Plant Evidence

by Karen R. Adams and Malaina L. Brown

1
As at all other sites excavated or tested by the Crow Canyon Archaeological Center, archaeologists and volunteers routinely collected plant materials from Castle Rock Pueblo for analysis. Larger pieces of vegetal material, such as pieces of charred wood, maize (Zea mays) cob segments and kernels, and common bean (Phaseolus vulgaris) cotyledons (seed halves), were collected by hand whenever found during excavation. By convention, these specimens are often called macrofossils because they are large enough to be seen and collected separately. Plant remains were also collected from a targeted set of archaeological contexts for a water-separation process termed flotation. This process involves the use of water to separate fragile organic material from sediment, allowing for the recovery of tiny items, called microfossils, which are identified under a microscope. Together, these two methods produced an extremely rich and informative record of plant resources used by the inhabitants of Castle Rock Pueblo—a record that also reveals the nature of the environment surrounding the Castle Rock community in the late A.D. 1200s.

Methods
Flotation Sample Processing

2
Castle Rock Pueblo flotation samples were a standard 1 liter in size. If more than 1 liter was available, the extra sediment was curated; if less than 1 liter was submitted, the exact amount was measured and recorded, and the entire sample was processed. Flotation samples were processed by two different methods, although the slight differences between them are not considered significant. For some samples, a small bucket partly filled with water was placed on top of a graduated series of five USGS geological sieves with mesh sizes of 4.75 mm, 2.80 mm, 1.40 mm, 0.71 mm, and 0.25 mm, respectively. A flotation sample was then poured into the bucket, and running water was added for three to five minutes while the sediments were stirred continuously. This process allowed floating organic materials to flow over the edge of the bucket and through the stacked sieves until being caught by the first sieve with a mesh size smaller than the individual pieces of organic material. The contents of each sieve were then dried separately and placed in individual plastic bags for later examination.

3
Other samples were water processed as well, though slightly differently. Each flotation sample was poured into a bucket of water, and the sediments were stirred to loosen the organic material. The material that floated to the top was poured off onto flotation cloth with a mesh size of 0.355 mm. Running water was then added to the bucket, and the sediments were once again stirred and the floating light fraction added to the flotation cloth. This process was repeated until the water in the bucket was relatively clear. The light fraction was dried on the flotation cloth on drying racks and then sieved through USGS geological sieves with the same mesh sizes given above.

Subsampling for Seeds and Other Microremains

4
Because the volumes of light-fraction material recovered from the processed samples were quite variable, some samples were subsampled while others were examined in their entirety. All material greater than 2.80 mm in size was examined from all samples, whereas the material from the 1.40-mm, 0.71-mm, and 0.25-mm sieve sizes was subsampled if the volume of material within a sieve exceeded 5 ml. Then, approximately 25 percent of the material in the 1.40-mm sieve and 10 percent from each of the 0.71-mm and 0.25-mm sieves were examined. For any given sieve size, if numerous plant taxa were identified within a subsample, additional material was analyzed until no new taxa were found. All material was examined under a binocular microscope with magnifications ranging from 7X to 45X. Standard texts aided identification (e.g., Martin and Barkley 1961*1), as did comparison with modern collections of plant parts from the region.

Wood Charcoal Analysis

5
Charcoal was examined from all 57 flotation samples that were analyzed for seeds and other minute plant parts. Twenty pieces of burned wood (primarily larger pieces and any that looked "different") were selected from the 4.75-mm sieve for each sample, and each piece was examined for anatomical structure, ensuring the best chance of confident identification (Minnis 1987*1:122). Flotation samples with fewer than 20 pieces of charcoal large enough to retain features necessary for identification were examined in their entirety. Charcoal was also examined from macrofossil samples, again with the aim of identifying the taxa represented by 20 pieces of charred wood per sample.

