Introduction

Shields Pueblo (5MT3807) is a large Puebloan site with a primary occupation dating between A.D. 1050 and 1300, and an extended occupation that began by at least A.D. 775. Shields is located on property owned by Colorado Mountain College (CMC) and James and Veda Wilson, immediately north of Goodman Point Monument, a unit of Hovenweep National Monument, in Montezuma County (Figure 1). This report summarizes the Crow Canyon Archaeological Center's research goals and the archaeological fieldwork conducted at Shields Pueblo to date, and it describes archaeological fieldwork conducted in 1999-our third field season. We also present preliminary results and interpretations about the occupation and abandonment of Shields developed on the basis of tree-ring dates and preliminary artifact analyses. The 1999 fieldwork was conducted from April until November under State of Colorado Archaeological Permit number 99-3. Upon completion of all fieldwork and analyses, detailed information reporting the work conducted at Shields Pueblo will be available in an electronic site report to be published over the Internet.

Project History

Shields Pueblo first came to the attention of the archaeological community upon publication of an article reporting the recovery of a copper bell with a burial (Hayes and Chappell 1962). From 1975 to 1977, Colorado Mountain College (CMC) conducted summer archaeological field schools at Shields. CMC excavated four masonry kivas, portions of a disturbed surface roomblock, and several arbitrary grid squares. Three of these kivas were tree-ring dated (Table 1, Adler 1990:Table 8), suggesting that the site was occupied from the A.D. 1100s into the early A.D. 1200s (Adler 1994; Adler and Varien 1994). Full-coverage survey of the Sand Canyon locality by Crow Canyon archaeologists resulted in the recording of several surface artifact and sandstone scatters within the boundaries of Shields Pueblo in 1987 (Adler 1988, 1992). Pottery noted during this reconnaissance also suggested a primary occupation dating between A.D. 1050 and 1225 (Adler 1990:260). Shields Pueblo was selected as the ideal location to initiate Crow Canyon's five-year research plan titled Communities Through Time: Migration, Cooperation, and Conflict; fieldwork began in 1997.

Research Questions

Communities through Time: Migration, Cooperation, and Conflict is a regional research design that examines the development and abandonment of ancient Puebloan communities in the Mesa Verde archaeological region from A.D. 900 to 1300, incorporating data gathered at the residential, community, and regional levels. The research being conducted at Shields Pueblo is designed primarily to collect artifact and ecofact assemblages from residential structures occupied and abandoned between A.D. 1050 and 1225, the interval during which Shields Pueblo is thought to have been a focal location within the larger Goodman Point community (Adler 1990:260, 1994; Adler and Varien 1994; Varien 1999). The focus of community-level research involves integrating Shields Pueblo into the larger framework of the surrounding natural environment and cultural landscape defined by the numerous prehistoric settlements surrounding both Goodman Point and Shields pueblos. Regional-level research will compare the Goodman Point Community with 26 other, similarly long-lasting communities in the Mesa Verde region (Varien et al. 1996; Varien 1999) and other Mesa Verde-region community centers investigated by Crow Canyon (Ortman et al. 2000).

Several research questions are oriented toward reconstructing the history of occupation and changing population levels at Shields Pueblo and in the immediately surrounding area. These data will allow us to assess the nature and timing of population aggregation into community centers and to evaluate the impact these populations had on their surrounding natural environment. One manner in which we are pursuing this line of inquiry is through intensive sampling of wood and charcoal for tree-ring analysis. Tree-ring dates obtained from samples from structures and trash deposits will be used to determine if Shields Pueblo was occupied continuously or simply at intervals punctuated by periods of abandonment. These data will also be used to assess the intensity of contemporaneous occupation of Shields during different periods. Currently, tree-ring dates from Shields indicate use of the location in the A.D. 770s, with a fairly continuous distribution of tree-ring dates from structures spanning the period between A.D. 1100 and 1260 (Table 2). Tree-ring species information, macrobotanical data derived from flotation analysis, palynological analysis, and analysis of faunal assemblages will also be used to reconstruct local environmental conditions and available resources, and to evaluate changes in the use or availability of these resources through time.

Within the history of occupation, we are particularly interested in the A.D. 1050-1225 period, the time during which Shields Pueblo appears to have served as a community center for the Goodman Point community. This span includes the period of Chacoan influence, from approximately A.D. 1050 to 1150, and the Post-Chacoan Period, from A.D. 1150 to 1225. A community center is defined by an area of dense residential and public architecture that was central to a cluster of individual settlements. We hope to determine if site organization at Shields, and the organization of the Goodman Point community, changed from the Chacoan to post-Chacoan periods, and if occupation at Shields was continuous across this transition. The center of the community is thought to have shifted from Shields to Goodman Point Pueblo, a large, aggregated village located just southeast of Shields, between A.D. 1225 and 1300 (Adler and Varien 1994).

An important goal at Shields is to identify and investigate public architecture. The presence of residential architecture at Shields Pueblo has been confirmed, but the extent of public architecture remains unknown. A prehistoric road, a prominently located and oversized roomblock that may have been a focal point within the community, and the recovery of a copper bell all suggest the local importance of Shields Pueblo and raise several key research questions. Are there activities represented at community centers that are not represented at other residential sites or site clusters within a community? What were the relations between the residents of the community center and those living in surrounding settlements? Were the residents of community centers differentiated from the other residents of the community? These questions will be addressed through comparison of material recovered from Shields and other excavated sites in the immediate area. Community centers may be expected to display evidence of "high status" individuals, or at least individuals with preferential access to limited resources. The recovery of a copper bell from a burial at Shields highlights the potential for high status individuals having resided at Shields. Are there additional sources of evidence for status differentiation? To address this question, material recovered from Shields will be compared with the artifact assemblages from neighboring settlements, but a detailed evaluation of artifact differentiation within the site will also be used to assess questions of status differentiation and access to nonlocal materials. Additionally, the temporal depth represented at Shields provides the opportunity to evaluate changes in access to such materials across the Chacoan and Post-Chacoan periods.

Comparative community information derives from neighboring sites known primarily through archaeological survey and the evaluation of surface material. However, excavation data are available for the Mustoe site, a unit pueblo located approximately 1 km southwest of Shields (Gould 1982), from numerous dispersed roomblock settlements in the neighboring Sand Canyon community (Varien, ed. 1999), from Sand Canyon Pueblo (Bradley 1992, 1993), and from Castle Rock Pueblo (Kuckelman 2000). These sources will be used to supplement the comparative database.

The relationship between residents of the Goodman Point community and those of other communities in the Mesa Verde region will also be examined. Comparison to contemporaneous excavated sites from throughout the Mesa Verde region will be used to place the Shields Pueblo excavations in a regional context. Finally, assessment of trends in community development evidenced by the data from Shields Pueblo will be compared with similar information from other Colorado Plateau regions to provide a comparative evaluation of the process of community formation and change during the Chacoan and Post-Chacoan periods.

History of Crow Canyon Research at Shields Pueblo

In advance of our first season, a detailed map of the surface remains at Shields Pueblo was prepared and an arbitrary grid system was established. Historic land use has obscured or removed most of the surface remains at Shields, and plowing has consistently mixed near-surface deposits. Mapping of the distribution of surface remains identified 18 high-density concentrations of artifacts and sandstone rubble, each thought to represent prehistoric habitation areas (Figure 2).

