Regional provenance of dolerite prehistoric objects through mineral analysis

Microchemical Journal 124 (2016) 167–174

Contents lists available at ScienceDirect

Microchemical Journal journal homepage: www.elsevier.com/locate/microc

Regional provenance of dolerite prehistoric objects through mineral analysis Gianni Gallello a,⁎, Teresa Orozco b, Agustin Pastor a, Miguel de la Guardia a, Joan Bernabeu b a b

Department of Analytical Chemistry, University of Valencia, 50 Dr. Moliner Street, 46100 Burjassot, Valencia, Spain Department of Prehistory and Archaeology, University of Valencia, 28 Blasco Ibáñez Street, 46010 Valencia, Spain

a r t i c l e

i n f o

Article history: Received 30 July 2015 Accepted 16 August 2015 Available online xxxx Keywords: Dolerite Rare earth elements (REE) Multivariate statistics Polished stones Prehistory Valencia region

a b s t r a c t A methodology based on the mineral analysis determination has been developed to identify the origin of dolerite stone outcrops collected to fabricate lithic objects during the Prehistoric period. The method is based on the use of inductively coupled plasma mass spectrometry (ICP-MS) to analyse rare earth elements (REE) and trace elements. Additionally a no destructive geochemical analysis based on X-ray fluorescence (XRF) was employed for major elements analysis. The aforementioned methodologies were applied to samples from different archaeological fields or natural outcrops located in the Mediterranean area of Spain, between Valencia and Alicante. Principal component analysis (PCA) was employed to interpret the dolerite geological provenance. These preliminary results show that statistical analysis permits to distinguish stone sample origins according to their REE profile at regional level and that Ti/Fe major element relation perhaps shows just coarse differences between samples collected on the extreme north and south of the studied region. The proposed method could be useful to discriminate the regional origin of lithic objects belonging to dolerite rocks and to interpret the primary material transport and exchange of lithic materials in Valencian Prehistory. © 2015 Elsevier B.V. All rights reserved.

1. Introduction Geochemical studies of lithic artefacts try to find a relationship between the archaeological materials and the primary source present in the nature using sample chemical profile. In the last decades the determination of the geologic source and provenance of lithic objects and west flakes has become a standard practice in archaeological research to evaluate prehistoric migrations and fluxes. Provenance information results are useful to rebuilt settlement patterns, and investigate stone tool technologies, exchange systems and territoriality [1–10]. Technological advances have recently improved the ability to determine the source of archaeological materials. Archaeological lithic provenance studies have typically focused on ground stone tools, flints and obsidian. Obsidian is a volcanic glass widely used for stone tools; it is an ideal material for reconstructing exchange systems since it occurs geologically in a limited number of location, and is frequently found in archaeological sites even far from a source and may be chemically fingerprinted, allowing source attribution of artefacts based on it which has been designed a multi-method exploratory approach to chemically characterise geological obsidian samples from Sardinia [9]. For the full chemical characterisation of geological samples a combination of electron microprobe analysis, X-ray fluorescence (XRF), neutron activation analyses (NAA) and inductively coupled plasma ⁎ Corresponding author. Tel.: +34 69 7636957; fax: +34 96 3544838. E-mail address: [email protected] (G. Gallello).

http://dx.doi.org/10.1016/j.microc.2015.08.018 0026-265X/© 2015 Elsevier B.V. All rights reserved.

mass spectrometry (ICP-MS) was employed. The results provided detailed view of the distribution of obsidian from the central Mediterranean Island sources. Some studies have been centred on polished stone axes that are important components in Neolithic manufacturing techniques being made with the petrological characterisation of Neolithic polished stone axes found in several archaeological sites in central Sardinia [1]. A petrographical description was given for all encountered lithologies, and bulk rock chemical analyses of minerals of 12 representative samples were given, with the aim of determining the origin of the raw materials. The mineralogical and petrological characterisation showed useful information to distinguish between exotic (“nephrite” and glaucophane schist) and local (phonolite) Sardinia polished stone axis raw material origins. Williams-Thorpe et al. [10] studied one of the best known petrological groups of polished stone implements found in England and Wales which comprise axes and related artefacts made of fine to mediumgrained quartz dolerite. These implements of Neolithic and Bronze age date are termed Group XVIII. The aim of this work was to establish geochemical and magnetic characteristics of Group XVIII implements using non-destructive methods X-ray fluorescence analysis (XRF) and magnetic susceptibility measurement, to compare the chemical and magnetic characteristics with those of potential sources including the Whin Sill, consisting on outcrops or secondary deposits which crops out extensively in northern England, additionally than other igneous outcrops in northern Britain. In this way it was tested the validity of source(s) of Group XVIII and produced a geochemical and magnetic

168

G. Gallello et al. / Microchemical Journal 124 (2016) 167–174

description for the Group which can help in non-destructive provenancing of dolerite artefacts. However the aforementioned instruments proved to be very effective in testing the proposed Group XVIII source implements, but did not always provided enough information to identify specific alternative source for non-Group XVIII implements. The methods were considered by the authors not suitable for very weathered artefacts which have not fresh surfaces available for measurements. Geochemical analysis were also applied in a world heritage site as Stonehenge [2,3] to explore the correlation between the rhyolitic and dacitic lithologies from the site and the two main volcanic sequences of Ordovician age belonging to the Fishguard Volcanic Group and the Sealyham Volcanic Formation exposed across north Pembrokeshire . The work of Bevins et al. [2] was innovative in the use of zirconium (Zr) chemistry using laser-ablation ICP-MS analysis in archaeopetrological provenancing and in a second work Bevins et al. [3] concluded that the only dacitic or rhyolitic lithology which can be matched with any degree of confidence between the Stonehenge landscape and a specific source area is the so-called ‘rhyolite with fabric’ lithology, which matches with foliated rhyolitic rocks exposed in the Pont Saeson area of north Pembrokeshire, and Craig Rhos-y-felin. This is correlated with some confidence in terms of petrography, whole rock geochemistry and mineral chemistry. In this preliminary study we have tried to identify outcrop raw materials collected for manufacturing lithic objects during the Prehistoric period employing ICP-MS to analyse rare earth elements (REE) and trace elements. No destructive geochemical analysis consisted on the use of XRF was also employed for determination of major elements. The ultimate goal was to develop a method to complete the reconstruction of transportation networks of prehistoric tools already developed in a comprehensive study of thin film analysis with polarizing microscope [11]. Oxides, major elements, trace elements and rare earth elements (REE) were determined in a total of 16 samples including 13 natural and 3 archaeological stone samples. All of them collected in archaeological fields or natural outcrops located in the Mediterranean area of Spain, between Valencia and Alicante region. The multi-element capability of ICP-MS was employed to identify changes in trace elements and REE between