6
Charred wood was oriented under the microscope and snapped to produce a clean transverse (cross) section. A collection of modern wood specimens representing the majority of tree and shrub species currently growing in the region provided material for comparison. This collection has been charred and is supported by specimens deposited in the University of Arizona herbarium. It includes smaller stems, twigs, branches, and trunk wood in order to represent the broad range of sizes of woody materials that might have been carried into a prehistoric village.

Other Macroremains

7
Although bags of macrofossils handpicked by excavators contained primarily wood charcoal, nonwood specimens were present as well. All macrofossils greater than 4.75 mm in size were examined with the unaided eye in order to locate nonwood plant parts; the nonwood specimens were then examined under the microscope. All macrofossil materials smaller than 4.75 mm were scanned under the microscope to locate any nonwood items.

Results
Overview

8
Fifty-seven flotation samples and 343 macrofossil samples from Castle Rock Pueblo were examined (Table 1). Although diverse contexts were sampled, the majority of flotation and macrofossil samples derived from secondary refuse and mixed cultural deposits. A number of flotation samples represented primary refuse in hearths or firepits within structures, including kivas, habitation rooms, and towers. Roof fall and wall fall were also well represented.

9
A diversity of reproductive plant parts and wood types had been preserved, representing both wild and domesticated plants (Table 2). At least 37 separate plant taxa were identified from Castle Rock Pueblo. Most specimens were charred. In Table 2 and all other tables in this report, taxonomic identifications are followed by the word "type," signifying that the specimen or specimens resemble the named taxon but might also compare well with other related (or at times unrelated) taxa. This conservative approach is used because of the similarity in appearance of parts of various Southwestern plants, especially when ancient specimens are carbonized or damaged. For ease of reading, the word "type" has been omitted from text discussions.

10
The entire archaeobotanical data set, including provenience information, will eventually be published on Crow Canyon's Web site. Also scheduled for on-line publication are two supporting documents designed to be used in conjunction with the analytic results presented in the database. The first document, a detailed description of the criteria used to identify prehistoric plant remains, will allow researchers to 1) corroborate any identifications we make, 2) identify unknowns from written descriptions, and 3) examine metric data in light of potential domestication models based on anatomical and morpohological traits. The second, a comprehensive summary of the ethnographic uses of various plants identified in the Mesa Verde region, may provide insights into possible prehistoric use of those same resources. (See Research Publications.)

Charred vs. Uncharred Plant Parts

11
When uncharred plant parts are preserved in archaeological sites, it is important to consider whether they were associated with cultural activities in the past or whether they entered the site via natural processes during site occupation or afterward. A conservative rule of thumb is to consider uncharred plant materials in archaeological sites as being unrelated to prehistoric human use of plants (Minnis 1981*1). Charred specimens, on the other hand, are often assumed to be related to intentional or unintentional prehistoric human use of fire. This assumption can, on occasion, be in error. For example, the recovery of a single charred Old World domesticated wheat (Triticum) grain from an ancient context at Troy's Tower in the Sand Canyon locality can be explained by the fact that modern dryland wheat-field stubble had been burned over the site (Adams 1999*1).

12
At Castle Rock Pueblo, a number of plant parts were recognized solely in uncharred condition either in flotation samples or as macrofossils. These included seeds of Celtis (hackberry), Cucurbita/C. pepo, and Prunus/Rosa, along with a Juniperus fruit/fruit coat/seed, Opuntia epidermis fragments, and Pinus edulis nutshells (Table 3). The Pinus edulis nutshells and the Prunus/Rosa fruit all displayed rodent tooth marks, suggesting the likely source of entry. The Opuntia epidermis was quite fragile, indicating relatively recent origin. A single Celtis seed and some Juniperus fruits and seeds in fairly good condition appeared modern to analysts. In contrast, the unburned Cucurbita/C. pepo seeds most likely relate to prehistoric use of domesticated squash, for three reasons: (1) squash is a known ancient domesticate in the area, (2) it was not likely grown as a dryland crop in historic times, and (3) the degraded condition of the specimens suggests that they are fairly old.