The 1997 Field Season

Archaeological fieldwork in 1997 combined surface collection and the excavation of randomly selected 1-x-1-m test units to provide a representative sample of surface and subsurface assemblages. Additionally, excavation was conducted within the only preserved surface architecture still standing at the site-three previously disturbed surface rooms thought to be part of a possible great house roomblock (Block 100, Figure 2). The northern portion of Shields was under active cultivation in 1997 (Figure 2), so fieldwork was restricted to the southern portion of the site. Each high-density artifact concentration was sampled with three to nine randomly selected, 1-x-1-m test units (Ward 1997).

The work conducted during the 1997 field season provided information regarding the extent of historic disturbance to the site. Excavation established that plowing had disturbed the sediment to an average depth of 30 to 35 cm below modern ground surface, mixing surface materials and underlying sterile sediment over much of the site. However, several intact structures and features were found immediately below the plow zone. Excavations in the Block 100 surface rooms defined the extent of previous disturbance, but also revealed undisturbed deposits, provided architectural information, and produced artifact assemblages useful for dating the occupation of this roomblock. Testing established that Block 700 represented the remains of a historic settlement within a low-density scatter of prehistoric artifacts. Several tree-ring samples were collected, some of which yielded dates.

Despite the extensive disturbance from agriculture and historic-collection activities, the 1997 fieldwork confirmed that Shields Pueblo retained a great deal of important archaeological data. At the conclusion of the 1997 field season, the National Geographic Society sponsored a remote-sensing project at Shields Pueblo. Remote-sensing techniques identify locations of differential soil disturbance; disturbances, called anomalies, can be natural or cultural in origin. Using electrical resistivity and magnetic field gradient, the Shields survey identified numerous subsurface anomalies thought to represent subterranean structures and features (Varien 1997). These data have been influential in guiding subsequent research at Shields and are described below.

Remote-Sensing Results

Remote sensing was conducted over most of the site, including the majority of the high-density artifact areas (Figure 3; Varien 1997:Figures 3 and 4). We identified 181 anomalies thought to represent subterranean structures or other cultural features (Varien 1997). Several anomalies were identified by both resistivity and magnetism, but the estimated number of potential unique structures indicated by remote sensing is approximately 100. Each structure discovered during test excavation in 1997 was associated with a remote-sensing anomaly, which strongly suggests that many of the additional anomalies are also associated with buried structures. Figure 3 is a graphic representation of the electrical resistivity data. Many of the anomalies in this figure, identified by the darkest locations, occur in roughly east-west alignments and are thought to represent pit structures located to the south of masonry roomblocks. Several anomalies create a southwest-to-northeast alignment that parallels a prehistoric road that runs from Casa Negra, the Chacoan-era center of the neighboring Sand Canyon Community, to Shields Pueblo.

The remote-sensing data provide a detailed picture of likely settlement structure and extent, but the interpretation of remote-sensing anomalies as buried pit structures required additional confirmation. The testing of remote-sensing anomalies was one of the primary objectives for the 1998 field season.

The 1998 Field Season

During the 1998 field season we completed excavations in unfinished units from the 1997 season and expanded excavations both inside and adjacent to previously discovered structures and features We also continued the strategy of surface collection and random excavation pursued in 1997, extending sampling coverage to the northern portions of the site. Five high-density concentrations (Blocks 1100-1500) were sampled with randomly selected test excavations, three of which were also supplemented with judgmental excavations designed to explore specific features or to collect robust assemblages. Mechanical and hand excavation techniques were used to define the extent of architecture and features associated with the three remaining surface rooms and to locate remaining intact deposits in the immediate area of this roomblock (Block 100 area). A kiva (Structure 123), located to the south of these surface rooms and previously excavated by Colorado Mountain College, was reexposed. For a more detailed report see The Shields Pueblo Research Project, Annual Report, 1998 Fieldwork.

We developed and refined a method for testing remote-sensing anomalies. This method combines mechanical and hand excavation (Duff and Ryan 1998; Duff et al. 1999). Mechanical excavation is used to expose deposits in areas of remote-sensing anomalies, to identify structures and determine their spatial extent, and to permit assessment of the depositional sequence within structures. Hand excavation is then implemented to collect artifacts and ecofacts from the fill of structures and to expose architectural features and floor contexts. The remote-sensing data proved to be highly accurate in predicting the locations of buried architectural features.

Nine remote-sensing anomalies were tested during the 1998 season; eight of these revealed buried prehistoric structures or features. The ninth, an anomaly associated with a large depression thought to be a possible great kiva (Adler 1990:Figure 26), proved to be a historic feature excavated for the disposal of prehistoric rubble. Seven of the eight structural anomalies revealed masonry-lined pit structures (kivas), and the eighth was a dense trash deposit thought (possibly) to overlie a structure. When combined with the structures found during random test excavations, a total of 12 anomalies had been tested by the end of fieldwork in 1998, 11 of which revealed buried prehistoric structures.

Many of the anomalies were tested toward the end of the 1998 field season, and limited excavation was conducted within these before the end of the field season. However, the limited excavation within structures during 1998 indicated that several had been burned upon abandonment, and wood samples were collected from numerous structures and submitted for tree-ring dating. The majority of the structures tested in 1998 were masonry-lined pit structures, generally thought to postdate A.D. 1150. All of the tree-ring dates from Shields are listed, by block area and study unit, in Table 2. These are also discussed with reference to particular structures below. However, tree-ring results indicate that many of the masonry-lined pit structures tested in 1998 were constructed and occupied after A.D. 1245, an altogether unexpected finding.

Field Methods, 1999 Season

The field methods employed during the 1999 season were a continuation of those of the previous season (Duff and Ryan 1998), although our research emphases changed. In 1999, we concentrated our effort on the testing of remote-sensing anomalies and on the excavation of structures and midden deposits discovered as a result. No surface collections or randomly selected test excavations were conducted during the 1999 season, although excavation in a few units begun in earlier years was completed. Instead, due to the unexpected finding that many of the structures discovered in 1998 postdated A.D. 1245, effort was directed toward the identification and excavation of earlier structures and contexts-primarily structures dating to the A.D. 1000-1100s. We continued to use the results of the remote-sensing project to select areas likely to contain subterranean structures. Remote-sensing anomalies were initially tested by excavating an east-west trench through plow-zone-disturbed deposits. Trenching was used to define structure boundaries and to reveal the sequence of structure deposition. Hand excavation within structures was then used to recover artifacts from structure fill, floor, and feature contexts.

Field Personnel, 1999 Season

The following people contributed to Shields Pueblo fieldwork in 1999: Andrew Duff, project director; Susan Ryan, assistant project director; Melissa Churchill and Kristin Kuckelman, research archaeologists; Mark Varien, director of research; Frances Black, Drew Kenworthy, Hugh Robinson, Josh Roffler, and Kari Schleher, research interns; Jane Dillard, research volunteer; the Crow Canyon education staff; and the many participants in Crow Canyon's research and education programs who made possible the work accomplished this year.

