Early Holocene coca chewing in northern Peru.

Dillehay, Tom D. ; Rossen, Jack ; Ugent, Donald 等

[ILLUSTRATION OMITTED]

Introduction

Coca production and the mastication of its leaves have long been part of the indigenous Andean economy, providing nutritional, medicinal and digestive properties. The origin and use of the coca plant in South America have long been debated in anthropology, botany, medicine and Latin American politics. Although the history of coca chewing extends back into pre-Columbian times (Plowman 1979; Plowman & Hensold 2004), several issues are poorly understood. They are the dates by which people began to routinely exploit this plant, the specific ancient technologies employed to extract its alkaloid stimulants and nutritional qualities (Rury & Plowman 1983; Pacini & Franquemont 1985), and the wider social and economic impact of the crop on early community development.

This report is of a study of the macrobotanical remains of archaeological coca leaves (Erythroxylum novagranatense var. truxillense) preserved in the buried house floors of early Holocene foragers and cultivators that provides direct evidence of the consumption of this plant in tropical dry and humid forests on the lower slopes of the Andes in north-western Peru. Two AMS radiocarbon dates on leaves indicate that coca chewing began by at least 8000 cal BP, the earliest use known to date. Our data also reveal evidence for baking calcium-bearing rocks to produce lime to extract alkaloids from the leaves. The production of lime was segregated spatially from individual domestic households located across the valley, suggesting it was a community activity.

Use of coca

The genus Erythroxylum includes at least 230 species of coca that are distributed from Mexico and the Bahamas to north-west Argentina (Plowman 1979; Plowman & Hensold 2004). Only the cultivated species contains enough alkaloid to be worthy of chewing as a stimulant (Rury & Plowman 1983; Pacini & Franquemont 1985). Today, the species E. novogranatense var. truxillense is cultivated in semi-tropical to tropical areas on the lower western slopes of the Ecuadorian and Peruvian Andes. This small leaf variety was called tupa coca by the Inca, and was considered royal coca because it had a high content of wintergreen oil and other compounds (Rostworowski 1988). This variety is drought resistant and adapted to arid conditions, but still depends on irrigation on the western slopes of the Peruvian Andes. Its principal habitat is the tropical chaupiyiunga zone located at 500-1500m in elevation, where the average temperature ranges from 18-25[degrees]C (Plowman 1979; Ugent & Ochoa 2006). There are no known wild progenitors of E. coca novagranatense on the western slopes of the Andes, suggesting that it was developed from E. coca coca somewhere on the eastern slopes (Plowman 1983), or possibly in the upper Maranon Valley, and transported to the Nanchoc Valley. E. coca novagranatense is a variety selected through cultivation probably for its aridity tolerance and its aromatic oils which provide a preferred taste.

Prior to our radiocarbon dates on coca leaves, it was thought that the late Preceramic inhabitants of the Ecuadorian and Peruvian coast were consuming coca by at least 5000 cal BP (Engel 1963; Klepinger & Kuhn 1973; Lathrap et al. 1976; Cohen 1978). A calcite (or lime) alkali is required to release coca alkaloids. Powdered lime has been found in contexts dating to c. 5000 cal BP and presumed to be associated with coca chewing. The importance of coca was emphasised and exploited during the Inca and later Spanish colonial periods (Cieza de Leon 1973; Chavez Velasquez 1977). It was used as a stimulant to reduce fatigue, hunger, high-altitude hypoxia and thirst and as a medicine and digestive. The alkaloids in coca improve the metabolism of starches, thus providing a surge in energy and a reduction of fatigue. But coca was not just for chewing. It was also a symbol of social status and ethnic identity and an element of oracles and rituals (Plowman 1983). During Inca times, coca fields on the western slopes of the Andes were the centre of conflict, tension and cooperation between ethnic groups vying for control of this resource (Dillehay 1979; Marcus & Silva 1988; Rostworowski 1988). In recent decades, the plant has been the centre of controversy primarily due to the modern day extraction of cocaine from it. The political implications of cocaine production and the social, economic and psychological effects of its use have been well publicised (Plowman 1979; Plowman & Hensold 2004; Bieri et al. 2006). Although cocaine is one of the 14 alkaloids of the coca leaf, there is no evidence to suggest that it was ever extracted in pre-Colombian times.