13
Other taxa and parts were found in both charred and uncharred condition at Castle Rock Pueblo. Because charred fragments of domesticated Zea mays cobs, cupules, and kernels were collected from multiple flotation and macrofossil samples, a single unburned Zea mays cob fragment associated with wall fall in Structure 206 is considered to have been of cultural origin and fortuitously preserved. In cases involving wild plants, however, where a single instance of an uncharred plant part is accompanied by examples of the same part in charred condition (Atriplex wood, Gramineae caryopsis, Pinus bark fragment, Stipa hymenoides caryopsis, Yucca/Y. baccata seed), we interpret the uncharred items conservatively as noncultural. These plants are all part of the modern flora of the region and could owe their presence to incidental intrusion of parts. Some uncharred Opuntia (prickly pear)/Opuntia seeds or fruit parts either had rodent tooth marks or appeared modern to analysts, and on these bases they, too, are excluded from further consideration.

14
One taxon whose wood seems able to withstand long burial in open sites without being burned is juniper (Juniperus). At Castle Rock Pueblo, although juniper wood was overwhelmingly preserved in charred condition, some specimens appeared "partially charred," revealing the species' ability to preserve for centuries. The uncharred sections were often darkly colored, as if aged. The wood is known today for its resistance to decay and for that reason is commonly used for fence posts. There is a reasonable chance that uncharred juniper wood in six archaeological contexts at Castle Rock Pueblo might be prehistoric. This seems especially likely for specimens preserved in roof fall (Structures 103 and 105) and on or above a possible use surface (Structure 401). Nonetheless, the conservative approach taken here has been to omit those pieces of wood from further consideration.

15
Perhaps the most perplexing situation concerns cheno-am seeds. It is often difficult to discern charring on these naturally black seeds, especially when they are not swollen or broken open for an interior view. These seeds represent members of both the goosefoot (Chenopodiaceae) and pigweed (Amaranthaceae) families, which grow in disturbed habitats, as archaeological sites often are, and produce copious quantities of seeds. At Castle Rock Pueblo, cheno-am seeds were found uncharred nearly as often as charred (Table 3). In one case (Structure 103), at least 100 uncharred seeds were found in a postabandonment deposit in fill above roof/wall fall, suggesting perhaps a rodent cache. Most often, however, they occurred in relatively low numbers, whether charred or not. Because of the weedy nature of the plants, in this report we omit all uncharred cheno-am seeds from discussions of ancient plant use.

Discussion
Foods

16
Judging from the entire set of 57 flotation and 343 macrofossil samples, it is clear that the residents of Castle Rock Pueblo were agriculturalists who grew maize (Zea mays) (Table 4). Maize occurred most often in the form of charred cupules and cob fragments in hearths and secondary refuse, and also as kernels and shank segments. The presence of cobs and shank segments suggests that fields might have been close enough for these secondary waste products to have entered the pueblo. The leftover cobs probably provided a source of fuel or tinder for hearths and firepits. A number of charred possible "meal cakes" on Bench 6 in the oversized kiva, Structure 105, are suspected of consisting of ground maize kernels. Measurements and key traits of maize and other well-preserved domesticates from Castle Rock Pueblo are summarized in Table 5.

17
Farmers also grew and/or had access to squash (Cucurbita pepo), common beans (Phaseolus vulgaris), and gourds (Lagenaria). Squash was preserved as seeds and rind fragments in five locations. Charred beans were preserved as cotyledons (seed halves) in three locations, including roof fall with de facto refuse in Structure 305, a tower. This suggests that beans might have been drying on the roof when the tower collapsed. A limited number of gourd rind fragments completes the list of recognized domesticates. One fragment was collected with de facto refuse in the kiva (Structure 402) associated with the D-shaped structure, and another was found in midden debris associated with either Structure 104 or Structure 105, both kivas. The discovery of gourd remains in the Sand Canyon locality only in late Pueblo III contexts at Sand Canyon Pueblo and Castle Rock Pueblo is noteworthy. This record has been interpreted elsewhere (Adams and Bowyer 1998*1) as evidence of one means of coping with food stress. That is, items such as gourds might have been grown in order to be kept and traded for food at some future time of hunger. The association of beans with Structure 305 (a tower) and of gourd rind with Structure 402 (a kiva) is also intriguing, considering the sparseness of these domesticates in Castle Rock deposits generally. Perhaps limited availability of, or access to, these foods brought them under some degree of social control.

