Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes

Journal of Archaeological Science xxx (2017) 1e10

Contents lists available at ScienceDirect

Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas

Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes Wayne Powell a, b, c, *, Ryan Mathur d, H. Arthur Bankoff e, Andrea Mason a, b, c, c f, Linda Godfrey g Aleksandar Bulatovi
c f, Vojislav Filipovi
a

Department of Earth and Environmental Sciences, Brooklyn College, Brooklyn, NY, USA Earth and Environmental Sciences Program, CUNY Graduate Center, New York, NY, USA Department of Earth and Planetary Science, American Museum of Natural History, New York, NY, USA d Department of Geology, Juniata College, Huntingdon, PA, USA e Department of Anthropology and Archaeology, Brooklyn College, Brooklyn, NY, USA f Arheoloski Institut, Srpska Akademija Nauka i Umetnosti, Belgrade, Serbia g Department of Earth and Planetary Sciences, Rutgers University, Piscataway, NJ, USA b c

a r t i c l e i n f o

a b s t r a c t

Article history: Received 15 February 2017 Received in revised form 2 June 2017 Accepted 27 June 2017 Available online xxx

Southeastern Europe is the birthplace of metallurgy, with evidence of copper smelting at ca. 5000 BCE. There the later Eneolithic (Copper Age) was associated with the casting of massive copper tools. However, copper metallurgy in this region ceased, or significantly decreased, centuries before the dawn of the Bronze Age. Archaeologists continue to be debate whether this hiatus was imposed on early metalworking communities as a result of exhaustion of workable mineral resources, or instead a cultural transition that was associated with changes in depositional practices and material culture. Copper isotopes provide a broadly applicable means of addressing this question. Copper isotopes fractionate in the near-surface environment such that surficial oxide ores can be differentiated from non-weathered sulphide ores that occur at greater depth. This compositional variation is transferred to associated copper artifacts, the final product of the metallurgical process. In the central Balkans, a shift from 65Cu-enriched to 65Cu-depleted copper artifacts occurs across the metallurgical hiatus at the Eneolithic-Bronze Age boundary, ca. 2500 BCE. This indicates that the reemergence of metal production at the beginning of the Bronze Age is associated with pyrotechnical advancements that allowed for the extraction of copper from sulphide ore. Thus copper isotopes provide direct evidence that the copper hiatus was the result of exhaustion of near-surface oxide ores after one-and-a-half millennia of mining, and that the beginning of the Bronze Age in the Balkans is associated with the introduction of more complex smelting techniques for metal extraction from regionally abundant sulphidic deposits. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Bronze age Eneolithic Serbia Balkans Copper Isotopes Metallurgy

1. Introduction The extent to which humans manipulated and used Earth's rock and mineral resources forms the basis for the classic subdivisions of prehistory (stone, bronze, and iron ages). Here we focus on the transition of technologies associated with the smelting of copper ores and their relationship to the beginning of the Bronze Age in southeastern Europe. The copper mineral malachite (Cu2CO3[OH]2), with its intense green color, has been used to manufacture beads

* Corresponding author. Department of Earth and Environmental Sciences, Brooklyn College, Brooklyn, NY, USA. E-mail address: [email protected] (W. Powell).

since the dawn of agriculture 10,000 years ago (Bar-Yosef Mayer and Porat, 2008). However, the first instance of intentional extraction of copper from this mineral through pyrotechnology appears to have occurred at 5000 BCE by the Vinca culture of Serbia (Radivojevi
c et al., 2010). There the Eneolithic began with the production of small jewelry pieces and rapidly progressed to the casting of massive copper shaft-hole axes (Fig. 1) for approximately 1500 years (Jovanovi
c, 2009; Sava, 2015). However, copper metallurgy ceased, or dramatically decreased, in the mountains of southern Serbia and the adjacent Pannonian Plain (Vojvodina, Hungarian Plain, Romanian Banat) in the mid-fourth millennium (Sava, 2015), centuries before the Bronze Age began (O'Brien, 2014; Sherratt, 1997; Spasi
c, 2010; Tasi
c, 2003) (Fig. 2). A similar hiatus in

http://dx.doi.org/10.1016/j.jas.2017.06.012 0305-4403/© 2017 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Powell, W., et al., Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes, Journal of Archaeological Science (2017), http://dx.doi.org/10.1016/j.jas.2017.06.012

2

W. Powell et al. / Journal of Archaeological Science xxx (2017) 1e10

Fig. 1. Variation in form and size of typical axes from the Eneolithic to the Early Bronze Age of Serbia.

copper production occurred at the end of the Eneolithic and Early Bronze Age in southwest Spain (Rothenberg and Blanco-Freijeiro, 1981), and the Adige Valley of northeastern Italy during the Middle Bronze Age (Cierny et al., 1998). Whether any, or all, of these lulls in metal production were the result of a cultural shift or depletion of mineral resources due to extended periods of mining remains a subject of debate (Kienlin, 2014, 2016; O'Brien, 2014). Eneolithic cultures in Serbia appear to have relied on oxide ores: malachite fragments and beads are common in settlements of that age (Glumac and Trigham, 1990; Radivojevi
c and Kuzmanovi
cCvetkovi
c, 2014); sulphide minerals were intentionally sorted from oxide ores at Eneolithic mine sites (Jovanovi
c and Ottaway, 1976); rare Eneolithic slag fragments are compositionally consistent with malachite-bearing ore (Glumac and Todd, 1991; Radivojevi
c et al., 2010). Smelting of copper oxides is a relatively simple, potentially slagless process that can be conducted in a modified hearth-like structure (Craddock, 2001; O'Brien, 2014; Timberlake, 2007), as has been found in several Balkan sites (Radivojevi
c and Rehren, 2016; Rehren et al., 2016). The same pyrotechnology can be used to extract copper from mixed oxide-

tetrahedrite (fahlore) ore (Bourgarit, 2007). This form of copper production supplied much of the copper of central Europe in the € ppner et al., 2005), and may have been Early Bronze Age (Ho employed at European Eneolithic sites where fahlore ores ([Cu,Ag]6Cu4(Fe,Zn]2[As,Sb]4S13) were present (Bougarit, 2007). However, this is not the case in Serbia where such deposits are rare, and the vast majority of early copper artifacts lack the As-Sb-Agbearing composition that is indicative of a fahlore ore source. The production of copper from the far more plentiful, purely sulphidic ore (predominantly chalcopyrite; CuFeS2) is a more complex, multi-stage process that requires contrasting oxidation conditions, specialized furnaces, and the controlled inclusion of fluxing agents (Bourgarit, 2007; Craddock, 2001). The earliest definitive evidence for multi-stage chalcopyrite smelting in continental Europe is associated with the ore deposits of Mitterberg, €llner, 2009, Austria in the Middle Bronze Age (1600-1200 BCE) (Sto as cited in O'Brien, 2014; Pernicka et al., 2016) where “the scale of production achieved during the Middle Bronze Age was at a level that can arguably be described as industrial” (O'Brien, 2014, p.164). However, the presence of slag with copper matte inclusions from Eneolithic sites in Bulgaria (Ryndina et al., 1999) suggest that mixed malachite-chalcopyrite ores may have been exploited to some extent at that time. Whether such mixed ores were processed by repeated cycles of crushing and smelting (Bourgarit, 2007), or the first introduction of a two-step, roasting and smelting process (Ryndina et al., 1999) is uncertain. It has been speculated that the marked reduction in metal production in southeastern Europe in the latest Eneolithic may have been due to social forcings (Kienlin, 2013; Weninger et al., 2009), or alternatively, to the exhaustion of surficial oxidebearing ores and the technical inability to smelt the underlying sulphide minerals (Papalas, 2008; Sherratt, 1997). However, definitive archaeological evidence associated with metallurgical processes (e.g., slag, crucibles) at the Eneolithic-Bronze Age transition in the Balkans is exceedingly rare, and so this hypothesis has remained conjectural and disputed. Although “lowly slag is rarer than gold” (Papalas, 2008, p. 93), an estimated 4.7 tonnes of Eneolithic copper artifacts survive in the Balkan region (Pernicka et al., 1997). Advances in our understanding of copper isotopes and their variance with respect to copper ore mineral composition provides a new approach to investigating ancient mining practices through analysis of copper artifacts, the final product of the metallurgical process. Copper isotope analysis of 120 artifacts (Table 1) from across Serbia, and adjacent sites in Bosnia and the Romanian Banat (Fig. 3), spanning a 3500-year age range from the middle Eneolithic to the Early Iron Age, indicate that not only were local copper oxide ores depleted by the end of the Eneolithic, but that the critical pyrotechnological advancement of the smelting of