samples [12,13]. XRF is a technique that permit direct, fast, cheap and safe analyses for archaeological sample compared with mass spectrometry one. We employed XRF to analyse specimens and elements trying to identify markers that permit us to observe differences between samples and identify archaeological objects raw material provenance avoiding the use of destructive ultra-trace (REE) analysis. Principal component analysis (PCA) was used to identify archaeological dolerite samples of raw material provenance employing REE as variables. Ti/Fe relations were pin-point as a coarse provenance marker for dolerite objects belonging to the studied region. 2. Materials and methods 2.1. The studied samples An amount of 16 dolerite samples were analysed; 13 from natural outcrops were exploited during the antiquity as quarries and 3 prehistoric utensil fragments (Fig. 1) from archaeological work-field samples were collected from different areas of Valencia Community (Fig. 1). Sample numbers and geographical origin are summarized in Table 1. Samples S1, S2, and S3, were collected in Finestrat (Orxeta, Alicante) outcrop employed as an ancient quarry, composed of a relativity fresh igneous rock defined as dolerite labrador and piroxene. Samples S4, S5, S6, S7, and S8 are from Pinós, XinorletFont d' Almorquí y Sax (Alicante) outcrops, following a certain structural tectonic alignment by the intersection of various failures. The basic rock outcrop of Pinos is located on the eastern slope of the saline dome of Cabezo de la Sal; materials here have two textures: one of the outcrop edges and one (ortophyric to subtrachytic) and the rock mass (monocrystalline hypidiomorphic dibasic). At NW of Sax into the Vinalopó (Santa Eulalia) depression, several hints of dolerite are found and some of them are difficult to locate, because they are small in size and are located at ground level. The larger quarry (CG) is dead today. These materials have been classified as trachyandesites and microgabbro with coarse andesina. Sample S9 comes from Sierra Orihuela, precisely from the hint call “el Tunel” and intense exploitation. The rock in this outcrop is relatively fresh, very compact, dark greenish colour and size of fine and homogeneous

Fig. 1. Collected sample locations. Dolerite polished stone. “Arenal de la Costa” object raw material provenance (brown colour). “Bancal de Satorre” object raw material provenance (blue colour). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

G. Gallello et al. / Microchemical Journal 124 (2016) 167–174 Table 1 Sample description. Sample

Proceeding

Origin

Group

S1 S2 S3 S4 S5 S6 S7 S8 S9 S10 S11 S12 S13 S14 S15 S16

Finestrat (Cantera) Finestrat Finestrat Pinoso-Xinorlet Vinalopó-Santa Eulalia Vinalopó-Santa Eulalia Vinalopó-Santa Eulalia (C.G.) Vinalopó-Santa Eulalia (C.G.) Sierra Orihuela (Tunel) Altura Almansa (Cerro Los Cuchillos) Almansa (Cerro Los Cuchillos) Almansa (Cerro Los Cuchillos) Arenal de la Costa Bancal de Satorre Ereta del Pedregal

M7 LD outcrop M4 LD, outcrop M5 LD, outcrop Outcrop M1 LD, outcrop Muestra 3 LD, outcrop Outcrop Outcrop Outcrop Outcrop Outcrop LD Lc1, outcrop LD Lc 3, outcrop Arch. 88.28.01 Arch. AC-76-24 Arch. 421 EP-21348 SIP

Dolerite Dolerite Dolerite Dolerite Dolerite Dolerite Dolerite Dolerite Dolerite Dolerite Dolerite Dolerite Dolerite Dolerite Dolerite Dolerite

grain. It is included by a petrological classification among the quartziferous dolerite. Sample S10 comes from Altura. The enclaves of Torás-Bejín, Altura and Soneja (Castellon), correspond to dolerite that present variation in the composition of the medium to basic plagioclase. Between those enclaves the one located at the W Altura was formerly exploited as quarries. Here the rock has an intense alteration and jointing and great textural variability between different areas of the same enclave. Samples S11, S12, and S13 are dolerites coming from El Cerro De Los Cuchillos located in the northeast of Almansa (Albacete) where in the 80s an archaeological site of the Bronze Age was discovered. Sample S14 came from the archaeological excavation of Arenal de la Costa. This site is located in Ontinyent (Valencia) near the city centre. It was discovered in the late 80s being exposed sections of various prehistoric structures for sand mining. Materials recovered and radiocarbon dating placed this excavation site in the early second millennium BC. The lithic material studied in this excavation is polished stone in high state of fragmentation. Sample S15 comes from Bancal de Satorre archaeological site which is located at the end of Benefallim (Alicante) in a place called Les Puntes. The chronology estimated by the pottery found on the site corresponds to the Early Neolithic I. The material analysed concerns polished object fragments. Sample S16 comes from Ereta del Pedregal archaeological site located in Navarres (Valencia). This site is one of the most remarkable archaeological open sites in Valencia Community due to it documented sequence (that covers from Neolithic IIB and HCT), such as the room structure and the variety and richness of the archaeological record found. Numerous archaeological excavations beginning in the 40s have been carried out in this site. The sequence of this deposit has facilitated the implementation of various studies (palynology, sedimentology, zooarchaeology) that have provided paleo-environmental and economic information. The lithic material analysed is part of the pieces preserved in the SIP (Prehistoric Research Service of Valencia) coming from different excavations carried out between the 40s and 80s [11]. The petrological study carried out by Orozco suggests that Ereta del Pedregal object was made employing raw material collected in a quarry at Quesa (Locality very close to Ereta del Pedregal). To proof our proposed methodology Quesa quarry material was not included in our analysed set of samples. The raw material provenance of the other two archaeological samples ( Arenal de la Costa and Bancal de Satorre) should be located between Orihuela and Altura outcrops (Fig. 1). 2.2. Chemical analysis of stone samples 2.2.1. ICP-MS analysis Sampling was carried out collecting different rocks belonging to different areas in the same region as was described above. Sample preparation and digestion were carried out pre-crashing the stone samples employing a jaw crusher, homogenized and pestle by using an agate