18
Occupants of Castle Rock Pueblo harvested wild foods, at least on occasion (Table 4). The majority of this record is preserved in flotation samples. Charred cheno-am seeds were found in 19 samples. These plants would have volunteered readily in gardens and other disturbed locations and were likely harvested for both greens and seeds. Currently, goosefoot plants come up in early summer, and pigweed seedlings germinate after the first summer rains. The seeds of these plants start to ripen in mid- to late summer and continue to be available until fall frost. Burned seeds of prickly-pear cactus were found in deposits at Castle Rock Pueblo; the fleshy fruits of this cactus ripen in late summer and can cling to the plants until damaged by animals or freezing temperatures. The grains (caryopses) of rice grass (Stipa hymenoides) were a resource in late spring and early summer. Seeds of purslane (Portulaca) would have been gathered from gardens and other disturbed places along with the cheno-ams. Broad-leaf yucca (Yucca baccata) fruits, with mature seeds, ripen in late summer. Reproductive parts of additional taxa are present in very small numbers at Castle Rock Pueblo, hinting at an even broader diversity of native foodstuffs.

Fuel and Construction Materials

19
Judging from the entire flotation and macrofossil record, it appears that at least 15 types of wood were brought into Castle Rock Pueblo and burned (Table 6). Although some of this material likely represents construction, much of it is probably left over from the use of wood as fuel. The types of wood found as charcoal the most often were juniper (Juniperus) and pine (Pinus and Pinus edulis). The presence of other juniper and pine parts, including charred juniper twigs and pine bark, bark scales, and cone scales, suggests that a pinyon-juniper woodland existed during the occupation of the pueblo. Additional types of wood that were commonly found include saltbush (Atriplex), cottonwood/willow (Populus/Salix), and sagebrush (Artemisia). Saltbush and sagebrush are common shrubs today in the pinyon-juniper woodland, and sagebrush flourishes in recovering agricultural fields. Cottonwood and willow trees can be found in canyon bottoms. The remaining types of Castle Rock charcoal are also components of the modern flora, representing both uplands and lowlands.

Subsistence and Wood Use

20
The distribution of plant remains in different types of structures (Table 7 and Table 8) might shed light on the functions of those structures or on the pueblo's social organization. The following discussion focuses on plant remains identified in flotation samples taken from 15 thermal features (hearths and firepits) inside structures. These features are all considered to have retained the primary refuse of the last fire(s) burned in them. It should be stressed that the samples represent only the last few uses of these hearths and firepits, rather than overall or year-round use. This record is occasionally supplemented with macrofossil information. Reproductive parts of plants are generally assumed to represent accidents of food processing, although there could be other reasons why such parts might have been preserved in hearths and firepits. For example, some seeds and fruits could have been brought in with moist vegetation gathered to provide a protective layer during the roasting of some foods, such as maize ears. None of the thermal features discussed here are thought to have been roasting pits. Use of flowers and fruits in medicine and ceremonies might also have resulted in the introduction of some reproductive plant parts into hearths or firepits. The charcoal is assumed to represent primarily burned pieces of fuelwood.