Fig. 2. Chronological framework for the study area comprising Serbia, including the Vojvodina and the contiguous Romanian Banat. Data compiled from Filipovi
c (2013), Kalafati
c troi (2013), Sava (2015), and Tasi
c (2003). (2006), Pa

Please cite this article in press as: Powell, W., et al., Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes, Journal of Archaeological Science (2017), http://dx.doi.org/10.1016/j.jas.2017.06.012

W. Powell et al. / Journal of Archaeological Science xxx (2017) 1e10

3

Table 1 Description and d65Cu composition of analyzed artifacts. Sample

Material

Description

Museum

Museum Number

Site

Age

del 65Cu

AR 01 AR 02 AR 06 AR 08 AR 13 AR 16 AR 18 AR 20 BI 01 BI 02 BI 03 BI 04 BI 05 BO 02 BO 03 BO 04 BO 05 BO 06 BO 11 BO 12 BO 13 BO 14  02 CA

Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Copper Copper Copper Copper Copper Copper Copper Copper Bronze

Pin Pin Wire Wire Saltaleone Wire Bracelet Bracelet Spearhead Spearhead Axe - Socketed Axe - Socketed Pin Pin Axe - Cross Axe - Cross Axe - Cross Axe - Cross Axe - Cross Needle Needle Needle Torque

Cx030 Cx008 Cx230 Cx084 Cx012 Cx047 Cx086

S¸agu S¸agu S¸agu S¸agu Pecica Pecica Pecica Pecica Batkovi
c Batkovi
c Batkovi
c Vrsinje Ku cerine Trnjane Voluja Bunar Seliste Veliki Gradac Donja Bela Reka Kmpije  Skjopuluji
Coka Lu Balas

LBA LBA LBA LBA MBA MBA MBA MBA LBA LBA LBA LBA LBA MBA Eneolithic Eneolithic Eneolithic Eneolithic Eneolithic Eneolithic Eneolithic Eneolithic MBA


0.14 0.35
0.15
0.27
0.58
0.21
0.19 0.11
0.33
0.39
0.60
0.27
0.05
0.44 0.90
0.16 0.86
0.09
0.24 0.33 0.44 0.16
0.31

 05 CA  06 CA DE 06 DE 09 JA 01 JA 02 JA 03 KI 01 KI 02 KI 07 KI 09 KI 10 KI 11 KI 12 KI 13 KI 14 KI 15 KI 16 KI 17 KI 18 KI 19 KI 20 KI 21 KI 22 KI 24 KI 26 KI 27 KI 28 KI 29 KI 30 KI 31 LE 11 LE 12 LE 13 LE 15 LE 16 LO 03 LO 10 LO 11 LO14 LO16 LO18 NI 02 NI 03 NI 05 NI 06 NI 07 NI 08 NI 09 NI 10

Bronze

Hair Ring

Bronze

Hair Ring

Arad, RO Arad, RO Arad, RO Arad, RO Arad, RO Arad, RO Arad, RO Arad, RO Bijelina, BH Bijelina, BH Bijelina, BH Bijelina, BH Bijelina, BH Bor, RS Bor, RS Bor, RS Bor, RS Bor, RS Bor, RS Bor, RS Bor, RS Bor, RS   Ca cak, RS   Ca cak, RS   Ca cak, RS

Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Copper Copper Copper Copper Copper Copper Copper Copper Copper Copper Bronze Bronze Bronze Bronze Bronze Copper Bronze Bronze Bronze Copper Bronze Bronze Bronze Bronze Copper Copper Bronze Bronze Bronze Bronze Bronze Bronze Copper Copper Copper

Bracelet Pin Bracelet Pin Knife Wire Knife Wire Knife Spiral Ear Ring Pin Knife Axe - Flat Axe - Flat Hair Accesory Pendant Ring Hair Accessory Bracelet Bracelet Pin Axe - Socketed Axe - Socketed Axe - Socketed Arrowhead Arrowhead Pin Pendant Pendant Arm band Wire Axe - Hammer Axe - Socketed Bracelet Melted Bronze Sword Axe - Cross Axe - Cross Belt Bead Buttons Hair Pin Axe - Flat Axe - Socketed Sword Needle Needle Needle

Deva, RO Deva, RO Jagodina, RS Jagodina, RS Jagodina, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Kikinda, RS Leskovac, RS Leskovac, RS Leskovac, RS Leskovac, RS Leskovac, RS Loznica, RS Loznica, RS Loznica, RS Loznica, RS Loznica, RS Loznica, RS Nis, RS Nis, RS Nis, RS Nis, RS Nis, RS Nis, RS Nis, RS Nis, RS

797 1518 2239 848 2191

I483

Dubac

I499

Dubac

MBA


0.20

I502

Dubac

MBA


0.14

C237 C035

S¸oimus¸ S¸oimus¸ Panjeva cki Rit Panjeva cki Rit Panjeva cki Rit Idjos Idjos Idjos Idjos Mokrin Ostoji
cevo Ostoji
cevo Ostoji
cevo Podlokanj Mokrin Mokrin Mokrin Mokrin Mokrin Mokrin Mokrin Mokrin Idjos Jablanji Jablanji Jablanji Jablanji Jablanji Ostoji
cevo Ma
cedonce Ma
cedonce Ma
cedonce Chance Find Hisar Paulje Paulje Paulje Detinji Potok Rmsko Groblje Paulje Medosevac Medosevac  Velika Humska Cuka Stani cenje Guvniste  Velika Humska Cuka Bunanj Bubanj

MBA MBA LBA LBA LBA LBA LBA LBA LBA EBA MBA MBA MBA Eneolithic EBA EBA EBA EBA EBA EBA EBA EBA LBA EIA EIA EIA EIA EIA MBA EIA EIA EIA Eneolithic LBA LBA LBA LBA Eneolithic Eneolithic EIA LBA LBA LBA LBA LBA Eneolithic Eneolithic Eneolithic

0.01
0.21
0.16
0.74 0.60
0.41
0.37
0.48
0.32
0.60
0.34
0.55
0.39 0.19 0.30 0.23
0.57
1.04
1.18 0.18
0.53
0.64
0.42
0.36
0.37
0.40
0.54
0.38
0.89
0.38
0.52
0.36
0.25 0.24
0.45
0.01 0.24 0.67 0.24
0.75
1.59 0.54
0.40
0.10
0.25 0.58 0.02 0.01