1999 Excavations and Preliminary Results

Excavation during the 1999 season was focused on six of the 18 high-density concentrations originally mapped, expanding or completing areas previously examined. The completion of a single, 1-x-1-m test unit remaining from the 1998 season (in Block 1200) marks the only exception to this strategy. Portions of Blocks 100, 200, 800, 1300, 1400, and 1500 were examined this season. Judgmental excavations were again concentrated in Blocks 100, 200, and 1400, and previous excavation in Blocks 1300 and 1500 was supplemented. Our primary excavation strategy involved the judgmental testing of remote-sensing anomalies and portions of structures exposed during this process, as well as excavation of 2-x-2-m units in areas of high artifact density associated with these structures. In practice, these test units were placed southeast of known structures, where midden deposits typically occur (Lipe 1989). Irregularly shaped units, generally associated with the exposure of buried structures, were used as conditions warranted. Table 3 lists the 1999 excavation units by block and presents information about unit size, average unit depth, the presence of features and structures, and the status of each unit at the end of the field season (complete or incomplete). Figure 4 illustrates the locations of all excavations conducted to date. Excavations, grouped by block, are discussed in greater detail below. We also provide a brief summary of each previously investigated block, in which we use tree-ring dates and/or decorated pottery to infer the time period of use and deposition. Dating of the pottery assemblages follows Wilson and Blinman (1991).

Block 100

Block 100 was again a focus of excavation during the 1999 season (Figure 5). This area is located at the highest point of the site and contains the only preserved, substantial above-ground architecture at the site. Multiple lines of evidence suggest that this roomblock may have been a community facility or was at least a focus of community attention (Adler 1990; Duff and Ryan 1998). Excavation of areas surrounding this roomblock suggests that this portion of the site has the greatest depth of use and range of occupation thus far noted at Shields Pueblo. In 1999, our efforts included continued excavation of incomplete units, exploration of structures revealed in trenches during the 1998 season, the testing of additional remote-sensing anomalies, and excavation of structures discovered during that testing.

Structures 102, 103, 104, and 121

These structures are four rooms in the preserved roomblock in Block 100. This roomblock occupies a prominent location at Shields and in the surrounding area, and the masonry is reminiscent of Chacoan-style masonry. Excavation in these areas in 1997 and 1998 produced a number of wood samples, a few of which yielded tree-ring dates. Backhoe Trench 114, used to expose Structure 121, produced a sample that yielded a noncutting date of A.D. 1110. Excavations in Structure 104 produced a wood sample with a cutting date of A.D. 1127. Although these samples derive from disturbed contexts, they suggest an early A.D. 1100s use of this roomblock.

Structure 110 and Nonstructure 101

Work in Structure 110 was completed in 1999. Structure 110 is an earthen pit structure tree-ring dated to the A.D. 770s, and it is overlain by a later, intact midden deposit (Nonstructure 101). Wood from the midden deposit, likely fuelwood, produced a noncutting date of A.D. 1127, suggesting possible deposition in this structure contemporaneous with the use of the Block 100 roomblock. This area has been examined using trenches, a 1-x-1-m unit, and two 2-x-2-m units. A single unit was excavated to the floor of Structure 110, exposing the hearth, which was then excavated. Unfortunately, no ash was present. A possible subterranean room was found in the northern portion of this excavation area, and this will be investigated in 2000.

Structure 122 and Structure 123

Colorado Mountain College excavated these structures in 1977, discovering a masonry kiva (Structure 123) constructed inside an earlier, larger kiva (Structure 122). In 1998, we excavated Backhoe Trench 115 across this depression to determine if this was the previously excavated structure, and then we completely reexcavated the interior structure (Structure 123). Additionally, a unit was placed between the two structures in what appeared to be intact deposits, and work in this unit continued in 1999. This unit revealed intact stratigraphy and burned roof deposits associated with Structure 122; numerous tree-ring samples should permit us to absolutely date its construction. It appears that Structure 122 was burned at abandonment and then was partly filled with trash. Sometime later, Structure 123 was constructed inside Structure 122, reusing the earlier structure's southern recess and ventilation system. Several wood samples from our 1998 trenching and reexcavation of Structure 123 appear to derive from the trash fill that collapsed into this structure after it was excavated in 1977. Tree-ring dates from CMC excavations include a noncutting date of A.D. 1117 (Table 1), but it is not clear from what context these samples derived. The latest noncutting tree-ring date from our excavations, likely from a trash deposit that postdates the earlier construction but predates the final kiva, is A.D. 1147. These data suggest that an early A.D. 1100s date for Structure 122 is likely, and this is consistent with the tentative dating of the associated preserved roomblock.

Structure 124

Structure 124 is an earthen-walled, subterranean room that was used for grinding. In 1999, continuing excavation in this area defined the edge of the structure and a portion of the floor and exposed the tops of two mealing bins. The fill consisted of ash-rich sediment, natural deposits, and trash deposits. Several ephemeral use-surfaces and pits were found as well. Excavation of the grinding features will be completed in the 2000 field season.

Structure 136

Backhoe Trench 133 was excavated to test a remote-sensing anomaly, exposing an earthen-walled pit structure (Structure 136). No additional excavation within this structure was conducted in 1999.

Structure 137

Backhoe Trench 135 was excavated to test a remote-sensing anomaly, exposing Structure 137. Two 2-x-2-m units were used to sample this structure and fill, and both were excavated to the floor of the structure. Fill consisted of dense midden deposits reaching to the floor. A section of the southwestern wall of Structure 137 incorporated masonry, but the remainder was earthen. A hearth, ventilator tunnel, wall aperture, and disassembled deflector were exposed or excavated, and ash was collected from the hearth for analysis. The decorated pottery assemblage from the trash fill is dominated by Mancos (approximately 60 percent) and McElmo (approximately 35 percent) black-on-white types, suggesting deposition in the A.D. 1100s.

Structure 138

Structure 138 is an earthen-edged pit structure revealed during the excavation of Backhoe Trench 128 in 1998. Excavation in a section of this structure indicates that it was filled with trash, and it may have been associated with the use of Structures 102 or 103. Decorated white ware pottery that could be assigned to type from the trash fill contains nearly equal percentages of Mancos and McElmo black-on-white sherds. This assemblage is consistent with deposition prior to A.D. 1140, approximately the same period suggested by several of the features and deposits surrounding the Block 100 roomblock.

Structure 139

Structure 139 was encountered in a judgmentally selected 2-x-2-m unit. An additional unit was excavated to further define the extent of this structure, and the southern recess and a portion of the ventilator shaft were exposed. The southern recess bench and walls are earthen, and a posthole in the southern recess appears to have anchored a roof-support post. This structure will be examined further in 2000.

Structure 140

A portion of Structure 140, an intensively burned wattle-and-daub wall, had been discovered at the western end of Backhoe Trench 128 in 1998. A wood sample from this wall produced a cutting date of A.D. 1103. In 1999, the trench was expanded west to define the extent of this wattle-and-daub surface room, revealing limited undisturbed fill. Further excavation revealed a floor surface and a slab-lined hearth, from which ash was taken and sampled for analysis. The location of this room suggests that its use could have been associated with any number of adjacent structures or features, but the A.D. 1103 cutting date makes Structure 140 the earliest absolutely dated Chacoan-era structure presently known at Shields Pueblo.

