Introduction

Nine pollen samples were collected and examined from stratigraphic deposits in Backhoe Trench 1 at site 48LA277 to identify vegetation history at the site. Previous radiocarbon ages reported for this site suggested a Plains Woodland occupation, and the previous stratigraphic pollen analysis indicated a wetter interval that may be correlated with lateral movement of Crow Creek. Another stratigraphic column was examined from the 1992 excavations to identify vegetation associated with the occupation of this site, as well as the postoccupational vegetation. In addition, pollen samples were examined from contexts inside and outside Feature 9, a large roasting basin that yielded a radiocarbon age of 1470 ± 60 B.P., to provide subsistence information. A flotation sample from the fill of Feature 9 was also examined to provide information concerning possible use of this feature for food-processing activities. Macrofloral and charcoal samples were examined from Feature 10 to identify any plants that may have been processed in this large roasting basin. A single flotation sample was examined from the fill of Feature 12, a pit recovered in the stratigraphic profile, to assist in identifying the function of this pit.

Methods

A standard chemical-extraction preparation technique based on flotation was used in this laboratory for the removal of the pollen from the large volume of sand, silt, and clay with which they were mixed. This particular process was developed for extraction of pollen from soils where preservation has been less than ideal and pollen density is low.

Hydrochloric acid (10%) was used to remove calcium carbonates present in the soil, after which the samples were screened through 150-micron mesh. The samples were rinsed until neutral by adding water, letting the samples stand for three hours, then pouring off the supernatant. A small quantity of sodium hexametaphosphate was added to each sample once it reached neutrality, then the beaker was again filled with water and allowed to stand for three hours. The samples were again rinsed until neutral, filling the beakers only with water. This step was added to remove clay prior to heavy liquid separation. Zinc bromide (density 2.0) was used for the flotation process. The samples were mixed with zinc bromide while still moist, immediately after centrifugation to remove the dilute hydrochloric acid and water. All samples received a short (10-minute) treatment in hot hydrofluoric acid to remove any remaining inorganic particles. The samples were then acetolated for three minutes to remove any extraneous organic matter.

A light microscope was used to count the pollen to a total of 100-200 pollen grains at a magnification of 500X. Pollen preservation in these samples varied from good to poor. Comparative reference material collected at the Intermountain Herbarium at Utah State University and the University of Colorado Herbarium were used to identify the pollen to the family, genus, and species level, where possible.

Pollen aggregates were recorded during identification of the pollen. Aggregates are clumps of a single type of pollen, and may be interpreted to represent pollen dispersal over short distances, or the actual introduction of portions of the plant represented into an archeological setting. Aggregates were included in the pollen counts as single grains, as is customary. The presence of aggregates is noted by an "A" next to the pollen frequency on the pollen diagram. 

Indeterminate pollen includes pollen grains that are folded, mutilated, and otherwise distorted beyond recognition. These grains are included in the total pollen count, as they are part of the pollen record. 

Macrofloral samples were floated using a modification of the procedures outlined by Matthews (1979). One liter of sample was added to approximately three gallons of water. The sample was stirred until a strong vortex formed, which was allowed to slow before pouring the light fraction through a 150-micron mesh sieve. Additional water was added and the process repeated until all visible macrofloral material was removed from the sample (a minimum of three times). The material remaining in the bottom (heavy fraction) was poured through a 1-mm-mesh screen. The floated portions were allowed to dry. The light fractions were passed through a series of graduated screens (U.S. Standard Sieves with 2-mm, 1-mm, 0.5-mm and 0.25-mm openings) to separate charcoal debris and to initially sort the seeds. The contents of each screen were then examined. Charcoal pieces larger than 2 mm in diameter were broken to expose a fresh cross section and examined under a binocular microscope at magnifications up to 80X. The material that remained in the 2-mm, 1-mm, and 0.5-mm sieves was scanned under a binocular stereo microscope at a magnification of 8X, with some identifications requiring magnifications of up to 40X. A portion of the finest material in the 0.25-mm screen was also examined under a magnification of 8X. The material that passed through the 0.25-mm screen was not examined. The coarse, or heavy, fractions were also examined. Macrofloral remains were identified using manuals (Martin and Barkley 1973; Musil 1978; Schopmeyer 1974) and by comparison with recent and archeological references. The term "seed" is used to represent seeds, achenes, caryopses, and other disseminules. Remains were recorded as charred or uncharred, and whole or fragmentary.