18,790

A00700

A01227 A11768 A00724 A02917 A02908 A02909 A02913 A02911 A02894 A02827 188

C34 1704 C73 C1067 C503

(continued on next page)

Please cite this article in press as: Powell, W., et al., Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes, Journal of Archaeological Science (2017), http://dx.doi.org/10.1016/j.jas.2017.06.012

4

W. Powell et al. / Journal of Archaeological Science xxx (2017) 1e10

Table 1 (continued ) Sample

Material

Description

Museum

Museum Number

Site

Age

del 65Cu

NI 11 NI 12 NI 13 NI 14 NI 15 PA 01 PA 02 PA 03 PA 04 PA 05 PA 06 PA 07 PA 08 PA 09 PO 16 PO 20 PO 21 PO 30 PO 31 PO 32 PO 33 PO 34 PO 35 PO 38 PO 39 PO 40 PO 41 RU 01 RU 02  01 SA

Copper Copper Copper Copper Copper Copper Copper Copper Bronze Bronze Bronze Bronze Copper Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Copper Copper Copper Copper Bronze Copper Bronze

Axe - Flat Axe - Hammer Axe - Hammer Axe - Cross Axe - Flat Wire Awl Axe - Cross Sawblade Sheet Saltaleone Pendant Axe - Cross Sword Knife Bracelet Spiral Pin Knife Bracelet Button Sword Axe - Socketed Ingot Ingot Axe - Flat Sickle Axe - Socketed Axe - Socketed Bracelet

1854 286 3523 3330 4490

Bubanj Sokobanja Trbovac Lazinja Stani cenje Jabuka Tri Hunke Jabuka Tri Hunke Jabuka Tri Hunke Gaj Gaj Gaj Gaj Bajbuk Chance Find Bradarac Bradarac Bradarac Vojilovo Ostava Vojilovo Ostava Vojilovo Ostava Vojilovo Ostava Vojilovo Ostava Vojilovo Ostava Suvi Do Suvi Do Suvi Do Selo Krupaja Budjanovci Budjanovci Metlik

EBA MBA MBA Eneolithic Eneolithic Eneolithic Eneolithic Eneolithic LBA LBA LBA LBA Eneolithic LBA LBA LBA LBA LBA LBA LBA LBA LBA LBA Eneolithic Eneolithic Eneolithic EBA Eneolithic LBA LBA


0.71
0.66
0.76 1.54 0.48 2.27
0.24
0.28 0.03
0.16
0.40
0.45 0.19
0.44
0.11
0.20
0.72
0.21
0.42
0.28
0.22
0.29
0.26
0.12
0.07 0.17
1.90 0.01
0.10
0.28

 02 SA  03 SA  06 SA

Bronze

Bracelet

Nis, RS Nis, RS Nis, RS Nis, RS Nis, RS Pan cevo, RS Pan cevo, RS Pan cevo, RS Pan cevo, RS Pan cevo, RS Pan cevo, RS Pan cevo, RS Pan cevo, RS Pan cevo, RS Po zarevac, RS Po zarevac, RS Po zarevac, RS Po zarevac, RS Po zarevac, RS Po zarevac, RS Po zarevac, RS Po zarevac, RS Po zarevac, RS Po zarevac, RS Po zarevac, RS Po zarevac, RS Po zarevac, RS Ruma, RS Ruma, RS  Sabac, RS  Sabac, RS

Bronze

Bracelet

Bronze

Axe - Flanged

 07 SA  08 SA  09 SA

Bronze

Spearhead

Bronze

Axe - Flanged

Bronze

SR 02 SR 04 SR 05 SR 06 SR 11 SU 02 SU 03 SU 07 VR 01 VR 02 VR 04

Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze Bronze

P2188


0.45

P2617

Mutnik Badanja Ka cer

LBA

 Sabac, RS  Sabac, RS  Sabac, RS

LBA


0.22

P2173

Banovo Polje

MBA


0.19

P0331

Dragojevac

EIA


0.46

P0393

Bela Reka

MBA


0.19

Axe - Flat

 Sabac, RS  Sabac, RS

P2172

MBA


0.20

Wire Hair Pin Bracelet Decorated Stick Torque Bracelet Spiral Knife Knife Knife Bracelet

Sremska Mitrovica, Sremska Mitrovica, Sremska Mitrovica, Sremska Mitrovica, Sremska Mitrovica, Subotica, RS Subotica, RS Subotica, RS Vrsac, RS Vrsac, RS Vrsac, RS

Bela Reka Pe cinci Pe cinci Pe cinci Pe cinci Dobrinci Gjurgen Gjurgen Gjurgen Vatin Vatin Vatin

LBA LBA LBA LBA LBA LBA LBA LBA MBA MBA MBA


0.22
0.10
0.38
0.47
0.83
0.29
0.40
0.28
0.33
0.25
0.98

P2189

RS RS RS RS RS

sulphidic ores ushered in the dawn of the European Bronze Age. 1.1. Late Eneolithic and early Bronze Age cultures of the central Balkans This study focusses on artifacts from sites across the mountainous terrain of central Serbia, the adjacent plains of the Vojvodina, and the contiguous Banat of western Romania (Fig. 3). The recent Eneolithic chronological synthesis of the Lower Mures¸ Basin by Sava (2015) includes the cultures of Romania, Hungary, and troi Serbia, and has been adopted herein, supplemented by Pa lcut¸a culture, and Filipovi
c (2013), Kalafati
c (2013) for the Sa (2006), and Tasi
c (2003) for subdivisions of the Bronze Age in Serbia. This chronological framework is summarized in Fig. 2. The terminology and the chronology, both relative and absolute, for the period that succeeds the Neolithic and precedes the Early Bronze Age in southeastern Europe are quite complex. In Serbia, the source for most of our samples, Eneolithic is the term of choice. for this roughly 2500-year period. For consistency, we have used the term “Eneolithic” throughout. Other scholars and scholarly traditions have also used the terms Copper Age, Kupferzeit, Chalcolithic,

1851 877 849 865 989

or Final Neolithic for this same time (Sava, 2015). The Eneolithic period in Serbia begins in the fifth, and spans the fourth and third millennia BCE (Sava, 2015). The late Neolithic/early Eneolithic of the mid-sixth to mid-fifth millennia in Serbia is the time of the development and flourit of the Vin ca culture (Bori
c, 2009; Chapman, 1981; Sava, 2015). These agriculturalists lived in settled villages on river terraces and steep slopes, in wattle and daub houses laid out in rows on intensively occupied, long-lasting stratified sites. The first copper-mining and copper-smelting people, they exploited the copper sources of east Serbia as early as 5000 BCE (Bori
c, 2009; Jovanovi
c, 1971; Radivojevi
c et al., 2010). Metal products of the younger Vin ca phase included large shafthole axe-adzes and chisels. Pottery was mostly black monochrome, with burnished surface and carinated shapes (Tringham and Krsti
c, 1990). Small ceramic female figurines attest to a complex ritual life. Their pottery, figurines, and metal artifacts are more widely known than any other Eneolithic culture. lcut¸a-Krivodol culThe material culture of the Bubanj Hum-Sa troi, 2013), resembled tural complex, which arose ca. 4300 BCE (Pa
r and that of the Vin ca. Along with the contemporaneous Tiszapolga Bodrogkeresztúr cultures of Hungary, these groups are associated

Please cite this article in press as: Powell, W., et al., Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes, Journal of Archaeological Science (2017), http://dx.doi.org/10.1016/j.jas.2017.06.012