Structure 141

Structure 141 was exposed when Backhoe Trench 116 was excavated in 1998. In 1999, a portion of this structure was excavated to floor. Fill consisted of natural deposits and culturally redeposited sterile sediment. Structure 141 is a roughly rectangular, earthen-walled subterranean room. A series of upright slabs defined a feature in the southeast corner of the room and a Chapin Gray vessel was recovered from the floor, indicating that Structure 141 predates A.D. 950, although the absence of other pottery types suggests that it could date as early as A.D. 600.

Structure 145

Backhoe Trench 143 was excavated to test a large remote-sensing anomaly, discovering Structure 145. An irregularly shaped portion of this structure was excavated to floor. Fill consisted of structural collapse with abundant sandstone blocks and caliche-rich sediment. A hearth that had been remodeled, a pit (disassembled deflector?), and bench surface were encountered and excavated. Ash from both hearths was collected for analysis. Structure 145 appears to have had its roofing elements removed at abandonment, and it may have been used for the deposition of construction rubble.

Structure 146

Backhoe Trench 144 was excavated to test a remote-sensing anomaly, revealing Structure 146, a roughly circular structure with masonry and earthen walls. The portion of this structure within the trench was excavated to floor. Fill consisted of caliche-rich soils with some intermixed, unburned roofing material. A diverse artifact assemblage was recovered from the floor and the few centimeters above, including numerous flakes and pottery, and an articulated turkey skeleton. A hearth was encountered and excavated, and ash was collected for analysis.

Block 100 2-x-2-m unit

A single 2-x-2-m unit was excavated in the Block 100 area in 1999. This unit was excavated through the plow zone, encountering Nonstructure 142, a midden deposit noted in Backhoe Trench 127 in 1998. Limited intact midden deposits were found in the 2-x-2-m unit before sterile sediment was reached.

Block 100 Summary

Portions of four masonry surface rooms, one wattle-and-daub surface room, three masonry kivas, three earthen and masonry-walled pit structures, four earthen-walled pit structures, two subterranean rooms, four 2-x-2-m units, and eight 1-x-1-m units have now been excavated in the Block 100 area (Table 3). Generally, the decorated white ware pottery assemblage is dominated by Mancos and McElmo black-on-white types, suggesting that deposition in this area primarily predates A.D. 1180. Tree-ring dates securely place the construction of Structure 110 in the A.D. 770s, but there appears to have been a hiatus of use that lasted almost 300 years. Several tree-ring dates, from a variety of contexts, suggest that the Block 100 area was again the focus of activity in the early to mid-A.D. 1100s. To date, there is limited evidence for A.D. 1200s deposition in the area, although additional tree-ring dates and pottery analyses should help to clarify this observation. Several currently undated structures have the potential to date to the apparent hiatus. Although preliminary, these data support the suggestion that the Block 100 roomblock-the only preserved, standing architecture on the site-may have been a focal structure within the Goodman Point community during the Chacoan era. The thin, tabular masonry walls, relatively large rooms, the location of the roomblock at the highest point within the site, and its association with the prehistoric road that runs to Casa Negra, all support this possibility. However, this building does not have all of the hallmarks of a Chacoan great house; chiefly, it lacks a great kiva. The investigation of structures and deposits in the Block 100 area will continue to be a research priority in 2000.

Block 200

In 1999, effort in Block 200 was devoted to the continued excavation of structures discovered in 1998 and to the testing of remote-sensing anomalies and their associated structures (Figure 6). Each area in which excavation was conducted in 1999 is described below, and results from structures examined in 1998 are also summarized.

Structure 205

Structure 205 is a masonry and earthen-walled subterranean room initially exposed in 1997 and completed in 1999. Fill consisted of mixed roof collapse and secondary trash deposits overlying a prepared floor. A single feature-a slightly elevated portion of the floor-appears to have served as a bench or shelf. No additional features or indications of structure function were found. Preliminary analysis of pottery from the fill provides little information about when the structure was used.

Structure 208

A section of Structure 208 was excavated to floor in 1998. This structure is a masonry-lined kiva with fill composed of burned roofing material. However, it appeared that primary roofing elements had been removed prior to structure abandonment and burning. A small section of a two-course masonry wall was found above the collapsed roof-an example of the circular stone structures built inside the depressions of collapsed, burned kivas at Shields Pueblo (described in detail below, see descriptions of Structures 229, 1411, and 1412). Six noncutting tree-ring dates were obtained from Structure 208, the latest of which is A.D. 1245vv.

Structure 221

Structure 221 is a masonry-lined kiva discovered in Backhoe Trench 216 late in the 1998 season. Excavation in the western portion of the structure began at the base of the trench (Segment 1) in 1998, encountering abundant burned roof beams. Tree-ring cutting dates of A.D. 1219 and 1243 were obtained from samples in this structure. To ensure that we would find the hearth, Backhoe Trench 230 was excavated southwards, perpendicular to the original trench. The south wall of the southern recess was exposed in this trench.

The remaining portion of Segment 1 was excavated until the floor was reached. A second unit was excavated beneath Backhoe Trench 230 to expose the hearth. Both units produced hundreds of potentially datable wood fragments and revealed a de facto floor assemblage that included portions of at least three reconstructible vessels. A masonry deflector and part of the hearth were exposed, and ash was collected for analysis.

Structure 222

Structure 222 was exposed by Backhoe Trench 217 late in the 1998 field season. The trench exposed portions of this masonry-lined structure, but no additional excavation was conducted. To ensure that we would encounter the hearth, Backhoe Trench 231 was excavated perpendicular to, and on the south side of, the original trench. The southern wall of the southern recess was exposed in this trench.

Two units were excavated from the base of the trenches to the structure floor; both contained natural fill and unburned roof fall. A coursed masonry deflector and a pilaster were exposed. The hearth, a bench surface, and part of a floor pit were also excavated, and ash was collected from the hearth for analysis. Structure 222 appears to have had its roof timbers and contents salvaged at the time of its abandonment; decorated pottery recovered from the fill suggests this abandonment took place prior to the late A.D. 1100s.

Structure 223

Structure 223 was exposed in Backhoe Trench 218 late in the 1998 season, and the entire portion of the structure within this trench was excavated to the floor. Structure 223 is a masonry-lined kiva with burned roof-fall deposits, although primary roofing elements appear to have been removed. Several samples yielding noncutting tree-ring dates were collected from this structure, including one with a date of A.D. 1250. An additional wood sample from Backhoe Trench 218, likely a piece of Structure 223 roof material, produced a noncutting date of A.D. 1256.

Structure 224

During the 1998 season, Backhoe Trench 219 exposed Structure 224, an intensely burned masonry kiva. Excavation within the west half of this trench (Segment 1) exposed several primary roof beams and numerous smaller structural elements. A tree-ring cutting date of A.D. 1238 was obtained from one of these samples. Additionally, wood samples collected during trench excavations produced noncutting dates of A.D. 1246 and 1250.