[FIGURE 1 OMITTED]

Context

The transition in prehistory from mobile lifestyles to sedentary agricultural communities is evidenced at 46 Preceramic sites in the semi-tropical Nanchoc Valley, a branch of the Zana Valley situated at 500-1200m above sea level (Dillehay et al. 2005, 2007; Piperno & Dillehay 2008; Figure 1). Paleoecological studies indicate that both the humid and dry forest, a setting conducive to coca growing (Dillehay et al. 1989, 2005, 2007; Dillon & Cadle 1991), also existed in late Pleistocene to early Holocene times (Vuilleumier 1971; Simpson 1975; Figure 2).

The macrobotanical and starch grain remains of squash (Cucurbita), peanut (Arachis), a quinoa-like chenopod (Chenopodium), manioc (Manihot), cotton (Gossypium) and other plants were recovered from sealed hearths and floors and from the calculus of human teeth in 13 excavated house structures in the study area. This evidence, along with the bones of various large and small animal species, reveals initial broad spectrum subsistence and later a primary farming economy in the seasonally dry forest radiocarbon dated from about 11 000-6200 cal BP (Dillehay in press).

During the late Paijan and Las Pircas phases (c. 11 300-7800 cal BP), an economy dominated by plant collecting and hunting transitioned to crop production practised in gardens close to small circular houses. At the end of the Las Pircas phase (9700-7800 cal BP), public architecture in the form of two tear-drop shaped, multi-tiered, earthen- and stone-lined mounds appeared on the north-east side of the river (Figure 1, site CA-09-04; Figure 3, Zones A & C) in direct association with an off-mound lime-producing area at the same site (Figure 3, Zone B) (Dillehay et al. 1989). Radiocarbon dates on charcoal from hearths in mound and off-mound areas place the initial construction of this site at c. 8200-7500 cal BP.

[FIGURE 2 OMITTED]

Methods

The present study concerns five sites, two at Las Pircas and two at Tierra Blanca (CA-09-52, CA-09-27, CA-09-71 and CA-09-77 in Figure 1), and one across the river from them, CA-09-04. The latter contained two ritual mounds (Zones A & C) and a working area (Zone B). Coca leaves and associated lime came from both the Las Pircas and Tierra Blanca sites, and evidence for lime extraction from the site CA-09-04 on the other side of the Nanchoc River.

Excavation recovery technique and preservation

The excavated sites are not located in plough zones but in isolated, undisturbed and uninhabited side canyons and alluvial fans (quebradas) and are intact. In the study area, the seasonally dry forested environment allows for the preservation of desiccated plant remains when they are sealed in buried house floors, overlaid by grinding stones or stones fallen from the side walls of the dwellings, and thus protected from seasonal rains. The coca leaves from the late Las Pircas and Tierra Blanca phase domestic sites were recovered from hard-packed floors buried at 30-85cm in depth in small elliptical (c. 2.1-2.3m diameter) stone-lined structures at CA-09-71 and CA-09-77. Other plant remains were carbonised or desiccated and recovered in screens and through water flotation.

[FIGURE 3a OMITTED]

Coca leaf analysis

The archaeological and modern coca leaves were examined under a stereoscopic microscope at a magnification of 20-400 x. A modern reference collection of coca leaves from different varieties of Erythroxylum and other species available at the Department of Botany, Southern Illinois University, Carbondale, the Laboratorio de Arqueobiologia at the Universidad Nacional de Trujillo, Peru, and online at the Field Museum of Natural History, Chicago, Illinois, were used to confirm the identity of the coca leaves. The late Timothy Plowman (pers. comm. 1988) preliminarily identified the taxa of the first two coca leaves excavated at Nanchoc Valley sites. Identification of more recently excavated leaves was made by our co-authors: botanists Ugent and Vazquez.