W. Powell et al. / Journal of Archaeological Science xxx (2017) 1e10

5

Fig. 3. Archaeological sites associated with analyzed artifacts. Site names correspond to those listed in Table 1.

with increased production and distribution of copper into the early fourth millenium (Sava, 2015; Spasi
c, 2010) (Fig. 2). The subsequent arrival of the Cernavoda III-Boleraz cultures from the east into the northern Balkans (Benac, 1983; Gimbutas, 1965; Kapuran and Bulatovi
c, 2012; Spasi
c, 2010; Tasi
c, 1983a) ca. 3600 BCE (Sava, 2015), coincided with the end of the Bodrogkeresztúr (Fig. 2) (Sava, 2015) and is followed by the appearance of the Baden, Kostalac, Cot¸ofeni, and Vu cedol cultures (Kapuran and Bulatovi
c, 2012). The material culture of this period includes new burial customs (tumulus burial), and new pottery shapes with a wide variety of incised decoration (Benac, 1983) which are interpreted as different “cultures” or “groups” (Bankoff and Winter 1990; Garasanin, 1973; Tasi
c, 1983b, 1983c). Aside from the ceramic differences, and possible observable geographic concentrations, the cultures of the latest Eneolithic in the northern Balkans are very similar (Tasi
c, 1983b). This may be, in part, due to the incomplete archaeological record and paucity of exposure in sites of this period, which makes distinctions other than ceramic styles difficult. Alternatively, it may reflect a true similarity among them. Pottery shapes indicate a shared heritage from the middle Eneolithic Baden culture (Dimitrijevi
c, 1983);

wattle and daub architecture is common in the few excavated late Eneolithic occupation levels; usually fortified settlements, some in dominant hilly positions, others on tell-like sites of earlier occupations, such as Gomolava (Petrovi
c and Jovanovi
c, 2002) or Bubanj (Bulatovi
c and Kapuran, 2017). Of greatest significance to this study, copper artifacts are rare in sites of this period, regardless of ceramic style which identify them as Cernavoda III, or Kostolac, or Cot¸ofeni or Vu cedol (Spasi
c, 2010; Tasi
c, 1983b, 1983c). The few metal objects that exist are small, including needles, awls, decorative wire spirals for hair ornamentation, and rare flat axes (Fig. 3) (Antonovi
c, 2014; Jovanovi
c, 1971; Spasi
c, 2010). The Serbian Early Bronze Age begins at 2500 BCE with the end of the Vu cedol (Tasi
c, 2003), and the appearance of the VinkovciSomodyvar culture (Kalafanti
c, 2006; Tasi
c, 1984, 2003). Extensive flat inhumation cemeteries such as Mokrin near Kikinda (Giri
c, 1971) and tumulus burials like Beloti
c in western Serbia (Garasanin, 1983a) provide some insight into the artifacts and societies of this period (O'Shea, 1996). Stratified settlements like Gomolava indicate the contemporaneity of the ceramic styles (Tasi
c, 2006). Although most of the metal artifacts (knot-headed pins, triangular daggers, small flat axes) from Early Bronze Age Mokrin are still made of

Please cite this article in press as: Powell, W., et al., Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes, Journal of Archaeological Science (2017), http://dx.doi.org/10.1016/j.jas.2017.06.012

6

W. Powell et al. / Journal of Archaeological Science xxx (2017) 1e10

copper (Giri
c, 1971), artifacts of tin-bronze appear at this time (Antonovi
c, 2014; Kienlin, 2013). The contemporaneous Vinkovci, Somodyvar, and Bubanj-Hum III cultures share the paucity of metal objects, the positioning of their rare settlements on river terraces or dominant heights, an assumed transhumant stock raising economy, and a basically similar inventory of ceramic shapes (Garasanin, 1983b). Except rarely in the Bubanj-Hum III context, there is a complete lack of ceramic decoration, certainly none of the deeply incised decoration that typifies the latest Eneolithic (KostolacCot¸ofeni-Vu cedol) pottery (Benac, 1983). It is not until the Middle Bronze Age that metal objects, now predominantly tin-bronze, reappear in abundance.

1.2. The Eneolithic copper hiatus in Serbia The onset of the hiatus in copper production in the late Eneolithic is synchronous with an appreciable cultural shift that may reflect the influx/influence of new population elements from the east. Accordingly, one school of thought on the cause of the metal hiatus is that metal use and deposition was one aspect of rapidly changing material and non-material culture (e.g., Kienlin, 2013), possibly connected with an incursion of new people (Tasi
c, 1983b). Others correlate the collapse of copper metallurgy in southeast Europe with social change initiated by an interval of rapid climate change at 4000-3200 BCE (Weninger et al., 2009; Weninger and Harper, 2015). Alternatively, depletion of workable oxide ores could have forced the cessation of large-scale copper smelting on the late Eneolithic people. Given that metallurgy has been cited as an important factor in shaping the structure of prehistoric society (e.g., O'Brien, 2014; O'Shea, 1996; Sherratt, 1993), the abrupt end of mining and metal working might be expected to engender large cultural changes. This argument is accepted in regard to the beginning of metallurgy, and is equally valid for the resumption of metal production after a hiatus of several centuries. A key distinction between the two hypotheses presented above is the nature of the ore when mining resumed in the Bronze Age. If the metallurgical hiatus was due to depletion of oxide ore, then

metallurgical activities would have resumed with the extraction and smelting of sulphide ores. In contrast, if diminished metal use was a cultural choice, then oxide ore reserves would have remained, and would have been the first ores mined when metal production resumed in the region. The behavior of copper isotopes in near-surface environments allows us to differentiate oxide ores that occur at Earth's surface from non-weathered sulphide ores that occur at greater depth (Mathur and Fantle, 2015), and so may be used to test these two alternative scenarios.

1.3. Copper isotopes and weathering processes Copper has two isotopes (65 and 63); 65Cu tends to favor stronger bonding environments associated with oxidation (Zhu et al., 2002). Thus, the oxidation of copper generates fluids and associated minerals that are enriched in the 65Cu isotope, as seen in watersheds impacted by weathering of sulphide minerals (Borok et al., 2008; Pokrovsky et al., 2008). Non-weathered sulphide ores lack appreciable fractionation (<1‰) (Larson et al., 2003). Thus, oxidative weathering of sulphide ores leads to the development of stratified isotopic reservoirs for copper depending on the climate and migration of mass. In desert regions the migration of the water table results in significant vertical mobilization of metal and overall enrichment of copper at depth, leading to the development of supergene ores. Due to their economic importance, supergene deposits have been studied extensively, and associated models have been applied to archaeometallurgical studies (e.g., Kienlin, 2016; O'Brien, 2014; Strahm and Hauptmann, 2009). In such deposits the following mineralogical and isotopic stratification develops (Fig. 4): 1) a surface leached zone of consisting Cu-bearing Fe-oxides that represent the residues of highly weathered copper sulphides, and which is depleted in 65Cu (Mathur and Fantle, 2015); 2) a highly fractionated (values of 3‰) enriched zone of copper oxides and sulphide minerals (bornite, chalcocite, covellite) which develop at the water table; and 3) non-fractionated, non-weathered primary sulphide ore below the water table (Asadi et al., 2015; Mathur et al.,

Fig. 4. Schematic model of for weathering and isotopic stratification of copper sulphide ores under arid (migrating water table) and humid (stationary water table) conditions. Generalized d65Cu values are based on a starting composition of 0‰ in the primary ore. Mineral abbreviations: Az ¼ azurite; Bn ¼ bornite; Cc ¼ chalcocite; Cov ¼ covellite; Cp ¼ chalcopyrite; Cup ¼ cuprite; Hm ¼ hematite; Lim ¼ limonite; Mal ¼ malachite; Py ¼ pyrite.

