Petro-mineralogy and geochemistry as tools of provenance analysis on archaeological pottery: study of Inka Period ceramics from Paria, Bolivia

(This is a sample cover image for this issue. The actual cover is not yet available at this time.)

This article appeared in a journal published by Elsevier. The attached copy is furnished to the author for internal non-commercial research and education use, including for instruction at the authors institution and sharing with colleagues. Other uses, including reproduction and distribution, or selling or licensing copies, or posting to personal, institutional or third party websites are prohibited. In most cases authors are permitted to post their version of the article (e.g. in Word or Tex form) to their personal website or institutional repository. Authors requiring further information regarding Elsevier’s archiving and manuscript policies are encouraged to visit: http://www.elsevier.com/copyright

Author's personal copy

Journal of South American Earth Sciences 36 (2012) 1e17

Contents lists available at SciVerse ScienceDirect

Journal of South American Earth Sciences journal homepage: www.elsevier.com/locate/jsames

Petro-mineralogy and geochemistry as tools of provenance analysis on archaeological pottery: Study of Inka Period ceramics from Paria, Bolivia V. Szilágyi a, f, *, J. Gyarmati b, M. Tóth c, H. Taubald d, M. Balla e, Zs. Kasztovszky a, Gy. Szakmány f a

Dept. of Nuclear Research, Institute of Isotopes, Hungarian Academy of Sciences, 29-33 Konkoly-Thege Str., Budapest H-1121, Hungary Museum of Ethnography, 12 Kossuth Sq., Budapest H-1055, Hungary c Institute of Geochemical Research, Hungarian Academy of Sciences, 45 Budaörsi Str., Budapest H-1112, Hungary d Isotope Geochemistry, University of Tübingen, 56 Wilhelmstr., Tübingen D-72074, Germany e } egyetem Rakpart, Budapest H-1111, Hungary Inst. of Nuclear Techniques, Budapest University of Technology and Economics, 9 Mu f Dept. of Petrology and Geochemistry, Institute of Geography and Earth Sciences, Eötvös Loránd University, 1/c Pázmány stny., Budapest H-1118, Hungary b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 18 February 2011 Accepted 15 November 2011

This paper summarized the results of comprehensive petro-mineralogical and geochemical (archeometrical) investigation of Inka Period ceramics excavated from Inka (A.D. 1438e1535) and Late Intermediate Period (A.D. 1000/1200e1438) sites of the Paria Basin (Dept. Oruro, Bolivia). Applying geological analytical techniques we observed a complex and important archaeological subject of the region and the era, the cultural-economic influence of the conquering Inkas in the provincial region of Paria appearing in the ceramic material. According to our results, continuity and changes of raw material utilization and pottery manufacturing techniques from the Late Intermediate to the Inka Period are characterized by analytical methods. The geological field survey provided efficient basis for the identification of utilized raw material sources. On the one hand, ceramic supply of both eras proved to be based almost entirely on local and near raw material sources. So, imperial handicraft applied local materials but with sophisticated imperial techniques in Paria. On the other hand, Inka Imperial and local-style vessels also show clear differences in their material which suggests that sources and techniques functioned already in the Late Intermediate Period subsisted even after the Inka conquest of the Paria Basin. Based on our geological investigations, pottery supply system of the Paria region proved to be rather complex during the Inka Period. Ó 2011 Elsevier Ltd. All rights reserved.

Keywords: Petrography Geochemistry Archaeological application Inka period pottery Bolivia

1. Introduction Aspects of geology and archaeology are similar because both investigate the limited and sporadic remnants of the past. Archaeological finds can be entirely natural materials (e.g. stone tools, building stones, bone artefacts) or artificial objects made with a technique which can be modelled with natural geological processes (e.g. ceramics, metal objects, glass finds, textile). On the one hand, pottery can be considered as a metasedimentary rock which is a result of an artificial technique: natural or non-natural mixture of sediments which is metamorphosed by

* Corresponding author. Dept. of Nuclear Research, Institute of Isotopes, Hungarian Academy of Sciences, 29-33 Konkoly-Thege Str., Budapest H-1121, Hungary. Tel.: þ36 1 392 2222x3214, fax: þ36 1 392 2584. E-mail addresses: [email protected] (V. Szilágyi), [email protected] (J. Gyarmati), [email protected] (M. Tóth), [email protected] (H. Taubald), [email protected] (M. Balla), [email protected] (Gy. Szakmány). 0895-9811/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jsames.2011.11.001

a low pressure e high temperature (firing) process. Hence, classical geological investigations are powerful methods to characterize the materials applied for ceramic making, to determine whether these materials could be natural or not, to establish the efficient field survey for potential raw material sources and to model the possible mixtures of the constituents. One the other hand, pottery is the most abundant find in the archaeological excavations and its appearance (vessel shape, style) is highly influenced by the cultural-social changes. Thus, getting detailed information about the raw materials and manufacturing techniques of such archaeological artefacts can help to better understand the social processes taking place in a certain region and era. The application of geological methodology for answering archaeological questions is the topic of the archaeometrical research. In Inka Period (A.D. 1438e1535) archaeological excavations from Ecuador to Chile there can be found an easily identifiable ceramic style, the Inka Imperial which is manufactured with the quality of Cuzco Inka ware (Rowe, 1944). This type of pottery is characterized with the same quality and standard (concerning

Author's personal copy

2

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

shape and decoration) in every Inka site through the Empire. This type of archaeological finds is one of the examples of the Inka standardizing intentions which can be traced in the whole Inka material culture (e.g. architecture, metalworking or textiles). It is a principal question of the Inka archaeology how this standardization took place in the colonized territories of the huge empire. Were these Inka Imperial style ceramics manufactured in some specially controlled workshops (Morris, 1978; D’Altroy, 1992) or all over the empire by local potters who were taught to follow the imperial standards (Hayashida, 1998)? To answer these questions a heterogeneous set of archaeological (ceramic, building materials) and comparative geological samples was investigated. Our aim was to clarify three interconnected questions corresponding to the principal objective of the analysis: (1) whether the raw material of the ceramics belonging to the Inka style is similar to or different from the raw material of Late Intermediate Period pottery, i.e. the raw material used for pottery making changed in the Inka Period or were the same sources used as in the former period; (2) whether the Inka style ceramics of the Paria Basin were made from local raw material; (3) whether the material of the vessels representing Inka Imperial and Inka local style is different. The studies of ceramic collections from individual Inka period sites similar to the research detailed in this paper are mainly focused on the coastal Peru region (Ixer and Lunt, 1991; Hayashida, 1999; Hayashida et al., 2002, 2003a, b; Velde and Druc, 1999a; Costin, 2001) or northwest Argentina (Bertolino and Fabra, 2003; Ratto et al., 2002, 2004, 2005; Plá and Ratto, 2003). A comprehensive study on special types of ritual ceramics from all around the Empire was also published (Bray et al., 2005). Some isolated researches on Inka pottery are known from Ecuador (Jamieson and Hancock, 2004) and Chile (Alden et al., 2006). However, the ceramic manufacture of the Inka provincial region of the present Bolivia has yet been investigated preliminary from archaeometrical point of view (Williams et al., 2006). This paper deals with the reconstruction of pottery manufacture of the Late Prehispanic Paria Basin of the Bolivian Altiplano investigated by the Paria Archaeological Project (PAP) in 2004e2006. The archaeometrical investigation of pottery from Paria started in 2004. Our preliminary results on petrographic observations and mineralogical analysis (Szilágyi et al., 2005, 2007; Szilágyi and Szakmány, 2009) provided a basis for the comprehensive study. The main goals of the here presented examinations are the petromineralogical and geochemical description of the ceramic artefacts (to provide fundamental raw data on Inka period pottery from Bolivia), and the identification of their possible provenance and technological characteristics by taking comparative geological samples into the investigation.

The special role of the Paria Basin appears clearly in the ethnohistoric sources (Vaca de Castro, 1908: 435) as the imperial Inka road e which started from Cuzco and passed along the western side of Lake Titicaca e bifurcated at the ancient site of Paria with the main road heading towards Chile/Argentina and another branch towards the Valley of Cochabamba (Gutiérrez Osinaga, 2005) and further to Incallacta (Hyslop, 1984; Guaman Poma de Ayala, 1980). At this intersection, the Inka state established an administrative centre, one of the capitals of the provinces to the south from Cuzco (Fig. 1). According to Cieza de León (1973, 1985), there were great buildings (depositories and lodgings for the Inka and the Temple of the Sun) to be constructed in the settlement. Other ethnohistorical accounts indicate that maize grown in the imperial state farms in the Valley of Cochabamba was transported through ancient Paria (Repartimiento, 1977). In this regard, the Paria Basin had a double function in the Inka Period. It formed the border between the different ecological levels and e at the same time e acted as an intermediate zone in the redistribution network of the Inka Empire. The PAP studied the interregional ecological relationships between the Andean Altiplano and the temperate valleys, as well as the political and economic interactions that existed between the Inka imperial centre and the peripheral zones. Until now, there are only unpublished project reports that discuss the archaeological records and

2. Background 2.1. Archaeological context The Paria Basin is located approximately 200 km to the southeast of La Paz, Bolivia at an altitude of 3700e3800 metres above m.s.l. The basin is situated at a strategically important geographical location on the margin of the Sierra de Azanaques. This range of the mountainous Andes forms a natural separation between the immense Altiplano and the more temperate valleys of Cochabamba and the eastern tropical foothills of the Andes. At this point, the Paria Basin forms an ecological doorway between the different zones and ecological levels. In this way, the territory played a vital role in the life of Andean ethnic groups and in the political economy of the Inka Empire, as was suggested in John Victor Murra’s investigations (1972, 1983).

Fig. 1. Location of Paria site in the Inka Empire (map is modified after Gyarmati and Varga, 1999), the satellite picture (source: Google Earth, 2008), sketch of the Ce 1 site and the view of structure BH (photo by J. Gyarmati).

Author's personal copy

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

conclusions. Archaeological systematic surface survey extended to a 95.5 km2 part of the Paria Basin, and located 113 sites ranging from the Formative Period (B.C. 1000eA.D. 600) until the Colonial Period (A.D. 1535e1825). The most important and extended archaeological site of the basin is Paria, an Inka administrative centre (its code is No.1 on the map). According to 14C dating of charcoal samples from one of the objects, the Inka conquest could happen at about 1400e1420. This date indicates an earlier occupation of the southern territory by Inkas than the classical archaeological interpretation. The ceramic sherds analyzed in this paper originated from the surface collection of the identified sites and the interior of a structure excavated at Paria in 2005.

3

caldera. Their appearance is very heterogenous concerning their grain size, consolidation/cementation, porosity and fabric/structure. However, their collective feature is that they are potentially suitable for pottery making either as paste (finer grained, clayey sediments) or as temper (coarser grained, silty-sandy sediments) raw materials (to favourably modify the physical properties of the plastic paste during shaping, drying and firing ceramics). All of them are related to former lacustrine, fluvial or recent eolian processes (GEOBOL, 1992, 1994). In the field, these formations appear in variously incised valleys (Fig. 2) or seasonally waterflooded plain areas. 3. Methods

2.2. Geological setting The known geological history of the Paria Basin and its surrounding started in the Palaeozoic. There are Palaeo-Mesozoic siliciclastic sedimentary rocks (Ordovician-Devonian shalessiltstones-sandstones, Jurassic-Cretaceous sandstones) in the area as products of flysch formation. Subsequently, PaleogeneeNeogene clastic sedimentary and volcanic rocks (volcanites-pyroclastics and ores connecting to subduction related Miocene effusive volcanic activity) and finally Quaternary sediments (fluvial and lacustrine clay-silt-sand-gravel) were formed (GEOBOL, 1992, 1994). From the point of view of our investigation, the most important formations for possible raw materials of the pottery are the Silurian shales-siltstones-sandstones, the Miocene pyroclastics and volcanites and those Quaternary clayey sediments which contain the fragments of the above mentioned rock types in large quantity or have the apparently adequate physical properties (clay contentplasticity, grain size distribution) for pottery making. In a proper sense, sediments of the Paria Basin have their direct source only from the Silurian siliciclastic sedimentary sequence since this is the only material which can be eroded within 5 km from the main archaeological site on the edges of the alluvial basin. The relatively monotonous sequence of the Silurian grey shalessiltstones-sandstones was subjected to a low grade metamorphism (GEOBOL, 1992, 1994). In the field, the Silurian rocks form eroded ridges built of stratified (weakly folded) structures (Fig. 2). The thick, cyclical sequence (alternating finer and coarser grained layers) forms a relatively homogeneous source for the alteration products. The Miocene pyroclastics and volcanites connect with two main structures and activities. The Morococala Volcanic Field is situated w15 km to the southeast from Paria and is in connection with the basin area through rivers (it is at the upper course, near the beginning of the main branch of Jacha Uma river). The complex was formed during the Miocene (25e5 Ma ago) (Morgan et al., 1998) and resulted in welded and non-welded rhyodacitic tuffs (ignimbrite) and lavas which form an extremely eroded plateau (Fig. 2). The Soledad caldera is on the plain of the Altiplano, w30 km to the west from Paria, the archaeological site. Its structure is a lowvolume, nonresurgent “ash-flow”-type caldera (Redwood, 1987) which was formed during the Miocene (15e5 Ma ago). The interbedded, air-fall and non-welded ash-flow, dacitic tuffs and dacitic lavas build up the present remnants (rims) of the collapsed caldera structure which rises above the plain of the Altiplano (Fig. 2). There is no permanent fluvial contact to the Paria Basin presently. However, the Quaternary lacustrine environment could possibly connect the two areas when different sediments could be mixed. The investigated Quaternary sediments are from the Paria Basin (with an erosional area of the Silurian sedimentary rocks and the Morococala Volcanic Field) and the surrounding of the Soledad