Discussion

Site 48LA277 is a Plains Woodland site located next to Crow Creek on the F. E. Warren Air Force Base, just west of Cheyenne, Wyoming. Crow Creek drains from the Laramie Mountains to the South Platte River in Colorado. Features recovered suggest intense use of the area, with on-site activities including lithic reduction, tool manufacture and resharpening, hunting, and food processing (Connor 1993:5, 18, 24). Local vegetation consists of prickly pear (Opuntia), yucca (Yucca), and a mixed-grass prairie including blue grama (Bouteloua gracilis), buffalo grass (Buchloe dactyloides), needle-and-thread grass (Stipa comata), linear-leaved sedge (Cyperus), and copper mallow (Malva) (Tetra Tech 1987:6-9). Plants found in the riparian community along Crow Creek include willows (Salix), cattail (Typha), rushes (Juncus), sedges (Cyperus), slough grass (Beckmannia), American manna grass (Glyceria grandis), cow parsnip (Heracleum), dock (Rumex), clover (Trifolium), death camas (Zigadenus nuttallii), plantain (Plantago), water smartweed (Polygonum), cinquefoil (Potentilla), and violet (Viola) (Tetra Tech 1987: 6-10-6-11). 

Pollen and macrofloral analyses identified remains of plants that may have been important economic resources for the occupants of site 48LA277. For macrofloral remains from prehistoric sites, it has become most acceptable to consider only charred seeds for the interpretation of a feature and utilization of resources (Minnis 1981). Few seeds live longer than a century, and most for a much shorter time period (Harrington 1972; Justice and Bass 1978; Quick 1961). It is presumed that once the seeds have died, decomposing organisms act to decay the seeds. Interpretation of uncharred seeds to represent presence in the prehistoric record is considered on a sample-by-sample basis. Extraordinary conditions for preservation are important in this interpretation. All uncharred remains recovered in flotation samples from site 48LA277 were components of the local, modern vegetation. 

It is also a commonly accepted practice in archeological studies to reference ethnological (historic) plant uses as indicators of possible or even probable plant uses in prehistoric times. It provides evidence of the exploitation, in historic times, of numerous plants, both by broad categories, such as greens, seeds, roots, and tubers, etc., and by specific example, e.g., seeds parched and ground into meal that was formed into cakes and fried in grease. Repetitive evidence of the exploitation of resources indicates a widespread utilization and strengthens the possibility that the same or similar resources were used in prehistoric times. Ethnographic sources do document that with some plants the historic use was developed in the distant past. A plant with medicinal qualities very likely was discovered in prehistoric times, and the usage persisted into historic times. There is, however, likely to have been a loss of knowledge concerning the utilization of plant resources as cultures moved from subsistence to agricultural economies, were introduced to European foods during the historic period, or both. The ethnobotanical literature serves only as a guide, indicating that the potential for use existed in prehistoric times-not as conclusive evidence that the resources were used. Pollen and macrofloral remains, when compared with the material culture (artifacts and features) recovered by archeologists, become indicators of use. Pollen and macrofloral analysis identified remains of plants that might have been important food items for the various occupants of this site. These plants will be discussed in the following paragraphs. 

Chenopodium (goosefoot) and Atriplex (saltbush) are part of a group of plants known collectively as Cheno-ams. These plants were exploited for both their greens (cooked as potherbs) and seeds. The seeds were ground and used to make a variety of mushes and cakes. The seeds are usually noted to have been parched prior to grinding. The greens are most tender when young, in the spring, but may be used at any time (Chamberlin 1964:366; Gallagher 1977:12-16; Gilmore 1977:26; Harrington 1967:55, 57, 71; Rogers 1980:43, 66; Schopmeyer 1974). Chenopodium leaves are rich in calcium and vitamin C and were eaten to treat stomachaches and to prevent scurvy. Leaf poultices were applied to burns, and a tea made from the whole plant was used to treat diarrhea (Angier 1986:191-193; Foster and Duke 1990:216; Krochmal and Krochmal 1973:66-67). Atriplex leaves and young shoots have a salty taste and may be used as a seasoning. A poultice of the chewed plant was applied to ant, bee, and wasp sting swellings. Atriplex canescens (four-wing saltbush) was used for stomach pain or as an emetic. Dried leaves were used as a snuff for nose trouble, and a poultice of the warm, pulverized root was applied to toothaches (Moerman 1986:85-86; Weiner 1972:75). 