Please cite this article in press as: Powell, W., et al., Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes, Journal of Archaeological Science (2017), http://dx.doi.org/10.1016/j.jas.2017.06.012

W. Powell et al. / Journal of Archaeological Science xxx (2017) 1e10

2009, 2010; Mirnejad et al., 2010). Well-established supergene models can be applied to copper ores from hyper-arid regions such as the southwest United States and Chile. However, supergene models are not applicable to the study of ores from the Balkan region where stable temperate and moist conditions have prevailed since the beginning of the Holocene (11,300 BP) (Bordon et al., 2009), other than during the rapid climate change event at 8300-7900 BP when cooler, drier steppelike conditions occurred (Aufgebauer et al., 2012; Bordon et al., 2009). Therefore, prior to their exploitation in the Eneolithic, Balkan copper ores would have weathered in a humid environment, and consequently, copper sulphides would have oxidized above a static water table. These conditions produce copper oxides near the surface, and result in only minor enrichment of copper at the water table. This in situ weathering leads to the following isotopic stratification that is distinctly different than that of supergene ores (Fig. 4): 1) oxides above the water table that are enriched in 65Cu (values of about 1e2‰); 2) residual weathered sulphides minerals at the water table that are depleted in 65Cu; and 3) non-weathered and non-fractionated (values of about 0‰) sulphide ore below the

7

water table (Aseal et al., 2012). In some ores where weathering is incomplete, malachite may occur with remnant sulphide assemblages above the water table. Rostocker et al. (1989) suggested that copper could have been extracted from such ores with the same one-step reduction process that was used for pure malachite ores. Alternatively, reactions with sulphur during the smelting of mixed ores could produce a copper matte, from which copper metal could be liberated by re-smelting of the crushed matte concentrate (Bourgarit, 2007). In both cases the isotopic composition of the metal produced would result from the mixing of the two isotopic reservoirs of copper, that is, between 0‰ and 2‰ depending on the relative proportions of malachite and chalcopyrite in the charge. Copper isotopic analysis provides no direct insight into the specific location from which copper ore was extracted (provenance). Rather, it identifies the nature of the minerals that comprised the ore, and aspects of the climate in which copper ores weathered. Given the behavior of Cu isotopes, it is expected that in the Balkans the transformative shift to metallurgy based on purely sulphidic ore would be delineated by a significant decrease in d65Cu

Fig. 5. Variation in d65Cu in Eneolithic to early Iron Age copper and bronze artifacts from Serbia and the surrounding region. The distinctly higher d65Cu values from the Eneolithic indicate mining and smelting of surface oxidized ore (malachite). Lower d65Cu values from the Bronze and Iron ages demonstrate the use of unweathered sulphide ores.

Please cite this article in press as: Powell, W., et al., Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes, Journal of Archaeological Science (2017), http://dx.doi.org/10.1016/j.jas.2017.06.012

8

W. Powell et al. / Journal of Archaeological Science xxx (2017) 1e10

in copper artifacts corresponding to the use of the 65Cu-depleted sulphide ore that occurs at the water table, or the first primary sulphide ore. Importantly, it has been demonstrated that neither ancient nor modern techniques for the processing of copper ores fractionate copper, and so the copper isotope composition of metallic artifacts corresponds to that of the ores from which they were made (Gale et al., 1999; Mathur et al., 2009). For any given deposit, the copper isotope values in all three reservoirs are controlled by the starting composition of the nonweathered ore, degree of weathering, and mineral species being weathered. This degree of variability from numerable ore deposits would likely result in the overlap of copper isotope composition between populations of artifacts. Therefore, shifts in the mean copper isotope values and associated standard deviations would best reflect changes in ores use. 2. Methods Details of sample preparation and measurement are fully documented in Mathur et al. (2005). Of significant importance for the analysis of copper artifacts, surficial oxidation can lead to inconsistent copper isotope measurements as demonstrated in the differences between the rinds and cores of native copper artifacts from Keweenaw, Michigan (Mathur et al., 2010). Therefore, surface patinas were removed from metal artifacts prior to sampling by grinding with a small carbide-tip drilling tool. A sample was then cut or ground from the clean section of the artifact. To further negate the surficial oxidation signal, at least 0.2 g of metal were dissolved for analysis. A total of 120 artifacts were obtained from the collections of museums in Serbia, eastern Bosnia, and western Romania, spanning approximately 3500 years from the Late lcut¸a-Krivodol) to the Early Iron Age Eneolithic (Bubanj Hum-Sa (Table 1). Artifacts were dissolved in 5 ml of heated (100  C) aqua regia (3:1 ultrapure 11 M HCl and 16 N HNO3) in a 15 ml telfon jar for 4 h. A small aliquot was taken from the jar and diluted to 100 ppb for isotopic analysis. No ion exchange chromatography was needed as copper was >80% of all ions present in solution. The solutions were measured on a Thermo Scientific Neptune Plus multicollector ICPMS at Rutgers University. The instrument was in low-resolution mode with on-peak blank subtraction, and 30 ratios were measured for each sample. Copper isotope values are reported relative to the NIST 976 standard, and mass bias is controlled with standard sample bracketing. Errors for each measurement were calculated to be ±0.09‰, (n ¼ 26, 2s) by monitoring the variation of the NIST 976 through four different sessions. An in-house standard, a USA penny from 1838 reported in Mathur et al. (2009), overlapped previous reported values at d65Cu ¼
0.02 ± 0.04‰, n ¼ 5, 2s. 3. Results The two most thoroughly studied Eneolithic mine sites in the Balkans, Rudna Glava and Aibunar, are both associated with near surface oxidized, epithermal vein systems (Cernych, 1978; Kienlin, 2014). Accordingly, a reasonable starting copper isotope composition for high-temperature ores in the region can be estimated from the mean value of
0.2‰ ± 0.5 (1s) from 164 published measurements from chalcopyrite and bornite from 8 epithermal and massive sulphide deposits. This value serves as a baseline for the sources of sulphide-derived copper and a comparison point to infer the composition of weathered copper surface oxides. Twenty-two (88%) of Eneolithic artifacts (n ¼ 25) have d65Cu values greater than this, whereas eight (73%) of the Early Bronze age artifacts (n ¼ 11) yield compositions less than
0.2‰ (Fig. 5). The mean of