In order to describe and compare the archaeological and geological samples, microscopic petrographic investigation (Dept. of Petrology and Geochemistry, Eötvös Loránd University of Budapest), X-ray powder diffraction analysis (Inst. of Geochemical Research, HAS, Budapest), instrumental neutron activation analysis (Inst. of Nuclear Techniques, Budapest University of Technology and Economics), X-ray fluorescence geochemical analysis (Dept. of Geochemistry, University of Tübingen) and prompt gamma activation analysis (Dept. of Nuclear Research, Inst. of Isotopes, HAS, Budapest) were used. The microscopic petrographic investigations were carried out on a Nikon ALPHAPHOT-2 polarizing microscope. The mineral phase analyses were done on a Philips PW 1730 diffractometer with a Bragg-Brentano alignment and graphite monochromator (other parameters: CuKa radiation, 45 kV tension, 35 mA intensity, 0.05e0.01 2Q step size, 1 s time constant). The chemical measurements provided concentrations for 11 major and 29 trace elements. Since the element data detected with various methods have overlapped in the case of the different samples the most sensitive method was used for certain elements in each case. The XRF and PGAA provided the major element concentrations (SiO2, TiO2, Al2O3, Fe2O3, MnO, MgO, CaO, Na2O, K2O, P2O5, H2O/LOI). Concerning the trace elements XRF results were applied for Rb, Sr, Ba, Zr, Nb, Y, V and Zn. For most of the other trace elements the INAA measurements were used (Th, U, Hf, Ta, La, Ce, Nd, Sm, Eu, Yb, Lu, Sc, Cr, Co, As, Sb, Cs), while PGAA provided B, Cl and Gd concentrations. If one of the methods could not be applied on a sample, the data were gained by the second most sensitive measurement. The INAA measurements were realized in the pool-type reactor of the Budapest University of Technology and Economics, Hungary. The samples were irradiated with a thermal neutron flux of 2.4  1012 n cm2s1 for 8 h. Gamma-spectrometric measurements were performed by an HPGe Well-type detector (resolution 1.95 keV, relative efficiency 20.5%). For the evaluation of spectra, Sampo 90 software was used. Standardization was made by the single-comparator method (De Corte, 1987), using gold as comparator. The thermal/epithermal flux-ratio was monitored by zirconium foils. The accuracy of the measurements was monitored by analysing samples of the NBS SRM 1633a Coal Fly Ash Standard Reference Material. The XRF analyses were done with a wavelength dispersive X-ray fluorescence analyser (Bruker AXS S4 Pioneer X-ray spectrometer, Rh tube at 4 kW, Hahn-Weinheimer et al., 1984) on homogenized fused beads with 1.5 g of dried (at 1050  C over night) sample powder mixed with 7.5 g of Li2B4O7. Loss on ignition was determined at 1050  C externally and is displayed as LOI. Analytical error and detection limits vary and depend on element and sample composition. The samples were measured using an internal rock calibration curve with 35 international standards, compiled in Govindarau (1989).

Author's personal copy

4

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

Fig. 2. Geological sketch of the Paria Basin and the eastern rim of the Altiplano with the caldera of Soledad. Stars mark the archaeological sites, while circles show the geological sampling points. Photos show the view of the main geographic units of the territory.

The PGAA facility of the Budapest Neutron Centre is operated on an external cold neutron beam at the 10 MW Budapest Research Reactor, Hungary (5  107 cm2s1 flux, for detailed description see Révay and Belgya, 2004; Lindstrom and Révay, 2004; Révay et al., 2008). Gamma-ray spectra were measured using a calibrated HPGe detector (Belgya and Révay, 2004; Fazekas et al., 1999; Molnár et al., 2002). For the spectrum evaluation, the Hypermet PC software was used (Phillips and Marlow, 1976; Fazekas et al., 1997; Révay et al., 2001a, 2005). The quantitative analysis is based on

the k0 principle (Molnár et al., 1998). The element identification was performed using the spectroscopic data libraries developed at the Institute of Isotopes, HAS (Révay and Molnár, 2003; Révay et al., 2000, 2001b; Choi et al., 2007; Révay et al., 2004). The composition was determined following the method described in Révay (2009), the uncertainties of the concentration values were determined according to GUM (1993) and Révay (2006). Since this method is able to measure hydrogen, H2O content (which is not equal with LOI) of the samples could be calculated.

Author's personal copy

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

4. Analyzed materials In the framework of our research 206 ceramic fragments (144 pieces from the excavation, 62 pieces collected from the surface), 8 building material remnants (4 adobes and 4 building stones) and 50 comparative geological samples were investigated (Table 1). 4.1. Archaeological samples The investigated archaeological finds derived from different stages of the archaeological sampling work (field survey and excavations). The sample set consists of 62 sherds collected at sites from the Formative to the Inka period (Fig. 2), while the majority of the analyzed potsherds (144 pieces) were excavated from a structure of Paria (site No.1). The selected ceramic material belonged to the Inka Imperial, pre-Inka local and transitional styles (Fig. 3) (transitions among stylistic groups are poorly observed, e.g. Bray, 2004). Inka Imperial ceramics (Fig. 3a) are manufactured with the quality of Cuzco Inka ware (Rowe, 1944), which is of the same quality and standard (concerning shape and decoration) in the whole Empire. Inka pacaje ceramics (Fig. 3d), which are thought to be regional variants of the Inka Imperial style, basically are bowls with special llama decoration. Pre-Inka local style (Fig. 3c) means a vessel type of the Late Intermediate period (LIP, A.D. 1000/1200e1438) utilized also in the Inka period. Transitional styles (Inka local (Fig. 3b) and mixed (Fig. 3e) style) follow a fashion mixing the imperial and local characteristics concerning the form and decoration but their quality never reach that of the Inka Imperial style. The assemblage also contained some fragments of a special white ware which shows regional rather than imperial characteristics. Classifying the excavated ceramics according to the vessel forms, the major types are the jars, bowls, plates, cooking pots and big dishes of coarse material, while there are only some fragments of pukus (straightside bowls) and queros (ceremonial cups). For the excavated Inka period structures, stone, dried and unfired brick (adobe) and clayey paste as plaster/daub were utilized as building materials. To sample the surely local materials used in the Inka period, aside from ceramic fragments, four adobe samples (No. 4/3-6) were taken from an adobe structure and four fragments of building stones (No. 4/1-2, 4/7-8) from another structure. 4.2. Geological samples Simultaneously with archaeological finds, 50 comparative sediments and hard rocks were also collected. Based on the above presented knowledge on the geological setting, the main targets of the sampling were the potential sources of the clay paste and siltsand sized tempering component raw materials for the pottery making process. These were the fine-grained alluvial sediments of the Paria Basin (22 samples: No. SED/01-04, No. 1/1-3, 2/1, 5/1-8, 7/ 1-2, 7/4-7) and the eastern rim of the Altiplano (4 samples from present brick clay mines: CLAY/01-04), the Silurian shale-siltstonesandstone ridges of the Eastern Cordillera (6 samples: No. 3/1-2a2b, 6/8, 7/3, 7/8) and the pyroclastics-volcanites of the south eastern Morococala Volcanic Field (2 samples: No. 6/4-5) and the western Soledad caldera (16 samples: No. 8/1-6 with subsamples). See Fig. 2 for the location of the sampling points. 5. Results 5.1. Petrographic investigations Petrographic examinations were carried out on almost the complete sample collection (249 thin sections). The basic

5

descriptive parameters were grain size distribution, shape, roundness, sphericity of grains, fabric and orientation, quantity and composition of aplastic inclusions, porosity and optical behaviour of the paste. The properties of geological samples comparable to that of the ceramics were completed with the description of the mineral and lithic fragments and the partly vitrified matrix in the case of volcaniclastics and volcanites. Based on the petrographic classification of pottery, it was possible to make a comparison to the potential raw materials e especially the tempering clasts. The main discriminative phenomenon for ceramics was the aplastic inclusion composition, but for detailed subgrouping, the fabric and grain size distribution was also taken into consideration. The investigated 206 ceramic fragments can be classified into three main petrographic groups: (I) pyroclastic and volcanic originated, (II) sedimentary rock originated and (III) metamorphic rock originated. Group I has predominance in the collection with 106 samples. It has 20e30% aplastic inclusion content which is mainly volcaniclastic/volcanic rock originated and its fine-grained paste is well sorted, weakly anisotropic or isotropic and sometimes the optical activity changes in bends. Further division of group I (four subgroups) is based on the difference in the quality of the volcaniclastic (I/A with pumice, I/B with glass shard fragments), volcanic rock (I/C) and mineral (I/D without rock) grains (Fig. 4aed)(see detailed qualitative and quantitative description in Szilágyi and Szakmány, 2009). Subgroup I/A contains fine (50 mm) to coarse grained (300 mm), hiatal fabric ceramics dominantly with low sphericity, angular to subangular, unweathered pumiceous tuff rock and mineral fragments. The ceramic paste is a micaceous, fine grained silt with variable optical behaviour (from homogeneous anisotropic and red through sandwich structure to homogeneous isotropic and grey paste). The pumiceous clasts are fresh, angular grains. Subgroup I/ A/a can be characterized as a fine grained ceramic group with 50e100 mm dominant pumice size and it contains quartz, plagioclase, similar quantities of coarse grained (w120 mm) biotite and fine grained (w30 mm) hornblende (rarely orthopyroxene) crystals and occasionally volcanite or siltstone fragments. Subgroup I/A/b is a coarser grained ceramic type with 100e275 mm dominant pumice size and with plagioclase, quartz, biotite (w700 mm), very rare hornblende or orthopyroxene, and in some cases volcanite, metamorphic quartzite or siltstone fragments. Accessories appear in limited amount in both subgroups (opaque minerals), in addition secondary phases (limonite) are present. Subgroup I/B collects fine (50e75 mm) to coarse grained (150e175 mm), hiatal fabric ceramics dominantly with pyroclastic, very angular, low sphericity, unweathered glass shard and pumiceous tuff rock (15e30 w%) and mineral fragments. The micaceous, fine silty paste is anisotropic and red or isotropic. Subgroup I/B/ a contains fine grained ceramics with almost only bone-shaped glass shards (with 50e75 mm dominant glass shard size) and felsic components (plagioclase and quartz). Subgroup I/B/b is a coarser grained (with 125e175 mm dominant glass shard size) pottery type mainly with bone-shaped glass shards and pumice (with biotite, hornblende and plagioclase phenocrysts) and subordinately siltstone fragments. Subgroup I/B/c is similarly a coarser grained (with 125e175 mm dominant glass shard size) ceramic type but with the predominance of bone-shaped glass shards and siltstone clasts and minor content of pumice (with biotite). Subgroup I/ B/d has glass shards as major constituents but with a less typical shape (not bone- but irregular shaped), the additional pumice clasts have only biotite content. Accessory minerals (opaque phases, rare titanite) are rare in all of the subgroups. Subgroup I/C gathers medium grained (100e125 mm) pottery with volcanic rock fragments as temper. This subordinate subgroup

Author's personal copy

6

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

Table 1 Chemical composition of the investigated samples. Results are gained by the combination of different methods: XRF data are with normal, PGAA with italics, INAA with underlined letters. Sample Ceramics P/12.20. I.19.8. I.20.18. I.29.1. PA/4 PA/6 P/12.17. P/34.88. P/92.4. I.10.24. I.16.24. I.15.7. I.20.6. P/34.1. P/34.2. P/34.143. I.14.6. I.15.1. BA/1 P/1.166. P/33.5. P/48.101. I.30.5. P/1.186. I.16.1. P/12.13. P/42.9. P/53.79. I.10.68. I.20.14. I.20.20. PA/1 P/1.21. P/53.108. P/63.7. 41.125. I.7.18. I.13.9. PA/5 Adobes 4/3 4/4 4/5 4/6 Sediments CLAY-01 CLAY-02 CLAY-03 CLAY-04 SED-01 SED-02 SED-03 SED-04 2/1 5/2