[FIGURE 3b OMITTED]

The few studies (Molina et al. 1989; Cortella et al. 2001; Johnson et al. 2003) available on coca starch grains and other microscopic features also were consulted. Cross- or X-shaped crystals were observed on the surface of both modern and archaeological leaves. A distinctive central panel, a prominent feature of modern leaves, was observed on the archaeological specimens.

We also undertook chemical analysis at all the sites, including the three zones at CA-09-04 (Figure 3), to detect, characterise and map the use of lime and calcite and investigate its production and its possible association with coca chewing (Tables 1-3; Technical Note).

Results

Coca leaves

A total of seven coca leaves were recovered from sealed floors in two late Las Pircas phase houses (Figures 4-6; CA-09-71 and CA-09-77 in Figure 1) and one Tierra Blanca phase house (CA-09-77, 7800-5000 cal BP). Five leaves were agglutinated, suggesting they had been chewed. The archaeological leaves are generally well preserved, although thin, fragile and diaphanous due to their age and buried contexts. The best preserved leaves reveal eucamptodromus to broquidromus vein patterns. This type of foliar venation is observed in modern coca specimens (Figure 6a), which were collected in Cascas, Cajamarca, located in a tropical dry forest about 80km from the Nanchoc Valley. While leaf vein patterns are often used to identify coca leaves to the genus level, some taxonomists disagree on the exact microscopic identification indicators of coca to species. Most of this difficulty relates to distinguishing E. coca var. coca and E. coca var. ipadu, two eastern tropical forest varieties (Molina et al. 1989; Cortella et al. 2001; Ugent & Ochoa 2006: 126-30). In contrast, the cellular inclusions of the archaeological specimens, including spiral and reticular vessels, pericytic fibres with calcium oxlate inclusions and pericytic stomas are typical of the species E. novogranatense var. truxillensis (Rusby) Plowman (cf. Rury & Plowman 1983; Cortella et al. 2001; Johnson et al. 2003; Bieri et al. 2006). These inclusions contain cubic crystals in the vein, which measure approximately two microns. Based on the combination of macroscopic and microscopic traits, especially with the geographic location of the leaves on the lower western slopes of the northern Peruvian Andes, we believe these specimens to be the variety that has been traditionally classified as var. truxillensis (Rusby) Plowman. We use this latter term, although there has been a recent reclassification of this variety as E. novogranatense vat. truxillensis (Rusby) Ugent (Ugent & Ochoa 2006: 128-9).

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

Two AMS dates on leaves from Las Pircas house floors were processed at 7950 cal BP (7120 [+ or -] 50 BP: Beta-226458) and 7920 cal BP (7080 [+ or -] 40 BP: Beta-226221). These dates agree with those arrived at previously for several food crops, such as squash, peanuts, pacay, beans and quinoa, recovered from these and other houses (Dillehay et al. 1989, 2007; Piperno & Dillehay 2008).

[FIGURE 6 OMITTED]

Lime processing

Directly associated with the leaves were 11 irregularly-shaped spheroids of burned and precipitated calcium, which represent lime evidently processed for use with the leaves.

Three other Las Pircas huts (CA-09-52, CA-09-27, CA-09-50) also yielded the remains of conically-shaped pieces of processed calcite dated as early as 9000-8300 cal BP (7600 [+ or -] 60 BP), but no coca leaves (Rossen 1991). The same processed lime, also without coca leaves, was excavated from the mounds and the off-mound activity areas across the river in CA-09-04, revealing a direct cultural connection between the mounds and the domestic sites.

The mounds on site CA-09-04 (Figures 1 & 3) were constructed on an isolated alluvial fan 2-3km away from the domestic sites containing coca leaves (Dillehay et al. 1989). The mounds (Zones A & C) are [sup.14]C dated between 8400 and 7800 cal BP and the off-mound area (Zone B) between 8100 and 7500 cal BP (Dillehay et al. 1997). The largest preserved mound is 1.2m high and 25m long (Figure 3, Zone A). Portions of the second mound (C) have been destroyed by a modern-day cemetery, which impeded our work there. We studied the early building phases of the mounds and the off-mound area by excavating trenches and block excavations. The mounds were constructed in multiple stages. Only unifacial flakes made of local andesite and basalt, some with burned calcite deposits on the used edges, were found in association with hearths and burned stains in the mound and off-mound areas.