In 1999, Backhoe Trench 232 was excavated parallel to the original trench on its south edge to facilitate access and to remove plow-zone deposits. Excavation within Segment 1 continued, and the remaining portion of Structure 224 within Backhoe Trench 219 was excavated. The fill consists of dense, burned roofing material with approximately 250 additional pieces of wood which were collected for tree-ring dating. Neither of these units were completed, and excavation within this structure will continue in 2000.

Structure 225 and Structure 229

The south end of Structure 225, including the tops of the two southernmost pilasters, was discovered in 1998 in Backhoe Trench 220. To facilitate excavation of the structure and exposure of the hearth, in 1999 Backhoe Trench 228 was excavated north from the original trench, and a masonry wall was discovered in its northern portion. Additional exposure of the south edge of the Trench 228 revealed two courses of masonry. This wall was part of a later, circular masonry structure (Structure 229) built inside the depression created after the roof of the masonry kiva (Structure 225) had collapsed. Similar structures were encountered nearby in Structure 208, in Block 1400 (see Block 1400 description below), and in Block 400 by Colorado Mountain College in 1977.

A unit excavated from the base of Backhoe Trench 228 was designed both to expose any features or surface associated with Structure 229 and to expose the hearth and floor of Structure 225. No features associated with Structure 229 were detected. The wall of Structure 229 consisted of two to three courses of dry-laid masonry (maximum height of 22 cm) in the southern portion of the excavation unit. The northern wall was more substantial, ranging from five to eight courses of dry-laid masonry and reaching a maximum height of 59 cm. It appears that Structure 229 was constructed atop an uneven surface within the depression created by the collapse of the underlying kiva (Structure 225), with the different courses designed to achieve a uniform final wall height. The function of this structure remains unknown.

The excavation of this segment continued to the floor of Structure 225. Fill consisted of burned roof materials intermixed with natural deposits, but the primary roofing elements appear to have been removed. A coursed masonry deflector was exposed, and a sipapu and a hearth were excavated. Ash was collected from the hearth for analysis.

Kiva Excavated by Colorado Mountain College in 1975

A kiva completely excavated during a Colorado Mountain College field school in 1975 was relocated using original field notes (Bagwell 1975) and a preserved site datum. No reexcavation was conducted. The structure is a masonry-lined kiva with tree-ring cutting dates of A.D. 1129 and 1131 (Table 1).

Structure 234

Backhoe Trench 227 was initially excavated along an east-west axis to test a large remote-sensing anomaly. The structure discovered in this trench-Structure 234-proved to be deeply buried, so the excavation was expanded to the south to facilitate access. The test pit was eventually excavated to slightly below the level at which the face of the masonry bench was found.

Structure 234 was examined with two excavation units that began at the base of the trench. Segment 1 exposed a portion of the southern recess pilaster and uncovered portions of the bench face and floor. Several additional features were encountered, including a slab deflector placed inside an earlier hearth, and a portion of the final hearth. A second unit was excavated specifically to expose more of the final hearth, and abundant ash was collected for analysis. Structure 234 incorporated masonry into the pilasters and bench, but the bench face was primarily plastered earth. Fill consisted of natural deposits and unburned roof fall, though the uppermost fill appears to have resulted from trash deposition. This trash deposit (Nonstructure 233) was probably associated with the use of several structures located immediately to the north, probably during the later A.D. 1200s, and will be examined in 2000. Preliminary analysis of pottery recovered from Structure 234, and its architectural characteristics, suggests original construction and use in the A.D. 1100s.

Structure 237

Backhoe Trench 236 was excavated to test a remote-sensing anomaly, and revealed Structure 237, an earthen-edged pit structure without a bench. Excavation of the portion of the structure within the trench exposed the floor and a hearth that had been remodeled twice. The hearths were excavated, and ash from the latest hearth was collected for analysis. The excavated structure fill consists of unburned roof collapse and natural deposits. A trash deposit defines the uppermost structure fill, which will be sampled in 2000. Limited pottery data and the earthen construction suggest that Structure 237 was used sometime prior to A.D. 1140.

Block 200 2-x-2-m Units

Several 2-x-2-m units, a few of which were begun in 1998, were excavated to collect assemblages associated with Block 200 structures. Six 2-x-2-m units were excavated during 1999, only one of which was completed. Three of these units encountered intact trash deposits below plow zone in the same vicinity (south of Structures 221-223); this deposit was designated Nonstructure 237. A fourth unit may also be associated with this deposit. Two additional 2-x-2-m units, immediately south of Structures 225 and 229, also encountered an intact midden, designated Nonstructure 238. These units will be completed in 2000.

Block 200 Summary

Portions of six masonry kivas, one earthen pit structure, one earthen and masonry pit structure, a subterranean room, and several 1-x-1-m and 2-x-2-m units have been excavated in Block 200 (Table 3). Four of the kivas have been tree-ring dated to the mid-A.D. 1200s. A previously excavated kiva with tree-ring dates in the A.D. 1130s is present, and other pit structures examined during Crow Canyon excavations suggest use of Block 200 may predate A.D. 1100. The decorated pottery assemblage excavated or collected from nonstructural contexts is dominated by Mancos Black-on-white (65 percent) and McElmo Black-on-white (27 percent), suggesting extensive deposition in Block 200 throughout the late A.D. 1000s and early 1100s, consistent with several of the excavated structures. Deposits associated with the late-1200s occupation are still being analyzed, but these appear to have been more localized within the Block 200 area.

Block 300

Four 1-x-1-m test units were excavated in the Block 300 area in 1997, but no features or structures were found. Decorated white ware pottery from these units is dominated by Mancos and McElmo black-on-white, suggesting that these deposits date primarily to A.D. 1000-1180.

Block 400

Block 400 contains several mounded areas of rubble that have been bulldozed together. Colorado Mountain College conducted excavations in this area; a sample from a kiva they excavated (Feature C) produced a tree-ring cutting date of A.D. 1212 (Table 1). A circular stone structure had been constructed atop this kiva after its roof had collapsed, similar to structures also encountered in Blocks 200 and 1400. Four 1-x-1-m test units were excavated in this area in 1997, but no features or structures were encountered. The decorated pottery assemblage from these excavations is dominated by Mancos and McElmo black-on-white, but it also contains Mesa Verde Black-on-white. Pre-A.D. 1100 red ware pottery is also present. The pottery suggests primary use between A.D. 1000 and 1180, with continuing use into the A.D. 1200s indicated by both pottery and tree-ring dates.

Block 500

Block 500 was examined with nine 1-x-1-m test units in 1997 and 1998, but only a single pit feature was encountered. A large depression appears to indicate the presence of a pit structure that had been excavated sometime prior to 1975 (Adler 1990:Figure 26), and aerial photos from 1975 suggest that it may have been a masonry-lined kiva. No information is available about this structure or its excavation. Decorated pottery that could be identified to type is dominated by Mancos (61 percent) and McElmo (27 percent) black-on-white, with Mesa Verde Black-on-white (11 percent) also present. These data suggest use of this area primarily from the later A.D. 1000s into the mid-1100s, with some post-A.D. 1180 deposition.