21
Two towers (Structures 305 and 401) at Castle Rock Pueblo preserved a record of reproductive plant parts suggestive of food preparation. These parts included cheno-am seeds, grass grains (including rice grass), and Yucca seeds. No maize was collected from either tower, not even those parts (cobs and cupules) suggestive of fuel use. The plant record from the towers reflects the presence of people in the late spring (Stipa hymenoides) and summer. Stickleaf (Mentzelia albicaulis) seeds, rarely found in archaeological sites in the Sand Canyon locality, were preserved in Structure 401. Historically, stickleaf seeds have been eaten by a number of Southwestern groups (Bohrer 1978*1:13). Charcoal of nine different tree and shrub types was also preserved in the two towers, suggesting that various woods were used as fuel.

22
Maize was recovered from primary refuse in thermal features in seven kivas (Structures 103, 107, 112, 125, 302, 304, and 402). The presence of only charred maize kernels in three of these (103, 107, 112) suggests that maize was the sole food product prepared in the last fires built in these structures. A potential food, Portulaca, was preserved in Structure 125. Plant remains collected from the hearth in Structure 302 were similar to those in the two towers. The thermal features in Structure 304 and Structure 402 (inside the D-shaped structure) lacked reproductive plant parts entirely. Structure 402, however, had preserved gourd (Lagenaria) rind on a surface with de facto refuse, and plant parts (maize cupules, stickleaf seeds) were also present in an associated midden.

23
The emphasis on domesticated plant use in kivas is confirmed in three additional locations. Although no hearth materials were examined from the oversized kiva, Structure 105, the discovery of a number of possible Zea mays "meal cakes" on Bench 6 suggests that maize was stored inside the kiva. A midden associated with Structure 101, another kiva, contained preserved maize remains (kernel, cupule, cob) and cheno-am seeds. Finally, a maize cob was found in a midden associated with Structure 405. The use of maize cobs as fuel is indicated in Structures 112, 125, and 304. In all kivas, charcoal from a variety of tree and shrub species, representing both uplands and lowlands, also indicates fuel use.

24
Of five surface structures thought to have been habitations, two (Structures 119 and 124) preserved nothing to suggest food preparation or use, except for some maize kernels in Structure 124. Structures 108 and 117 retained only two potential wild plant foods each, along with maize kernels. The plant record in Structure 108 revealed nothing that would help explain its unusual semisubterranean position. Only Structure 111, a small room built against the southern face of the butte, contained numerous plant taxa suggestive of food preparation—seeds of cheno-am, prickly-pear cactus (Opuntia), husk tomato (Physalis), and purslane, along with grass grains and maize kernels. As in the case of towers and kivas, a number of tree and shrub types (12) were carried into habitations as fuel sources, along with leftover maize cobs.

25
In an overview of the plant record from primary thermal features presumed to relate to the last fires in many structures, some patterns emerge. More than half of these features preserved remains of maize kernels, and a third of them preserved cheno-am seeds. Maize is a domesticate, and cheno-ams are known to proliferate in disturbed areas such as gardens, so in one sense, both plants may be considered the products of agricultural harvests. Their presence in so many thermal features suggests that farming provided most of the last foods eaten. This inference is supported by the presence of maize cobs as fuel or tinder in many of the same locations. The presence of prickly-pear cactus (Opuntia) seeds in 20 percent of the hearths and firepits reveals use of a fairly predictable wild food. Eight additional wild plant taxa were preserved in only one or two thermal features, revealing rather limited reliance on them as foods.

26
Plant remains recovered from thermal features in the two towers, a single kiva (Structure 302), and a single habitation room (Structure 111) indicate that a diversity of foodstuffs had been prepared, at least during the last uses of these features. The towers, however, lacked evidence of maize as either food or fuel. Fewer foods were prepared in the other excavated structures at the site. The food product that was preserved most often in primary thermal features in kivas, as well as in middens clearly associated with kivas, was maize. Maize and other food products were also present in structures thought to have been habitations.