Type

SiO2

TiO2

Al2O3

Fe2O3

MnO

MgO

CaO

Na2O

K2O

P2O5

LOI(H2O)

Sum(%) ppm

Rb

Sr

Ba

Th

U

Zr

Hf

I/A/a I/A/a I/A/a I/A/a Average of I/A/b I/A/b I/A/b I/A/b I/A/b I/A/b I/A/b Average of I/B/a I/B/a I/B/b I/B/b I/B/b I/B/b I/B/b Average of I/C/a I/C/a I/C/a I/C/a I/C/a Average of I/C/b I/C/b Average of I/D I/D Average of II/A II/A II/A II/A Average of II/B or C II/C II/B II/B II/B II/C II/B Average of III

64.00 59.48 61.00 60.29 61.19 62.00 62.21 63.00 61.04 55.45 62.65 60.95 61.04 67.78 59.64 63.36 62.41 61.65 64.57 64.11 63.36 58.00 57.99 59.42 58.75 59.71 58.77 60.36 61.35 60.85 63.27 61.23 62.25 61.09 60.06 57.73 57.31 59.05 64.00 58.08 58.80 58.84 61.48 61.16 60.57 60.42 56.00

0.85 0.75 0.82 0.83 0.81 0.64 0.73 0.62 0.77 0.86 0.80 0.88 0.76 0.59 0.57 0.68 0.68 0.67 0.68 0.71 0.65 0.97 0.76 0.97 0.73 0.78 0.84 0.79 0.87 0.83 0.73 0.76 0.74 0.79 0.82 0.75 0.74 0.78 0.94 0.78 0.80 0.75 0.82 0.79 0.69 0.80 1.18

18.80 19.45 18.89 18.67 18.95 19.40 17.98 18.70 18.55 22.08 20.22 22.38 19.90 17.11 17.04 17.57 19.01 17.92 18.77 19.21 18.09 21.00 18.41 17.75 19.25 18.66 19.01 18.64 20.89 19.76 18.32 17.62 17.97 20.82 21.53 21.95 21.73 21.51 19.00 22.03 20.65 22.57 21.39 22.04 19.70 21.05 28.70

6.70 6.78 6.31 6.27 6.52 4.80 4.95 5.10 5.13 6.40 5.39 5.92 5.38 4.79 5.19 5.29 5.85 5.27 5.36 5.54 5.33 7.10 6.27 6.35 6.31 6.22 6.45 5.97 4.29 5.13 5.36 5.82 5.59 7.04 7.45 8.66 8.77 7.98 6.70 8.00 7.28 7.20 5.93 8.02 7.07 7.17 2.43

0.08 0.10 0.10 0.09 0.09 0.06 0.05 0.08 0.08 0.06 0.06 0.06 0.07 0.08 0.09 0.04 0.04 0.04 0.04 0.04 0.05 0.19 0.11 0.10 0.07 0.14 0.12 0.06 0.06 0.06 0.10 0.09 0.10 0.10 0.07 0.06 0.07 0.08 0.11 0.10 0.08 0.10 0.05 0.10 0.06 0.09 0.03

1.90 2.54 2.01 1.98 2.11 2.20 1.49 2.00 1.68 2.14 1.85 1.85 1.89 1.61 1.90 1.72 2.00 1.82 1.97 2.20 1.89 2.50 3.21 3.15 2.38 3.80 3.01 1.73 1.91 1.82 2.27 1.88 2.07 2.07 2.26 1.79 1.79 1.98 2.00 2.46 1.97 1.91 1.64 1.39 2.14 1.93 4.80

1.00 1.88 4.01 4.03 2.73 1.90 2.08 1.40 1.74 2.72 1.86 1.62 1.90 1.60 4.15 1.26 0.97 1.31 1.20 1.29 1.68 2.60 3.26 2.99 2.08 3.47 2.88 1.86 2.60 2.23 3.02 3.31 3.16 1.16 1.70 1.71 1.71 1.57 0.38 1.39 0.56 0.52 1.62 0.78 1.52 0.97 0.33

1.29 1.72 1.86 1.82 1.67 2.13 1.93 2.51 1.45 1.87 1.86 1.66 1.92 1.93 1.52 1.10 1.08 1.00 1.21 1.21 1.29 1.86 1.93 1.85 1.70 2.20 1.91 1.31 3.19 2.25 1.13 1.02 1.07 0.88 0.69 0.75 0.74 0.77 1.01 0.99 0.75 0.77 0.61 0.68 0.98 0.83 0.91

4.00 4.55 4.05 4.02 4.16 4.80 4.87 5.20 4.70 3.64 4.04 3.83 4.44 3.75 5.04 4.09 4.57 4.30 4.84 4.54 4.45 4.50 4.36 3.67 4.25 4.18 4.19 4.09 3.57 3.83 3.81 3.72 3.76 4.68 4.13 4.35 4.23 4.35 4.70 4.64 4.17 4.60 4.24 3.92 3.99 4.32 4.00

n.d. 0.33 0.35 0.25 0.31 n.d. 0.39 n.d. 0.29 0.19 0.20 0.22 0.18 0.24 0.26 0.15 0.15 0.17 0.23 0.18 0.20 n.d. 0.25 0.39 0.20 0.34 0.23 0.34 0.44 0.39 0.17 0.15 0.16 n.d. 0.25 0.97 0.97 0.73 n.d. 0.26 0.18 0.16 0.25 0.14 0.51 0.25 n.d.

0.81 2.14 0.79 0.79 1.13 2.15 n.d. 1.40 n.d. n.d. 1.03 0.77 1.34 0.76 3.90 n.d. n.d. n.d. 1.19 0.89 1.69 1.07 n.d. n.d. n.d. 1.07 1.07 n.d. 1.05 1.05 n.d. n.d.

99.43 99.96 100.39 99.26

1.27 1.21 1.28 2.14 1.48 1.28 n.d. n.d. n.d. 2.21 1.09 2.96 1.89 1.43

99.90 100.35 100.22 100.50

165 194 136 132 157 216 251 228 208 146 220 210 148 169 201 197 223 209 220 217 205 187 137 119 185 138 153 224 91 158 191 178 185 183 206 197 198 196 202 215 199 232 207 216 175 207 189

n.d. 466 377 363 402 n.d. 400 n.d. 384 485 327 309 381 233 200 251 236 238 207 242 230 n.d. 580 628 371 623 551 439 1084 762 483 377 430 n.d. 142 270 272 228 n.d. 237 136 142 187 138 196 173 n.d.

890 918 818 824 863 1220 1132 810 1829 1743 763 850 1192 718 839 868 818 838 595 626 757 1600 1257 1152 834 1304 1229 1069 1630 1350 811 829 820 730 658 760 752 725 1100 827 804 872 1255 1237 625 960 850

16.4 n.d. n.d. n.d. 16.4 20.1 19.7 19.8 19.9 23.3 n.d. n.d. 20.6 n.d. n.d. 21.1 21.6 19.9 n.d. n.d. 20.9 15.5 15.5 15.7 15.9 n.d. 15.7 16.9 n.d. 16.9 15.2 14.9 15.1 14.6 n.d. n.d. n.d. 14.6 18.5 n.d. n.d. 16.1 n.d. n.d. n.d. 17.3 24.7

4.5 n.d. n.d. n.d. 4.5 6.1 5.5 4.5 4.2 2.4 n.d. n.d. 4.5 n.d. n.d. 6.7 6.7 6.0 n.d. n.d. 6.5 3.2 3.4 3.8 4.0 n.d. 3.6 4.5 n.d. 4.5 3.2 3.3 3.2 2.9 n.d. n.d. n.d. 2.9 4.4 n.d. n.d. 3.1 n.d. n.d. n.d. 3.7 6.6

n.d. 155 173 171 166 n.d. 206 n.d. 208 294 173 178 212 162 189 254 205 221 217 216 209 n.d. 190 219 183 187 195 157 250 204 212 220 216 n.d. 196 194 190 193 n.d. 177 191 209 221 224 207 205 n.d.

5.2 n.d. n.d. n.d. 5.2 6.0 5.7 3.6 5.5 8.2 n.d. n.d. 5.8 n.d. n.d. 6.8 6.1 6.5 n.d. n.d. 6.5 4.4 5.5 5.9 4.7 n.d. 5.1 5.0 n.d. 5.0 5.6 5.8 5.7 5.9 n.d. n.d. n.d. 5.9 7.2 n.d. n.d. 5.9 n.d. n.d. n.d. 6.6 5.4

Adobe Adobe Adobe Adobe Average of adobe

62.07 46.27 59.78 58.20 56.58

0.83 0.46 0.69 0.80 0.69

16.13 21.82 18.65 19.47 19.02

4.92 7.09 6.16 6.12 6.07

0.09 0.07 0.12 0.09 0.09

1.61 2.24 1.63 1.85 1.83

2.55 4.56 0.89 1.53 2.38

1.31 0.72 1.21 1.04 1.07

3.55 4.18 3.82 4.01 3.89

0.19 0.22 0.20 0.22 0.21

6.26 11.91 6.41 5.28 7.47

99.51 99.56 99.55 98.62

172 220 181 193 191

205 282 211 179 219

612 2297 757 712 1094

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

203 140 194 183 180

n.d. n.d. n.d. n.d.

Brickclay Brickclay Brickclay Brickclay River sediment, sand River sediment, sand River sediment, sand River sediment, sand River sediment, sand River sediment, silt

71.91 54.17 75.14 56.83 77.00

0.94 0.92 0.62 0.83 0.53

13.24 23.41 11.90 21.05 10.20

5.17 8.05 4.53 7.27 4.20

0.07 0.12 0.04 0.12 0.07

1.02 1.75 1.06 2.11 0.80

0.35 0.45 0.90 1.90 0.91

1.14 0.50 1.33 0.85 1.49

2.78 4.82 2.12 3.66 2.41

0.20 0.18 0.11 0.16 n.d.

3.17 5.94 2.87 5.84 2.42

100.14 100.50 100.79 100.77 100.03

129 239 103 194 77

130 263 197 94 n.d.

358 561 376 566 500

n.d. n.d. n.d. n.d. 7.2

n.d. n.d. n.d. n.d. 2.2

379 154 372 154 n.d.

n.d. n.d. n.d. n.d. 3.6

75.00

0.60

10.60

4.20

0.07

1.40

0.95

1.33

2.52

n.d.

2.97

99.64

95

n.d.

550

9.0

2.8

n.d.

5.3

88.00

0.34

4.70

2.78

0.32

0.31

0.30

0.44

1.13

n.d.

1.72

100.04

84

n.d.

370

6.7

2.1

n.d.

3.9

70.56

0.35

7.45

3.61

0.05

0.52

0.54

1.11

1.83

0.12

n.d.

86.14

92

146

422

5.1

1.5

163

7.0

74.03

0.53

11.99

4.57

0.05

1.04

0.92

1.42

2.62

0.17

2.53

99.87

118

238

575

n.d.

n.d.

200

n.d.

50.62

0.44

21.39

7.13

0.07

2.30

2.82

0.66

4.03

0.20

10.21

99.87

215

276

2175

n.d.

n.d.

167

n.d.

I/A/a

I/A/b

I/B/a

I/C/a

I/C/b

I/D

II/A

II/B-C

100.08 96.91 100.01 95.75 95.73 100.16 100.36 100.42 99.48 95.46 96.97 94.34 100.25 100.11 99.79 96.82 96.92 95.93 100.84 95.37 100.60 98.41 95.81

100.12 98.94 95.42 97.63 100.47 100.36 100.38 99.81

Author's personal copy

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

Nb

Ta

Y

La

Ce

Nd

n.d. 14.0 n.d. n.d. 14.0 n.d. 16.2 n.d. 17.2 18.4 19.0 19.0 18.0 18.0 18.0 15.4 14.4 15.4 15.0 17.0 16.2 n.d. n.d. n.d. 14.0 n.d. 14.0 16.6 n.d. 16.6 14.2 14.6 14.4 n.d. 21.0 15.0 16.0 17.3 n.d. 15.6 16.8 18.0 19.0 21.0 16.0 17.7 n.d.

1.6 n.d. n.d. n.d. 1.6 1.6 1.4 1.7 1.5 1.3 n.d. n.d. 1.5 n.d. n.d. 1.3 1.4 1.5 n.d. n.d. 1.4 1.1 0.9 n.d. 1.4 n.d. 1.1 1.7 n.d. 1.7 1.2 1.5 1.4 1.5 n.d. n.d. n.d. 1.5 1.6 n.d. n.d. 1.5 n.d. n.d. n.d. 1.6 2.4

n.d. 30 28 28 29 n.d. 18 n.d. 30 25 23 29 25 28 28 29 26 25 29 25 27 n.d. 27 29 28 26 27 19 25 22 33 31 32 n.d. 40 38 37 38 n.d. 37 38 36 39 51 38 40 n.d.