169

mortar. The digestion method and concentration ranges of the dilutions from the digested solution were evaluated in order to provide reproducible and comparable results compatible with the sensitivity of the analytical methods employed. The digestion method consisted in the addition of 1.35 ml HCl and 0.45 ml HNO3 to 0.15 g of sample in glass tubes placing them in a water bath at 100°C for 40 min. Subsequently, the digested solutions were carefully poured into plastic tubes of 15 ml, bringing the volume to 15 ml with purified water. This concentrated solution was used to measure trace elements such as Ba, Bi, Cd, Cr, Co, Cu, Pb, Li, Mn, Mo, Ni, Sr, V, and Zn and REEs (La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu,) Sc and Y, solution. A multi-element stock solution containing the mentioned elements at a concentration of 100 μg/ml was used to prepare the calibration standards; 5 ml volumetric flasks were used adding 0.15 ml of HNO3, 0.45 ml of HCl and the corresponding volume of standard solution and filling up to volume with pure water. The prepared dilutions were analysed by ICP-MS with Perkin Elmer Elan DRCII (Concord, Ontario, Canada). To avoid the obstruction of the nebulizer system samples were filtered employing filter paper (Whatman™ N.1 of 70 mm). Concentration ranges between 1 and 600 μg/l were used for trace elements (Ba, Bi, Cd, Cr, Co, Cu, Pb, Li, Mn, Mo, Ni, Sr, V, Zn, La, Ce, Pr, Nd), and concentration ranges between 1 and 100 μg/l for Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, and Lu. All standards were acquired from Sharlab S.L. (Barcelona). Soil GBW07408 was used as standard reference materials for evaluating the analytical quality of the method. Rh was used as internal standard for ICP-MS analyses. Once sample preparation had been developed as described in the methodology, analyses have been performed by ICP-MS employing the analytical parameters showed in Table 2. Thirty-one elements were analysed including mayor elements, trace elements and REEs. The analytical mass isotope instrumental detection and quantification limit (LOD and LOQ, respectively) and R2 are listed in Table 3. 2.2.2. XRF analysis Homogenized and pestle samples were directly analysed by X-ray fluorescence. Spectra were obtained using a portable model S1 Titan energy dispersive X-ray fluorescence spectrometer from Bruker (Kennewick, Washington, USA) equipped with a rhodium X-ray tube and XFlash® SDD detector. For instrument control S1RemoteCtrl (Geochemtrace programme) and S1Sync software from Bruker were employed to measure percentage of Al2O3, SiO2, K2O, CaO, Ti, and Fe and for spectra treatment, the ARTAX software from Bruker was used. The standard error of readings during the analysis ranged from 1% to 9% for oxides and elements. Again GBW07408 was used as standard reference materials for evaluating the quality of the employed method (Table 4). 2.3. Data analysis For statistic analysis 16 dolerite rock samples were employed. All variables (i.e., REE, Sc and Y) being used for modelling. PCA was used

Table 2 ICP-MS employed parameters. Vacuum pressure Nebulizer Gas Flow RF power Nebulizer pump Tubes internal diameter Lens voltage Analog stage Pulse stage Read Delay Sample flush Reading parameters

5 × 10−5Torr 0.92 L/min 1100 W 20 rpm 0.76 mm 6.5–8.5 V −1950 V 1050 V 15 s 60 s • Dwell time UMA: 50 ms • Sweeps: 20 • Readings: 1 • Replicates: 3

170

G. Gallello et al. / Microchemical Journal 124 (2016) 167–174

Table 3 ICP-MS analysis. Mass, detection limits (LOD), quantification limits (LOQ) and R2 of 31 elements. Element

Mass [Da]

LOD

LOQ

R2

La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Sc Y Sr Zn Cu Ba Mn Bi Cd Cr Co Pb Li Mo Ni V Rha

139 140 141 142 152 151 158 159 162 165 166 169 172 175 45 89 88 64 63 138 55 209 111 52 59 207 7 95 60 51 103

0.0004 0.0005 0.00010 0.0003 0.0003 6E−05 0.00015 5E−05 1.1E−05 3E−05 0.00013 1.6E−05 7E−05 1.7E−05 0.013 0.0005 0.0012 0.0015 0.0014 0.0010 0.002 0.0006 0.00017 0.01 0.0004 0.0007 0.0002 0.0011 0.007 0.03

0.0014 0.0018 0.0003 0.0010 0.0011 0.00018 0.0005 0.00017 4E−05 0.00011 0.0005 5E−05 0.0002 6E−05 0.04 0.0016 0.004 0.005 0.004 0.003 0.007 0.002 0.0006 0.3 0.0014 0.002 0.0008 0.004 0.02 0.11

0.9997 0.9997 0.9997 0.9985 0.9999 0.9998 0.9998 0.9977 0.9998 0.9983 0.9999 0.9985 0.9999 0.9991 0.9998 0.9996 0.9999 0.9998 0.9999 0.9999 0.9997 0.9999 0.9995 0.9986 0.9986 0.9996 0.9994 0.9998 0.9996 0.9999

Note: LOD and LOQ expressed as μg/g for all elements. a Internal standard.

to explore large geochemical datasets reducing the number of variables and providing a deeper insight into the structure of the variance of the dataset (Jolliffe 2002). For PCA in a dataset with 16 samples and 16 variables autoscaling was employed as pre-processing step prior to modelling. Data analysis was carried out using the PLS Toolbox 6.5 for Eigenvector Research Inc. (Wenatchee, WA, USA), running in Matlab R2014b from Mathworks Inc. (Natick, MA, USA). 3. Results and discussion 3.1. The elemental composition of dolerite stones analysed by ICP-MS As described above samples were measured by ICP-MS and the results obtained are reported in Tables 5a–5b and 6a–6b. Being expressed the mineral concentration in μg/g. The obtained mean concentrations with their standard deviations pointed differences in geochemical composition between samples. Concentrations of Ba, Bi, Cd, Pb, Mo, Co and Mn are similar between samples S4 (Pinoso), S10 (Altura) and S9 (Sierra Orihuela) outcrops. In these samples differences in elements such as Cr, Cu, Li, Ni, Sr, V and Zn could be also appreciated. Table 4 XRF analysis. CRM (GBW07408) obtained values (ob. value) and certified values (cert. value) of the analysed oxides and elements. Element

Ob. value

Cert. value

Al2O3 SiO2 K2O CaO Ti Fe

11.86 ± 0.49 58.2 ± 1.7 2.264 ± 0.014 7.83 ± 0.09 0.3829 ± 0.0010 4.07 ± 0.002

11.92 ± 0.15 58.61 ± 0.13 2.42 ± 0.04 8.27 ± 0.12 0.380 ± 0.012 4.0817 ± 0.0738

Note: Values expressed as percentage (%).

