Copyright © 2002 by Crow Canyon Archaeological Center. All rights reserved.

List of Tables
List of Illustrations
Research Objectives and Methods
Architecture and Site Layout
Population Estimates
Faunal Remains
Archaeobotanical Remains
Human Skeletal Remains
Water Control and Subsistence
Abandonment and Emigration
Appendix A

Research Objectives and Methods

by Melissa J. Churchill

The Crow Canyon Archaeological Center conducted research at Woods Canyon Pueblo (Site 5MT11842) to improve our understanding of settlement aggregation in the Mesa Verde region during the late Pueblo II and Pueblo III periods (A.D. 1000–1300) and to contribute to an ever-growing body of research into the depopulation of the region by the end of the latter period. The excavations at Woods Canyon Pueblo were conducted as part of Crow Canyon's Village Testing Project, which is summarized in Ortman et al. (2000*1). For a discussion of the research design that guided our investigations at Woods Canyon Pueblo specifically, see Wilshusen (1994*2).

Crow Canyon archaeologists designed the Village Testing Project to complement the Center's earlier research in the nearby Sand Canyon locality (Database Map 338). Our research in the Sand Canyon locality suggested that large villages had formed rapidly in the mid-1200s as people moved from small upland settlements to large villages in canyon environments. The Village Testing Project focused on the investigation of large, late sites outside the locality in an effort to determine whether this aggregation might have been part of a larger phenomenon in the central Mesa Verde region (see Ortman et al. 2000*1). The testing program conducted at Woods Canyon Pueblo and Woods Canyon Reservoir (Site 5MT12086) examined whether the changes through time in settlement patterns, subsistence practices, and the use of the cultural landscape that were identified in the Sand Canyon locality also occurred in the Woods Canyon community.

On the basis of our limited excavations, it appears that Woods Canyon Pueblo was a small village compared with other late Pueblo villages in the Mesa Verde region, but that its occupation was much longer, spanning almost 150 years. The departure of the residents from Woods Canyon coincided with the final emigrations out of the region in the late A.D. 1200s.

Research Questions

The research questions at Woods Canyon Pueblo focused on three broad topics: (1) site chronology, (2) village layout and organization, and (3) agriculture practices and water and soil management. In the lists of specific questions that follow, the chapters of this report that address each question are noted in parentheses.

Site Chronology

Village Layout and Organization

Agricultural Practices and Water and Soil Management

Other Research Questions


Seventy-five small units, less than one percent of the 18.5-acre site, were excavated as part of Crow Canyon's three-year testing program. Test excavation of a small number of kivas and their associated middens was our primary focus, but we examined several possible public areas and water-control features as well. The sampling strategies employed allowed us to address many of the research questions posed above; however, because the testing was limited in scope, further excavations would be required to refine the chronological history of the village, make a stronger case for how the rim complex may have been used, and identify whether gardens or small agricultural fields were present within the pueblo.

Kiva Testing

The objectives of the kiva-testing program were threefold. Our primary goal was to gain a better understanding of when the site was constructed, which required the collection of as many tree-ring dating samples as possible. Because dateable specimens are much more likely to be preserved in kivas—and particularly in kivas that burned—our excavations focused on the test excavation of these subterranean structures rather than of surface rooms. Second, we wanted to evaluate how kivas at Woods Canyon Pueblo were used and abandoned. For this, we needed to examine kiva stratigraphy and sample enough floor and roof-fall assemblages to reconstruct activities and abandonment processes. The third objective of the testing program was to estimate the population of the village. The test excavation of a subset of the possible kivas identified during mapping allowed us to determine how many of the suspected kivas were in fact kivas, and this information, extrapolated across the entire site, provided us with the data necessary for estimating the population of the village as a whole.

Eight circular depressions or flat areas thought to perhaps be kivas were randomly selected from what were tentatively identified as the three main residential sections of the site (the upper west side, the canyon bottom, and the east talus slope). Each depression or flat area was tested by excavating a 2-x-1-m unit in its approximate northwest quadrant, and six of the eight (Structures 1-S, 2-S, 3-S, 4-S, 5-S, and 7-S) turned out to be kivas. Two additional kivas (Structures 6-S and 8-S) were selected "judgmentally," that is, they were specifically chosen for excavation because we wanted to quickly expand our kiva sample and it was clear from evidence visible on the modern ground surface that these two structures were, in fact, kivas. One of these, Structure 6-S, was located in the rim complex. An additional kiva (Structure 9-S) was unexpectedly found below Nonstructure 1-N, for a total of nine kivas that were tested during Crow Canyon's excavations.