47 57 39 39 46 61 60 45 61 69 61 57 59 46 42 43 48 44 49 50 46 43 49 51 49 61 51 50 108 79 44 43 44 44 45 51 51 48 58 52 44 46 37 45 48 47 72

103 108 78 75 91 121 122 82 126 137 109 116 116 94 98 92 101 93 104 111 99 84 98 112 103 119 103 98 182 140 92 93 93 89 105 103 95 98 130 112 92 95 113 129 97 110 159

38 51 36 52 44 50 52 33 74 62 41 50 52 30 57 35 35 39 47 40 40 43 52 47 49 44 47 45 77 61 38 41 40 38 51 42 36 42 48 45 48 42 41 47 45 45 80

19.2 n.d. 13.3 16.5 16.3

n.d. n.d. n.d. n.d.

37 26 33 34 32

41 34 48 47 43

86 93 106 106 98

28.0 22.0 19.0 18.0 n.d.

n.d. n.d. n.d. n.d. 0.9

57 35 37 38 n.d.

35 62 29 36 23

n.d.

1.0

n.d.

n.d.

0.9

n.d.

Sm

7

Eu

Tb

Yb

Lu

Sc

V

Cr

Co

Ni

Zn

As

Sb

Cs

B

Cl

Gd

8.1 8.7 4.8 5.3 6.7 8.8 9.3 5.8 9.6 9.2 7.5 8.2 8.3 7.2 9.2 7.3 7.8 7.0 6.5 8.7 7.7 7.6 7.9 8.7 8.1 6.9 7.8 7.4 9.2 8.3 7.3 7.3 7.3 7.6 6.2 6.7 8.4 7.2 9.7 7.2 7.1 7.8 6.1 9.1 6.8 7.7 8.9

1.5 1.6 1.2 1.2 1.4 1.7 1.5 1.1 1.8 1.9 1.1 1.2 1.5 1.0 1.1 1.3 1.3 1.2 0.8 1.1 1.1 1.5 1.8 2.1 1.8 1.8 1.8 1.4 2.9 2.1 1.5 1.4 1.4 1.4 0.7 1.0 1.2 1.1 1.8 1.0 0.8 1.5 0.8 1.0 0.9 1.1 2.5

n.d. n.d. n.d. n.d.

3.4 2.8 2.5 2.5 2.8 2.2 1.6 1.7 2.5 1.6 2.4 2.7 2.1 2.6 2.6 2.8 2.8 2.3 2.6 2.4 2.6 2.7 2.4 2.6 2.5 2.2 2.5 1.8 1.9 1.9 2.7 2.7 2.7 3.1 3.7 3.5 3.4 3.4 4.1 3.4 3.6 3.0 3.5 4.6 3.4 3.7 5.1

0.36 n.d. n.d. n.d. 0.36 0.32 0.21 0.26 0.36 0.22 n.d. n.d. 0.27 n.d. n.d. 0.36 0.37 0.35 n.d. n.d. 0.36 0.42 0.32 0.31 0.35 n.d. 0.35 0.24 n.d. 0.24 0.38 0.39 0.39 0.46 n.d. n.d. n.d. 0.46 0.57 n.d. n.d. 0.42 n.d. n.d. n.d. 0.50 0.62

16.1 n.d. n.d. n.d. 16.1 12.6 9.8 12.5 13.8 13.3 n.d. n.d. 12.4 n.d. n.d. 13.5 15.5 14.2 n.d. n.d. 14 22.6 15.8 16.6 15.9 n.d. 17.7 13.0 n.d. 13 14.0 14.2 14.1 16.6 n.d. n.d. n.d. 16.6 17.8 n.d. n.d. 17.8 n.d. n.d. n.d. 17.8 25.4

n.d. 120 134 129 128 n.d. 82 n.d. 82 97 116 140 103 85 83 104 116 105 111 124 104 n.d. 111 134 110 126 120 103 92 97 108 106 107 163 141 144 148 149 n.d. 137 129 128 149 158 130 138 n.d.

102 56 34 35 57 51 62 65 70 68 64 85 66 38 43 72 88 74 47 51 59 118 60 118 120 32 90 52 71 62 70 79 75 93 82 75 73 81 92 74 80 87 70 75 64 77 138

16.4 16.0 16.0 15.0 15.9 11.2 9.2 10.2 12.5 13.5 12.0 15.0 11.9 11.0 10.0 11.5 13.8 10.4 12.0 13.0 11.7 19.0 14.8 19.6 16.5 16.0 17.2 6.9 9.0 8.0 13.2 13.7 13.5 16.1 20.0 17.0 18.0 17.8 17.5 20.0 16.8 15.3 14.0 21.0 18.0 17.5 8.5

n.d. 44.0 37.0 39.0 40.0 n.d. n.d. n.d. n.d. 15.8 42.0 70.0 42.6 53.0 35.0 8.8 20.0 n.d. 63.0 49.0 38.1 n.d. 8.4 1.3 n.d. 46.0 18.6 n.d. 39.0 39.0 28.9 13.6 21.3 n.d. 89.0 14.0 22.0 41.7 n.d. 49.8 15.3 54.5 82.0 105.0 82.0 64.8 n.d.

104 143 121 120 122 98 109 138 83 95 83 95 100 77 170 92 103 86 87 89 100 118 124 170 111 123 129 123 80 101 87 85 86 110 101 407 409 257 120 133 92 117 57 89 106 102 n.d.

13 n.d. n.d. n.d. 13 13 72 30 13 21 n.d. n.d. 30 n.d. n.d. 133 154 205 n.d. n.d. 164 23 14 10 16 n.d. 16 53 n.d. 53 6 9 7 23 n.d. n.d. n.d. 23 7 n.d. n.d. 26 n.d. n.d. n.d. 17 10

22.2 n.d. n.d. n.d. 22.2 3.2 2.9 30.1 2.5 1.7 n.d. n.d. 8.1 n.d. n.d. 9.5 7.2 18.8 n.d. n.d. 11.9 8.3 1.4 1.6 3.6 n.d. 3.7 3.4 n.d. 3.4 2.0 9.2 5.6 6.3 n.d. n.d. n.d. 6.3 1.5 n.d. n.d. 3.8 n.d. n.d. n.d. 2.7 1.6

60 n.d. n.d. n.d. 60 42 30 29 22 12 n.d. n.d. 27 n.d. n.d. 37 40 41 n.d. n.d. 39 15 12 18 32 n.d. 19 61 n.d. 61 260 163 212 43 n.d. n.d. n.d. 43 12 n.d. n.d. 46 n.d. n.d. n.d. 29 19

117 n.d. n.d. n.d. 117 109 118 143 n.d. 104 n.d. n.d. 95 n.d. n.d. n.d. n.d. 205 n.d. n.d. 205 94 n.d. n.d. n.d. n.d. 94 n.d. n.d.

600 n.d. n.d. n.d. 600 350 400 430 n.d. 582 n.d. n.d. 441 n.d. n.d. n.d. n.d. 295 n.d. n.d. 295 154 n.d. n.d. n.d. n.d. 154 n.d. n.d.

6.8 n.d. n.d. n.d. 6.8 6.0 6.0 5.0 n.d. 6.4 n.d. n.d. 5.9 n.d. n.d. n.d. n.d. 5.1 n.d. n.d. 5.1 6.4 n.d. n.d. n.d. n.d. 6.4 n.d. n.d.

149 n.d. 149 136 n.d. n.d. n.d. 136 91 n.d. n.d. 130 n.d. n.d. n.d. 110 125

78 n.d. 78 86 n.d. n.d. n.d. 86 60 n.d. n.d. 94 n.d. n.d. n.d. 77 n.d.

6.0 n.d. 6.0 6.4 n.d. n.d. n.d. 6.4 9.0 n.d. n.d. 6.2 n.d. n.d. n.d. 7.6 10.6

33 30 39 38 35

6.5 4.6 5.7 7.3 6.0

0.9 0.9 0.9 0.9 0.9

n.d. n.d. n.d. n.d.

3.4 2.4 3.0 3.1 3.0

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

112 134 118 134 125

46 51 49 62 52

12.6 13.8 16.4 17.3 15.0

51.9 67.8 58.6 49.4 56.9

67 97 89 85 84

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

91 124 76 90 49

48 53 28 39 17

6.7 9.2 3.7 7.6 3.5

0.8 1.2 0.7 0.8 0.9

n.d. n.d. n.d. n.d. 0.4

5.0 3.3 3.1 3.5 1.3

n.d. n.d. n.d. n.d. 0.19

n.d. n.d. n.d. n.d. 6.2

93 157 82 148 n.d.

51 86 40 76 30

39.0 37.0 11.0 19.0 7.8

83.0 83.0 88.0 87.0 n.d.

54 80 56 88 48

n.d. n.d. n.d. n.d. 12

n.d. n.d. n.d. n.d. 18.2

n.d. n.d. n.d. n.d. 23

n.d. n.d. n.d. n.d. 61

n.d. n.d. n.d. n.d. 69

n.d. n.d. n.d. n.d. 4.2

28

59

21

4.4

1.0

0.6

1.7

0.25

7.3

n.d.

40

9.0

n.d.

55

16

14.9

24

93

400

4.9

n.d.

19

40

14

3.7

0.8

0.6

1.7

0.23

5.4

n.d.

24

7.2

n.d.

61

22

4.6

17

74

95

2.8

0.6

14

15

31

13

2.5

0.6

0.4

1.5

0.23

4.1

38

21

5.2

n.d.

42

34

9.5

13

n.d.

n.d.

n.d.

n.d.

n.d.

24

32

76

21

6.6

1.0

n.d.

2.2

n.d.

n.d.

74

38

11.2

60.2

49

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

25

28

89

33

5.0

0.9

n.d.

2.3

n.d.

n.d.

131

54

13.4

77.3

102

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

n.d. n.d. n.d. 1.2 0.8 n.d. n.d. 1.0 n.d. n.d. 0.9 1.1 0.8 n.d. n.d. 0.9 n.d. n.d. n.d. 1.2 n.d. 1.2 n.d. n.d. 0.9 1.0 0.9 1.0 n.d. n.d. n.d. 1.0 1.2 n.d. n.d. 1.0 n.d. n.d. n.d. 1.1 1.5

(continued on next page)

Author's personal copy

8

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

Table 1 (continued) Sample

Type

SiO2

TiO2

Al2O3

Fe2O3

MnO

MgO

CaO

Na2O

K2O

P2O5

LOI(H2O)

5/3

River sediment, silty clay River sediment, gravely sand River sediment, sand River sediment, silt River sediment, silty sand River sediment, sand River sediment, sand River sediment, silt River sediment, clay Average of sediments

66.08

0.55

15.09

6.53

0.08

1.25

0.60

0.90

2.57

0.14

70.41

0.57

10.83

9.56

0.10

0.92

0.24

0.57

1.99

72.85

0.57

12.06

5.22

0.07

1.21

1.03

1.06

61.32 76.15

0.73 0.60

16.58 10.43

8.14 5.22

0.14 0.06

1.29 0.78

1.45 0.63

78.24

0.51

9.84

3.68

0.05

0.90

80.68

0.45

7.16

5.82

0.06

66.12 60.56

0.65 0.90

16.57 19.77

4.63 6.03

69.77

0.61

13.38

39.43 65.48 57.67 67.84 57.61

0.82 0.54 0.87 0.77 0.75

67.74

5/4 5/5 5/7 5/8 7/1 7/2 7/5 7/6

Sedimentary rocks 3/2a Claystone/shale 3/2b Sandstone 6/8 Siltstone 7/3 Siltstone/shale Average of sed. rocks Volcanoclastic-volcanic rocks 6/4 Morococala rhyodacitic tuff 6/5 Morococala rhyodacitic tuff Average of Morococala tuff 8/1 Soledad prim. weath. sand 8/2b Soledad dacitic tuff 8/2d Soledad dacitic tuff 8/3 Soledad dacitic tuff 8/5b Soledad dacitic tuff 8/6a Soledad dacite 8/6b Soledad dacite 8/6c Soledad dacitic tuff 8/6d Soledad dacitic tuff 8/6g Soledad dacitic tuff 8/6h Soledad dacite Average of Soledad rocks

Sum(%) ppm

Rb

Sr

5.74

99.53

141

146

0.18

3.87

99.24

100

2.25

0.12

3.49

99.93

0.87 1.04

3.18 1.95

0.68 0.12

4.74 2.69

0.97

1.35

2.15

0.16

0.45

0.68

1.31

1.59

0.06 0.06

1.42 1.72

1.30 0.68

1.91 1.02

5.60

0.09

1.17

0.93

16.85 11.95 21.25 14.33 16.10

23.31 8.33 7.39 5.23 11.06

0.11 0.17 0.05 0.06 0.10

2.52 2.38 2.61 2.07 2.39

0.46

16.00

2.61

0.04

66.96

0.46

16.45

2.52

67.35

0.46

16.23

68.24

0.92

62.82 63.14 68.46 49.38 68.61 66.31 66.67 67.13 65.16 66.52 64.77

0.82 0.82 0.50 0.61 0.69 0.69 0.63 0.69 0.68 0.69 0.70

Th

U

Zr

Hf

484

n.d.

n.d.