Comparing samples S1, S2, and S3, preceding from the same site (Finestrat) differences in their trace analysis contents of Cd, Cr, Cu, Pb, Li, Mn, Ni, Sr, V, and Zn could be appreciated. Samples S11, S12, and S13 from Almansa outcrops are also different between them in Ba, Bi, Cd, Cr, Cu, Li, Mn, Mo, Ni, and Sr contents. Concentrations of Ba, Bi, Cd, Cr, Cu, Li, Mn, Mo, Ni, Sr, V and Zn reveal dissimilarity between Vinalopó samples S5, S6, S7, and S8. Achaeological samples S14 (Arenal Costa), S15 (Bancal of Satorre) and S16 (Ereta del Pedregal) results are also diverse in their Ba, Bi, Cr, Cu, Pb, Li, Mn, Mo, Ni, Sr, and V values. REE sample contents can be observed in Tables 6a–6b. REE and Y values, except Sc in samples S4 and S10 are different. The opposite happened for trace elements of values in the same samples. Finestrat samples S1, S2, and S3 REE and Y values are similar, especially this could be observed among samples S2 and S3. Samples S11, S12, and S13 Almansa also have similarities in REE and Sc values except Y. Vinalopó samples S5, S6, S7, and S8 presented similar concentrations of REE Sc and Y between S5 and S8 on one side and S6–S7 on the other site. Important difference in their geochemical profiles (REE Sc and Y) between the two pairs of samples can be appreciated despite all comes from the same locality. Archaeological samples S14 (Arenal de la Costa) and S15 (Bancal de Satorre) are very similar in their REE Sc and Y values. The last archaeological sample (S16, Ereta del Pedregal) has higher values in those elements and so is clearly different from the other two archaeological samples. In short results show that trace elements do not allow clearly discriminate samples from different groups or to highlight similarities of the samples belonging to the same group. Probably due to the strong partition of rear earth elements into the particulate phase, their coherent behaviour during weathering, erosion and fluvial transportation and their high resistance to chemical mobilization [14–16], REE, Sc and Y and especially some of them depending on the case, clearly put in evidence geochemical differences between dolerites collected from differences area of the same region. 3.2. PCA analyses of sample origin Principal component analyses (PCA) has been applied to the whole set of samples. Fig. 2a shows the score plots of 16 dolerite samples. The first two principal components explain the main part of the variance of data PC1 (85.99%) and PC2 (9.19%) respectively. Score plots represents as data points (samples) projected into the PC space. In the scores plot it can be observed that the group of samples from “PINOSO”, “FINESTRAT”, “ALMANSA” and “VINALOPÓ” (S5 and S8 Vinalopó samples) coming from natural outcrops are similarly distributed and also are related to the archaeological samples “ARENAL DE LA COSTA” and “BANCAL DE SATORRE”. On the other hand "ORIHUELA" sample is farther from the aforementioned group and big differences can be appreciated between “ALTURA” and archaeological sample from Ereta de Predegal. “VINALOPÓ” samples S6 and S7 are different than S5 and S8 and other samples. In the loading plot shown in Fig. 2b, it can be seen that the contribution of each variable (elements) to the PC1 being the absolute intensity of the loading of each variable directly correlated with its contribution to the model and the sign with its direction. As explained above, PC1 contains useful information for differentiating dolerite samples proceeding from different geographical areas. Interpreting magnitude and signs of the loading from PC1 clearly show that La, Ce, Pr, Nd, Sm, Eu, Gd, Tb, Dy, Ho, Er, Tm, Yb, Lu, and Y are the most important variables for model building and high element values could be appreciated in S10 “ALTURA” and S16 “ERETA DEL PEDREGAL” that were collected in two different geographical areas and also show variation between each other (see Fig. 2a). On the other hand Sc is not a representative variable in the model. Looking carefully at the statistic relationship between natural outcrops and archaeological materials S1, S2, and S3 samples from Finestrat

G. Gallello et al. / Microchemical Journal 124 (2016) 167–174

171

Table 5a Trace elements values and standard deviations expressed as μg/g. ID

Ba

Bi

Cd

Cr

Co

Cu

Pb

S4 S10 S9 S1 S11 S12 S5 S6 S7 S8 S2 S3 S13 S14 S15 S16

44 ± 8 33 ± 8 35.52 ± 1.04 54 ± 8 133 ± 22 124 ± 2 783 ± 11 383 ± 85 118 ± 18 88 ± 4 59 ± 5 71 ± 4 116 ± 7 67 ± 4 125 ± 7 590 ± 30

0.0408 ± 0.0007 0.08 ± 0.02 0.018 ± 0.002 0.061 ± 0.005 0.0329 ± 0.0014 0.0115 ± 0.0012 0.030 ± 0.003 0.005 ± 0.003 0.018 ± 0.003 0.02 ± 0.005 0.029 ± 0.003 0.029 ± 0.003 0.025 ± 0.006 0.07 ± 0.02 0.022 ± 0.008 0.014 ± 0.005

0.076 ± 0.004 0.068 ± 0.010 0.0223 ± 0.0008 0.102 ± 0.006 0.142 ± 0.011 0.122 ± 0.002 0.0570 ± 0.0004 0.108 ± 0.012 0.089 ± 0.005 0.113 ± 0.006 0.045 ± 0.004 0.027 ± 0.002 0.055 ± 0.006 0.111 ± 0.002 0.0902 ± 0.0011 0.0895 ± 0.0008

200 ± 50 89 ± 16 108 ± 7 8.9 ± 0.6 29 ± 2 24.5 ± 0.9 61 ± 10 62 ± 16 42 ± 12 61 ± 3 97.34 ± 0.03 148 ± 9 110 ± 10 180 ± 40 50 ± 10 100 ± 20

