Recent Advances in Provenance Research Along the Middle Valley Using Time of Flight Laser Ablation-Inductively Coupled Plasma-Mass Spectrometry (TOF-LA-ICP-MS) Ida Redbird, Pee Posh potter J. Andrew Darling1, Hector Neff2, B. Sunday Eiselt3, Linda Newman1 Effigy vessel by Background Dorothea Sunn- Paste Study Avery Table 1. List of Source Samples Mica Schist Temper Study Chemical-based provenance research is new to ceramic wenty-six clays from four ethnographic clay pits and one geological deposit were submitted studies, which have long been dominated by petrographic techniques. Source Area (Site Name) See Map GRIC-CRMP Miksa Grand Total T for TOF-LA-ICP-MS (Table 1). Two of the sample locations were situated on alluvial One of the challenges of the petrographic approach is characterization of Gila Butte (Pinal Schist) 10 4 14 The mica schist temper study included 33 Pinal Schist specimens from sources mica schist-tempered pottery and source identification. Recent research Pima Butte (Pinal Schist) 7 5 12 floodplains bordering the Gila River (at Casa Blanca) and the Santa Cruz Wash (at Komatke). Enid (Pinal Schist) 10 10 The remaining three included argillic clays on the bajadas or foothills of Pima Butte, the sampled by Beth Miksa for the Center for Desert Archaeology and 23 sampled by in the Gila River Indian Community integrates consultation with Native the Cultural Resource Management Program, Gila River Indian Community potters to interpret trace element patterns using Time of Flight-Laser Rattlesnake Hill (Pinal Schist) 6 6 Estrella Mountains at Santa Cruz, and Crooked Red (Figure 1). These clays were compared to Montezuma Peak (Pinal Schist) 5 5 (Table 1). Eleven source area outcrops were included, although only four of Ablation-Inductively Coupled Plasma-Mass Spectrometry (TOF-LA-ICP-MS). 29 ceramic specimens, including two ethnographic sherds (Table 2). Analyses of the sherd Chandler Heights (Pinal Schist) 3 3 these (Pima Butte, Gila Butte, Rattlesnake, and Enid) were sampled This study is unique for its application in a Native American research pastes and ceramic test tiles targeted the clay matrix, while avoiding any temper or natural Florence Junction (Pinal Schist) 2 2 extensively (Figure 3). Source samples were compared to 71 ceramic sherd environment and for its potential contribution as an adjunct to labor non-plastics in the paste. Five spots or locations were analyzed on each sample, and the five Florence Northeast (Pinal Schist) 1 1 replicate measurements were averaged prior to quantitative analysis. This practice specimens containing schist temper and 5 schist rock artifacts from seven intensive petrographic analyses. Sacaton Butte (Pinal Schist) 1 1 minimizes instrumental noise and compositional variation within the clay fraction, simulating Hohokam sites (Table 2). Analyses targeted muscovite at five separate locations Blackwater (Pinal Schist) 1 1 on each specimen. Unlike the paste study, the five replicate muscovite analyses LA-ICP-MS has seen increasing use in archaeology during the past five Sacaton Butte (Pinal Schist) 1 1 a bulk analysis of raw clay or ceramic paste. years (e.g., Speakman and Neff, eds. 2005). Advantages of the technique were treated as independent samples. This practice eliminates the possibility of Grand Total 23 33 56 hree main geochemical groups were identified on the basis of raw clay and ceramic include minimal sample preparation, speed of analysis, and the potential T combining muscovite fragments (tempoer) from different compositional sources analyses: the Calcareous Group, which is rich in caliche, calcium, and strontium; Group 1 into a single average value for a specimen. This can happen in ceramics, for spatially resolved microanalysis of specific matrix components. Source Area (Site Name) See Map GRIC-CRMP Miksa Grand Total Conclusion Linda Newman (Gila River Indian Community) collecting particularly if tempers from different geological sources were combined and Previous LA-ICP-MS of micaceous pottery and raw materials from the Gila (Casa Blanca) which is enriched in several elements especially cesium, and Group 2 (Santa clay at Crooked Red Casa Blanca (Alluvial Clay, Gila River, O'odham Historic Pit) 9 9 Cruz/Estrella) which is depleted in cesium; and other trace elements (Figure 2). added to the same ceramic paste (Cogswell et al. 2005; Miksa 2001). River area by Cogswell et al. (2005) demonstrate that schist sources show Komatke (Alluvial Clay, Santa Cruz Wash, Pee-Posh Modern Pit) 3 3 TOF-LA-ICP-MS analysis of the matrix component of ceramics, ceramic raw materials, schist rocks, and mica