Four Corners Research-Archaeology in the Mesa Verde Region

POSTER PRESENTATION GIVEN BY DAVID DOVE AT THE 2019 PECOS CONFERENCE IN CLOUDCROFT NEW MEXICO
DM Dove, S Di Naso, W Hurst, W Lucius


THE SCIENCE BEHIND THE METHOD
USING PXRF, NAA, AND ICP-MS TO IDENTIFY CERAMIC PRODUCTION AND TRADE OF SAN JUAN REDWARE 


ABSTRACT
Geochemical sourcing studies conducted on San Juan Redware pottery and clay from Montezuma, Recapture, Cottonwood, and Alkali Canyons in southeast Utah and several large villages in southwestern Colorado have revealed insightful information about this widely exchanged ware.  
We present our latest discoveries; new evidence of intra-, inter-, and extra-canyon movement of redware pottery from a major production center in central Montezuma Canyon to numerous communities within and outside the production zone. We also introduce two potential production zones in Recapture and Cottonwood Canyon as defined by new additions to the assemblage of PXRF-analyzed clay and ceramics.

INTRODUCTOIN
Montezuma Canyon is a 73 km-long erosional incision through Cretaceous and Jurassic rocks. It was home to numerous Puebloan habitation sites during the Basketmaker III to PIII era. Canyon sites dating to the AD 750-1050 period contain abundant San Juan Redware pottery. The clay resources selected for their production can be geochemically sourced to specific locales where pedogenic processes have reduced mudstones to weathered clay, some of which can be used to make pottery. Questions that we hope to answer through these ongoing analyses follow:
Can red ware found in Montezuma canyon be linked to the geological landscape?
Which sites in southeastern Utah were producing San Juan Redware?
What mechanism was used to exchange red ware pottery across the MV region; direct     exchange from producer to end user or down-line trading where pots moved from site to     site until reaching a final destination?

METHODS
Using a systematic sampling strategy, clay samples were collected from the Jurassic Morrison Formation and include additional samples gathered by the Blanding Redware Project. Thus far, the geological reconnaissance and clay sample array covers approximately 2,380 square kilometers. All clays were subjected to technological ceramic analysis. Refiring analysis was used to evaluate usability and identify those clays that matched the Munsell values of our archaeological sherd assemblage.
Clay and sherd samples were analyzed using Inductively Coupled Plasma - Mass Spectrometry (ICP-MS), and Portable X-RAY Fluorescence (PXRF). Analytical data selected from our database includes 108 clay and 220 sherd samples, including a subset of 59 sherd samples from 163 sherds analyzed using Instrumental Neutron Activation Analysis (INAA). RQ-Mode Principal Components Analysis (RQ-PCA) and ESDA (Exploratory Spatial Data Analysis) were used to examine the point patterning (cluster morphology) of samples within PCA compositional space. 

RESULTS & DISCUSSION
Some of the clays that are exposed throughout Montezuma Canyon can be used to make viable pots. Pottery found on these sites refire to a comparable Munsell color to refired clays that were collected from outcrops found proximal to them. Gray clay associated with the lower Brushy Basin and Westwater Canyon facies at the Nancy Patterson site are viable, however, their refired clay color is not observed in the analyzed red ware assemblage. Not all clays were found to be useable. Samples with relatively high percentages of 2:1 type clay (e.g. montmorillonite) commonly deformed; those containing excessive silt or sand often crumbled or broke easily. 
The resulting PCA biplots produced similar punctiform (point) patterning in independent analyses of elemental data for sherds using both PXRF and INAA. Despite the lower detection limits of INAA compared to PXRF, most of the elements analyzed using PXRF were within two-sigma of their certified values for SRM U.S.G.S Cody Shale Sco-1. Measurements above two sigma of the certified values nonetheless demonstrated good precision. An analysis of a subset of sherds (n-59) by PXRF and INAA using the same elements was used to confirm that BOTH methods produce impressive results. Unarguably, whether using INAA or PXRF, and using the same or even different variables (elements), the two methods produced identical ceramic groups (Figures 3, 4, and 5) demonstrating the viability of PXRF as an analytical sourcing method provided it is implemented properly.  
Although our clay database for the various members of the Morrison Formation is extensive, no San Juan Redware found at villages within Montezuma Canyon was linked to clays located outside of the canyon. Clay geochemistry and geostatistical analyses demonstrate that the lower Morrison exposures found within central Montezuma Canyon are chemically different from the upper Morrison exposures that predominate in the canyons to the west and the upper Morrison Fm exposed in Montezuma Canyon.