263

n.d.

61

503

n.d.

n.d.

210

n.d.

108

185

522

n.d.

n.d.

375

n.d.

99.11 99.67

157 89

141 145

520 572

n.d. n.d.

n.d. n.d.

230 356

n.d. n.d.

2.09

99.94

93

233

512

n.d.

n.d.

296

n.d.

0.14

1.64

99.97

66

204

529

n.d.

n.d.

303

n.d.

3.21 4.01

0.16 0.12

3.63 4.46

99.65 99.31

171 229

289 140

635 667

n.d. n.d.

n.d. n.d.

245 191

n.d. n.d.

1.07

2.68

0.16

3.89

132

152

600

7.0

2.2

254

5.0

1.44 2.57 0.27 0.48 1.19

0.54 1.14 0.86 1.95 1.12

2.81 1.27 4.52 2.85 2.86

0.10 0.17 0.14 0.16 0.14

11.03 5.93 4.21 2.68 5.96

98.96 99.93 99.83 98.40

147 65 227 140 144

64 53 47 68 58

362 213 609 388 393

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

141 167 154 256 179

n.d. n.d. n.d. n.d.

0.90

2.04

3.44

4.75

0.26

1.16

99.40

258

517

894

n.d.

n.d.

196

n.d.

0.03

1.26

2.10

3.24

4.64

0.26

1.52

99.45

231

526

941

n.d.

n.d.

194

n.d.

2.56

0.04

1.08

2.07

3.34

4.69

0.26

1.34

244

521

918

15.54

2.60

0.04

0.94

4.13

3.32

2.50

0.16

0.97

99.37

57

658

1023

n.d.

n.d.

328

n.d.

15.77 15.50 15.08 11.69 15.38 15.51 15.14 16.09 16.39 15.42 15.23

4.05 4.22 2.26 14.44 3.09 3.33 3.13 3.56 4.29 3.38 4.39

0.06 0.06 0.03 0.07 0.04 0.05 0.05 0.04 0.05 0.04 0.05

2.08 1.85 0.96 1.17 1.32 1.39 1.47 0.83 1.64 1.39 1.37

3.20 3.19 2.39 2.15 2.47 2.82 2.92 3.00 2.62 2.73 2.87

2.57 2.72 3.35 2.72 3.51 3.27 2.97 3.48 2.29 2.94 3.01

4.16 4.33 3.43 3.77 4.84 4.06 4.58 4.43 3.97 4.79 4.08

0.32 0.32 0.18 0.37 0.10 0.27 0.26 0.31 0.26 0.27 0.26

4.11 3.53 2.52 12.18 0.43 1.88 1.88 0.48 1.77 1.92 2.88

99.94 99.66 99.15 98.56 100.70 99.84 99.69 100.02 99.12 100.35

152 156 122 123 151 165 166 164 154 164 143

547 553 490 539 525 545 515 568 522 535 545

1287 1305 1246 1286 862 1264 1337 1284 1576 1370 1258

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

256 248 154 181 234 226 212 227 223 230 229

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

contains very varied samples with anisotropic and yellowishreddish, silty paste and different originated volcanic clasts. Subgroup I/C/a has low sphericity, angular to subangular, unweathered, hialopilitic volcanites as aplastics with different composition (from dacitic through andesitic to basaltic andesitic e based on mineralogical and fabric observations). Subgroup I/C/ b has an aplastic content of medium sphericity, subrounded, weathered, vitrophyric volcanic rock fragments. Subgroup I/D has an almost serial fabric and contains no rock fragments (dominant grain size 50e100 mm). The major constituents of these ceramics are quartz, plagioclase, biotite, hornblende and rare orthopyroxene. All of these mineral clasts are medium sphericity, subangular-subrounded and unweathered. Limited opaque phases are present. Group II (with sedimentary origin) is the other dominant petrographic group of the pottery assemblage with 93 representatives. It has medium-coarse grained (175e200 mm and 300e1300 mm) hiatal fabric, 30e35% aplastic inclusion content. This group can be characterized on the base of the presence of subrounded-well rounded, low-medium sphericity, clastic sedimentary rock fragments which

Ba

195

are natural components of the anisotropic and brown, coarse silty paste. Division of the group into three subgroups (II/A-B-C) is based on the dominant grain size of the aplastic sedimentary rock clasts (Fig. 4eeg). Subgroup II/A contains pottery fragments with fine grained (claystone/shale-siltstone) sedimentary (sometimes low grade metamorphic) rock and mineral fragments. The aplastic inclusions themselves contain quartz, muscovite-sericite (sometimes biotite) and plagioclase in a weakly-well oriented fabric. Subgroup II/B has similar composition to the previous one as regard the aplastic inclusions but their dominant grain size is coarser (claystone-siltstone together with sandstone) and their fabric do not shows any metamorphic feature. It contains both rock and mineral clasts. Subgroup II/C can be characterized with the coarsest dominant grain size of the aplastics since this group of pottery has mainly sandstone rock and mineral fragments as clastic constituents. Regarding the mineralogical composition, this subgroup is also similar to the previous ones. Group III (with metamorphic origin) is the smallest group in the investigated assemblage with 7 samples (Fig. 4h). This group incorporates coarse grained (200e300 mm), hiatal fabric ceramics

Author's personal copy

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

9

Table 1 (extended) Nb

Ta

Y

La

Ce

Nd

14.6

n.d.

32

38

90

34

15.2

n.d.

33

24

69

17.9

n.d.

36

35

16.8 18.1

n.d. n.d.

57 34

n.d.

n.d.

n.d.

Sm

Eu

Tb

Yb

Lu

Sc

V

Cr

Co

Ni

Zn

As

Sb

Cs

B

Cl

Gd

4.6

0.7

n.d.

2.8

n.d.

n.d.

101

55

14.4

72.2

75

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

27

4.6

0.7

n.d.

2.9

n.d.

n.d.

91

57

18.5

79.3

90

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

82

36

5.2

0.8

n.d.

2.9

n.d.

n.d.

87

42

12.1

83.9

58

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

67 31

96 71

53 27

8.6 5.3

1.0 0.8

n.d. n.d.

5.2 2.8

n.d. n.d.

n.d. n.d.

120 79

63 38

17.5 11.5

70.3 77.0

74 52

n.d. n.d.

n.d. n.d.

n.d. n.d.

n.d. n.d.

n.d. n.d.

n.d. n.d.

27

28

71

25

5.8

0.9

n.d.

2.3

n.d.

n.d.

62

34

7.3

65.8

34

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

24

30

71

23

3.9

0.8

n.d.

1.8

n.d.

n.d.

65

38

7.2

76.1

39

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

19.4 20.0

n.d. n.d.

30 38

43 44

86 107

40 43

8.0 7.6

1.1 0.8

n.d. n.d.

2.6 3.6

n.d. n.d.

n.d. n.d.

75 144

27 64

9.2 40.1

48.5 50.5

73 100

n.d. n.d.

n.d. n.d.

n.d. n.d.

n.d. n.d.

n.d. n.d.

n.d. n.d.

19.0

0.8

34

34

77

31

5.6

0.9

0.5

2.8

0.23

5.8

97

46

15.7

73.5

65

21

11.8

19

76

188

4.0

17.4 11.5 16.3 16.6 15.5

n.d. n.d. n.d. n.d.

39 32 36 37 36

52 16 39 34 35

91 46 94 86 79

38 16 37 42 33

12.4 3.8 5.8 5.5 6.9

1.5 0.7 0.5 0.6 0.8

n.d. n.d. n.d. n.d.

3.3 2.9 3.4 3.4 3.3

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

144 85 169 94 123

116 45 86 48 74

43.6 22.9 13.4 12.6 23.1

206.0 81.5 79.1 40.6 101.8

136 68 78 116 99

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d.

n.d.

n.d.

14

65

115

42

7.5

1.4

n.d.

1.2

n.d.

n.d.

47

n.d.

n.d.

19.9

53

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

14

69

118

45

6.3

1.4

n.d.

1.2

n.d.

n.d.

47

n.d.

n.d.

20.4

51

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

14

67

117

43

6.9

1.4

20.2

52

1.2

47

n.d.

n.d.

46

691

1123

381

67.4

6.4

n.d.

3.8

n.d.

n.d.

66

n.d.

1.4

27.6

52

n.d.

n.d.

n.d.

n.d.

n.d.

n.d.

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

19 21 18 15 12 15 19 17 26 17 20

51 59 59 49 62 64 55 43 57 53 113

114 120 109 91 105 121 112 97 124 115 203

39 51 39 34 36 52 35 33 54 49 73

6.2 7.8 6.5 7.4 7.4 7.4 6.2 4.1 8.9 6.2 12.3

1.5 1.7 1.4 1.7 1.6 1.6 1.4 1.4 1.7 1.5 2.0

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

1.4 1.7 1.7 1.0 0.8 1.2 1.5 1.3 2.3 1.3 1.6

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

91 91 55 81 65 68 73 59 77 66 72

23 27 19 42 20 5 18 27 13 6 20

6.5 4.4 1.5 8.7 5.0 3.0 2.1 2.9 9.8 3.0 4.4

34.7 28.3 4.2 53.0 48.0 26.0 25.2 40.6 5.1 35.0 29.8

82 88 71 107 56 75 62 67 213 73 86

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d.

with well rounded, medium sphericity and e in contrast to group I and II e metamorphic mica schist rock and mineral fragments (muscovite, quartz, opaque phases) in a white, isotropic clay paste. The eight building material remnants are divided into two types (adobes and rocks) from the point of view of the petrographic description. The adobe samples represented four types of adobes according to their colour and internal structure: (1) loose fabric, grey adobe with a high quantity of plant remnants (No. 4/3), (2) loose fabric, dark red adobe with a high quantity of plant remnants (No. 4/ 4), (3) loose fabric, light red adobe with a high quantity of plant remnants (No. 4/5), (4) loose fabric, yellowish brown plaster with a high quantity of plant remnants (No. 4/6). The most abundant building stones are sandstones and siltstones, which clearly characterize the local geology. The selected building stone samples were: (1) fine-grained, dark grey, schistose siltstone (No. 4/1), (2) mediumgrained, fine bedded reddish brown sandstone (No. 4/2), (3) medium-grained, green, micaceous sandstone with undulate bedding planes (No. 4/7), (4) greenish grey shale (No. 4/8). The adobes are basically fine grained (from 10e30 mm to 50 mm average grain size), almost serial and quite dense fabric, limonitic,

partly micaceous, secondarily carbonized silty clays. Their secondary carbonatic and limonitic content is present in the form of 75e175 mm sized nodules and dispersed mottles. The average mineralogical composition is quartz, plagioclase, carbonate (most probably calcite), limonite, mica (rather biotite than muscovite), opaque phases and rare tourmaline. Occasionally argillaceous rock fragments are found too. The remnants of the plant tempering can be observed as characteristic shaped pores or sometimes phytolithes. The building stone samples are relatively monotonous, finely layered, clastic sedimentary rock fragments with similar mineralogical composition (normal or slightly undulatory extinction quartz, fine grained muscovite or sericite (rare limonitized mica) and weathered plagioclase, tourmaline, opaque phases, zircon and secondary limonite) but with variable dominant grain size (from 10e50 mm to 100e150 mm). The investigated geological samples (35 thin sections) were separated according to the sampling considerations and the geological setting. The alluvial sediments of the Paria Basin and the eastern rim of the Altiplano (18 samples: No. SED/01-04, No. 1/2-3,

Author's personal copy

10

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

Fig. 3. Archaeological styles of ceramic appearing at Paria Basin exemplified by plates (aeb) and bowls (cee).

