis. However, the determination of the concentration of these metals is also important. It shows the impact of changes in the soil. The analysis of metals in archaeological bones cannot be only used as a way of reconstruction of population’s habits, but can be useful complement to other archaeological and anthropological studies. In order to make these results more authentic, it is imperative to analyze the composition of the soil, and to observe the extent of diagenesis in archaeological bone. Many factors determine the exposure of archaeological bones to diagenesis (length of the historical period, type and part of the bone, groundwater, soil acidity, etc.). Taking into consideration not only the above-mentioned, but also the fact that different metals behave
400
A. Stipisic et al. / Journal of Archaeological Science 46 (2014) 393e400
differently in the soil, the relationship between metal in bones and diet of archaeological populations was shown. In future studies, it would be interesting to make a comparative analysis of metals in the bones and teeth of a certain archaeological population. Namely, teeth contain more inorganic substances than archaeological bones and are more resistant to changes in the soil, and thus expected to be less exposed to diagenesis than bones. Acknowledgments The authors would like to thank dr. Vedrana Delonga and dr. Tonci Buric from the Archaeological Museum in Split for enabling the access to skeletal material from the Naklice burial site. This study was financially supported by a grant from the Ministry of Science, Education and Sports of the Republic Croatia (Grant number 216-2160800-0655). References Ambrose, S., Krigbaum, J., 2003. Bone chemistry and bioarchaeology. J. Anthropol. Archaeol. 22, 193e199. Analytik Jena AG, Jena, Germany, 2001. Operating Manual AAS Vario 6. Baranowska, I., Czernicki, K., Aleksandrowicz, R., 1995. The analysis of lead, cadmium, zinc, cooper and nickel content in human bones from the upper Silesian industrial district. Sci. Total Environ. 159, 155e162. Carvalaho, M.L., Marques, A.F., Lima, M.T., Reus, U., 2004. Trace elements distribution and post-mortem intake in human bones from middle Age by total reflection X-ray fluorescence. Spectrochim. Acta B 59, 1251e1257. CEM Corporation, Illinos, USA, 2006. Operating Manual CEM Mars 5. Delonga, V., Buric, T., 1998. Ostrovica near the Bribir. In: Archaeological and Historial Sketch. The Museum of Croatian Archaeological Monuments, Split. Dobrovolskaya, M.V., 2005. Upper palaeolithic and late stone age human diet. J. Physiol. Anthropol. Appl. Hum. Sci. 24, 433e438. Donno, A., Santoro, V., Fazio, A., Corrado, S., Urso, D., Baldassarra, S.L., 2010. Analysis of neolithic human remains discovered in southern Italy. J. Archaeol. Sci. 37, 3482e3487. Goldstein, I., 1995. Hrvatski srednji vijek. Novi Liber, Zagreb, Croatia, pp. 84e116. Gonzalez-Reiemers, E., Arnay-de-la-Rosa, M., Velasco-Vazquez, J., GalindoMartin, L., Satolaria-Fernandez, F., 2005. Bone cadmium and lead in the ancient population from El Hierro, Canary Islands. Biol. Trace Elem. Res. 105, 37e51. Hedges, R.E.M., 2002. Bone diagenesis: an overview of processes. Archacometry 44, 319e328. Hedges, R.E.M., Millard, A.R., 1995. Bones and groundwater, towards the modeling of diagenetic processes. J. Archaeol. Sci. 22, 155e164. Jurkiewicz, A., Wiechula, D., Nowak, R., Gazdzik, T., Loska, K., 2004. Metal content in femoral head spongious bone of people living in regions of different degrees of environmental pollution in Southern and Middle Poland. Ecotoxicol. Environ. Saf. 59, 95e101. Lambert, J.B., Vlasak, S., Szpunar, C.B., Bukistra, J.E., 1985. Bone diagenesis and dietary analysis. J. Hum. Evol. 14, 477e482. Martinez-Garcia, J.M., Moreno, J.M., Moreno-Clavel, J., Vergara, N., GarciaSanchez, A., Guillamon, A., 2005. Heavy metals in human bones in different historical epochs. Sci. Total Environ. 348, 51e72.