consistent elemental differences and that individual temper particles can Pima Butte (Residual Clay, Bajada, Geologic Sample) 3 3 Nearly all of the Calcareous Group clays were collected from residual deposits at Pima Mica schist geochemistry varies from east to west based on TOF-LA-ICP-MS Dorothea Sunn-Avery Pee Posh pottery consultant temper in ceramics has produced a substantial database for the identification of ceramic production locations be characterized within sherds (See also Abbott 2001; Miksa 2001). Santa Cruz (Residual Clay, Bajada, Pee-Posh Modern Pit) 9 9 Butte and Crooked Red without the aid of potters, and only three sherds were attributed to analysis. The westernmost schist sources (Enid and Montezuma Peak) are Crooked Red (Residual Clay, Bajada, O'odham Historic Pit) 2 2 and intra-regional exchange applicable over the course of the Hohokam pre-Classic and Classic periods (ca. AD There is also precedent for the provenance study of buff ware via LA-ICP- this group. All three sherds contained high levels of caliche, including the two ethnographic characterized by muscovite that is relatively depleted in cesium and rubidium, Grand Total 26 26 500-1450) to the Historic Period (AD 1450-1950). Chemical patterning and provenance assignments are clearest MS analysis for the study of ceramic pastes (Beck and Neff 2006). sherd wasters made from Komatke alluvial clays. These results indicate that high levels of while other sources located further east show greater enrichment of these two caliche may dilute other trace element abundances and hinder provenance assignments. in the paste study, which identified two ceramic production zones, one near Casa Blanca and one in the Santa his study expands the raw material source database of Pinal Schist and elements. Rubidium and cesium concentrations in schist rock artifacts (from T High levels of caliche also contribute to failures during firing as the Komatke wasters Cruz Wash area of the Gila River reservation. locally available raw clays in south-central . Alluvial and residual Table 2. List of Artifact Samples GR-893) are consistent with the depleted concentrations that characterize the suggest. It is likely that potters may have used red-firing, caliche rich clays, such as those he mica temper study identified consistent chemical differences between source outcrops, separating the Plain Paste (Clay Schist Tempered westernmost outcrops. Canonical discriminant analysis of the four well- T raw clays were collected from locations on the reservation in the middle Site Number Decoration Grand Total Targeted) (Mica Targeted) from Crooked Red, for slip or paint rather than paste. westernmost sources at Montezuma Peak and Enid from schist temper sources to the east. Assignment of Gila River Valley south of Phoenix, Arizona. Samples of Pinal Schist were sampled outcrops show that the axis of greatest separation (Discriminant GR-1430 (Pee-Posh Village) Historic Red-on-buff 1 1 ceramics, however, to eastern sources was sometimes ambiguous due to the small number of micaceous schist obtained from rock outcrops and from archaeological sites also along the The Casa Blanca Group includes clays obtained from the bank of the Gila River upstream Function 1) separates Enid from the other three sources, while subsequent axes Historic Buff 1 1 discriminate the three eastern groups (Pima Butte, Gila Butte, and samples from sources including Blackwater, Florence, and Chandler Heights. In addition, the possibility that middle Gila River drainage. from Pima Butte near the Hohokam ceremonial center at Snaketown. The Santa Cruz Group Historic Plain 2 2 Rattlesnake) (Figure 4). mica temper in a single vessel might derive from multiple sources cannot be eliminated. Evidence for mixing consists of residual clays found on the western slopes of the Estrella Mountains above the The analyzed ceramic samples consist of prehistoric Hohokam, and both GR-1441 (Pee-Posh Village) Historic Red-on-buff 1 1 temper, if confirmed, would have important implications for understanding ceramic production practices Santa Cruz wash. The Casa Blanca Group is dominated by Red Smudged and Red-on-brown chist temper acquisition and utilization clearly emphasizes eastern source historic and modern O’odham/Pee Posh plain, red, and buff wares from Historic Red-on-brown 1 1 S including temper acquisition, distribution, and exchange. Historic Buff 4 4 sherds (n = 6), whereas the majority of the buff, red-on-buff, plain, and plain smudged groups. Probabilities of group membership for ceramic temper were calculated sites along the Gila River and Santa Cruz