5/1-3, 5/5-8, 7/1-2, 7/4-5, 7/7) were ranging from silty clay to fine gravely sand and all of those could have originated from the fluvial weathering of the Palaeozoic clastic sedimentary unit of the Eastern Cordillera. These weakly-medium layered, medium sphericity, rounded-well rounded, medium-well sorted sediments mainly consist of rock and mineral fragments of the above mentioned Palaeozoic sedimentary rocks (shales/claystones-siltstones-sandstones) and subordinately varying quantities of weathered volcanic-volcaniclastic rock clasts (their content depends on the locality and its distance from the nearest volcanic sources). The major mineral phases of sedimentary origin are quartz, feldspars, mica (muscovite-biotite) and clay minerals. The volcanic related constituents are mainly devitrified, weathered volcanic glass with crystallites and phenocrystals of feldspar, biotite (rarely hornblende) and quartz. Samples from the Silurian sedimentary ridges of the Eastern Cordillera (2 samples: No. 3/1-2a-2b) are weakly-well layered, medium-well sorted shales-siltstones-sandstones mainly consisting of quartz, feldspars, muscovite-biotite and probably clay minerals invisible under the microscope. The pyroclastics of the south eastern Morococala Volcanic Field (2 samples: No. 6/4-5) are weakly welded dacitic tuffs (ignimbrites) with highly vitric, pumiceous matrix and feldspar (mainly plagioclase and less K-feldspar), quartz and biotite phenocrysts. Low quantity of accessories is represented: opaque minerals, apatite and zircon. The tuffs and volcanites of the western Soledad caldera (13 samples: No. 8/1-3 and 8/5-6 with subsamples) have acidicintermediate composition too. Welded dacitic-rhyodacitic tuffs have feldspar (mainly plagioclase), biotite, hornblende and minor quartz content in the vitric, pumiceous matrix. Volcanites are dacitesrhyodacites with vitroporphyric fabric and similar phenocrysts as in the volcaniclastics. The most abundant accessory is titanite. 5.2. Chemical analyses Three different methods for chemical analyses of solid materials (XRFþ/-INAAþ/-PGAA) were carried out on 39 representative ceramic fragments e chosen on the basis of microscopic studies e ,

four adobe remnants and 36 geological comparative samples (Table 1). Two separate ways of processing the raw data were applied: 1) plotting of element concentrations in normalized, multi-elemental abundance (so called spider-) diagrams and 2) displaying the concentrations in bivariate correlation diagrams, using element ratios instead of absolute concentrations. This way, it was possible to better outline the similarities and differences among the individual samples. The normalization of the multielemental abundance patterns was done relative to an average value for Post Archaean Australian Shale (PAAS) which is a preferred standard material in geochemical sedimentary rock investigations for fine grained siliciclastic sediments (Nance and Taylor, 1976; Taylor and McLennan, 1985; McLennan, 1989, 2001). The interpretation of the spider diagrams using groups defined by petrographic characteristics made it possible to check and prove the coherence of the groups by an independent method. For this reason see Fig. 5 for each group and subgroup. Group I can be generally characterized with basically average (similar to PAAS) major and trace element concentrations (Fig. 5aed). This pattern is accompanied with a slight enrichment of specific cations in the silicate phases (MgO, CaO, Na2O), P2O5, mobile trace elements (Rb, Sr, Ba), Zn and light rare earth elements, while TiO2, Fe2O3, Yb, V, Cr and Co show lower concentrations compared to the reference material. In the cases of I/A, I/B and I/D subgroups geochemical composition supports the petrographic classification, the members of the groups are similar to each other from the point of view of the distribution of the immobile major and trace elements. In contrast with other subgroups of petrographic group I, subgroup I/C is incoherent from both a petrographic and a geochemical point of view. This feature expresses the heterogeneity of the applied raw materials of this subgroup concerning the mineralogy of both the clay paste and the aplastics. Although some immobile elements (Zr, Nb, Y) show similar relative content in subgroup I/C, most of the silicate cation position filling major elements vary. This characteristic makes it very probable that the individual members of this subgroup originated from different sources. Both major and trace element distribution of ceramics of group II are more similar to the average composition of PAAS (Fig. 5eef).

Author's personal copy

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

11

Fig. 4. Thin section photomicrographs of the main petrographic types of pottery.

The most variable elements are Ca, P and Eu, Zn. The complete group II is quite coherent especially in the case of immobile trace elements. Concerning this petrographic group, on the base of geochemistry the subgroups cannot be distinguished as unequivocally as by petrography. However, some tendencies can be outlined in connection with the average grain size and clay content of

aplastics: i.e. finer grained subgroup II/A (with claystone/shale fragments) shows partly higher CaO and lower Eu concentrations than the coarser grained samples. The major and trace element distribution of petrographic group III is basically different from that of the previous types (Fig. 5geh). The high TiO2, Al2O3 and average rare earth element content, and

Author's personal copy

12

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

Fig. 5. Major and trace element distribution of the main petrographic types of pottery.

Author's personal copy

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

13

Fig. 6. Discrimination of the main petrographic types (I, II, III) of pottery by geochemically significant element ratios. (a) K2O/Na2O vs. SiO2/Al2O3 and (b) TiO2/Al2O3 vs. Y diagrams. Legend: 1 e fine-grained alluvial sediments (clay-silt), 2 e coarse-grained alluvial sediments (sand-gravely sand), 3 e adobes, 4 e Palaeozoic sedimentary rocks, 5 e alluvial clay from recent brick mine, 6 e Soledad tuffs, 7 e Morococala tuffs, 8 e petrographic type I of ceramics, 9 e petrographic type II of ceramics, 10 e petrographic type III of ceramics; A e cluster of potential tempering materials of volcaniclastic origin, B e cluster of potential clayey paste materials of sedimentary origin, C e cluster of 1st group ceramics which could be modelized.

the low Fe2O3, MgO, CaO and Co content of the sample underlines a peculiar chemical character. This supports the idea of a different geological origin of the third group, although this suggestion is based on one measurement only. To present the geochemical characteristics of ceramic and comparative sediment/rock groups, absolute concentrations as well as ratios of elements of geochemical significance in provenance analysis were applied. This led us to the conclusion that immobile and partly incompatible major and trace elements should be preferred. Fig. 6 shows bivariate correlation diagrams of major (Fig. 6a) and trace (Fig. 6b) element ratios. The main clusters of ceramics in Fig. 6a group to the lower left part of the diagram which can be explained by the predominance of the finer grained and more clay containing (with higher Al2O3 concentrations) raw materials (grouping in field B below the red broken line). Those can be separated into petrographic group I with lower and group II and III with higher K2O/Na2O ratios. The single representative of group III has the lowest SiO2/Al2O3 value. Considering the potential raw material samples, it can be stated that major element ratios cannot distinguish the two volcaniclastic sources: both groups of tuffs and volcanites have relatively low major element ratios (marked with field A). In contrast, chemical composition of the Silurian sedimentary rocks and sediments of the Paria Basin and of the adobe samples disperse in a wide range. This complex data set can be separated into two main clusters: a low SiO2/Al2O3 e high K2O/Na2O one and a high SiO2/Al2O3 e low K2O/Na2O one. The former is identical with the Al2O3-rich, clayey (fine grained) sediments/rocks (marked with field B below the red broken line), while the latter resembles the coarser grained siliciclastic materials (above the red broken line). The trace element values and ratios in Fig. 6b outline a quite compact grouping of ceramic samples with medium TiO2/Al2O3 (0.03e0.05) and medium Yb (20e40 ppm) values. Samples of petrographic group II shows a bit higher Y and lower TiO2/Al2O3 values than that of group I on the average. In this bivariate diagram, the tuff related samples of the two volcaniclastic sources (lowmedium TiO2/Al2O3 and low Y values) can be better distinguished from each other (though they still do not have discriminative importance) since Morococala’s samples have lower TiO2/Al2O3 ratios than Soledad’s ones. When we use this combination of element ratios, sediments and Silurian sedimentary rocks of the Paria Basin and the adobe samples can be separated much better and form much clearer distinctive groups than in the major

element diagram. However, the coarser grained sediments/rocks cluster closer to the higher TiO2/Al2O3 and lower Y values. Using the bulk chemistry for the characterization of the ceramics, the separation of the specimens into certain subgroups is not that clear, easy and evident as with petrography. However, the petrographic groups I and II can be quite clearly distinguished and their relation to the potential raw material sources can be assumed, additionally, almost all of the immobile major and trace elements suggest the same relationships. This outlines the important role of aplastic inclusions in the bulk composition of ceramics. The sample set of petrographic group I (marked with field C) is between the clusters of the pyroclastics (both Morococala and Soledad, marked with field A) and of the finer grained sediments of the Paria Basin (marked with field B). On the other hand, the cluster of the petrographic group II overlaps well with the set of the finer grained sediments of the Paria Basin, and especially to the archaeological adobes from Paria and partly the brick clays of the eastern rim of the Altiplano. It was not possible to make a distinction between the two pyroclastic sources based on the bulk chemistry. Utilizing the major element ratios, the finer (clayish-silty) and the coarser (sandy-gravely) sediments of the alluvial plain of the Paria basin could be separated. It was an important feature that all of the investigated ceramics clustered rather closer to the finer than to the coarser grained sediments. 5.3. Phase analyses Instrumental mineralogical examination was used to determine the crystalline and weakly crystalline phases of the fine grained paste of the ceramics. Interpreting these data, the estimation of the maximum firing temperature and the basic characterization of the firing atmosphere could be done. In addition, further examination of the firing conditions is planned. The XRD analyses were performed on the complete ceramic assemblage (206 pieces). Fig. 7 shows typical representatives of each petrographic ceramic group. The basic mineral phases of group I subtypes (the lower curves in Fig. 7) are quartz, different feldspar types (plagioclase and K-feldspar) with variable crystallization stages and 10Åtype phyllosilicate phase (illite-sericite, sometimes muscovitebiotite with (002) basal reflection at 9.98 Å, weak (110) reflection at 4.48 Å, weak triplet between 2.70 and 3.00 Å). Weak elevation of the baseline between 3 and 4 Å indicates the presence of amorphous phase, especially in the case of subgroups I/A and I/B. In

Author's personal copy

14

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

Fig. 7. Mineralogical compositions of the main petrographic types of pottery by XRD.

addition, amphibole is present in subgroups I/A and I/D, and goethite and hematite are in quantities of accessories. Based on the mineral assemblage, no significant phase changes could be detected, so the estimated maximum firing temperature could not exceed the 650e700  C. Group II ceramic (second curve from the top in Fig. 7) has high quartz content too, but its 10Å-type phyllosilicate is rather illite (with expressed (110) reflection at 4.49 Å, moderated (002) basal reflection at 9.98 Å and (004) reflection at 4.95 Å) and its feldspar phase is more monotypical (plagioclase). Some dolomite and trace of hematite are also present. Similarly to the samples of group I of pottery, no high-temperature phase formation during the firing was observed. The maximum firing temperature is estimated to be between 650 and 700  C (since the carbonate content is secondary, it is irrelevant in the calculation). Group III of pottery (uppermost curve in Fig. 7) has an absolutely different XRD pattern. Together with the predominant quartz, phyllosilicate phase with basal reflection at 9.32 Å, mullite phase (with peaks at 5.39 Å, 4.66 Å, 3.43e3.46 Å) and an unidentifiable phase at 3.11 Å are detectable. In addition, a small amount of K-feldspar and hematite could be identified. Due to the neoformed Al-silicate (mullite) phase, the mineral assemblage indicates high-temperature

recrystallization of the matrix at about 850e950  C maximum firing temperature. Variable but permanent presence of hematite in the ceramic material signifies the basically oxidative character of the firing atmosphere, even in the case of the entirely white walled vessels (e.g. petrographic group III). 6. Discussion 6.1. Provenance Petrographic observation of ceramic fragments resulted in the classification of the assemblage into three main aplastic constituent material groups: (I) pyroclastic and volcanic originated, (II) sedimentary rock originated and (III) metamorphic rock originated. Based on the aplastics’ petrology-mineralogy and the fabric features, subgroups could be separated inside those groups. Confronting this result with the geochemical data of the local rock types and the knowledge on the geological setting, the applied raw material (temper) sources of the first ceramic group have to be the pyroclastic areas of both the Morococala Volcanic Field and the Soledad Caldera (10e30 km far from the site, so these are near