Middle Bronze Age, Late Bronze Age and Early Iron Age (n ¼ 84) cluster near
0.2‰ (Fig. 5). A one-way ANOVA was performed with d65Cu as the dependent variable across five time-period groups: Eneolithic, Early Bronze Age, Middle Bronze Age, Late Bronze Age, and Early Iron Age. There was a significant difference between groups (F[4122] ¼ 9.622, p < 0.001), and a priori planned post hoc LSD contrasts found that d65Cu levels were higher in the Eneolithic than in all other groups, with a trend for Early Bronze Age to be lower than Late Bronze Age (p ¼ 0.067). 4. Discussion Isotopic compositions indicate that copper sources for Eneolithic artifacts were derived from surface occurrences of copper mineralization, with these artifacts displaying the largest variation and highest mean value of d65Cu. Multiple studies demonstrate that near surface oxidation of copper minerals can lead to variable copper isotope values and the generation of Cu oxides that are isotopically enriched in 65Cu. For instance, the oxidation of chalcopyrite ore generates fluids that are 1.5‰ higher than the starting chalcopyrite (Kimball et al., 2009; Pokrovsky et al., 2008). Given that reported chalcopyrite values from similar copper deposits is
0.2‰, oxide minerals derived from in situ transport of copper would be approximately 1‰. Only artifacts from the Eneolithic have values in this range. Therefore, both archaeological observations (Glumac and Todd, 1991; Jovanovi
c and Ottaway, 1976; Radivojevi
c et al., 2010), and isotopic composition of copper artifacts indicate that surface deposits composed of oxides of copper or mixtures of Cu-oxides and Cu-sulphides were mined extensively during the Eneolithic, and that such ores were used exclusively to manufacture copper artifacts in Serbia at that time. Furthermore, the 65Cu values indicate that these weathered ores used in the Eneolithic Balkans developed under humid conditions, consistent with the Balkan climate throughout the Holocene. Extraction of near-surface copper oxides and mixed oxidesulphide assemblages would have ultimately exposed the zone of residual, 65Cu-depleted weathered sulphides. Oxide-based smelting techniques would have failed to produce metal from this ore. With the regional exhaustion of oxide ore, metal production would have ceased until local cultures acquired the pyrotechnology associated with sulphide smelting (Mitterberg process). According to the isotopic data, that transformative step occurred in the Early Bronze Age. Metals from the Middle Bronze Age through the Early Iron Age were extracted predominantly from larger tonnage nonweathered sulphide ores, as is evident from the tighter clustering and nearly identical means of the copper isotope values around
0.3‰ (Fig. 5). None of the values from the Late Bronze Age to Iron Age artifacts display significant fractionation above 1‰ (Fig. 5). Such values correspond to copper sourced from hydrothermal mineralization that was not impacted by surface weathering. Rare finds of slag from the later Bronze Age in Romania are consistent with the isotopic data, indicating that sulphide smelting was an established practice by this time in the study area (Papalas, 2008). Furthermore, the lack of 65Cu-enriched artifacts in the Bronze Age suggests there was little or no recycling of Eneolithic metal, and that there was no apparent import of copper from distant regions where weathered ores continued to be extracted well into the Bronze Age. The redox state of copper ores that were extracted in Balkan prehistory, as indicated by the Cu isotopic composition of artifacts, are consistent with available archaeological evidence, that is, oxidized ore (malachite) was mined in the Eneolithic, and sulphides (chalcopyrite-bornite) were the predominant ore minerals mined by the Middle Bronze Age. The shift from 65Cu-enriched to 65Cudepleted copper in artifacts across the Eneolithic-Bronze Age

Please cite this article in press as: Powell, W., et al., Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes, Journal of Archaeological Science (2017), http://dx.doi.org/10.1016/j.jas.2017.06.012

W. Powell et al. / Journal of Archaeological Science xxx (2017) 1e10

boundary at 2500 BCE indicates that accessible near-surface oxide ore reserves were depleted after approximately one-and-a-half millennia of mining, and that the beginning of the Bronze Age in the Balkans corresponded to the acquisition of pyrotechnology approximately 1000 years later. This allowed for the extraction of metals from ores composed solely of sulphide minerals (i.e., absence of malachite) and the resumption of copper mining activity in the region. The hiatus in copper production at the end of the Eneolithic and the resumption of copper smelting from chalcopyrite ore, together with the appearance of common tin-bearing bronze in the Early Bronze Age and the sustained tin-bronze production thereafter, argue for the transference of more advanced metallurgical lore, possibly from the Near East where it was discovered approximately 1000 years earlier (Heyd, 2013), or from another European culture which may have already used sulphide ores as subsidiary sources for copper in the Eneolithic (Kienlin, 2016; O'Brien, 2014). Given that the Cu isotope-based investigative approach is based on the copper artifacts themselves, which are abundant, it can be applied globally to fully characterize the Eneolithic-Bronze Age transition, the transfer of ancient pyrotechnology, and the general characterization of copper-ore-type (oxide, supergene, unweathered) used for ancient metal production. Acknowledgments We wish to thank the Institute for Aegean Prehistory Renewal Research Grant Program and the PSC-CUNY Research Award program (67704-00 45) for financial support for this project. This research would not have been possible without the help and hospitality of our colleagues in the museums of Serbia, the Vojvodina region, and Romania. We acknowledge with thanks: Snje zana Anti
c  (Bijelina); Moma Cerovi
c (Sabac); Catalin Critescu (Deva); Jelena   Djordjevi
c (Pan cevo); Katarina Dmitrovi
c (Ca cak); Rada Gligori
c (Loznica); Dragan Ja canovi
c (Po zaravac); Miroslav Jesreti
c (Sremska Mitrovica); Dragan Jovanovi
c (Vrsac); Igor Jovanovi
c (Bor); Smilja Jovi
c (Leskovac); Lidija Milasinovi
c (Kikinda); Slobodan Miti
c (Nis); Sonja Peri
c (Jagodina); Victor Sava (Arad); and Agnes Szekeres (Subotica). We also thank two anonymous reviewers whose comments and insights improved the manuscript. References € Antonovi
c, D., 2014. Kupferzeitlichen Axte und Beile in Serbien. Pr€ ahistorische Bronzefunde. In: Akademie der Wissenschaften und der Literatur, Vol 9, Band 27 Mainz. Asadi, S., Mathur, R., Moore, F., Zarasvandi, A., 2015. Copper isotope fractionation in the Meiduk porphyry copper deposit, Northwest of Kerman Cenozoic magmatic arc. Iran. Terra Nova 27, 36e41. Asael, D., Matthews, A., Bar-Matthews, M., Harlavan, Y., Segal, I., 2012. Tracking redox controls and sources of sedimentary mineralization using copper and lead isotopes. Chem. Geol. 310e311, 23e35. Aufgebauer, A., Panagiotopoulos, K., Wagner, B., Schaebitz, F., Viehberg, F., Vogel, H., Zanchetta, G., Sulpizio, R., Leng, M., Damaschke, M., 2012. Climate and environmental change in the Balkans over the last 17 ka recorded in sediments from Lake Prespa (Albania/F.Y.R. of Macedonia/Greece. Quat. Int. 274, 122e135. Bankoff, H.A., Winter, F., 1990. The later Aeneolithic in southeastern Europe. Amer. J. Arch. 94, 175e191. Benac, A., 1983. In: Praistorija jugoslavenskih zemalja, vol. 3 (Eneolit). Akademija Nauka I Umjetnosti Bosne i Hercogovine, Sarajevo.
zine, A., Brewer, S., Fouache, E., 2009. Pollen-inferred late Bordon, A., Peyron, O., Le glacial and holocene climate in southern Balkans (Lake Maliq). Quat. Int. 200, 19e30. Bori
c, D., 2009. Absolute dating of metallurgical innovations in the Vin ca culture of the Balkans. In: Keinlin, T., Roberts, B. (Eds.), Metals in Society, Studies in Honour of Barbara S. Ottaway, Bonn, pp. 191e245. Borrok, D.M., Nimick, D.A., Wanty, R.B., Ridley, W.I., 2008. Isotopic variations of dissolved copper and zinc in stream waters affected by historical mining. Geochim. Cosmo. Acta 72, 329e344. Bourgarit, D., 2007. Chalcolithic copper smelting. In: La Niece, S., Hook, D., Craddock, P. (Eds.), Metals and Mines. Archetype Publications, London,