19 ± 5 17 ± 5 29.43 ± 0.12 25 ± 2 20.8 ± 0.4 17.6 ± 0.4 17.3 ± 0.7 32 ± 9 54 ± 11 21.1 ± 0.8 21.1 ± 0.7 25.7 ± 0.7 22.33 ± 0.96 23 ± 3 24 ± 4 18 ± 3

9±2 19 ± 5 132 ± 4 131 ± 14 92 ± 2 90 ± 4 56 ± 2 78 ± 22 118 ± 25 104 ± 2 68 ± 4 90 ± 2 80 ± 3 116 ± 12 62 ± 7 16 ± 2

3.0 ± 0.5 2.4 ± 0.7 2.93 ± 0.15 3.5 ± 0.5 5.3 ± 0.8 7.56 ± 0.15 10.0 ± 0.6 7.4 ± 1.4 13 ± 2 4.1 ± 0.6 2.04 ± 0.07 3.1 ± 0.2 8±2 43 ± 9 9±2 4.0 ± 0.9

Table 5b Trace elements values and standard deviations expressed as μg/g. ID

Li

Mn

Mo

Ni

Sr

V

Zn

S4 S10 S9 S1 S11 S12 S5 S6 S7 S8 S2 S3 S13 S14 S15 S16

63 ± 8 115 ± 8 23 ± 5 15 ± 3 37 ± 9 34 ± 6 42 ± 7 24 ± 3 28 ± 2 26 ± 2 19.6 ± 0.8 23.0 ± 0.3 25.8 ± 0.4 66 ± 8 23 ± 5 101 ± 12

420 ± 80 1690 ± 340 390 ± 30 830 ± 10 650 ± 20 710 ± 30 530 ± 50 1190 ± 320 770 ± 160 580 ± 60 600 ± 10 610 ± 40 750 ± 80 900 ± 180 690 ± 160 990 ± 200

0.12 ± 0.02 0.238 ± 0.003 0.120 ± 0.007 0.286 ± 0.012 0.169 ± 0.004 0.070 ± 0.004 0.290 ± 0.007 0.034 ± 0.007 0.28 ± 0.03 0.102 ± 0.003 0.2640 ± 0.0006 0.5 ± 0.007 0.6 ± 0.013 0.359 ± 0.010 0.205 ± 0.011 0.6 ± 0.03

56 ± 17 34 ± 10 43.4 ± 0.2 28.8 ± 1.5 90 ± 3 25.5 ± 0.7 35.90 ± 1.03 57 ± 17 51 ± 11 25.63 ± 1.12 31.9 ± 1.2 40.31 ± 1.10 45 ± 2 35 ± 5 41 ± 7 13 ± 3

88 ± 6 29.61 ± 0.10 53 ± 3 49 ± 3 155 ± 13 164 ± 3 254 ± 14 310 ± 3 202 ± 4 56 ± 2 58 ± 4 98 ± 8 51.1 ± 0.9 32.7 ± 1.0 39 ± 2 150 ± 14

200 ± 50 130 ± 30 58 ± 3 219 ± 5 157.1 ± 1.2 135 ± 2 140 ± 10 100 ± 30 200 ± 50 130 ± 10 140 ± 2 200 ± 10 140 ± 10 160 ± 30 220 ± 50 190 ± 40

130 ± 30 250 ± 10 61 ± 2 189 ± 5 159 ± 8 144 ± 4 130 ± 10 140 ± 30 130 ± 30 164 ± 8 105 ± 2 99 ± 6 130 ± 10 190 ± 30 180 ± 40 110 ± 20

(Orxeta), S4 from Pinos (Xinorlet), S5 and S8 from Vinalopó (Alicante) and S11, S12, and S13 from Cierro de Los Cuchillos (Almansa) belong all to outcrops located at few kilometres of distance each other. This fact could justify similarities in their REE contents. Also it is possible to observe the relation that exists between the mentioned outcrops and S14 and S15 archaeological samples. S14 is a tool object proceeding from Arenal de la Costa archaeological site next to Ontinyent that

seems very similar in it REE Sc and Y values to the Almansa samples geographically closer to the site than the other outcrops. Therefore hypothetically the primary employed materials to manufacture Arenal de la Costa tool could be extracted in any outcrop very close to Almansa. The tool fragment S15 from Bancals de Satorre archaeological site is located next to Benefallim. Observing the PCA model this sample (S15) is very similar to “ALMANSA” and “VINALOPÓ” (S5 and S8) values than

Table 6a Rare earth elements values and standard deviations expressed as μg/g. ID

La

Ce

Pr

Nd

Sm

Eu

Gd

Tb

S4 S10 S9 S1 S11 S12 S5 S6 S7 S8 S2 S3 S13 S14 S15 S16

15.77 ± 1.13 32 ± 7 4.3 ± 0.2 12.2 ± 0.7 16 ± 2 13.9 ± 0.5 11.7 ± 0.4 18 ± 3 19 ± 2 11.43 ± 1.13 9.7 ± 0.2 9.00 ± 0.04 13 ± 2 15 ± 2 15 ± 2 41.3 ± 2.2

20 ± 3 47 ± 12 7.1 ± 0.2 20 ± 2 27 ± 4 23.55 ± 0.09 19.16 ± 0.07 30 ± 5 35 ± 5 21 ± 2 16.3 ± 0.6 16.0 ± 0.2 23 ± 2 26 ± 3 25 ± 3 71.2 ± 3.6

2.8 ± 0.4 6.6 ± 1.3 1.10 ± 0.03 3.1 ± 0.3 4.1 ± 0.5 3.65 ± 0.06 2.95 ± 0.06 4.7 ± 0.9 5.1 ± 0.6 3.3 ± 0.3 2.46 ± 0.13 2.43 ± 0.05 3.5 ± 0.4 4.0 ± 0.4 3.9 ± 0.4 10.61 ± 0.99

12.9 ± 1.3 29 ± 5 5.4 ± 0.1 14.8 ± 1.2 20 ± 2 17.4 ± 0.6 13.9 ± 0.5 22 ± 4 24 ± 2 16 ± 2 11.6 ± 0.7 11.7 ± 0.3 17 ± 2 19 ± 2 19 ± 2 45.3 ± 5.6