Author's personal copy

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

sources). On the one hand, this conclusion contradicts Arnold’s idea (Arnold, 2005) that pottery raw material comes from not more than a 7 km distance (if we presume that the place of manufacture was in the archaeological site of Paria and not outside of the settlement). On the other hand, it is in agreement with recent publications which identify the source area of ceramic materials at a longer distance (Valera Guarda, 2002; Alden et al., 2006). The provenance statement for the second ceramic group is that the utilized raw material (untempered silty clay paste) came from the alluvial plain of the Paria basin (max. 5 km distance from the site, so it is a local source). There is no potential source of the third ceramic group, neither in the vicinity nor in the wider surroundings (w50 km) of the site. This information and the fact that the third ceramic group is presented with a very low ratio in the pottery assemblage suggest a foreign (imported) origin of the petrographic group III. There was a purposive selection of the sedimentary rock derived, fluvial sediments as raw materials both for pottery and adobe production. The adobe samples clearly agree with the geochemical composition of the fine grained sediments of the Paria Basin. Since these adobes show expressed similarity to the material of the second ceramic group, this provides further evidence of the local origin of the ceramic group II. It is not clear that the collected local fine grained sediments could provide the raw material (clay paste) for petrographic group I. Based on the fact that tuff and partly volcanic lithofragments in group I of ceramics are mostly quite unweathered, a raw material site near to the original, physically unweathered rock outcrop has to be assumed. However, for some subgroup I/C samples it has to be mentioned that: on the one hand, the grain size and the quantity of certain lithic clast types does not always correlate, while on the other hand, the quantity and the unweathered condition of the clasts directly correlates. As was already discussed by others (Bray et al., 2005), it is very probable that it was not only the shape and decoration that represented value in the classical Inka imperial pottery but the material itself (and its source) too. In addition to this, it is a common feature throughout the Andes in space and time that volcanic and tuff rocks were used as temper of the ceramics (Ixer and Lunt, 1991; Velde and Druc, 1999a) and this can be traced in the present-day traditional handicraft too (Arnold, 1972). Based on these suggestions, it is possible that in the case of Paria the usage of volcaniclastic rocks was not a chance of the Inka ceramic making potters but a conscious selection of a more highly esteemed material. Combining archaeometrical results with the basic archaeological data on the ceramic artefacts leads to new insights. It is very probable that the material used relates rather to the vessel styles (quality and stylistic features) and e in this way e to the age of the artefacts than to the function. This idea is supported by the fact that potteries of Inka and Inka-Colonial Period are entirely classified into the petrographic group I, while potteries of the Late Intermediate Period are only from the petrographic group II. Another important observation relates to the Inka Period ceramic assemblage. On the one hand, Inka style (high-quality Inka Imperial and Inka pacaje) potteries can almost entirely (96%) be connected to the petrographic group I (and about two-third to subgroup I/B). On the other hand, lower-quality Inka-imitating local potteries, artefacts with a mixed style (of both Late Intermediate and Inka) and Late Intermediate Period ceramics excavated from Inka buildings are almost entirely (93%) classified into the petrographic group II. These observations suggest that the local alluvial sediments of the Paria Basin were utilized for pottery manufacturing from the earlier periods (Late Intermediate) and stayed in use during the Inka Period for the making a local product imitating the imperial fashion (i.e. the Inka local style). In addition, this raw material is

15

basically used for lower quality coarse ware production. In contrast, pyroclastic rocks of the surroundings of the Paria Basin became commonly used as tempering materials of fine wares during the Inka period and especially for the Inka Imperial style pottery. 6.2. Technology Information on the manufacturing technology of pottery could be obtained from the fabric investigations by petrographic microscope and phase analysis by XRD. The results of the former method have shown a basic difference concerning the three petrographic groups of ceramics. The first group of ceramics was formed from a mixture of artificially added temper and a relatively pure (elutriated?) clay paste which proved to be a high quality, fine to medium-grained raw material. The pumiceous clasts of subgroup I/A are artificially added (temper) since fresh, angular grains could not survive as natural components of a clayey sediment (weathering product). Unweathered, angular clasts of subgroup I/B are also added to the paste as temper. The only exceptions are the representatives of subgroup I/D (these samples have an almost serial fabric, so it is not very probable that it was created intentionally). The utilization of unweathered rock fragments has modern analogies (Velde and Druc, 1999b; Ixer and Lunt, 1991). The method of the preparation could be e similarly to the present-day techniques e the exploiting, crushing, sieving and mixing with the clay paste. The second ceramic group contains fragments of vessels manufactured from untempered silty clay which was a natural mixture and potters used this lower quality, coarser grained material without any significant elaboration (e.g. elutriation). Clastic components of group III are natural parts of the mixture, and the fabric suggests the usage of the primary weathering product near to the rock outcrop as the raw material. It means that the third group of pottery was made from a natural (untreated), coarse grained, sandy-silty clay which could be exploited from the immediate vicinity of the rock outcrop. The XRD technique provided information on the firing conditions of the ceramic production. In contrast with the case of the raw material preparation types, it is not possible to make significant distinctions between the first and second petrographic groups of ceramics. Characterizing these pottery groups together, it can be said that the majority of the ceramics of Paria was fired at the same or similar circumstances, namely at 700e900  C maximum firing temperature and in a dominantly oxidizing but varying atmosphere. These statements are just partly in agreement with the former conclusions on Inka pottery. For coastal and Lake Titicaca region ceramics Wagner et al. (1989) suggested a higher, 850e1000  C firing temperature by Mössbauer spectroscopic investigations. The atmosphere of the pottery firing in Paria is similar to the typical Inka imperial ceramic style, which is hard reduction fired with surface oxidation (sandwich structured) or oxidized throughout in cross section (Menzel and Riddell, 1986). In contrast, the third petrographic group of pottery can be distinguished from the former ones concerning even the firing conditions. The presence of the neoformed, high-temperature phase (mullite) suggests a higher (850e950  C) maximum firing temperature. The final surface colour is white and in cross section white, light grey or pale rose. These differences e together with the foreign material utilization e suggest a pottery making technology isolated from the Inka imperial fashion or the typical highland traditions (Ixer and Lunt, 1991). However, the vessel shapes and principally their decorations are typical Inka forms (plates) which can rather indicate a separated, well defined Inka pottery manufacturing workshop. The above estimated firing temperatures do not contradict the idea of open pit firing suggested by others (Ixer and Lunt, 1991; Hayashida, 1998; Sillar, 1997). Although some pottery workshop

Author's personal copy

16

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17

sites have been identified by finding equipments of manufacturing (e.g. Hayashida, 1998; Donnan, 1997), no actual Inka kiln remnants have ever been found. 7. Conclusions The first detailed archaeometrical investigations on Inka Period ceramics from the territory of Bolivia were carried out on excavated and surface finds of the Paria Basin (Dept. of Oruro). To get information on the raw material usage and the manufacturing techniques applied by the potters, we made a comparative analysis of the ceramic fragments, other archaeological artefacts (adobes/plasters, building stones) and potential raw materials (fine grained sediments for the clay paste and clastic materials for the temper). Applying classical geological methodology (petrology, instrumental mineralogical and chemical investigations), answers to the three main questions raised in this paper are the followings: (1) Raw materials and sources used during the Late Intermediate Period continued to be utilized during the Inka Period. However, some additional raw materials, sources and techniques also came in fashion. The reason of the introduction of the “new” materials and techniques could be the demand for a larger amount and higher quality of vessels in the imperial system. (2) Neither Inka Imperial nor other Inka Period ceramics e except for the special white ware e proved to be an import product. Their raw material sources could be identified in the surroundings of the archaeological site of Paria or the Paria Basin. This statement supports the idea that e instead of transporting vessels e imperial handicraft techniques and/or artisans were brought to Paria from other regions of the Empire to supply the demands of the high level standardized pottery manufacture, as also occurred in the region of Lake Titicaca (Murra, 1978; Spurling, 1992). (3) Inka Imperial and local-style vessels show clear differences in their material. For Inka local pottery manufacturing, the nearest (max. 5 km distant), local raw material sources (petrographic group II) were used. These sources had functioned already in the Late Intermediate Period and the tradition stayed alive even after the Inka conquest of the Paria Basin, and during the standardization process which influenced the local ceramic manufacture. For Inka Imperial pottery making, two additional near (10e30 km distance) sources supplied tempering materials (petrographic group I). These volcanic related materials (just like the tempering technique itself) gained an important role in the ceramic handicraft during the Inka period most probably because of the quality requirements and/or the quantity demand of the newly introduced, Empire-related ceramic style. An important result is that the limited group of pottery with unique outlook (group III of ceramics) proved to have foreign origin, suggesting a long distance connection with a different region of the Inka Empire. Taken altogether, we find that the pottery supply system of the Paria region is rather complex. Our applied geological study provided useful and fundamentally new information about Inka Period pottery making system of the Paria Basin for the archaeological research. Acknowledgements The archaeological project was carried out with the permission of the Dirección Nacional de Arqueología, Bolivia, and was

financially supported by the OTKA Hungarian Scientific Research Fund (Grant T 047048), the Curtiss T. & Mary G. Brennan Foundation and the Heinz Foundation. The PGAA experiments have been done at the Budapest Neutron Centre, with the support of NAP VENEUS05 Contract No. OMFB 00184/2006. The geological field work was supported by the Hungarian Geological Society. The authors are thankful to Carola Condarco (project Co-Director), Alvaro Condarco Castellón and Mile Vargas Rosquellas (Oruro, Bolivia). A special thank to Margit Csömöri (Dept. of Petrology and Geochemistry, Eötvös Loránd University of Budapest) for giving efficient help in the sample preparation. For the careful reading and linguistic corrections of the manuscript, many thanks to dr. Jesse L. Weil and dr. Katalin Gherdán. References Alden, J.R., Minc, L., Lynch, T.F., 2006. Identifying the sources of Inka period ceramics from northern Chile: results of a neutron activation study. Journal of Archaeological Science 33, 575e594. Arnold, D.E., 1972. Mineralogical analyses of ceramic materials from Quinua, Department of Ayacucho, Peru. Archaeometry 14/1, 93e102. Arnold, D.E., 2005. Linking society with the compositional analyses of pottery: a model from comparative ethnography. In: Livingstone Smith, A., Bosquet, D., Martineau, R. (Eds.), Pottery Manufacturing Processes: Reconstruction and Interpretation. BAR International Series, vol. 1349. Archaeopress, Oxford, pp. 15e21. Belgya, T., Révay, Zs, 2004. Gamma-ray spectrometry. In: Molnár, G.L. (Ed.), Handbook of Prompt Gamma Activation Analysis with Neutron Beams. Kluwer Academic Publisher, Dordrecht, pp. 71e111. Bertolino, S.R., Fabra, M., 2003. Provenance and ceramic technology of pot sherds from ancient Andean cultures at the Ambato valley, Argentina. Applied Clay Science 24, 21e34. Bray, T., 2004. La alfafería imperial Inka: una comparación entre la cerámica estatal del area de Cuzco y la cerámica de las provincias. Chungará e Revista de Antropología Chilena 36, 365e374. Bray, T.L., Minc, L.D., Ceruti, M.C., Chávez, J.A., Perea, R., Reinhard, J., 2005. A compositional analysis of pottery vessels associated with the Inca ritual of capacocha. Journal of Anthropological Archaeology 24, 82e100. Choi, H.D., Firestone, R.B., Lindstrom, R.M., Molnár, G.L., Mughabghab, S.F., PaviottiCorcuera, R., Révay, Zs., Trkov, A., Zerkin, V., Chunmei, Z., 2007. Database of Prompt Gamma Rays from Slow Neutron Capture for Elemental Analysis. International Atomic Energy Agency, Vienna. Cieza de León, P., 1973. La crónica del Perú [1553]. Promoción Editorial Inca S.A., Lima. Cieza de León, P., 1985. Crónica del Perú. Segunda parte [1553]. Pontificia Universidad Católica, Lima. Costin, C.L., 2001. Production and exchange of ceramics. In: D’Altroy, T.N., Hastorf, C.A. (Eds.), Empire and Domestic Economy. Kluwer Academic/Plenum, New York, pp. 203e240. D’Altroy, T.N., 1992. Provincial Power in the Inka Empire. Smithsonian Institution Press, Washington, D.C., 272 pp. De Corte, F., 1987. The k0-standardization method e a move to optimization of NAA. PhD Thesis, University of Gent. Donnan, C.B., 1997. A Chimu-Inka ceramic-manufacturing center from the north coast of Peru. Latin American Antiquity 8, 30e54. Fazekas, B., Östör, J., Kis, Z., Molnár, G.L., Simonits, A., 1997. The new features of Hypermet-PC. In: Molnár, G., Belgya, T., Révay, Zs (Eds.), Proceedings of the 9th International Symposium on Capture Gamma-Ray Spectroscopy and Related Topics, Budapest, Hungary, October 8e12. Springer Verlag, Budapest, pp. 774e778. Fazekas, B., Révay, Zs., Östör, J., Belgya, T., Molnár, G., Simonits, A., 1999. A new method for determination of gamma-ray spectrometer nonlinearity. Nuclear Instruments and Methods A 422, 469e473. GEOBOL, 1992. Carta Geológica de Bolivia, Hoja Oruro (1:100 000) Publicación SGB Serie I-CGB-11 (Página 6140). GEOBOL, 1994. Carta Geológica de Bolivia, Hoja Bolivar (1:100 000) Publicación SGB Serie I-CGB-27 (Página 6240). Govindarau, K., 1989. Compilation of working values and sample description for 272 geostandards. Special Issue of Geostandards Newsletter, XIII. Guaman Poma de Ayala, F., 1980. El primer nueva corónica y buen gobierno. [1613]. In: Murra, J.V., Adorno, R., (Eds.), Vol. 3, Siglo XXI, México D.F. GUM, 1993. Guide to the Expression of Uncertainty in Measurement. ISO, Geneva. Gutiérrez Osinaga, D.J., 2005. Avances en la arqueología de caminos precolombianos en Bolivia tramo: Paria-Tapacarí (Sitios asociados y características formales de construcción del camino). Nuevos Aportes 3, 93e114. Gyarmati, J., Varga, A., 1999. The Chacaras of War. An Inka State Estate in the Cochabamba Valley, Bolivia. Museum of Ethnography, Budapest. Hahn-Weinheimer, P.F., Hirner, A.V., Weber-Diefenbach, K., 1984. Grundlagen und praktische Anwendung der Roentgenfluorezenzanalyse (RFA). Vieweg-Verlag, Braunschweig.