9

pp. 3e13. Bulatovi
c, A., Kapuran, A., 2017. The Carpathian Basin and the Northern Balkans between 3500 and 2500 BC: common aspects and regional differences. , Annales Universitatis Apulensis: Series Historica, 20/II.  Cernych, E.N., 1978. Aibunarea Balkan copper mine of the fouth millenium BC. Proc. Prehist. Soc. 44, 203e217. Chapman, J., 1981. In: The Vin ca Culture of South-east Europe: Studies in Chronology, Economy and Society, vol. 117. Oxford,BAR. Cierny, J., Marzatico, F., Perini, R., Weisgerber, G., 1998. Prehistoric copper metallurgy in the southern Alpine region. In: Mordant, C., Pernot, M., Rychner, V.  cle Avant Notre Ere. (Eds.), L'Atelier du Bronzier en Europe du XX au VIII Sie Actes du Colloque International Bronze ’96, vol. II, pp. 25e34. Paris. Craddock, P.T., 2001. From hearth to furnace: evidences for the earliest metal smelting technologies in the eastern Mediterranean. Paleorient 26, 151e165. Dimitrijevi
c, S., 1983. Vu cedolska kultura i vu cedolski kulturni kompleks. In: Benac, A. (Ed.), Praistorija jugoslavenskih zemalja, Vol. 3 (Eneolit). Akademija Nauka I Umjetnosti Bosne i Hercogovine, Sarajevo, pp. 267e342. Filipovi
c, V., 2013. New researches of the Late Bronze Age necropolises in northwestern Serbia region: chronological and terminological questions. J. Serb. Arch. Soc. 29, 51e83. Gale, N.H., Woodhead, A.P., Stos-Gale, Z.A., Walder, A., Bowen, I., 1999. Natural variations detected in the isotopic composition of copper: possible applications to archaeology and geochemistry. Int. J. Mass. Spectrom. 184, 1e9. Garasanin, M., 1973. Praistorija na tlu S.R. Srbije. Srpska knji zevna zadruga, Beograd. Garasanin, M., 1983a. Grupa beloti
c-bela crkva. In: Benac, A. (Ed.), Praistorija jugoslavenskih zemalja, Vol. 4 (Bonzano Doba). Akademija Nauka I Umjetnosti Bosne i Hercogovine, Sarajevo, pp. 705e718. Garasanin, M., 1983b. Podunavsko-Balkanski kompleks ranog bronzanog doba. In: Benac, A. (Ed.), Praistorija jugoslavenskih zemalja, Vol. 4 (Bonzano Doba). Akademija Nauka I Umjetnosti Bosne i Hercogovine, Sarajevo, pp. 463e470. Gimbutas, M., 1965. Bronze Age Cultures in Central and Eastern Europe. De Gruyter, The Hague. Giri
c, M., 1971. In: Mokrin; Nekropola ranog bronzanog doba, vol. 1. Smithsonian, Washington. Glumac, P.D., Todd, J.A., 1991. Early metallurgy in southeast Europe: the evidence for production. Recent trends. In: Glumac, P.D. (Ed.), Archaeometallurgical Research, MASCA Research Papers in Science and Archaeology. University of Pennsylvania, Philadelphia, pp. 9e19. Glumac, P., Tringham, R.E., 1990. The exploitation of copper minerals. In: Tringham, R.E., Krsti
c, D. (Eds.), Selevac, a Neolithic Village in Yugoslavia. University of California Press, Los Angeles, pp. 549e563. Heyd, V., 2013. Europe 2500 to 2200 BC: Between Expiring Ideologies and Emerging Complexity. In: Fokkens, H., Harding, A. (Eds.), The Oxford Handbook of the European Bronze Age. University Press, Oxford, pp. 47e67. €ppner, B., Barthelheim, M., Huijsmans, M., Krauss, R., Martinek, K.-P., Pernicka, E., Ho Schwab, R., 2005. Prehistoric copper production in the Inn valley (Austria), and the earliest copper in central Europe. Archaeometry 47, 293e315. Jovanovi
c, B., 1971. Metalurgija eneolitskog perioda. Srpska Akademija Nauka i Umetnosti, Beograd. Jovanovi
c, B., 2009. Beginning of the metal age in the central Balkans according to the results of the archeometallurgy. J. Min. Metal. B Metal. 45, 143e148. Jovanovi
c, B., Ottaway, B.S., 1976. Copper metallurgy and mining in the Vin ca group. Antiquity 50, 104e113. Kalafati
c, H., 2006. A Vinkovci culture urn grave from the site at 40 Duga Ulica in Vinkovci. Pril. Inst. Arheol. Zagrebu 23, 17e28. Kapuran, A., Bulatovi
c, A., 2012. Kulturna grupa Kocofeni-Kostolac na teritoriji Isto cne Srbije. Starinar 62, 1e19. Kienlin, T., 2013. Copper and bronze: bronze age metalworking in context. In: Fokkens, H., Harding, A. (Eds.), The Oxford Handbook of the European Bronze Age. University Press, Oxford, pp. 414e436. Kienlin, T., 2014. Aspects of metalworking and society from the Black Sea to the Baltic Sea from the fifth to the second millennium BC. In: Roberts, B., Thorton, C. (Eds.), Archaeometallurgy in Global Perspective: Methods and Syntheses. Springer, Heidelberg, pp. 447e472. Kienlin, T., 2016. Some thoughts on evolutionist notions in the study of early metallurgy. In: Bartelheim, M., Horejs, B., Krauß, R. (Eds.), Von Baden Bis Troia: Ressourcennutzung, Metallurgie Und Wissenstransfer. Verlag Marie Leidorf, pp. 123e138. Kimball, B.E., Mathur, R., Dohnalkova, A.C., Wall, A.J., Runkel, R.L., Brantley, S.L., 2009. Copper isotope fractionation in acid mine drainage. Geochim. Cosmo. Acta 73, 1247e1263. Larson, P.B., Maher, K., Ramos, F.C., Chang, Z., Gaspar, M., Meinert, L.D., 2003. Copper isotope ratios in magmatic and hydrothermal ore-forming environments. Chem. Geol. 201, 337e350. Mathur, R., Fantle, M.S., 2015. Copper isotopic perspectives on supergene processes: implications for the global Cu cycle. Elements 11, 323e329. Mathur, R., Ruiz, J., Titley, S., Liermann, L., Buss, H., Brantley, S.L., 2005. Cu isotopic fractionation in the supergene environment with and without bacteria. Geochim. Cosmo. Acta 69, 5233e5246. Mathur, R., Titley, S., Hart, G., Wilson, M., Davignon, M., Zlatos, C., 2009. The history of the United States cent revealed through copper isotope fractionation. J. Arch. Sci. 36, 430e433. Mathur, R., Dendas, M., Titley, S., Phillips, A., 2010. Patterns in the copper isotopic composition of minerals in porphyry copper deposits in the Southwestern United States of America. Econ. Geol. 105, 1457e1467.