Block 600

Block 600 was examined with nine 1-x-1-m test units in 1997 and 1998. No features or structures were found. Two excavated pit structures are visible in 1975 aerial photos of this area, and currently visible depressions suggest their location (Adler 1990:Figure 26). They appear to have been masonry-lined pit structures, but no information about their contents or characteristics is available. This appears to be the vicinity in which the burial with a copper bell was discovered (Hayes and Chappell 1962). Decorated pottery from Block 600 includes Mancos (35 percent), McElmo (39 percent), and Mesa Verde (26 percent) black-on-white. The composition of this assemblage is most consistent with deposition that occurred throughout the A.D. 1100s and into the A.D. 1200s.

Block 700

Eight 1-x-1-m test units were excavated in this area in 1997. A scatter of sandstone and other debris appear to be the remains of a historic homestead, and no prehistoric features or structures were found in this area. Overall, artifact density was quite low and several of the areas excavated encountered sterile sediment close to modern ground surface. The Block 700 materials appear to derive from sporadic prehistoric deposition, erosion, and historic disturbance.

Block 800

This area is truncated on the south by a county road. An unfinished 1-x-1-m unit from 1997, one 2-x-2-m unit, one backhoe trench, and a portion of a masonry kiva were excavated and completed during the 1999 season (Figure 7) (Table 3). One randomly selected 1-x-1-m test unit (431N 586E) remained to be completed. Most of the unit contained sterile sediment, but an irregular area of the unit contained cultural deposits. This section had been excavated to as great a depth as was possible without expanding the area of excavation. Excavation revealed a contact between sterile and cultural deposits that was thought to potentially represent the edge of an earthen-walled pit structure. The remote-sensing data supported this possibility, but the definition of the associated anomaly was less distinct than in other areas of the site. Because of the poor definition, it was not considered an excavation priority in 1998.

However, the identification of earth-walled pit structures, which generally are earlier than masonry-lined pit structures, became a higher research priority in 1999. Backhoe Trench 802 extended north from the 1-x-1-m unit, revealing the masonry pilasters, a southern recess bench face, and a northern bench face of Structure 803. The feature originally encountered in the 1-x-1-m unit was associated with the ventilator shaft of this structure.

A single, slightly irregular rectangular test unit was excavated in the northern portion of Structure 803. Natural accumulations and unburned roof fall were removed before reaching the floor. Twelve features were recorded for Structure 803, including two wall niches, one masonry bench, one masonry deflector, three masonry pilasters, one sipapu, one hearth (remodeled twice), and a wall that appears to have been a later construction. The sipapu and the hearths were excavated, and ash from the latest hearth was collected for analysis. The masonry consisted of finely coursed, thin tabular sandstone slabs. The later wall (Feature 12) was constructed on natural fill across the northern edge of the southern recess sometime after the structure had been abandoned, possibly to create a subterranean room out of the southern recess of Structure 803. The roof of Structure 803 appears to have been dismantled and the kiva was allowed to fill in naturally. The construction of the later wall attests to continued use, although the lack of trash fill suggests the possibility of a hiatus in the use of this immediate area.

One 2-x-2-m unit was excavated southeast of Structure 803 to obtain associated midden deposits. The unit was excavated until sterile deposits were encountered, with artifact density suggesting disturbed midden deposits.

Decorated pottery from surface collections and the 1-x-1-m units includes nearly equal numbers of Mancos, McElmo, and Mesa Verde black-on-white. The limited number of sherds from Structure 803 includes a Mesa Verde Black-on-white sherd. The pottery assemblage suggests deposition of material dating throughout the A.D. 1100s and early 1200s, with Structure 803 tentatively thought to have been used in the A.D. 1100s.

Block 900

Block 900 is also truncated by the county road. Three 1-x-1-m test units were excavated in this area in 1997. No prehistoric features or structures were found. Decorated pottery from surface collections and the 1-x-1-m units includes nearly equal numbers of Mancos, McElmo, and Mesa Verde black-on-white, suggesting A.D. 1100-1200s deposition in this location.

Block 1000

Four 1-x-1-m test units were excavated in this area in 1997. Additionally, a large depression thought to be a possible great kiva was examined with a trench in 1998, proving it was a historic pit into which rubble was deposited. A few pieces of decorated pottery from these units could be assigned to type, but Mancos and McElmo black-on-white sherds dominate. This small sample suggests that use of this area of the site generally predates A.D. 1180.

Block 1100

Four 1-x-1-m and five 2-x-2-m test units were excavated in this area in 1998. An intact midden and a slab-lined pit feature were discovered. Charcoal samples from the pit feature, possibly associated with an extramural food preparation area, produced noncutting tree-ring dates of A.D. 986 and 1199. The decorated white ware pottery assemblage from excavations in the Block 1100 area is dominated by McElmo Black-on-white (54 percent) and Mesa Verde Black-on-white (28 percent). Mancos Black-on-white and earlier white and red wares are also present. The pottery assemblage suggests use of this area throughout the A.D. 1100s and into the 1200s, consistent with the tree-ring dates.

Block 1200

In 1999, excavation continued to completion in one unfinished 1-x-1-m unit (594N 355E). Both disturbed and intact (Nonstructure 1202) midden deposits were exposed before sterile sediment was reached. Most of the pottery from this area that could be assigned to type was found in the intact portion of this midden. Approximately two-thirds of the pottery assemblage consists of McElmo Black-on-white, with the bulk of the remaining pottery classified as Mesa Verde Black-on-white, suggesting a mid-A.D. 1100s to early 1200s deposit.

Block 1300

The density of remote-sensing anomalies and artifacts from 1-x-1-m test pits in Block 1300 suggested the likelihood of both long-term use and the presence of earlier structures. Nine 2-x-2-m units, one 10-x-12-m surface-stripping unit, three backhoe trenches, and portions of two structures were excavated during the 1999 season (Figure 8) (Table 3).

Structure 1307

Backhoe Trench 1305 was excavated to test a remote-sensing anomaly, and revealed Structure 1307, an earth-walled pit structure that had been remodeled, with fill consisting of disturbed plow zone, natural accumulations, and portions of intact midden. Hand excavation within Structure 1307, in western and eastern segments, began at the base of the trench and continued until the structure floor was reached. Structure 1307 began as a relatively large circular structure that fell out of use, after which it was remodeled into a smaller structure, reusing the original western wall. The floor level of the original building was encountered in the eastern segment (Segment 2), and the hearth (Feature 2) associated with this use was excavated. A small pit had also been excavated into the original structure floor, but its function is unknown. The floor of the remodeled construction had been excavated through the original floor, and the hearth encountered on this lower floor appears to have been located near the center of the remodeled building. Both hearths were excavated, and ash from each was collected for analysis. Neither building appears to have had a bench. The pottery assemblage from the fill of the western (later) segment suggests a use that predates A.D. 1100.