Stains in the off-mound areas were produced by the process of precipitating lime from calcite. These stains measure c. 50cm in diameter and are 1-2cm in depth. Flakes with burned lime residues, chunks of lime, as well as clusters of rock crystals, were associated with only the mounds. It is the spatial isolation of the mounds, their multi-tiered design, and the presence of the rock crystals that lead us to believe that the production process was probably ritual related. Rock crystals are exotic to the valley and considered sacred items. They are found only in the mounds and in one furowed garden plot at CA-09-27 (Rossen 1991; Dillehay et al. 1989, 1997; Dillehay 2004). They occur in later Formative ceremonial platform mounds (Dillehay 2004) throughout the valley but never in domestic structures.

Soil chemical analysis of the stains at the mound site (CA-09-04) showed unusually high amounts of potassium and calcium in comparison to non-cultural control sediments in off-site areas (Table 1). The calcium enrichment probably resulted from the burning and precipitation of the lime from calcium-bearing rocks, which are available in the nearby mountains. Additional evidence for processing lime is shown on the edges of several unifacial flaked tools, which exhibit scraped, burned and unburned calcite powder. The same powder is recovered in the form of compressed conical and round chunks of lime (Figures 7 & 8) from the mounds and the stains and hearths in off-mound areas (Zone B) of CA-09-04 and from buried floors containing coca leaves in the dwellings (CA-09-52, CA-09-50 and CA-09-27).

Due to the destructive nature of some analyses, mineralogical and total elemental (TE) studies were performed on only two of the leaves shown in Figure 4b, which indicate moderate calcium (86 130 and 69 880 in mg/Kg) and potassium (120 990 and 108 450 in mg/Kg) content possibly derived by using lime to masticate them. Moderately high amounts of calcium and potassium were recovered from domestic house floors dated to both the Las Pircas and Tierra Blanca phases (Table 3) (Figure 1, CA-09-27, 50, 52, 71). We interpret the high content of calcium and potassium in the leaves and on the hut floors as evidence of an alkali extractive agent activated by coca chewing.

Today, the Aymara and Quechua people of Bolivia and southern Peru make lime or katawi by burning calcium-rich rocks and grinding the residue into a powder (Baker & Mazess 1963; Duke et al. 1975). The powder is mixed with water, ash and salt and pressed into small concretions for use as an alkali and masticated with coca leaves to release alkaloids. Although direct archaeological evidence is lacking, it is possible that some lime was used as a mineral supplement to tubers and chenopodium grains, both of which have been recovered from hut floors in the Nanchoc domestic sites.

Mineralogical, XRF, and TE analyses were performed on samples of burned, but unprocessed, calcite from off-mound activity areas in CA-09-04 (Sample A, Figure 7), and precipitated and compressed calcite from a house floor at CA-09-77 (Sample B, Figure 8), a domestic site located across the valley (Figure 1). The analysis suggested differences between the two samples (Table 2). Even though both contained about the same amount of total carbon, sample B had 15 times higher organic carbon than sample A. The high potassium content of sample B is associated with its ash content, which was evident from its darker colour before and after ignition. The dark bluish green hue of sample B after oxidation may indicate the presence of coca leaf extract material. This suggestion is consistent with the higher potassium, magnesium, sulphur, nitrogen and trace metal levels in the sample. The presence of sulphur is associated with gypsum, which was apparently used as a salt to stabilise the coca extract following its extraction with calcite lime. The salt may have been added as gypsum or it may have been formed as a precipitate following the reaction of calcium contained in the lime and sulphur contained in the coca leaves during the grinding process. The poor crystalline nature of gypsum in sample B also supports the latter interpretation. Overall, the findings suggest that calcite from sample A in the off-mound area of CA-09-04 was the source for lime extraction used in the process, while the sample B material from excavated house CA-09-77 represents the processed product.