27
With regard to fuel use, the presence of Juniperus and Pinus charcoal in more than 80 percent of primary thermal features from all structure types suggests that the pinyon-juniper woodland still served as a major source of fuel for the final fires built. The recovery of small parts such as twigs and bark scales confirms that the inhabitants of Castle Rock had access to whole trees, not just to major trunk wood (for example, construction beams). Most of the remaining wood types were probably common components of the pinyon-juniper woodland, including two that increase with agricultural disturbance (Artemisia, Chrysothamnus). A few small trees or shrubs that grow in lowlands (Fraxinus, Populus/Salix, Sarcobatus) complete the list of burned woods. The presence of maize cob parts in more than 50 percent of the thermal features suggests that leftover cobs were used as fuel or tinder for the last fires burned.

28
Two "chief kiva fuels" (Atriplex, Rhus) of the Hopi Indians (Whiting 1966*1:38) were found only in the refuse of primary thermal features in towers and kivas, and one additional type (Chrysothamnus) was almost exclusive to these structures. The fourth chief kiva fuel (Sarcobatus) was present in all structure types. This ancient pattern resembles that of historic Hopi wood use.

Plant Remains in Roof Fall and Wall Fall

29
Macrofossils handpicked from roof fall and wall fall strata in 11 structures have been analyzed for Castle Rock Pueblo (Table 9). The large majority of specimens are charred, with a few notable exceptions. Cucurbita seeds on a bench in a kiva (Structure 103) were uncharred, yet are considered prehistoric because of their degraded condition and because they represent a domesticate. That the bench provided an exceptional preservation situation is indicated by the presence of partly charred pieces of juniper (Juniperus) wood.

30
Roof fall and wall fall strata can include materials with more than one origin. Certainly, components of the roof are likely to be present in roof fall. Items left on top of the roof or suspended from the rafters for drying or storage might also be included. Items on kiva benches or possibly on floors might become mixed with collapsed roofing materials. It appears that the Castle Rock plant record encompasses both materials chosen for roof construction and domesticated crops that were associated with the use of roofs or the surfaces onto which the roofs collapsed. The patterns discussed below are demonstrated in structures and elsewhere in contexts designated "surface contact or fill above surface" and "upper fill" layers.

31
Of the 15 types of charcoal preserved in all deposits examined from Castle Rock Pueblo, a smaller subset of six types appears to be most often associated with roof fall and wall fall. Juniper (Juniperus) was preserved in nine of the 11 structures for which samples from roof fall and wall fall were analyzed, complementing the tree-ring record of regular use of juniper for construction timbers. Occasional use of Pinus wood is suggested by its presence in four of the roof fall strata; Pinus charcoal identified at the low magnifications used in this study potentially includes pinyon (P. edulis), ponderosa pine (P. ponderosa), douglas fir (Pseudotsuga), and spruce (Picea), some of which grow on the northern flanks of nearby Ute Mountain. The juniper and pine construction timbers likely provided the major roof elements (vigas) and supports, although other large timbers might have been used as well.

32
In addition to the major construction elements of juniper and pine, some of the smaller layers of roof construction—possibly the closing layers—might be represented by charcoal from shrubs. These shrubs include sagebrush (Artemisia), which was associated with seven roofs, and saltbush (Atriplex), rabbitbrush (Chrysothamnus), and cottonwood/willow (Populus/Salix), each collected from roof fall in three structures. At the Duckfoot site, a Pueblo I hamlet in the region, Artemisia was a possible roof-closing material, as well as a flammable material used in the intentional burning of structures (Adams 1993*1).