2.15 ± 0.4 4.59 ± 1.10 1.055 ± 0.013 2.6 ± 0.3 3.7 ± 0.6 3.206 ± 0.003 2.936 ± 0.011 4 ± 1.0 5.0 ± 0.7 3.0 ± 0.3 2.04 ± 0.12 2.12 ± 0.05 3.1 ± 0.4 3.6 ± 0.5 3.4 ± 0.4 7.08 ± 0.7

0.69 ± 0.11 1.3 ± 0.3 0.292 ± 0.003 0.38 ± 0.05 1.0 ± 0.2 0.799 ± 0.006 0.90 ± 0.02 1.3 ± 0.3 2.0 ± 0.2 0.58 ± 0.05 0.40 ± 0.03 0.472 ± 0.011 0.75 ± 0.08 0.70 ± 0.08 0.98 ± 0.11 1.58 ± 0.15

2.3 ± 0.4 4.63 ± 1.08 1.12 ± 0.02 2.7 ± 0.3 3.8 ± 0.7 3.28 ± 0.02 2.67 ± 0.04 4.52 ± 0.99 5.3 ± 0.8 3.2 ± 0.3 2.14 ± 0.13 2.24 ± 0.03 3.2 ± 0.3 3.8 ± 0.4 3.5 ± 0.4 6.5 ± 0.7

0.36 ± 0.06 0.7 ± 0.2 0.1975 ± 0.0009 0.45 ± 0.06 0.65 ± 0.12 0.561 ± 0.009 0.451 ± 0.006 0.8 ± 0.2 0.9 ± 0.2 0.55 ± 0.05 0.34 ± 0.02 0.371 ± 0.009 0.54 ± 0.06 0.65 ± 0.08 0.58 ± 0.07 1.00 ± 0.11

172

G. Gallello et al. / Microchemical Journal 124 (2016) 167–174

Table 6b Rare earth elements values and standard deviations expressed as μg/g. ID

Dy

Ho

Er

Tm

Yb

Lu

Sc

Y

S4 S10 S9 S1 S11 S12 S5 S6 S7 S8 S2 S3 S13 S14 S15 S16

1.7 ± 0.3 3.0 ± 0.8 1.02 ± 0.01 2.2 ± 0.3 3.3 ± 0.6 2.81 ± 0.04 2.26 ± 0.03 3.9 ± 0.9 4.7 ± 0.8 2.7 ± 0.3 1.62 ± 0.11 1.84 ± 0.04 2.7 ± 0.3 3.2 ± 0.4 2.8 ± 0.4 4.9 ± 0.6

0.37 ± 0.07 0.55 ± 0.15 0.2261 ± 0.0007 0.45 ± 0.06 0.71 ± 0.13 0.597 ± 0.009 0.480 ± 0.006 0.8 ± 0.2 1.0 ± 0.2 0.58 ± 0.06 0.33 ± 0.02 0.390 ± 0.008 0.59 ± 0.07 0.68 ± 0.10 0.58 ± 0.09 0.97 ± 0.13

0.9 ± 0.2 1.3 ± 0.4 0.601 ± 0.002 1.1 ± 0.2 1.8 ± 0.3 1.53 ± 0.02 1.236 ± 0.014 2.2 ± 0.5 2.5 ± 0.4 2.00 ± 0.14 0.84 ± 0.05 1.01 ± 0.02 1.5 ± 0.2 1.7 ± 0.2 2.0 ± 0.2 2.6 ± 0.3

0.13 ± 0.02 0.16 ± 0.05 0.0831 ± 0.0005 0.15 ± 0.02 0.25 ± 0.05 0.200 ± 0.003 0.163 ± 0.002 0.31 ± 0.08 0.33 ± 0.06 0.20 ± 0.02 0.104 ± 0.006 0.131 ± 0.002 0.20 ± 0.03 0.22 ± 0.03 0.18 ± 0.03 0.33 ± 0.05

0.71 ± 0.15 0.9 ± 0.3 0.4894 ± 0.0008 0.84 ± 0.13 1.4 ± 0.3 1.11 ± 0.02 0.904 ± 0.011 1.87 ± 0.47 1.84 ± 0.33 0.08 ± 0.10 0.57 ± 0.03 0.743 ± 0.008 1.13 ± 0.15 1.20 ± 0.18 0.9 ± 0.2 1.9 ± 0.3

0.09 ± 0.02 0.12 ± 0.04 0.0735 ± 0.0008 0.12 ± 0.02 0.20 ± 0.04 0.158 ± 0.006 0.12867 ± 0.00007 0.29 ± 0.08 0.26 ± 0.05 0.20 ± 0.013 0.079 ± 0.005 0.106 ± 0.002 0.16 ± 0.02 0.16 ± 0.03 0.13 ± 0.02 0.28 ± 0.04

9.6 ± 3.2 10.6 ± 3.5 4.837 ± 0.103 6.7 ± 0.6 4.8 ± 0.3 3.52 ± 0.13 5.6 ± 0.2 23.3 ± 8.4 23.1 ± 7.2 3.3 ± 0.2 3.39 ± 0.09 5.75 ± 0.07 3.29 ± 0.11 2.7 ± 0.5 2.8 ± 0.6 18.5 ± 3.3

8.2 ± 0.4 12.2 ± 0.5 5.01 ± 0.10 10.37 ± 0.09 16.0 ± 0.2 13.4 ± 0.5 10.09 ± 0.02 17.91 ± 1.13 19.7 ± 1.7 11.9 ± 0.2 7.07 ± 0.05 7.7 ± 0.2 11.26 ± 0.13 13.3 ± 0.4 11.9 ± 0.6 18.7 ± 0.9

a ALMANSA ALTURA

3

S10

ARENAL ARCH. BANCAL ARCH. ERETA ARCH.

Scores on PC 2 (9.19%)

S16

FINESTRAT

2

ORIHUELA PINOSO VINALOPÓ 95% Confidence Level

1

S1

S3

S9

S15

S4

S2

0

S12

S5

S8 S13

S14

-1

S11

-2

S7

S6

-3

-10

-8

-6

-4

-2

0

2

4

6

8

10

Scores on PC 1 (85.99%)

b

Variables/Loadings Plot

0.28 0.27

Gd

Sm

Tb

Dy Ho

0.26

Er

PC 1 (85.99%)

Tm

0.25

Y Yb Lu

Nd Eu Pr

0.24

Ce

0.23 0.22

La

0.21 0.2 Sc

0.19 2

4

6

8

10

12

Variable Fig. 2. a–b. Results PCA. Scores (a) and loadings (b) on PC1.