[FIGURE 7 OMITTED]

[FIGURE 8 OMITTED]

Although we performed multiple technical studies on only two samples, we believe that they represent the content and function of other raw and processed lime specimens, based on similar type, context and form, and we argue that the CA-09-04 mounds were built to spatially separate the production of lime as a community effort. The spatial isolation of the mounds and the presence of rock crystals in and around them suggest a ritual related lime-extraction technology. More detailed chemical analysis involving chromatographic separation and mass spectrometric identification are required to make additional comparisons between all specimens.

Conclusion

Coca chewing accompanied by the specialised provision of its adjunct, lime, was a feature of the rise of agriculture and social complexity in the region. Our evidence suggests that by 8000 cal BP there was an effective farming system in the Nanchoc Valley, employing a range of seed, tree, vegetable and root crops, coupled with the exploitation of animals and wild plants, which provided a balanced, nutritious and stable diet for the inhabitants of the tropical western slopes of the northern Peruvian Andes. As part of the diet, coca probably provided nutritional and medicinal benefits. The current evidence suggests that coca chewing took place in the households and not in communal ceremonies. Contemporary Preceramic sites in other areas of the valley do not yield coca and lime. Our data also continue to demonstrate that the practice of crop production was not spread evenly across all households and all areas at the same time and that certain crops were introduced with particular cultural traits (Dillehay in press). Cotton was an industrial crop introduced with small-scale irrigation technology; and peanuts, squash, a quinoa-like chenopod and manioc were adopted earlier at a time when a foraging way of life was shifting to an agricultural one (Dillehay et al. 2005, 2007; Piperno & Dillehay 2008). Different social and economic processes must have been associated with each of these developments and their uneven spread across the landscape.

In the chronology of the Nanchoc Valley, e. 8000-7800 cal BP (e. 7000 BP) represents the end of the Las Pircas phase and the beginning of the Tierra Blanca phase. This was a significant transitional time in the region's prehistory. The addition of local coca (and later cotton) to a previously existing suite of exotic cultivated plants, along with the construction of the mounds at Nanchoc, were accompanied by other significant cultural trends, including irrigation and lime extraction technologies. Preceramic peoples were becoming less cosmopolitan, in terms of exchange and outside contacts, and more localised in resource use and procurement. They became more pragmatic in their manufacture of unifacial stone tools, including a greater reliance on expedient stone tools (Rossen & Dillehay 2002). Furthermore, a variety of ritual and production activities was moving from individual households to a public setting, and domestic sites were shifting downward from smaller lateral valleys toward the main valley floor where more fertile soils for crop production existed. These were the beginnings of trends that would later culminate in the development of aggregated, multiple household agricultural field systems in the main valley and a wide range of cultural intensifications, including monumental architecture, which began to characterise Andean prehistory about 5000 cal years ago (Dillehay in press).

Acknowledgements

Research for this project was funded by the National Science Foundation, the University of Kentucky, Vanderbilt University, the Earthwatch Foundation, and the Heinz Foundation. We thank the Instituto Nacional de Cultura of Lima, Peru, for granting us the permits to carry out the research. We are also grateful to the late Timothy Plowman for encouraging our research through the 1980s and for classifying the first leaves excavated at sites.

Received: 4 January 2010; Accepted: 25 March 2010; Revised: 30 March 2010

References

BAKER, P.T. & R.B. MAZESS. 1963. Calcium: unusual sources in the highland Peruvian diet. Science 142: 1466-7.

BIERI, S., A. BRACHET, J.L.VEUTHEY & P. CHRISTEN. 2006. Cocaine distribution in wild Erythroxylum species. Journal of Ethnopharmacology 103: 439-47.

CHAVEZ VELASQUEZ, N.A. 1977. La materia medica en el Incanato. Lima: Editorial Mejia Baca.

CIEZA DE LEON, P. DE. 1973. La cronica del Peru (1553). Lima: Ediciones PEISA.

COHEN, M.N. 1978. Archaeological plant remains from the central coast of Peru. Nawpa Pacha 16: 36-7.

CORTELLA, A.R., M.L. POCHETTINO, A. MANZO & G. RAVINA. 2001. Erythroxylum coca: microscopical identification in powdered and carbonized archaeological material. Journal of Archaeological Science 28: 787-94.