CONCLUSIONS
Our growing geochemical database consists of clays found proximal to production villages and sherds recovered from them. Analysis of the PXRF data using PCA indicates that ceramics found on production sites have complimentary geochemical signatures with the surrounding landscape. Clays analyzed with ICP-MS and PXRF, and subsequent PCA analyses, indicate that both methods produce complimentary results and can readily identify unique geochemical signatures geographically (Di Naso, S.M. 2018).  
PCA analyses of NAA data on 163 sherds, and a subset of this assemblage using PXRF, produces a densely clustered morphology we identify as the ‘Nancy Patterson’ agglomeration. The Nancy Patterson agglomeration is dissimilar from the clays sampled elsewhere in southeastern Utah (Figure 2). This highly-clustered group consists of clay found proximal to the site and sherds from Nancy Patterson, Cave Canyon Village, Monument Village, Champagne Spring, and Mitchell Springs (Figures 3 and 4).
This analysis has successfully revealed the origin of many of the sherds in our sample subset recovered from sites in Montezuma Canyon and southwestern Colorado. Through systematic sampling of the geologic landscape the origin of the pottery found in this canyon and others will be revealed as we further examine the material consequence of a cultural relationship that existed between and among sites in southeastern Utah and southwestern Colorado (Dove, D.M., Di Naso, S.M., Hurst, W., and Lucius, W.A., manuscript in preparation 2019).
San Juan Redware Sourcing and Exchange Project
Jackson's Bench and its massive prehistoric belt-loop road may have been a producer and exporter of Deadmans Black-on-Red vessels. This ceremonial complex was occupied during two periods circa AD 950-1050 and mid 1200s. A late greathouse contains the westernmost known D-shaped bi-wall. This structure and surrounding structures and features are aligned with the summer and winter solstice and lunar and celestial alignments.
Circa AD 1100 greathouse built over the top of an Abajo/Bluff Phase C-shaped pueblo house in Montezuma Canyon circa AD 750-850. 
San Juan Redware types including provisional category of transitional early Deadmans B/R
Defensively perched village and potential SJ Redware production source village (circa AD 750-850) for Abajo and Bluff B/R vessels had long views. Sleeping Ute Mountain in background.  Red firing clays that were capable of being used to make red ware pottery were collected from nearby.
Ceramics and clays in a PCA (Principle Components Analysis) that reveals production locales in Montezuma Canyon in Southeast Utah. Over the last two years, other production locales have been identified at drainages further to the west.
Ceramic interaction and exchange diagram showing  production zones, production locales, and consumers/importers. Nancy Patterson was a particularly productive producer as denoted by the purple markings.
Allen Canyon from Red Top. Beautiful setting!
Better safe than sorry. The assent to the top.
The beauty of Red Top. Early defensive site with a great kiva and some of the best views in the world.
Blue Mountain in the background.  Defensive site (circa AD 750-850 and AD 1200-1250) with long views.  To date we have accomplished a complete full-documentation or redocumentation of approximately 40 village-level sites in SE Utah that have been identified as potential San Juan Redware production sources.  The geochemical signatures of SJRW pottery at these sites are being compared to the geochemical signatures of clay arrays that lie within 2 km of each village.  Several production locales that center on major villages have been identified.
Massive Recapture ceremonial complex with occupation from circa AD 750-850, AD 1000-1050 and circa AD 1200-1230.  The San Juan Redware production period lasted from about AD 750-1050.  Most of the villages in our study area were occupied for only one or two generations and were often reoccupied later for another generation or two.  Redware pottery produced at several identified locales appear to have been the source of the clays and the potters who were producing vessels for export to southwestern Colorado, and throughout the Four Corners region.
Principle Components Analysis (PCA) showing how our sample of San Juan Redware from Mitchell Springs, Haynie, and Cave Canyon Village cluster. PCA is one of many different types of data synthesis methods that allow us to plot the multiple components/elements representing the bulk material composing each sherd, to be synthesized to one data point which can then be plotted in a bi-plot. This clustering pattern suggests that Mitchell Springs and Haynie, though located well outside of the production zone where red firing clays that can make viable and properly colored pottery are readily available, used favored source sites for a large percentage of the red pots that were traded in to those sites. Nancy Patterson Village in Montezuma Canyon was a principle supplier of red pots for both Mitchell Springs and Haynie. Other production sources are being identified including one in Recapture Canyon and another in Cottonwood Canyon. Alkali Ridge Site 13 produced Abajo Red on Orange pottery that was exported to other sites inside the production zone as well as to Mitchell Springs and Haynie. 
Collecting data from one of 57 village-sized habitation sites. Not a bad gig if you can get it ! 
Looking west from on top of Red Top.  Beautiful scenery and wonderful archaeology.