Chenopodium is commonly found in cultivated fields, waste places, open woods or thickets, and on stony hills. It is an opportunistic weed, often establishing itself rapidly in disturbed areas (Fernald 1950:592-596; Martin 1971). Atriplex (saltbush) occurs as both an annual herb and perennial shrub. The seeds do not ripen until mid-fall and may remain on the shrubs throughout the winter into the next growing season. Atriplex is native to widely scattered areas in the western United States, in waste places and fields, growing in arid, alkaline, or saline soils (Kirk 1975:59; Muenscher 1987:180). 

Members of the morphological pollen group of High-spine Compositae group, such as Helianthus (sunflower), were used in a variety of ways, including medicinally and as food. Sunflower seeds are very rich in oil, and may be ground into paste for batter or roasted and eaten (Harrington 1967:312-314). Rabbitbrush may be used as fuel. Most composite seeds ripen in the late summer and fall.

A stratigraphic profile was sampled for pollen at Stadia Station 39 near Feature 9 (Table B1). The top of Feature 9 was recorded at a depth between 45 and 55 centimeters below surface at the top of Soil Unit 4. This unit dips to near the base of the trench in the area where the pollen column was collected. Assuming consistency of soil units, this date effectively dates the base of the pollen record to approximately 1500 B.P. The base of the pollen record may correspond roughly to the period of occupation. All sediments above this level indicate deposition following use of Feature 9. Interfingering of fluvial and colluvial deposits is noted (Appendix A). 

Sample 12, collected from the humus zone at a depth between 0 and 10 cm, most closely approximates the present ground surface. It is interesting to note that this sample is dominated by Cheno-am pollen, which also exhibits aggregates (Figure B1, Table B2). It is probable that the increase in Cheno-am pollen in this sample has displaced a major portion of the regional Pinus pollen wind transported to this area. The very large quantity of Cheno-am pollen recovered from this sample may reflect, at least in part, local stands of Atriplex (saltbush). The presence of relatively dense stands of saltbush is postulated from the pollen record, since saltbush is not noted at present in the local vegetation.

Association of the occupation represented by Feature 9 with the upper portion of Soil Unit 4, a colluvial layer, indicates that the majority of the sediment represented in this stratigraphic record accumulated very rapidly and represents only the last approximately 1,500 years. The base of the stratigraphic record in Backhoe Trench 1 represents the unconsolidated bedrock and appears to contain pollen from Holocene vegetation.¹ It exhibits a relatively high frequency of arboreal pollen, most of which is Pinus. The presence of small quantities of Juniperus and Picea pollen indicates that other trees were also present in the region, including at least juniper and spruce. It is likely that much of the tree pollen is present due to long-distance transport by the wind from mountainous communities to the west or northwest of the site. Only a small quantity of Cheno-am pollen was recovered from this sample, suggesting that saltbush was not an important member of the plant community at this time. Artemisia, High-spine Compositae, and Tubuliflorae pollen are recorded in moderate quantities in this sample, indicating that sagebrush and a variety of composites were probably common in the local vegetation. The elevation of Liguliflorae pollen suggests proximity to a drainage, because many members of this group of the Sunflower family grow best in the moister soils associated with drainages. 

Samples 17-19 exhibit elevated Gramineae pollen frequencies, suggesting that grasses grew on this colluvial sediment. Fluctuations in other pollen frequencies do not indicate stability in vegetation communities. Pollen transport from mountainous communities fluctuates as do both the Artemisia and High-spine Compositae frequencies. It is important to note that the combined Artemisia and High-spine Compositae frequencies fluctuate in opposition to the Pinus pollen frequencies, indicating that when the Artemisia and High-spine vegetation assemblage increased locally, it may have produced a sufficient quantity of pollen to cause a decline in relative frequency of Pinus pollen (present through long-distance transport) in these levels. The Cheno-am frequencies are moderately low in these samples, fluctuating only slightly. Sarcobatus pollen is also noted in Samples 18-20. Sarcobatus is a plant that prefers the slightly alkaline soils that often occur in drainages. Presence of small quantities of Sarcobatus in these samples suggests the possibility that the riparian zone along Crow Creek supported greasewood.