Structure 1308

Backhoe Trench 1306 was excavated to test a remote-sensing anomaly, and it revealed Structure 1308, an earthen pit structure. The western portion of Structure 1308 within the trench was excavated with the goal of determining the form of the walls, exposing the floor, and locating the hearth. Structure fill consists of an intact midden deposit, natural accumulations, and burned roof fall. An earthen bench and the structure floor were exposed, but the hearth was not found. Pottery from the trench-probably from trash that postdates the use of the structure-appears to date to the first half of the A.D. 1100s.

Structure 1309

Backhoe Trench 1304 was excavated to test a remote-sensing anomaly, and Structure 1309, an earthen pit structure, was identified in profile. Fill consists of disturbed plow zone, natural accumulations, sandstone debris, and a midden deposit. No additional excavation was conducted within this structure.

Other Block 1300 units

A 10-x-12-m area in the southwestern portion of Block 1300 was mechanically scraped to remove the plow zone in the area of several remote-sensing anomalies. The goal of this experimental procedure was to determine if features and structures, and their spatial relationships, were preserved beneath the plow zone. Three features were identified, two of which are likely to be structures. No excavation below the stripped area was conducted in 1999; exploration of these features will begin in 2000. Additionally, nine judgmentally selected 2-x-2-m units were excavated south and east of structures to collect associated midden deposits. Two of these units were completed in 1999. The completed units were excavated until sterile deposits were reached; no features or structures were present. Seven 2-x-2-m units, and a single 1-x-1-m unit (543N 642E) begun in 1998, will be completed in 2000.

Block 1400

Block 1400 continued to be a focus of our investigations, with four 2-x-2-m units, two backhoe trenches, an intact midden, and portions of six structures being excavated during the 1999 season (Figure 9). In 1998, portions of two intensively burned kivas (Structures 1402 and 1408) had been investigated; each contained later, circular stone structures that had been constructed in the depressions created after the roofs of these kivas collapsed. An additional area characterized by a dense midden deposit had also been exposed and partially examined (see Duff and Ryan 1998). Each of these areas was intensively excavated in 1999.

Structure 1402 and Structure 1412

In 1998, the northern third and the southern recess of Structure 1402 were excavated to their floor and bench surfaces (Duff and Ryan 1998). Structure 1402 was burned at abandonment, with roofing timbers and material preserving a de facto artifact assemblage. Natural sediment accumulated over the collapsed kiva roof, and then the circular masonry wall of Structure 1412 was constructed inside this depression. Tree-ring dates from the 1998 excavations of Structure 1402 include two near-cutting dates of A.D. 1258v, indicating that this structure was in use, and was probably constructed, soon after this date. The presence of Structure 1412 in the depression of this late, collapsed kiva became the focus of excavation in 1999.

The circular masonry wall of Structure 1412 was evident in the profiles of Backhoe Trenches 1403 and 1404 excavated in 1998 (Duff and Ryan 1998). That same year, a large segment had been excavated along the southeast of this area, exposing the top of the eastern wall of Structure 1412. At the beginning of the 1999 field season, a similar area along the southwest portion of Structure 1402 was exposed to facilitate the complete excavation of Structure 1412. The entire area within the wall of the structure was excavated. Natural and cultural fill was removed before encountering the surface upon which Structure 1412 was constructed. Shortly after the abandonment of Structure 1402, fill rich in charcoal flecks and artifacts was deposited over accumulating natural deposits, creating a platform on which Structure 1412 was constructed. Between six and 13 courses of dry-laid masonry (with a maximum height of 57 cm) were preserved, and this appears to have been close to the original height of the structure. No use-surface or internal features were present. The function of this structure is unknown.

Upon completion of the excavation of Structure 1412, a unit (Segment 12) was excavated through burned roof fall to the floor of Structure 1402 to uncover the hearth. The floor, the north portion of a hearth, and a sipapu were exposed. Portions of a reconstructible corrugated jar were recovered from the floor. Both features were excavated and ash was collected from the hearth for analysis.

Structure 1408, Structure 1411, and Structure 1413

Backhoe Trench 1406 was excavated to test a remote-sensing anomaly, and it revealed an intensively burned, masonry-lined kiva (Structure 1408) that also contained a circular stone structure (Structure 1411) constructed above roof fall and natural sediment. Excavation in Structure 1408 (Segment 1) in 1998 encountered the floor and collected a number of roof beams for tree-ring dating. The tree-ring dates from Structure 1408 include several noncutting dates and a single cutting date of A.D. 1254, suggesting construction or use at this time and demonstrating its contemporaneity with neighboring Structure 1402. An excavation strategy similar to that used in the excavation of the Structure 1402/Structure 1412 area was used.

A large, rectangular area on the north and south sides of Backhoe Trench 1406 was mechanically stripped to remove the plow zone revealing the outline of Structure 1408 and facilitating the complete excavation of Structure 1411. This exposed the entire outline of Structure 1408 and walls of the southern recess. Another structure (Structure 1413) was revealed in the southwestern corner of this area. Excavations within Structure 1408, Structure 1411, and Structure 1413 proceeded simultaneously.

A second segment within Backhoe Trench 1406 was excavated to expose additional floor materials and find the hearth of Structure 1408. Fill consisted of burned roof fall. Portions of at least two reconstructible vessels were recovered, the hearth was excavated, and ash was collected for analysis. The southern recess of Structure 1408 was also excavated. Fill consisted of natural deposits and burned roof fall atop the bench surface. A ventilation shaft and a tunnel leading to Structure 1413 were exposed (see below).

Structure 1411, a circular wall built within the depression created by the collapse of the roof of Structure 1408, had been bisected by Backhoe Trench 1406 in 1998. Complete excavation of the fill within Structure 1411 on either side of the trench was conducted in 1999. It appears that after the collapse of Structure 1408 there was a relatively brief interval of natural accumulation that filled in the depression and then Structure 1411 was constructed on this ephemeral surface. Structure 1411 had between two and five dry-laid masonry courses (with a maximum height of 40 cm) and no internal features. The function of this structure is unknown.

Structure 1413, a square, masonry-lined subterranean room, was detected after the area south of Backhoe Trench 1406 had been mechanically stripped. A north-south trench (Backhoe Trench 1410) was excavated to the south of this stripped area, encountering sterile deposits and the edge of a feature (Feature 2, Structure 1413) immediately below the plow zone. Two additional units were excavated between the trench and the stripped area to define the extent of Structure 1413, exposing approximately one-third of the eastern portion of the structure. Natural fill and unburned roof-fall deposits were excavated before reaching the floor. Several features were associated with this structure. A partially sealed doorway and masonry-lined entry or shaft were documented in the southern wall and exterior of the room. A tunnel entrance and tunnel in the east wall connected Structure 1413 with the southern recess of Structure 1408. No floor features or other indications of room function were found. The presence of the tunnel suggests contemporaneity with Structure 1408, which is tree-ring dated to A.D.1254.

Nonstructure 1409 and Structure 1414

Backhoe Trench 1407 was excavated during the 1998 season to test a remote-sensing anomaly; excavation ceased when dense midden deposits (Nonstructure 1409) were reached. Excavation within this trench continued in 1999 and a subterranean kiva (Structure 1414) was encountered beneath the midden. Units were placed on either side of the trench to sample the midden more intensively. Both units were excavated to the base of the Nonstructure 1409 deposit, at which point the north unit was terminated. The south unit was continued into Structure 1414.