14

16

G. Gallello et al. / Microchemical Journal 124 (2016) 167–174

173

Table 7 Oxides and major elements values and standard deviation expressed as percentage (%). ID

Al2O3%

SiO2%

K2O%

CaO%

Ti%

Fe%

S4 S10 S9 S1 S11 S12 S5 S6 S7 S8 S2 S3 S13 S14 S15 S16

10.6 ± 0.3 13.6 ± 0.7 10.0 ± 0.4 11.38 ± 0.79 11.5 ± 0.3 12.0 ± 1.6 14.11 ± 0.014 11.23 ± 0.14 10.67 ± 0.10 13.84 ± 0.06 11.957 ± 0.008 13.1 ± 0.5 13.73 ± 0.15 13.04 ± 0.09 14.3 ± 0.3 14.03 ± 0.11

51.2 ± 0.2 44.7 ± 1.2 49.3 ± 1.2 47.3 ± 0.8 44.2 ± 2.8 52.7 ± 0.4 53.0 ± 0.2 52.70 ± 0.02 45.60 ± 1.04 50.95 ± 0.07 47.2 ± 0.2 49.3 ± 0.4 51.9 ± 0.5 48.26 ± 0.06 52.6 ± 1.7 51.3 ± 0.4

0.277 ± 0.010 3.54 ± 0.12 0.373 ± 0.008 0.871 ± 0.016 2.61 ± 0.11 5.48 ± 0.03 2.16 ± 0.02 0.951 ± 0.006 1.54 ± 0.04 2.09 ± 0.04 0.80 ± 0.002 0.677 ± 0.021 1.06 ± 0.02 2.94 ± 0.03 1.57 ± 0.05 1.84 ± 0.03

4.8 ± 0.07 1.8 ± 0.04 8.4 ± 0.18 7.8 ± 0.11 6.1 ± 0.39 5.4 ± 0.07 5.5 ± 0.12 8.268 ± 0.003 6.5 ± 0.09 7.2 ± 0.08 8.6 ± 0.06 9.3 ± 0.10 7.8 ± 0.15 6.2 ± 0.06 6.7 ± 0.25 4.2 ± 0.09

0.652 ± 0.007 0.98 ± 0.03 0.487 ± 0.009 0.64 ± 0.04 0.67 ± 0.04 0.586 ± 0.009 0.648 ± 0.006 0.755 ± 0.003 0.836 ± 0.008 0.729 ± 0.012 0.631 ± 0.009 0.565 ± 0.009 0.6271 ± 0.0014 0.632 ± 0.004 0.65 ± 0.03 0.571 ± 0.003

6.87 ± 0.11 8.16 ± 0.11 8.485 ± 0.012 9.22 ± 0.08 8.95 ± 0.14 8.62 ± 0.14 7.95 ± 0.06 9.892 ± 0.004 9.67 ± 0.25 9.63 ± 0.11 9.12 ± 0.05 9.31 ± 0.02 9.89 ± 0.19 9.13 ± 0.06 8.56 ± 0.14 6.60 ± 0.16

“FINESTRAST” poorer in REE. Bancals de Satorre is geographically located between Finestrat and Vinalopó (Santa Eulalia) in Almansa area (Fig. 1). Therefore the fragment primary material origin could be related to outcrops located between Almansa and Finestrat on the Vinalopó inner area. Sample S4 from Pinós is located on the eastern slope of the saline dome of Cabezo de la Sal and seems to be very similar to the “FINESTRAT” samples. Sample S9 from Orihuela mountains (Murcia) contains lower level of REE Sc and Y compared to the other samples. Sample S10 from Altura (Castellón) outcrops is clearly different from the other samples due to its higher REE contents, and maybe because its geographical distance from the other outcrops. S16 is an archaeological fragment from Ereta del Pedregal archaeological simple located next to Navarres (Valencia). This sample cannot be related to any of the studied outcrop samples and is different than the other archaeological samples due to its higher REE, Sc and Y values. Two outliers preceding from Vinalopó (Santa Eulalia) outcrops have been found (S6 and S7). Those two have REE concentrations higher than samples S5 and S8 from the same outcrop. Samples S6 and S7 are positioned in PCA model closer to Almansa samples (S13, S11, and S12) than the other two samples (S5 and S8). A possible hypothesis is that this anomaly may be related to the model of exploitation quarries in Vinalopó, which have been exploited until few years and completely exhausted and maybe rocks from a different geochemical origin have been collected in this case.

Resuming, the results show that REE analysis enables to distinguish between different outcrops and also permit to reconstruct the origin of the primary material for the manufacturing of prehistoric tools.

3.3. XRF analysis of natural and archaeological stones As described above, samples were measured by XRF and the results obtained are reported in Table 7. Concentration values are expressed in percentage of oxides for Al2O3, SiO2, K2O and CaO and in terms of elements for Ti and Fe. Mean concentrations with their standard deviations have been reported confirming that the obtained results do not allow discriminate samples collected in different geographical areas. However we have employed Ti/Fe relationship [17] to detect differences between dolerite samples. In Fig. 3 it can be observed that the values of this parameters are very similar between all the samples except sample S10 (Altura) and S9 (Sierra Orihuela). In fact these two samples were collected in extreme north (Altura) and south (Orihuela) of the sampled area. So it can be seen that this Ti/Fe factor is able to discriminate sample provenance in a wider geographical areas compared to the REE values that can help to distinguish sample provenance from narrow range areas.

1

S10 ALTURA

0.9

S7 VINALOPÓ

0.8

S6 VINALOPÓ

0.7

S4 PINOSO

Ti

S5 VINALOPÓ

S16 ERETA ARCH

0.6

S8 VINALOPÓ S11 ALMANSA S15 BANCAL ARCH S1 FINESTRAT S13 ALMANSA S14 ARENAL ARCH S12 ALMANSA S2 FINESTRAT S3 FINESTRAT

S9 ORIHUELA

0.5 0.4 0.3 0.2 0.1 0

5

5.5

6

6.5

7

Fe Fig. 3. Ti/Fe relation.