DILLEHAY, T.D. 1979. Pre-Hispanic resource sharing in the central Andes. Science 204: 24-31.

--2004. Social integration, public landscape, and uncertainty in formative Peru. Journal of Social Archaeology 4: 268-92.

--(ed.) In press. From foragers to farmers in the Andes: new perspective on food production and social organization. Cambridge: Cambridge University Press.

DILLEHAY, T.D., P.J. NETHERLY & J. ROSSEN. 1989.

Early Preceramic public and residential sites on the forested slope of the western Andes, northern Peru. American Antiquity 54: 733-59.

--1997. The Nanchoc tradition: the beginnings of Andean civilization. American Scientist 85: 46-56.

DILLEHAY, T.D., H. ELING & J. ROSSEN. 2005. Preceramic irrigation canals in the Peruvian Andes. Proceedings of the National Academy of Sciences USA 102: 17241-4.

DILLEHAY, T.D., J. ROSSEN, T. ANDRES & D. WILLIAMS. 2007. Preceramic adoption of peanut, squash and cotton in northern Peru. Science 316: 1890-93.

DILLON, M.O. & J.E. CADLE. 1991. Bosque Monteseco: a cloud forest above the Peruvian desert. Bulletin of the Field Museum of Natural History 1: 14.

DUKE, J.A., D. AULIK & T. PLOWMAN. 1975. Nutritional value of coca. Botanical Museum Leaflets 24: 113-9.

ENGEL, F. 1963. A Preceramic settlement on the central coast of Peru: Asia, unit 1. Transactions of the American Philosophical Society 53: 1-139.

JOHNSON, E.L., J.A. SAUNDERS, S. MISCHKE, C.S. HELLING & S.D. EMCHE. 2003. Identification of Erythroxylum taxa by AFLP DNA analysis. Phytochemisty 64: 187-97.

KARATHANASIS, A. & B.F. HAJEK. 1992. Revised methods for rapid quantitative determination of minerals in soil clays. Journal of the Soil Science Society of America 46: 419-25.

KARATHANASIS, A. & D.L. SPARKS (ed.). 1996. Methods of soil analysis. Part 3, Chemical methods (Soil Science Society of America book series 5). Madison (WI): Soil Science Society of America: American Society of Agronomy.

KLEPINGER, L. & J. KUHN. 1973. Prehistoric dental calculus gives evidence for coca in early coastal Ecuador. Nature 269: 506-507.

LATHRAP, D., D. COLLIER & H. CHANDRA. 1976. Ancient Ecuador: culture, clay, and creativity, 3000-300 BC. Chicago (IL): Field Museum of Natural History.

MARCUS, J. & J. SILVA. 1988. The Chillon Valley coca lands: archaeological background and ecological context, in M. Rostworowski de Diez Canseco (ed.) Conflicts over coca fields in XVIth-century Peru (Memoirs of the Museum of Anthropology, University of Michigan 21): 1-32. Ann Arbor (MI): University of Michigan, Museum of Anthropology.

MOLINA, Y., T. TORRES, E. BELMONTE & C. SANTORO. 1989. Uso y posible cultivo de coca (Eryhtroxylum spp.) en epocas prehispanicas en los valles de Arica. Chungara 23: 37-49.

PACINI, D. & C. FRANQUEMONT (ed.). 1985. Coca and cocaine: effects on people and policy in Latin America (Cultural survival report 23). Cambridge (MA): Cultural Survival Inc.

PIPERNO, D.R. & T.D. DILLEHAY. 2008. Starch grains on human teeth reveal early broad crop diet in northern Peru. Proceedings of the National Academy of Sciences 105 (50): 19622-7.

PLOWMAN, T. 1979. The identity of Amazonian and Trujillo coca. Botanical Museum Leaflets 27: 45-68.

--1983. The origin, evolution and diffusion of coca, Erythroxylum spp., in D. Stone (ed.) Pre-Columbian plant migration: 125-63: Cambridge (MA): Harvard University Press.

PLOWMAN, T. & N. HENSOLD. 2004. Names, types, and distribution of neotropical species of Erythroxylum ( Erythroxylaceae). Brittonia 56: 1-53.