Structure 1414 is a pit structure with both earthen and masonry walls. Fill consisting of natural sediment and unburned roof deposits was excavated to floor within the original trench unit and the midden unit abutting this trench to the south. Three wall niches, an ash pit, and a hearth were excavated, and a masonry pilaster was exposed. Ash was collected from the hearth and the ash pit for analysis. In the portion of the structure excavated, masonry was only used to construct the wall niches and the pilaster. The pottery assemblage recovered from this structure is dominated by McElmo Black-on-white. The significant accumulation of natural fill and overlying trash deposition, and the largely earthen construction, suggest a pre-A.D. 1200 use of this building.

Structure 1416

Backhoe Trench 1415 was excavated to test a remote-sensing anomaly on the southwestern edge of Block 1400 and encountered Structure 1416, an earth-walled pit structure. Structure fill consists of disturbed plow zone, natural accumulations, and a dense trash deposit. Pottery from this trench is dominated by McElmo and Mesa Verde black-on-white, suggesting that this structure was filled with deposits in the late A.D. 1100s and early 1200s. Portions of this structure and trash deposit will be excavated in 2000.

Block 1400 2-x-2-m Excavation Units

Four judgmental 2-x-2-m units were excavated in 1999. Two were placed south of Structure 1408 (begun in 1998), and two were placed south of Structure 1402; all were intended to sample associated midden deposits. The four units were excavated until sterile sediment was reached.

Block 1400 Summary

The following were excavated in Block 400 to date: portions of two masonry-lined, burned kivas, two circular stone structures built within the depressions formed by the collapse of these kivas; a masonry-lined subterranean room; a masonry and earthen-walled pit structure; an earthen-walled pit structure; an intact midden; several 2-x-2-m units; and four 1-x-1-m units (Table 3). Additional excavation in Structure 1416 will be done in 2000, but the remainder of the Block 1400 excavations were completed in 1999. Tree-ring dates securely place the use and, likely, the construction of the two burned kivas in the A.D. 1250s. The circular stone structures constructed atop the kivas obviously postdate kiva abandonment. If the kivas were used for a decade or more after construction, these circular stone structures were constructed at about the time the Mesa Verde region was being depopulated. McElmo and Mesa Verde black-on-white pottery is well represented in the Block 1400 assemblage, consistent with the late structures. However, Mancos Black-on-white pottery is also abundant, suggesting extensive deposition in the later A.D. 1000s and throughout the 1100s, potentially associated with the uses of Structures 1414 and 1416.

Block 1500

Two backhoe trenches and portions of two masonry kivas were excavated in Block 1500 in 1999 (Figure 10). In 1998, four 1-x-1-m and four 2-x-2-m units were excavated in this area.

Structure 1504

Backhoe Trench 1502 was excavated to test a remote-sensing anomaly, exposing the masonry pilasters on either side of Structure 1504, a masonry-lined kiva. Excavation within the eastern two-thirds of the trench was conducted until the floor was exposed. Fill consisted of natural deposits and unburned roof fall. A hearth that had been remodeled was excavated, and ash from the latest hearth was sampled for analysis. This structure appears to have had its roof timbers salvaged when it was abandoned.

Structure 1505

Backhoe Trench 1503 was excavated to test a remote-sensing anomaly. Structure 1505, a masonry-lined kiva that had been burned at abandonment, was found in the trench. The entire portion of Structure 1505 within the trench was excavated to floor in western and eastern halves. Fill consisted of natural deposits overlying intensively burned roof fall. A masonry deflector with a small niche occupied the central part of the trench. A small portion of the hearth encountered in the north wall of the trench was excavated, and ash was collected for analysis. Numerous wood samples were collected for tree-ring analysis.

Block 1500 Summary

Pottery assemblage data from the units excavated in 1998, primarily the four 2-x-2-m units, contain approximately equivalent percentages of Mancos and McElmo black-on-white, and a smaller proportion of Mesa Verde Black-on-white sherds. The intensively burned Structure 1505 resembles structures elsewhere on the site that date to the mid-A.D. 1200s. The pottery assemblage indicates that Block 1500 was used throughout the A.D. 1100s and into the A.D. 1200s.

Summary of the 1999 Field Season

The 1999 field season at Shields Pueblo has proved to be our most successful season to date. We have now excavated portions of approximately 35 different structures, representing about one-third of the structures believed to be present at the site on the basis of remote-sensing data. Results from our work are beginning to provide new insights into the structure of the Goodman Point community. At least six masonry-lined kivas at Shields have been tree-ring dated between A.D. 1245 and 1258, and three of these had circular structures constructed atop them. These data suggest that people continued to occupy mesa-top settlements in the late A.D. 1200s, probably until the region was abandoned. This pattern contrasts with information from the neighboring Sand Canyon community, where mesa-top settlements appear to have been routinely abandoned after A.D. 1225 (Varien, ed. 1999). The Shields project provides additional information about the range of variability in the organization of late-Pueblo III communities in the Mesa Verde region. The types of structures encountered and examined in 1999-especially those with partial or completely earthen walls-and the pottery data analyzed to date confirm that Shields Pueblo was intensively occupied throughout the A.D. 1100s. The dominance of Mancos Black-on-white pottery in assemblages from nearly all areas of the site-usually between half and two-thirds of the typed sherds-suggests the likelihood of substantial occupation throughout the A.D. 1000s as well. Systematically collected data from structures and deposits spanning at least three centuries of (probably continuous) occupation will be a substantial and unique contribution to our understanding of Mesa Verde-region prehistory. These data will provide us the opportunity to successfully address questions related to the research design, Communities Through Time: Migration, Cooperation, and Conflict.

Goals for the 2000 Field Season

The majority of the units and structures investigated to date were completed, documented, and backfilled in 1999. The 2000 field season will be our last at Shields Pueblo. We plan to complete surface collections in all of the high-density areas of the site, and we will also make surface collections from many of the areas between the 18 identified blocks. Randomly selected 1-x-1-m test units will be excavated in the three untested high-density areas (Blocks 1600, 1700, and 1800), and these will be supplemented with judgmental units as warranted. Unfinished units will be completed. We will continue to excavate trenches in areas of remote-sensing anomalies to look for buried structures, including those areas for which we currently have no tested structures. Once the type and extent of structures have been identified, hand excavation will be conducted to collect artifact, tree-ring, and ecofact samples. Excavation of surface units adjacent to detected structures will continue, as these units provide the bulk of the artifacts used for assemblage comparisons. These data will aid in the reconstruction of the occupational history of Shields and the environmental history of the immediate area. Pottery data and dates resulting from more than 1,000 tree-ring samples submitted for analysis in 1999 will be used to select temporal intervals from which additional information or samples are needed.

How to cite this publication.

REMINDER: Archaeological resources are protected by federal laws, and archaeological research is guided by a set of professional ethics. See Archaeological Ethics and Law.


Crow Canyon Archaeological Center, 23390 Road K, Cortez, CO 81321. 970-565-8975 or 800-422-8975 Fax 970-565-4859
© Copyright 2000 by Crow Canyon Archaeological Center
All rights reserved