7.5

8

8.5

174

G. Gallello et al. / Microchemical Journal 124 (2016) 167–174

3.4. Parameters to evaluate the regional origin of prehistoric dolerite stone objects

then developing provenance studies employing REE analysis inside previously defined regions.

Differences in REE values between the study samples seem to be related to the geographic origin. This relation should be verified and this geographical discrimination capability of these elements has been confirmed with the analysed samples and could be completed analysing additional natural rocks together with archaeological polished stones. The disadvantage of REE analysis is that to do the determination by ICP-MS the use of destructive techniques need to be carried out. On the other hand major element analysis don't allow us to observe fine nuances but just coarse differences between samples collected in the different parts of the studied region. If we have a look at Fig. 1, REE values allowed us to observe that the archaeological objects found at Bancal de Satorre and Arenal de la Costa archaeological sites were made by raw materials collected in quarries located between Almansa and Vinalopó (Sant Eulalia) outcrops. On the other side, major element analysis employing the correlation of Ti/ Fe just shows differences between Altura in the extreme north and Orihuela in the extreme south of the region.

Acknowledgements

4. Conclusions We have identified dolerite rock outcrops relating them with lithic archaeological samples through geochemical analysis of oxides, major element trace elements and rare earths, testing an experimental method for reconstructing networks and transports of prehistoric tools. REE have proven to be useful in relating the primary material of the quarries with the lithic objects, due probably to the particular characteristics of these chemical elements. In this first study, the REE seem to be good discriminators at regional level to relate quarries with prehistoric lithic objects found in the archaeological sites at regional level. We have been able to associate the dolerite rocks proceeding from natural outcrops with archaeological lithic objects coming from archaeological sites. Archaeological samples have been associated to natural outcrop samples for their similarity in REE, Sc, and Y contents. REE analysis has shown to have the potential to contribute to the reconstruction of exchange of dolerite lithic materials. On the other hand major elements specifically the relation between Ti/Fe may be used to determine differences between dolerite outcrops located in the extreme of the geographical studied areas. Therefore our method may be able to help to discern the regional origin of lithic objects belonging to the same classes of rocks (dolerite), bringing a new proposal in the reconstruction of dolerite primary material transports and exchange of lithic materials in prehistory. However future study should be developed, first employing major elements to discriminate outcrops in wider geographical areas and

Authors acknowledge the financial support of the PROMETEO project (II/2014/077) from the Generalitat Valenciana. References [1] G. Bertorino, M. Franceschelli, C. Luglié, M. Marchi, S. Columbu, Petrographic characterisation of polished stone axes from Neolithic Sardinia: archaeological implications, Per. Miner. 71 (2002) 87–100. [2] R.E. Bevins, N.J.G. Pearce, R.A. Ixer, Stonehenge rhyolitic bluestone sources and the application of zircon chemistry as a new tool for provenancing rhyolitic lithics, J. Archaeol. Sci. 38 (2011) 605–622. [3] R.E. Bevins, R.A. Ixer, P.C. Webb, J.S. Watson, Provenancing the rhyolitic and dacitic components of the Stonehenge landscape bluestone lithology: new petrographical and geochemical evidence, J. Archaeol. Sci. 39 (2012) 1005–1019. [4] J.W. Eerkens, J.R. Ferguson, M.D. Glascock, G.E. Skinner, S.A. Waechter, Reduction strategies and geochemical characterization of lithic assemblages: a comparison of three case studies from Western North America, Am. Antiq. 72 (2007) 585–597. [5] A.A. Evans, R.E. Donahue, The elemental chemistry of lithic microwear: an experiment, J. Archaeol. Sci. 32 (2005) 1733–1740. [6] A. Olofsson, Provenancing flint artefacts with ICP-MS using REE signatures and Pb isotopes as discriminants: preliminary results of a case study from northern Sweden, Archaeometry 53 (2011) 1142–1170. [7] M.S. Shackely, Gamma rays X-rays and stone tools: some recent advances in archaeological geochemistry, J. Archaeol. Sci. 25 (1998) 259–270. [8] M.S. Shackely, Archaeological petrology and the archaeometry of lithic materials, Archaeometry 50 (2008) 194–215. [9] R.H. Tykot, Chemical fingerprinting and source tracing of obsidian: the central Mediterranean trade in black gold, Acc. Chem. Res. 35 (2002) 618–627. [10] O. Williams-Thorpe, P.C. Webb, M.C. Jones, Non-destructive geochemical and magnetic characterisation of Group XVIII dolerite stone axes and shaft-hole implements from England, J. Archaeol. Sci. 30 (2003) 1237–1267. [11] T. Orozco, Aprovisionamiento e intercambio. Analisis petrológico del utillaje pulimentado en la prehistoria reciente del País Valenciano (España), British Archaeological Reports, I.S. 867, Oxford, 2000. [12] G. Gallello, A. Pastor, A. Diez, N. La Roca, J. Bernabeu, Anthropogenic units fingerprinted by REE in archaeological stratigraphy: Mas d'Is (Spain) case, J. Archaeol. Sci. 40 (2013) 799–809. [13] G. Gallello, A. Pastor, A. Diez, J. Bernabeu, Lanthanides Revealing Anthropogenic Impact Within a Stratigraphic Sequence, J. Archaeol. (2014)http://dx.doi.org/10.1155/ 2014/767085. [14] L. Arena, M. Ortega, M.J. García-Martínez, E. Querol, J.F. Llamas, Geochemical characterization of the mining district of Linares (Jaen, Spain) by means of XRF and ICPAES, J. Geochem. Explor. 108 (2011) 21–26. [15] C.N. Munksgaard, K. Lim, D. Livingstone, Rare earth elements as provenance indicators in North Australian estuarine and coastal marine sediments, Estuar. Coast. Shelf Sci. 57 (2003) 399–409. [16] J. Zhangdog, L. Fuchun, C. Junji, W. Sumin, Y. Jimin, Geochemistry of Daihai lake sediments, Inner Mongolia, north China: implications for provenance, sedimentary sorting, and catchment weathering, Geomorphology 80 (2006) 147–163. [17] M. Lago, E. Arraz, C. Galé, J. Bastida, Las doleritas Toleiticas Triásicas del sector SE de la Cordillera Ibérica: petrología y geoquímica, Estud. Geol. 55 (1999) 223–235.