33
Nonwoody material associated with roof fall and wall fall inside structures includes the domesticates maize (Zea mays), beans (Phaseolus vulgaris), and squash (Cucurbita). In some cases the items were found in association with kiva benches, where they might have been sitting when the roof collapsed; it is also possible that they were associated with the roof and landed on the bench as the roof fell in. Maize was found in eight of the 11 structures as cobs, kernels, a "fused mass," and possible "meal cakes" of ground maize kernels. The "fused mass" on Bench 6 in Structure 105 consisted of cob, stalk, and kernel fragments. Nearly 2,000 charred kernels, some still attached to cobs, were preserved in roof fall in Structure 206. Possible maize "meal cakes" were found in Structures 105 (the oversized kiva) and 304 (a small kiva). Collectively, this record suggests that maize was being stored either on kiva benches or in containers suspended from roof rafters. Other domesticates present included uncharred squash (Cucurbita) seeds on Bench 2 in Structure 103 (a kiva) and charred common bean (Phaseolus vulgaris) cotyledons in Structures 206 (a kiva) and 305 (a tower). The storage of domesticates in kivas and towers might indicate a certain level of social control over foodstuffs.

Proximity of Agricultural Fields

34
Agricultural fields were near enough to Castle Rock Pueblo that secondary waste products such as maize cobs were carried in regularly and burned in hearths and firepits. Shank segments were also brought in on occasion. Cheno-am and Portulaca, weedy plants that thrive in agricultural fields, appear to have provided food in the form of seeds. Shrubs known to play a role in the recovery of abandoned agricultural fields (Artemisia, Chrysothamnus) were among the top 10 charcoal types identified in all deposits, suggesting that the inhabitants of Castle Rock had fairly regular access to disturbed land.

Seasonality of Site Use

35
Given what we know about plant seasonality and agricultural scheduling needs, the plant record at Castle Rock Pueblo places occupants in the area from the spring through the fall. Plants harvested during the late spring and early summer include rice grass (Stipa hymenoides) and woolly wheat (Plantago). The remaining wild plants produced reproductive parts available during the summer and fall, as did the domesticates maize, beans, squash, and gourds. The need to prepare fields and rework simple water-control structures in the spring, to watch over crops in the summer, and to harvest and process plants in the fall would have required some people to be close at hand throughout many months of the year.

Plant Use Through Time

36
Castle Rock Pueblo was built and inhabited during the last half of the thirteenth century. Refuse accumulations in three structures (108, 304, and 308) suggest that these buildings were abandoned while people continued to live in other areas of the village. The majority of structures, however, appear to have still been occupied when the attack that ended the habitation of the village occurred (see "The Final Days of Castle Rock Pueblo"). Structure 401, a tower, is one structure that might have been used after that event, at least for a short while, as suggested by the presence of a use surface above its burned roof.

37
Some plant refuse accumulated in Structures 108 and 304 while people still occupied other parts of the village. Plant material in the hearths of these structures seems not to vary noticeably from that found in general site debris. Remains of the same foods and wood charcoal were mixed into the postabandonment cultural refuse that found its way into these structures. No differences in plant use between these structures and the rest of the site are obvious.

38
As mentioned above, many of the structures at Castle Rock Pueblo were still in use at the time of the attack. The presence of maize and squash in two kivas (Structures 104 and 206) that were open at abandonment indicates that crops were available at this time. Most of these plant remains were found in roof fall or wall fall or else were on a surface or in fill above it. This implies that the plants were associated with the kiva roofs in some way (for example, the plants were drying on the roof or suspended in containers hanging from rafters) or were on the floor or a bench (in the case of Structure 206) or somehow were within a ventilator system (Structure 104). The very small number of preserved food plant remains present in postabandonment fill is consistent with natural deposition when people no longer lived at Castle Rock Pueblo.

39
The plant record in these and other structures might reveal something about the season during which the attack that ended the occupation of the village occurred. For example, plant remains from thermal features and from roof fall and wall fall deposits in two towers (Structures 305 and 401) and a kiva (Structure 302), supplemented by materials from just above the structure surfaces, suggest several possibilities. Two of these structures (302 and 305) contained evidence of domesticated plants along with cheno-am seeds. This implies human presence in the area in the late summer and fall, when crops were ripe and required processing for storage. Similarly, the presence of broad-leaf yucca (Yucca baccata) and prickly-pear cactus (Opuntia) seeds suggests late summer and fall. Because these items can be stored indefinitely, however, the season of harvest does not necessarily equate with the season of use. The presence of people in the area during spring and summer is implied by the discovery of rice grass (Stipa hymenoides) remains and the seeds of other later-ripening plants such as Malvaceae and Mentzelia, all of which can also be stored. Two scenarios could reasonably account for these data. First, it is possible that the attack on Castle Rock came in late spring/early summer and that some domesticates and garden weeds were still in storage from a previous fall harvest. Alternatively, the village may have been attacked during or following the fall harvest, in which case the seeds of wild plants found in the structures may have been stored from earlier in the growing season. Assuming that it would have been difficult to acquire crop surpluses large enough to store food for lengthy periods, the second scenario seems the more reasonable of the two.