Sample 16 is sandier than other samples in the column, probably as a result of its fluvial origin. It exhibits a reduced Pinus pollen frequency, suggesting that pollen transportation from the pine communities in the mountains to the west and/or northwest might have been reduced as a result of reduction in local pine population. Alternatively, the reduced Pinus pollen frequency may indicate displacement of the record by an abundance of High-spine Compositae. Elevation in the High-spine Compositae pollen frequency suggests that members of this group, which includes rabbitbrush, probably increased locally at the site. The Gramineae pollen frequency is reduced, and Sarcobatus is absent from this sample. Only Alnus pollen may reflect a tree commonly associated with drainages. This pollen type travels well and does not necessarily indicate that the Crow Creek channel was closer to the site area at this time. Reduction in the Gramineae frequencies may be the result of rapid deposition of fluvial sediments and may not indicative be of local vegetation on the colluvial sediments farther from the channel.

Samples 14 and 15 again note an increase in Gramineae pollen frequencies, suggesting that once again the sediments may have been wetter through occasional overbank episodes from proximity to the waters of Crow Creek, without receiving large quantities of sand during fluvial deposition. The wetness of the sediments appears to have encouraged an increase in the local grass population. The small quantity of Liguliflorae-type pollen recovered from Sample 14 strengthens this interpretation, because members of this group are often observed growing in wet areas along streams. The elevation of Pinus pollen noted in Sample 14 appears to reflect an increase in local pine populations in the mountainous areas to the west or northwest of this site. An increase in Artemisia pollen is noted in Sample 15, but it declines rapidly, indicating that any local increase in sagebrush community was not long-lived. An increase in Ephedra pollen is also noted in Sample 15, suggesting an increase in the local population of this species. 

Sample 13 exhibits an increased quantity of Cheno-am pollen, accompanied by declines in the Artemisia and High-spine Compositae pollen frequencies. Gramineae pollen also declines. It appears that immediately prior to the modern time period vegetation in the vicinity of this site had begun to change and that sagebrush, rabbitbrush, other composites, and grasses all declined as saltbush increased. This may have been the result of increasing aridity or alkalinity of the soil. The small quantity of Typha pollen noted in Sample 13 represents wind transport from the riparian community associated with Crow Creek.

Small quantities of solid starch and starch granules with hila are noted infrequently in the pollen record. This indicates that starchy roots and tubers were probably not abundant in the area represented in this stratigraphic column.

Feature 9 was a large roasting basin near Backhoe Trench 1 that yielded a radiocarbon age of 1470 ± 60 B.P. This feature exhibited a dense concentration of rock and an undulating bottom. Pollen samples were collected from both the feature fill (Sample 10), which was heavily laden with charcoal, and from the probable occupation surface outside of the feature (Sample 11). A flotation sample was also collected from the fill of this feature (Sample 21). There was so much charcoal contained in Sample 10 from the fill of this feature that it was impossible to concentrate the pollen sample sufficiently to obtain a pollen count. Sample 11, recovered from the occupation surface outside of Feature 9, did yield sufficient pollen for analysis. This sample exhibited relatively little Pinus pollen, and a greatly elevated High-spine Compositae pollen level, accompanied by a single small aggregate. The elevated High-spine Compositae frequency suggests that a member of this group of plants may have been processed or used in the vicinity of this roasting pit. 

Macrofloral Sample 21 contained one charred Chenopodium seed (Tables B3 and B4), suggesting processing of goosefoot seeds. Seeds may have spilled while being parched in preparation for being ground into a flour. Alternatively, the charred Chenopodium seed may also represent a local weed at the time of occupation whose seeds blew into the roasting basin and were subsequently charred. Uncharred Chenopodium seeds represent modern local vegetation. Salicaceae and probable Betula charcoal indicate that wood from trees along Crow Creek was being collected and used as fuel in the roasting basin. 

Feature 10 was another large roasting basin that yielded a radiocarbon age of 1160 ± 90 B.P. A macrofloral and charcoal sample were examined from the fill of this feature. Macrofloral Sample 22 contained a charred Atriplex fruit fragment and a charred probable Atriplex seed, suggesting that saltbush was utilized and processed in this roasting basin. Numerous uncharred remains represent modern intrusion of local plants. Salicaceae and probable Betula charcoal again indicate use of wood from riparian trees as fuel. Uncharred bone fragments were also recovered.

Sample 23 consisted of charcoal from the fill of Feature 10. The sample contained two pieces of probable Betula charcoal, one piece of Salicaceae charcoal, and one piece of charred unidentifiable bark.