Author's personal copy

V. Szilágyi et al. / Journal of South American Earth Sciences 36 (2012) 1e17 Hayashida, F., 1998. New insight into Inka pottery production. MASCA Research Papers in Science and Archaeology, Supplement to vol. 15(Andean Ceramics), pp. 313e335. Hayashida, F., 1999. Style, technology, and state production: Inka pottery manufacture in the Leche Valley, Peru. Latin American Antiquity 10, 337e352. Hayashida, F., Glascock, M., Häusler, W., Neff, H., Riederer, J., Wagner, U., 2002. Technology and organization of Inka pottery production: archaeometric perspectives. In: Jerem E.T., Biró K., (Eds.), Proceedings of the 31st International Symposium on Archaeometry, Central European Series 1, II, BAR International Series 1043, Archaeopress-Archaeolingua, Oxford, pp. 573e580. Hayashida, F., Häusler, W., Wagner, U., 2003a. Technology and organisation of Inka pottery production in the Leche Valley. Part I: study of clays. Hyperfine Interactions 150, 141e151. Hayashida, F., Häusler, W., Riederer, J., Wagner, U., 2003b. Technology and organisation of Inka pottery production in the Leche Valley. Part II: study of fired vessels. Hyperfine Interactions 150, 153e163. Hyslop, J., 1984. The Inka Road System. Academic Press, Orlando. Ixer, R.A., Lunt, S., 1991. The petrography of certain pre-Spanish pottery from Peru. In: Middleton, A., Freestone, I. (Eds.), Recent Developments in Ceramic Petrology. Occasional Paper 81. British Museum Publications, London, pp. 137e164. Jamieson, R.W., Hancock, R.G.V., 2004. Neutron activation analysis of colonial ceramics from Southern Highland Ecuador. Archaeometry 46, 569e583. Lindstrom, R.M., Révay, Zs, 2004. Beams and facilities. In: Molnár, G.L. (Ed.), Handbook of Prompt Gamma Activation Analysis with Neutron Beams. Kluwer Academic Publishers, Dordrecht, pp. 31e58. McLennan, S.M., 1989. Rare earth elements in sedimentary rocks: influence of provenance and sedimentary processes. In: Lipin, B.R., McKay, G.A. (Eds.), Geochemistry and Mineralogy of Rare Earth Elements. Reviews in Mineralogy 21, 169e200. McLennan, S.M., 2001. Relationships between the trace element composition of sedimentary rocks and upper continental crust. Geochemistry, Geophysics, Geosystems 2, 2000GC000109, 24 pp. Menzel, D., Riddell, F.A., 1986. Archaeological Investigations in Tambo Viejo, Acarí Valley, Peru 1954. Institute for Peruvian Studies, Sacramento. Molnár, G.L., Révay, Zs., Paul, R.L., Lindstrom, R.M., 1998. Prompt-gamma activation analysis using the k0 approach. Journal of Radioanalytical and Nuclear Chemistry 234, 21e26. Molnár, G.L., Révay, Zs., Belgya, T., 2002. Wide energy range efficiency calibration method for Ge detectors. Nuclear Instruments and Methods A 489, 140e159. Morgan, G.B., London, D., Luedke, R.G., 1998. Petrochemistry of Late Miocene Peraluminous Silicic volcanic rocks from the Morococala field, Bolivia. Journal of Petrology 39, 601e632. Morris, C., 1978. The archaeological study of Andean exchange systems. In: Redman, C.L. (Ed.), Social Archaeology: Beyond Subsistence and Dating. Academic Press, New York, pp. 289e310. Murra, J.V., 1972. El “control vertical” de un máximo de pisos ecológicos en la economía de las sociedades andinas. In: de Iñigo Ortíz de Zuñiga (Ed.), Visitas de la Provincia de León de Huánuco en 1562, Tomo II, pp. 429e468. Murra, J.V., 1978. Los olleros del Inka: hacia una historia y arqueología del Qollasuyu. In: Quesada, F.M.F., Pease, D. (Eds.), Historia Problema y Promesa: Homenaje a Jorge Basadre Sobrevilla. Pontificia Universidad Católica del Perú, Lima, pp. 415e423. Murra, J.V., 1983. La organización económica del Estado Inca, third ed. Siglo XXI, México D.F. Nance, W.B., Taylor, S.R., 1976. Rare earth patterns and crustal evolution e I. Australian post-Archean sedimentary rocks. Geochimica et Cosmochimica Acta 40, 1539e1551. Phillips, G.W., Marlow, K.W., 1976. Automatic analysis of gamma-ray spectra from germanium detectors. Nuclear Instruments and Methods 137, 525e536. Plá, R., Ratto, N., 2003. Provenance archaeological studies of ceramic raw materials and artefacts using instrumental neutron activation analysis: the cases of Chaschuil and Bolsón de Fiambalá (Catamarca, Argentina). In: Nuclear Techniques in Archaeological Investigations. IAEA, Technical Reports Series No. 416, Vienna, pp. 45e69. Ratto, N., Orgaz, M., De La Fuente, G., Plá, R., 2002. Ocupación de pisos de altura y contexto de producción cerámica durante el Formativo: El caso de la región puneña de Chaschuil y su relación con el Bolsón de Fiambalá (Depto. Tinogasta, Catamarca, Argentina). Estudios Atacameños 24, 51e69. Ratto, N., Orgaz, M., Plá, R., 2004. La explotación del alfar de La Troya en el tiempo: causalidad o memoria (Departamento Tinogasta, Catamarca, Argentina). Chungará e Revista de Antropología Chilena 36 (2), 351e363. Ratto, N., De La Fuente, G., Plá, R., Orgaz, M., Moreno, M., 2005. Compositional characterization of ceramic artefacts and clays: the utility of INAA to evaluate the prehispanic integration between the Puna and the Bolsón de Fiambalá Area, Northwestern Argentina. In: Kars, H., Burke, E. (Eds.), Geoarchaeological and

17

Bioarchaeological Studies 3, Proceedings of the 33rd International Symposium on Arcaheometry. Amsterdam, pp. 245e248. Redwood, S.D., 1987. The Soledad caldera, Bolivia: a Miocene caldera with associated epithermal Au-Ag-Cu-Pb-Zn mineralization. Geological Society of America Bulletin 99, 395e404. Repartimiento de Tierras por el Inca Huayna Capac [1556]. Versión Paleográfica de Don Adolfo de Morales, 1977. Universidad Mayor de San Simon, Cochabamba. Révay, Zs, 2006. Calculation of uncertainties in prompt gamma activation analysis. Nuclear Instruments and Methods A 564, 688e697. Révay, Zs, 2009. Determining elemental composition using prompt gamma activation analysis. Analytical Chemistry 81, 6851e6859. doi:10.1021/ac9011705. Révay, Zs, Belgya, T., 2004. Principles of PGAA method. In: Molnár, G.L. (Ed.), Handbook of Prompt Gamma Activation Analysis with Neutron Beams. Kluwer Academic Publishers, Dordrecht, pp. 1e30. Révay, Zs., Belgya, T., Ember, P.P., Molnár, G.L., 2001a. Recent developments in Hypermet-PC. Journal of Radioanalytical and Nuclear Chemistry 248, 401e405. Révay, Zs., Molnár, G.L., Belgya, T., Kasztovszky, Zs, Firestone, R.B., 2001b. A new gray spectrum catalog and library for PGAA. Journal of Radioanalytical and Nuclear Chemistry 248, 395e399. Révay, Zs., Belgya, T., Molnár, G.L., 2005. Application of Hypermet-PC in PGAA. Journal of Radioanalytical and Nuclear Chemistry 265, 261e265. Révay, Zs., Belgya, T., Szentmiklósi, L., Kis, Z., 2008. Recent developments in prompt gamma activation analysis in Budapest. Journal of Radioanalytical and Nuclear Chemistry 278, 643e646. Révay, Zs., Firestone, R.B., Belgya, T., Molnár, G.L., 2004. Catalog and atlas of prompt gamma rays. In: Molnár, G.L. (Ed.), Handbook of Prompt Gamma Activation Analysis with Neutron Beams. Kluwer Academic Publishers, Dordrecht, pp. 173e364. Révay, Zs., Molnár, G.L., 2003. Standardisation of the prompt gamma activation analysis method. Radiochimica Acta 91, 361e369. Révay, Zs., Molnár, G.L., Belgya, T., Kasztovszky, Zs, Firestone, R.B., 2000. A new gamma-ray spectrum catalog for PGAA. Journal of Radioanalytical and Nuclear Chemistry 244, 383e389. Rowe, J.H., 1944. An Introduction to the Archaeology of Cuzco. Papers of the Peabody Museum of American Archaeology and Ethnology, Vol. XXVII, No. 2. Harvard University, Cambridge, Massachusetts. Sillar, B., 1997. Reputable pots and disreputable potters: individual and community choice in present-day pottery production and exchange in the Andes. In: Cumberpatch, C.C., Blinkhorn, P.W. (Eds.), Not So Much a Pot, More a Way of life: Current Approaches to Artefact Analysis in Archaeology. Oxbow Monograph Series, vol. 83, Oxford, pp. 1e20. Spurling, G., 1992. The Organization of Craft Production in the Inka State: The Potters and Weavers of Milliraya. Ph.D. Thesis, Department of Anthropology, Cornell University, Ithaca. Szilágyi, V., Szakmány, Gy., . Comparison of volcaniclastic-tempered Inca Imperial ceramics from Paria, Bolivia with potential sources. In: Quinn, P.S. (Ed.), Interpreting Silent Artefacts: Petrographic Approaches to Archaeological Ceramics. Archaeopress, Oxford, pp. 211e225. Szilágyi, V., Szakmány, Gy, Gyarmati, J., 2005. Inka kori kerámiák petrográfiai } zetes eredményei (Paria, Bolívia). (Preliminary results of the vizsgálatának elo petrographic investigation of Inka Period ceramics (Paria, Bolivia)). Arche} hely 2005/2, 42e47 (in Hungarian). ometriai Mu Szilágyi, V., Szakmány, Gy., Gyarmati, J., Tóth, M., 2007. Preliminary comparative archaeometric results of colonial and Inka pottery in Paria (Oruro, Bolivia). In: Waksman, Y.S. (Ed.), Archaeometric and Archaeological Approaches to Ceramics: Papers Presented at EMAC’05, 8th European Meeting on Ancient Ceramics, Lyon 2005. BAR International Series, vol. 1691. Archaeopress, Oxford, pp. 195e199. Taylor, S.R., McLennan, S.M., 1985. The Continental Crust: Its Composition and Evolution. Blackwell Scientific Publications LTD, Oxford. Vaca de Castro Cavallero, C., 1908. Ordenanzas de tambos. [1543]. In: Revista Histórica, 3, pp. 427e492. Valera Guarda, V., 2002. Enseñanzas de alfareros toconceños: tradición y tecnología en la cerámica. Chungará e Revista de Antropología Chilena 34, 225e252. Velde, B., Druc, I.C., 1999a. Prehistoric Peru. In: Velde, B., Druc, I.C. (Eds.), Archaeological Ceramic Materials e Origin and Utilization. Springer, Berlin, pp. 225e237. Velde, B., Druc, I.C., 1999b. Modern ceramic production in the Andes. In: Velde, B., Druc, I.C. (Eds.), Archaeological Ceramic Materials e Origin and Utilization. Springer, Berlin, pp. 237e247. Wagner, U., Brandis, S.V., Marticorena, B., Salazar, R., Schwabe, R., Riederer, J., Wagner, F.E., 1989. Mössbauer studies of ceramics from the Inca period. In: Maniatis, Y. (Ed.), Archaeometry: Proceedings of the 25th International Symposium on Archaeometry. Elsevier, Amsterdam, pp. 159e168. Williams, V.I., Vargas, C.S., Romero, A., Speakman, R.J., Glascock, M.D., 2006. Inka Pottery Production and Consumption in NW Argentina, Northern Chile, and Bolivia. In: Abstract for the 71st Annual Meeting of the Society for American Archaeology, San Juan, Puerto Rico, 26e30 April 2006.