40
Above the burned roof in Structure 401, a tower, excavators uncovered a surface, which suggests that there was limited use of the site following the attack. Because no plant remains have been analyzed from this context, the nature of plant use during this time is unknown. Structure 302, a kiva, also might have been used for a short while following the attack, though no plant remains can be clearly tied to this use.

Resource Depletion and Food Stress

41
The plant remains preserved in structures abandoned before the attack were essentially the same ones preserved in structures occupied up to the end of the village's occupation, which provides us with no evidence of resource depletion or food stress. For comparison, a large archaeobotanical data set covering the last 100 years of occupation in the Sand Canyon locality (Adams and Bowyer 1998*1) revealed some possible food stress at Sand Canyon Pueblo. There, during the late A.D. 1200s, people may have relied more on Opuntia (prickly-pear) seeds than did others in the region during the late 1100s and early 1200s, and maize (Zea) kernels occurred in relatively low frequencies in Sand Canyon Pueblo hearths and firepits (seven of 38). At Castle Rock Pueblo, the presence of maize kernels in eight of 15 thermal features with primary refuse (Table 7) represents a relatively high rate for the region for the thirteenth century (Adams and Bowyer 1998*1). This suggests that food stress, if present, was perhaps less of a problem at Castle Rock than at some of the other pueblos in the region. That maize cobs and cupules were found in some of these same hearths and firepits suggests that people were burning cobs as fuel and not relying on them as a starvation food, as has been recorded in the Southwest ethnographically (Hill 1938*1:46).

42
Another ethnographically known buffering mechanism for protecting a household against food shortfalls might have been operating at Castle Rock Pueblo—the production of valuable goods to trade later for food (Lightfoot 1983*1:193). Although gourds were present in the Mesa Verde region from approximately A.D. 600, gourd remains were found in only limited quantities at Sand Canyon and Castle Rock pueblos. The finding of gourd (Lagenaria) rind in two locations at Castle Rock might reflect the growing of gourds to make ceremonial rattles to trade with others for food.

The Past Environment

43
The prehistoric plant record suggests that many of the plants present in the region today were also available to the ancient occupants of Castle Rock Pueblo. Although it is difficult to comment on relative proportions of taxa in past vegetation, the evidence suggests that opportunistic weeds (cheno-ams, Portulaca) took advantage of agricultural fields and were in turn harvested for food. The presence at Castle Rock Pueblo of wood of two shrubs (Artemisia and Chrysothamnus) known to fill a successional role in fallow agricultural fields supports the idea that some portions of the landscape had been opened for crops. That some of these fields were in the vicinity of Castle Rock is confirmed by the finding of agricultural waste products within the pueblo.

44
Generally, a pinyon-juniper woodland covered the area and supported many of the same shrubs found in the modern woodland. Although it has been difficult to locate two modern specimens of herbaceous taxa (Mentzelia albicaulis and Plantago) recovered at Castle Rock, both do occur occasionally in the modern environment and could owe their scarcity to heavy historic grazing. The Castle Rock plant remains, although heavily biased by cultural choice and practices, nevertheless reflect an environment with many of the modern plants present today.

References cited | To borrow, cite, or request permission
Copyright © 2000 by Crow Canyon Archaeological Center. All rights reserved.