Wash (Ravesloot et al 2007). Several instances of ceramic movement from production locales are worth noting. Based on the paste study, Historic Plain 4 4 sherds were associated with the Santa Cruz Group sources (n = 20). This pattern suggests using Mahalanobis distances from scores on the three discriminant functions. TOF-LA-ICP-MS was conducted on raw clays, mica schist (used as red smudged pottery from the vicinity of Casa Blanca was moved west to sites along the Santa Cruz Wash. Historic Red 1 1 that Classic Period red smudged ceramics were produced within the Hohokam core not far Two to seven replicates from each sample provided multiple independent lines tempering material), the clay component of archaeological ceramic from Snaketown (at Casa Blanca) and were likely transported to villages along the Santa Cruz Plain and buff sherds, on the other hand, were made from local clays in the Santa Cruz area. The mica pastes, and mica temper particles. GR-522 (Hohokam Village, Lower Santan) Gila Butte Red-on-buff 6 6 of evidence for assigning provenance. In some cases, all analyses from a given Santa Cruz Red-on-buff 4 4 wash. The bulk of the remaining assemblage was made locally using Santa Cruz clays. sample show low probabilities of membership in any groups, suggesting a temper study demonstrated that nearly all micaceous temper in pottery from sites along the Santa Cruz Wash Indigenous knowledge regarding the location and identification of clay Sacaton Red-on-buff 13 13 Historic ceramics also source to Santa Cruz and Casa Blanca, which further suggests probable unknown source(s). There also are cases in which the temper and Gila River (GR-522) derive from schist outcrops to the east (Pima Butte, Gila Butte or Rattlesnake), with almost no temper coming from the western sources at Enid or Montezuma Peak. In contrast, schist rock resources and ceramic production is significant to this research. Native Casa Grande Red-on-buff 6 6 continuity in the pattern of clay source utilization over time. particles from a single sherd are grouped with more than one source including samples collected from the Santa Cruz sites (groundstone and raw schist) were obtained from western criteria were used to refine sampling methodologies by targeting Red/Smudged 1 1 Pima Butte, Gila Butte, or Rattlesnake. This indicates that tempers were Prehistoric Plain 1 1 sources. This suggests that mica temper pottery was produced in the vicinity of eastern schist sources, or, materials that were most likely used for ceramic production in the past. Figure 1. Map showing locations of clay source sample areas and combined or mixed in the same paste and that potters either exploited multiple Prehistoric Red 1 1 that temper was transported to the west where it was incorporated into local ceramic pastes. Future analysis Gila River potters select raw clay based on texture, color, smell, taste, archaeological sites schist sources for their temper or that temper was traded among pottery of clay pastes from micaceous ceramics will determine if spatial trends in the use of clay resources mimic and accessibility with the majority of the alluvial clays coming from the GR-892 (Hohokam Village, Santa Cruz Wash) Red/Smudged 1 1 producing communities using different sources. GR-893 (Hohokam Village, Santa Cruz Wash) Gila Butte Red-on-buff 12 12 those identified for schist temper. riverbank and the first terrace above the valley floor. Residual (iron- Santa Cruz Red-on-buff 5 5 There are some clear patterns in the data. A total of 32 out of 71 ceramic rich) clays used in the production of utilitarian wares are recovered from Ethnographic clay pits at Santa Cruz Finally, consultation with local potters and ethnographic sources (Fontana et al. 1962; Russell 1908) was Sacaton Red-on-buff 1 4 5 the foothills (bajadas) associated with surrounding mountain ranges. samples (45%) sourced to eastern schist outcrops in nearly equal abundances critical in clay sample selection as well as highlighting production trends that link modern practices with Red/Smudged 4 7 11 Potters also select schist rock for temper based on quantity of mica, (Pima Butte = 11, Gila Butte = 11, and Rattlesnake = 10), whereas all of the prehistoric patterns identified by TOF-LA-ICP-MS. This includes clay and temper acquisition practices that Plain/Smudged 2 1 3 foliation, and friability. schist rock artifacts sourced to western outcrops (i.e. Montezuma Peak, Enid, avoid calcareous clays, target red, high iron or “earthy hematite”, clays for slips and paint, and emphasize Prehistoric Buff 1 1 2 or other unsampled outcrops around the Santa