Please cite this article in press as: Powell, W., et al., Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes, Journal of Archaeological Science (2017), http://dx.doi.org/10.1016/j.jas.2017.06.012

10

W. Powell et al. / Journal of Archaeological Science xxx (2017) 1e10

Bar-Yosef Mayer, D.E., Porat, N., 2008. Green stone beads at the dawn of agriculture. Proc. Natl. Acad. Sci. U.S.A. 105, 8458e8551. Mirnejad, H., Mathur, R., Einali, M., Dendas, M., Alirezaei, S., 2010. A comparative copper isotope study of porphyry copper deposits in Iran. Geochem. Expl., Environ. Anal. 10, 413e418. O'Brien, W., 2014. Prehistoric Copper Mining in Europe 5500-500 BC. Oxford University Press, Oxford. O'Shea, J.M., 1996. Villagers of the Maros: A Portrait of an Early Bronze Age Society. University of Michigan Press, Ann Arbor Michigan. Papalas, C.A., 2008. Bronze Age Metallurgy of the eastern Carpathian Basin. Unpublished PhD thesis. Arizona State University. Pernicka, E., Begemann, F., Schmitt-Strecker, S., Todorova, H., Kuleff, L., 1997. Prehistoric copper in Bulgaria: its composition and provenance. Eurasia Antiq. 2, 41e180. €llner, T., 2016. Bronze Age copper produced at Mitterberg, Pernicka, E., Lutz, J., Sto Austria, and its distribution. Archaeol. Austriaca 100, 19e55. Petrovi
c, J., Jovanovi
c, B., 2002. Gomolava. In: Settlements of the Late Eneolithic, vol. 4. Matica Srpska, Novi SadeBelgrade. Pokrovsky, O.S., Viers, J., Emnova, E.E., Kompantseva, E.I., Freydier, R., 2008. Copper isotope fractionation during its interaction with soil and aquatic microorganisms and metal oxy(hydr)oxides; possible structural control. Geochim. Cosmo. Acta 72, 1742e1757. lcut¸a Eneolithic culture. Ann. d’Universite
Valahia P atroi, C.N., 2013. About the Sa Targoviste, Sect. d’Archeologie d’Histoire 15, 117e140. Radivojevi
c, M., Kuzmanovi
c-Cvetkovi
c, J., 2014. Copper minerals and archaeometallurgical materials from the Vin ca culture sites of Belovode and Plo cnik: overview of the evidence and new data. Starinar 64, 7e30.  Radivojevi
c, M., Rehren, T., Pernicka, E., Slijvar, D., Brains, M., Bori
c, D., 2010. On the origins of extractive metallurgy: new evidence from Europe. J. Arch. Sci. 37, 2775e2787. Radivojevi
c, M., Rehren, T., Black, Paint It, 2016. The rise of metallurgy in the Balkans. J. Arch. Meth. Theory 23, 200e237. Rehren, T., Leshtakov, P., Penkova, P., 2016. Reconstructing Chalcolithic copper smelting at Akladi cheiri, Chernomorets, Bulgaria. In: Nikolov, V., Schier, W. (Eds.), Der Schwarzmeerraum vom Neolithikum bis in die Früheisenzeit (6000e600 v.Chr.). Kulturelle Interferenzen in der zirkumpontischen Zone und Kontakte mit ihren Nachbargebieten. Rahden/Westf.: Verlag Marie Leidorf GmbH. Rostocker, T., Piggott, V., and Dvorak, J., Direct reduction of copper by oxide-sulfide mineral interaction. Archaeomaterials, 3, 69e87. Rothenberg, B., Blanco-Freijeiro, A., 1981. Studies in Ancient Mining and Metallurgy in South-west Spain. Institute of Archaeolometallurgical Research, London. Ryndina, N., Indenbaum, G., Kolosova, V., 1999. Copper production from polymetallic sulphide ores in the northeastern Balkan Eneolithic culture. J. Arch. Sci. 26, 1059e1068. Sava, V., 2015. Neolithic and Eneolithic in the Lower Mures¸ Basin. Mega, ClujNapoca.

Sherratt, A., 1993. Who are you calling peripheral? Dependence and independence in European prehistory. In: Healy, F., Scarre, C. (Eds.), In Trade and Exchange in Prehistoric Europe. Oxbow, Oxford, pp. 145e255. Sherratt, A.H., 1997. Economy and Society in Prehistoric Europe: Changing Perspectives. Princeton University Press, Princeton. Spasi
c, M., 2010. Cot¸ofeni communities at their southwestern frontier and their relationship with Kostolac population in Serbia. Dacia 56, 157e175. € llner, T., 2009. Die zeitliche Einordnung der pra €historischen Montanreviere in Sto den Ost- und Südalpen: Anmerkungen zu einem Forschungsstand. In: Oeggl, K., Prast, M. (Eds.), Die Geschichte des Bergaus in Tirol und Seinen Angrenzenden Gebieten. Innsbruck University Press, Innsbruck, pp. 37e60. Strahm, C., Hauptmann, A., 2009. The metallurgical developmental phases in the old world. In: Kienlin, T.L., Roberts, B.W. (Eds.), Metals and Societies. Studies in €tsforschungen zur pr€ Honour of Barbara S. Ottaway. Universita ahistorischen Arch€ aologie 169, Bonn, pp. 116e128. Tasi
c, N., 1983a. Jugoslovensko podunavlje od Indoevropske seobe do prodora Skita. Matica Srpska, Beograd. Tasi
c, N., 1983b. Cot¸ofeni kultura. In: Benac, A. (Ed.), Praistorija jugoslavenskih zemalja, Vol. 3 (Eneolit). Akademija Nauka I Umjetnosti Bosne i Hercogovine, Sarajevo, pp. 115e128. Tasi
c, N., 1983c. Kostola cka kultura. In: Benac, A. (Ed.), Praistorija jugoslavenskih zemalja, Vol. 3 (Eneolit). Akademija Nauka i Umjetnosti Bosne i Hercogovine, Sarajevo, pp. 183e234. Tasi
c, N., 1984. Kulturen der Frühbronzezeit des Karpatenbeckens und Nordbalkans. Balkanoloski Institut SANU, Beograd. Tasi
c, N., 2003. Historical picture of development of bronze age cultures in Vojvodina. Starinar 53, 23e34. Tasi
c, N., 2006. Bronze and iron age sites in Srem and the stratigraphy of Gomolava. Balkanica 36, 7e16. Timberlake, S., 2007. The use of experimental archaeology, archaeometallurgy for the understanding and reconstruction of early bronze age mining and smelting. In: La Niece, S., Hook, D., Craddock, P. (Eds.), Metals and Mines. Archetype Publications, London, pp. 27e36. Tringham, R., Krsti
c, D., 1990. Selevac, a Neolithic Village in Yugoslavia. University of California, Institute of Archaeology, Los Angeles. Weninger, B., Harper, T., 2015. The geographic corridor for rapid climate change in €ol Eurasien 31, 475e505. Southeast Europe and Ukraine. Archa € hner, U., Budja, M., Bundschuh, M., Weninger, B., Lee, C., Rohling, E., Bar-Yosef, O., Bo €ris, O., Linst€ Feurdean, A., Gebe, H.G., Jo adter, J., Mayewski, P., Mühlenbruch, T., Reingruber, A., Rollefson, G., Schyle, D., Thissen, L., Todorova, H., Zielhofer, C., 2009. The impact of rapid climate change on prehistoric societies during t! he Holocene in the Eastern Mediterranean. Doc. Praehist. 36, 7e59. Zhu, X.K., Guo, Y., Williams, R.J., O'Nions, R.K., Matthews, A., Belshaw, N.S., Canters, G.W., de Waal, E.C., Weser, U., Burgess, B.K., Salvato, B., 2002. Mass fractionation processes of transition metal isotopes. Earth Plan. Sci. Lett. 200, 47e62.

Please cite this article in press as: Powell, W., et al., Digging deeper: Insights into metallurgical transitions in European prehistory through copper isotopes, Journal of Archaeological Science (2017), http://dx.doi.org/10.1016/j.jas.2017.06.012