Midden Testing

Our initial goal in testing middens was to better understand changes in pottery over time. Our strategy was to identify and test burned kivas whose dates of construction could be determined on the basis of tree-ring dates, and then excavate a limited portion of the middens associated with these same structures. The scarcity of burned roof fall in the tested kivas, however, made it difficult to estimate construction dates, which in turn prevented us from studying pottery change through time. Nonetheless, the midden assemblages helped us address questions about site chronology, intrasite organization and occupation, and subsistence strategies.

An area was defined as a midden if moderate quantities of artifacts were present on the modern ground surface or if the area was located immediately south of a kiva, where middens are typically found. Six midden areas were tested (Nonstructures 3-N, 4-N, 5-N, 6-N, 7-N, and 8-N). The refuse from four of the areas (Nonstructures 4-N, 6-N, 7-N, and 8-N) is believed to have originated from specific tested kivas, because of the proximity of the kivas to the midden areas and the absence of other nearby structures. The other two midden areas (Nonstructures 3-N and 5-N) appear to be generally associated with tested kivas, but it is clear that the refuse in them also could have originated from other structures in the immediate vicinity. Three to four 1-x-1-m units were randomly selected from each possible midden area. If these units did not produce an adequate sample of refuse, "judgment" units were placed in areas that were thought to contain more trash.

All the deposits in a given "midden" area, whether they consisted of midden deposits, natural deposits, extramural surfaces, or native sterile deposits, were designated parts of the same large nonstructure. As a result, a nonstructure as defined at Woods Canyon Pueblo contained multiple deposits (and different kinds of deposits), rather than a single cultural deposit as is standard at other sites excavated by Crow Canyon. For example, a nonstructure at Woods Canyon Pueblo could include several extramural surfaces, several midden deposits, and postoccupational fill. In an effort to separate the multiple deposits included in a particular nonstructure, we retroactively assigned every cultural deposit within the nonstructure a subunit number. For example, a secondary refuse deposit and an extramural surface in Nonstructure 5-N became Nonstructure 5.1-N and Nonstructure 5.2-N, respectively. The designations for natural deposits and mixed deposits were not changed; rather, they were grouped with the original nonstructure (in this example, Nonstructure 5-N) and are identified as "noncultural" in the database. In some of these nonstructural areas, limited amounts of midden were exposed, and often the midden deposits were no longer intact because they had been naturally redeposited from upslope.

Testing of Public Space

The rim complex and a suspected plaza in the canyon bottom (Nonstructure 1-N) were tested as possible public areas at Woods Canyon Pueblo. Investigations were aimed at understanding how these areas were used and what they might reflect about village organization, social differentiation, and community integration.

Testing in the rim complex involved exposing sections of the enclosing wall that defined this space, exposing the exterior faces of buildings contained in this space, partly excavating a kiva, and looking for evidence of a plaza surface. The plaza area (Nonstructure 2-N) was divided into two sampling strata. The first sampling stratum included all the visible open space, and the second sampling stratum included areas covered by rubble. Five 1-x-1-m units were randomly selected from Sampling Stratum 1, and three 1-x-1-m units were selected from Sampling Stratum 2. One unit was expanded into a 2-x-1-m unit because more space was needed for excavators to maneuver after a wall was exposed. Five judgment units were also opened adjacent to architectural features and buildings. Another judgment unit was excavated in a kiva (Structure 6-S). Lastly, all standing walls and wall features in the rim complex were recorded.

The possible plaza in the canyon bottom (Nonstructure 1-N) was tested by excavating 10 1-x-1-m judgment units. Testing revealed that this area was not used as a plaza; rather, it might have served as a garden, as well as a place where trash was discarded.

Testing of Water-Control Features

In an attempt to better understand the construction, use, and age of water-control features at Woods Canyon Pueblo, we tested several checkdams in the main drainage that bisected the site. Judgment units were placed adjacent to possible checkdams visible on the modern ground surface on top of the cliff (Nonstructure 9-N). Sections of five checkdams were exposed during testing. A prehistoric reservoir (Site 5MT12086) located northeast of the pueblo was also tested before excavations at Woods Canyon Pueblo began.

References cited | To borrow, cite, or request permission

Woods Canyon Pueblo Home