Cruz Wash). This indicates that the addition of highly foliated mica temper to improve firing success and vessel performance. Underlying major and minor (trace) geochemical patterns of the clay Prehistoric Plain 1 1 schist rock was procured locally for the production of groundstone and other Prehistoric/ Historic Plain 1 1 durable artifacts, but that production of micaceous pottery emphasized temper identified through TOF-LA-ICP-MS correlate with potters’ criteria for raw References Cited material selection including richness in iron, caliche, and salt. These Mica Schist Rock (artifact) 5 5 from sources upstream along the Gila River in the vicinity of Casa Blanca and patterns also document spatial patterning in clay and temper acquisition GR-894 (Hohokam Village, Santa Cruz Wash) Red/Smudged 1 1 Snaketown. The latter observation applies not only to ceramics from sites, such and continuity in raw material selection over time (prehistorically to the GR-895 (Hohokam Village, Santa Cruz Wash) Snaketown Red-on-buff 1 1 as GR-522, but also to downstream sites closer to Santa Cruz Wash, including present). Casa Grande Red-on-buff 2 2 GR-893. Abbott, D. R. (2001) Electron Microprobe Analyses of the GARP Pottery and Raw Clays . In The Grewe Red/Smudged 1 1 2 Archaeological Research Project Volume 2: Material Culture Part I: Ceramic Studies, edited by D. R. Abbott, Tanque Verde Red-on-brown 1 1 pp. 146-153. Anthropological Papers No. 99-1. Northland Research, Tempe. Historic Buff 1 1 Beck, M. and H. Neff (2006) Hohokam and Patayan Interaction in Southwestern Arizona: Evidence from GR-9099 (Hohokam Village, Santa Cruz Wash) Red/Smudged 1 1 Figure 3. Map showing locations of schist source sample areas and Ceramic Compositional Analysis. Journal of Archaeological Science, in press. GR-9106 (Hohokam Village, Santa Cruz Wash) Plain/Smudged 1 1 Close-up of Casa Blanca clay in-situ archaeological sites Cogswell, J. W., D. R. Abbott, E. J. Miksa, H. Neff, R. J. Speakman, M. D. Glascock (2005) Provenance Komatke (Modern Pee-Posh Village) Modern Pee-Posh Sherd 2 2 Study of Hohokam Schist-tempered pottery and Raw Materials from the Middle Gila River Valley, Arizona: Grand Total 29 76 105 Figure 4. Bivariate plot of discriminant function Techniques and Prospects. In Laser Ablation ICP-MS in Archaeological Research, edited by Speakman, R. and scores in TOF-LA-ICP-MS/schist source data set H. Neff. University of New Mexico Press, Albuquerque. Fontana, B. L., W. J. Robinson, C. W. Cormack, and E. E. Leavitt (1962) Papago Indian Pottery. University of Washington Press, Seattle. 800 µm Figure 2. Bivariate plot of Calcium and Cesium in ppm in TOF-LA-ICP-MS/clay source data set Miksa, E. J. (2001) Temper Provenance Studies. In The Grewe Archaeological Research Project Volume 2: Material Culture Part I: Ceramic Studies, edited by D. R. Abbott, pp. 7-45. Anthropological Papers No. 99-1. Northland Research, Tempe. Ravesloot, J. C., M. K. Woodson, and M. J. Boley (editors) 2007 [Draft] SFPP East Line Expansion Project, Arizona Portion: Final Data Recovery Report. William Self Associates, Inc., Tucson, AZ Russell, F. (1908) The Pima Indians. In Twenty-Sixth Annual Report of the Bureau of American Ethnology, 1904-1905, Smithsonian Institution, Washington, D. C. Reprinted by University Speakman, R and H. Neff (2005) Laser Ablation ICP-MS in Archaeological Research. University of New Mexico Press, Albuquerque. Gila River Indian Community lab staff collecting clay with Dorothea Sunn-Avery at the Komatke clay source (L-R: Linda Newman, Dorothea Sunn-Avery, Reylynne Williams, Lorie Sinclair) Electron image of schist sample Acknowledgements The authors wish to thank Dorothea Sunn-Avery, Wally Oliver, Reylynne 1. Gila River Indian Community, Cultural Resources Management Williams, Lorie Sinclair, Lynn Simon, Tom Herrschaft, William Doelle, and Program (GRIC CRMP) Beth Miksa. This research was conducted by the GRIC CRMP, in Prehistorically mined schist outcrop at collaboration with the Bureau of Reclamation-Pima-Maricopa Irrigation 2. Institute for Integrated Research in Materials, Environments, and Gila Butte Society, California State University, Long Beach (IIRMES) Project (BOR P-MIP) under the Tribal Self Governance Act of 1994 [P.L. 103- 413]. We also wish to thank IIRMES, the Center for Desert Archaeology, and 3. Southern Methodist University, Department of Anthropology William Self Associates, Inc. TOF-LA-ICP-MS was conducted as part of the Santan Data Recovery Project (P-MIP) and the SFPP East Line Expansion Gila River Indian Community, Cultural Resource Management Program Project.