ROSSEN, J. 1991. Ecotones and low risk intensification: the middle Preceramic habitation of Nanchoc, northern Peru. Unpublished PhD dissertation, University of Kentucky.

ROSSEN, J. & T.D. DILLEHAY. 2002. Tecnologla y ritos en el arcaico medio del Valle de Zana, Peru. Boletin de Arqueologia de la Universidad Pontificia Catolica del Peru 3: 24-41.

ROSTWOROWSKI DE DIEZ CANSECO, M. 1988. Conflicts over coca fields in XVIth-century Peru (Memoirs of the Museum of Anthropology, University of Michigan 21). Ann Arbor (MI): University of Michigan, Museum of Anthropology.

RURY, P. & T. PLOWMAN. 1983. Morphological studies of archaeological and recent coca leaves (Erythroxylum spp.). Botanical Museum Leaflets 29: 297-341.

SIMPSON, B.B. 1975. Pleistocene changes in the flora of the high tropical Andes. Palaeobiology 1: 273-94.

UGENT, D. & C.M. OCHOA. 2006. La etnobotanica del Peru: desde la prehistoria al presente. Lima: Consejo Nacional de Ciencia y Tecnologia.

VUILLEUMIER, B.S. 1971. Pleistocene changes in the fauna and flora of South America. Science 173: 771-80.

Technical Note: mineralogical, total elemental, X-ray diffraction and thermal analyses of raw and processed materials.

Although several raw and processed lime specimens were excavated in CA-09-04 and the house structures, respectively, we subjected only two samples to multiple technical analyses. Mineralogical analysis of finely ground samples A and B (< 100[micro]m) was conducted by X-ray diffraction (XRD) and thermal analysis. XRD analysis (2-60[degrees] 2[theta]) was performed with powder mounts using a Philips PW 1729 X-ray generator employing Co-K[alpha], and a PW 1840 diffractometer at settings of 40kV and 30mA. Thermogravimetric analysis (TGA) was run with approximately 10mg of each sample using a Thermal Analyst 2000 (TA Instruments) and a 951 Thermogravimetric Analyzer (DuPont Instruments). Samples were heated to 1000[degrees]C at a rate of 20[degrees]C per minute in a nitrogen atmosphere. TE analysis was conducted on powdered samples of samples A and B oxidised at 900[degrees]C for 2 hours. The oxidised samples were then fused with Li meta-borate powder in a platinum crucible by heating to approximately 1000[degrees]C with a Bunsen burner. The fused samples were cooled, dissolved in a mixture of 4% HN[O.sub.3] and analysed for K, Ca, Mg and trace metals (Cu, Ni, Pb and Zn) using a Thermo Elemental Solar M-5 atomic absorption spectrometer and total S by Ba[Cl.sub.2] turbidimetry (Karathanasis & Hajek 1992; Karathanasis and Sparks 1996). Carbon and nitrogen were measured on powdered samples using a Flash Elemental Analyzer 1112. Organic carbon was determined from the difference in total C content before and after acid dissolution with 12M HCl for 18 hours to remove carbonates.

Soil chemistry measurements were performed by different laboratories employing different measuring techniques: mg/[dm.sup.3] (Table 1), mg/Kg (Table 2), and parts per million (Table 3). All express chemicals as parts per million (ppm). However, the results in Table 1 and Table 3 were recorded on mg per unit volume ([dm.sup.3] = 1L) of extract basis. This is a relative value for comparison purposes, but it does not provide an exact quantitative measurement unless the precise mass of each analysed sample is known to make the conversion on mass per mass basis. This is not known for samples in Table 1 and Table 3, both of which refer to mass per volume of liquid and not to mass of solid. Samples A and B are more quantitative because the solid mass was known and the conversion was made. Comparisons can still be made as high medium or low with all measurements, but not on the same scale basis.