Mays, S., 2003. Bone strontium:calcium ratios and duration of breastfeeding in a Mediaeval skeletal population. J. Archaeol. Sci. 30, 731e741. Nakashima, T., Matsuno, K., Matsushita, M., Matsushita, T., 2010. Severe lead contamination among children of samurai families in Edo period Japan. J. Archaeol. Sci. 38, 23e28. National Institute of Standard and Technology, Gaithersburg, Maryland, USA, 2009. Certificate of Analysis Standard Reference Material Montana I Soil 2710a. Nielsen-Marsh, C.M., Hedges, R.E.M., 2000. Patterns of diagenesis in bone I: the effects of site environments. J. Archaeol. Sci. 27, 1139e1150. Nielsen-Marsh, C., Gernaey, A., Turner-Walker, G., Hedges, R., Pike, A., Collins, M., 2006. The chemical degradation of bone. In: Cox, M., Mays, S. (Eds.), Human e Osteology: In Archaeology and Forensic Science. Cambridge University Press, New York, pp. 439e453. Novak, M., Slaus, M., 2010. Health and disease in a Roman walled city: an example of Colonia Iulia Iader. J. Anthropol. Sci. 88, 189e206. Pate, F.D., Hutton, J.T., 1988. The use of soil chemistry data to address post-mortem diagenesis in bone mineral. J. Archaeol. Sci. 15, 729e739. Price, T.D., Blitz, J., Burton, J., Ezzo, J.A., 1992. Diagenesis in prethistoric bone: problems and solutions. J. Archaeol. Sci. 19, 513e529. Pyatt, F.B., Pyatt, A.J., Walker, C., Sheen, T., Grattan, J.P., 2005. The heavy metal content of skeletons from an ancient metalliferous polluted area in southern Jordan with particular reference to bioaccumulation and human health. Ecotoxicol. Environ. Saf. 60, 295e300. Runia Lex, T., 1987. Strontium and calcium distribution in plants effect on Palaeodietary studies. J. Archaeol. Sci. 14, 599e608. Schutkowski, H., Herrmann, B., 1999. Diet, status and decomposition at Weingarten: trace element and isotope analyses on early mediaeval skeletal material. J. Archaeol. Sci. 26, 675e685. Shafer, M.M., Siker, M., Overdier, J.T., Ramsl, P.C., Teschler-Nicola, M., Farrell, P.M., 2008. Enhanced methods for assessment of the trace element composition of Iron Age bone. Sci. Total Environ. 401, 144e161. Slaus, M., 1997a. Discriminant function sexing of fragmentary and complete femora from medieval cites in continental Croatia. Opvscvla. Archeol. 21, 167e175. Slaus, M., 2000b. Biocultural analysis of sex differences in mortality profiles and stress levels in the late medieval population from Nova Raca, Croatia. Am. J. Phys. Anthropol. 111, 193e209. Slaus, M., 2006c. In: Bioarcheology-demography, Health, Nutrition and Trauma Ancient Croatian population. Skolska knjiga, Zagreb, Croatia, pp. 61e98. Slaus, M., 2008d. Osteological and dental markers of health in the transition from the Late Antique to the Early Medieval period in Croatia. Am. J. Phys. Anthropol. 136, 455e469. Sutlovic, D., Stipisic, A., Marusic, J., Pavlov, J., Versic, M., Definis-Gojanovic, M., 2010a. Elemental status in individual from Naklice Burial site (Southern Croatia): Mediaeval Diet reconstruction. Croat. Chem. Acta 83, 217e221. Sutlovic, D., Ljubkovic, J., Stipisic, A., Knezovic, Z., Drmic-Hofman, I., Andjelinovic, S., 2014b. Skeletal remains from late Roman period: “As old as Diocletian’s Palace”. Coll. Antropol. (in press). Vuorinen, H., Tapper, U., Mussalo-Rauhamaa, H., 1990a. Trace and heavy metals in infants, analysis of long bones from Ficana, Italy, 8e6th Century BC. J. Archaeol. Sci. 17, 237e254. Vuorinen, H., Pihlman, S., Mussalo-Rauhamaa, H., 1996b. Trace and heavy metal analyses of a skeletal population representing the town people in Turku (Abo), Finland in the 16e17th centuries: with special reference to gender, age and social background. Sci. Total Environ. 177, 145e160. Wissenchaftlich Technishe Werkstatten-WTW, Weilheim, Germany, 2004. Operating Manual inoLab-pH720. Yamada, M., Minami, T., Yamada, G., Tohno, Y., Tohno, S., Ikeda, Y., 1997. Different element ratios of red cosmetics excavated from ancient burials of Japan. Sci. Total Environ. 199, 293e298.