The XRD and thermal characterisations suggested contrasting crystallinity and composition for samples A and B. Sample A from the off-mound production area produced sharp diffraction peaks indicating a relatively rich crystalline material dominated by calcite (diffraction maximum at 34[degrees] 2[THETA]) with minor quantities of quartz (26 and 31[degrees] 2[THETA]). In contrast, sample B from the domestic hut produced a diffused diffraction pattern indicative of a poor crystalline material low in mineral and high in organic constituents. Identifiable mineral components of sample B included calcite, gypsum, and quartz. TGA indicated that sample A contained approximately 85% calcite (weight loss 650-800[degrees]C), while sample B had only 20% calcite and 25% gypsum (weight loss 100-200[degrees]C). The gradual weight loss of 300-700[degrees]C in sample B is associated with organic C oxidation.

Tom D. Dillehay (1,2), *, Jack Rossen (3), Donald Ugent (4), Anathasios Karathanasis (5), Victor Vasquez (6) & Patricia J. Netherly (1)

(1) Department of Anthropology, Vanderbilt University, Nashville, TN37235, USA

(2) Instituto de Ciencias Sociales, Universidad Austral de Chile, Valdivia, Chile

(3) Department of Anthropology, Ithaca College, Ithaca, NY 14850, USA

(4) Department of Botany, Southern Illinois University, Carbondale, IL 62901, USA

(5) Department of Plant and Soil Sciences, University of Kentucky, KY40506, USA

(6) Laboratorio de Arqueobiologia, Universidad Nacional de Trujillo, Peru

* Author for correspondence (Email: tom.d.dillehay@vanderbilt.edu)

Table 1. Soil chemical analyses of sediments from Site CA-09-04
and off-site control areas *.

Soil provenience # P K pH Ca Mg

CA-09-04:

Mound A

Stratum A-III/00 14 57 124 6.92 6800 935
Stratum A-IV/0015 130 256 6.92 7260 486
Stratum B-II/00 16 137 267 7.11 12150 505
Stratum B-III/0017 57 720 8.00 13000 341

Off-mound activity
area B

Stratum B-II 152 830 7.90 18910 876
Stratum B-III 175 910 8.20 19020 945

Off-site control area:

Non-site B-II/0018 25 194 5.31 2375 735
non-site B-III/0019a 7 110 5.24 2045 664

* Processed by the Soil Test Laboratory, College of Agriculture,
University of Kentucky. Parts are measured in mg/[dm.sup.3]
extracted by Mehlich 111.

Table 2. Chemical analyses for archaeological calcite samples A
and B in mg/Kg *.

Sample K Ca Mg S Total C Org C

A 44 380 339 350 4740 70 10 2800 5000
B 165 410 111 990 39 760 16 580 98 000 75 000

 Total Trace
Sample N metals

A 0 0-200
B 125 10-1360

* Processed by the Department of Plant and Soil Sciences,
University of Kentucky.

Table 3. Soil chemistry measurements for domestic sites in parts
per million from domestic house floors of the Las Pircas phase *.

 pH P Ca Mg K

CA09-27:

Dwelling floor 8.29 84 8300 317 1155
Dwelling floor 8.24 53 9850 466 945
Dwelling floor 8.21 44 7900 405 1055
Dwelling floor 8.11 97 8050 402 870
Outside dwelling 8.22 51 8100 335 1025
Outside dwelling 8.28 71 7500 356 1045
Outside dwelling 8.44 58 8450 400 1090
Above floor 8.34 71 8200 302 1105
Stone structure 8.38 76 7450 283 1825
Adobe 8.39 89 6750 311 1650

CA09-28

Flexed burial 8.42 76 7800 338 820
Main block 8.57 155 5350 375 1330

CA09-52

North block 8.30 70 7850 295 1330
North block 8.32 52 7500 247 1455
Central block 8.28 63 8200 290 1525
Central block 8.41 55 12400 321 1582
Central block 8.43 74 8800 278 2420
Central block 8.46 93 7400 294 1280
Central block 8.47 56 7250 308 1245
South block 8.30 110 7400 254 1635
South block 8.30 110 7700 294 1635
South block 8.31 125 7350 330 1525
South block 8.36 115 7500 267 1410
South block 8.39 135 7500 339 1085
Off-site non-cultural control 7.65 n/a 4500 110 90
 samples

* Samples processed by the Soil Test Laboratory, College of
Agriculture, University of Kentucky.