Feature 12 was a pit noted in the north wall of a backhoe trench. No radiocarbon age was obtained for this feature. A single macrofloral sample was examined from the fill of this pit in an effort to identify economic activities that may assist in defining utilization of the pit. Sample 24 contained a charred Chenopodium seed, suggesting that goosefoot seeds were processed. Seeds from local weeds could also have been accidentally charred. Salicaceae charcoal indicates that a member of this family, most likely willow, was burned as fuel.

Examination of the macrofloral records at this site from recent and previous excavations reveals that processing of Chenopodium seeds and burning of Salicaceae wood as fuel appear to have been common practices at this site. Charred Chenopodium and/or Cheno-am seeds were recovered in all macrofloral samples from previous excavations (Cummings and Puseman 1991) and from two of the three samples from recent excavations. Salicaceae charcoal was recovered in all macrofloral samples examined from this site. Differences in the macrofloral records include possible processing of berries and Galium (bedstraw) and Rumex (dock) seeds, and the burning of Prunus wood as evidenced in previous excavations, as well as possible processing of Atriplex and burning of probable Betula wood seen in samples from recent excavations.

Summary and Conclusions

Stratigraphic samples from site 48LA277 were examined for pollen to identify occupational and postoccupational vegetation. Pollen and macrofloral samples were also examined from features at this site to provide information on past subsistence.

The pollen record appears to represent approximately the past 1,500 years. Interfingering of fluvial and colluvial deposits noted in the geomorphic record affected local vegetation. In general, the proximity of Crow Creek appears to have encouraged local growth of grasses. Only in Sample 16, which exhibited larger quantities of sand than did other fluvial deposits, is there a significant decline in Gramineae pollen frequencies. This is probably related to the deposition of the sandy sediments, rather than representing a vegetation change in the habitat outside the Crow Creek channel. 

Until relatively recently, local vegetation appears to have been a mixed grassland/shrubland, with the dominant shrubs being sagebrush and a member of the High-spine Compositae. The latter become much less abundant in the local vegetation beginning in Sample 15. Cheno-ams were relatively unimportant in the local vegetation, or were abundant only in much drier sediments removed from this area. By Sample 13, Cheno-am frequencies are increasing dramatically, and sagebrush (Artemisia) frequencies have fallen to their modern level. Grasses appear to have been more abundant for much of the past 1,500 years than they are today. Reduction in Gramineae pollen frequencies are noted in Sample 13, along with the dramatic rise in Cheno-am frequencies. Both appear to be in response to drier sediments.

The macrofloral record from features at site 48LA277 suggests that Chenopodium and Atriplex were processed at this site. Previous excavations at site 48LA277 also provided evidence for processing of Cheno-ams, suggesting that they were an important part of the diet for the occupants of this site. The pollen record of sediments outside Feature 9 does not contribute to interpretation of subsistence activities. Salicaceae and probable Betula charcoal indicate that wood from trees in the riparian habitat along Crow Creek were burned as fuel. 

Angier, Bradford
-1986 Field Guide to Medicinal Wild Plants. Stackpole Books, Harrisburg, Pennsylvania.

Chamberlin, Ralph V.
-1964 The Ethnobotany of the Gosiute Indians of Utah. Reprinted. Kraus ReprintCorporation, New York. Originally published 1911, Memoirs Vol. 2, Pt. 5, American Anthropological Association, Lancaster, Pennsylvania.

Connor, Melissa A.
-1993 1991 Test Excavations at 48LA277: A Plains Woodland Site on Crow Creek, Wyoming.   National Park Service, Midwest Archeological Center, Lincoln, Nebraska.

Cummings, Linda Scott, and Kathryn Puseman
-1991 Pollen and Macrofloral analysis at Site 48LA277, Crow Creek, Wyoming. Manuscript on file, Midwest Archeological Center, USDI National Park Service, Lincoln, Nebraska.

Fernald, M. L.
-1950 Gray's Manual of Botany. 8th ed. American Book Company, New York.

Foster, Steven, and James A. Duke
-1990 A Field Guide to Medicinal Plants. Houghton Mifflin Company, Boston.

Gallagher, Marsha V.
-1977 Contemporary Ethnobotany Among the Apache of the Clarkdale, Arizona Area Coconino and Prescott National Forests. Archeological Report No. 14. USDA Forest Service, Flagstaff, Arizona.

Gilmore, Melvin R.
-1977 Uses of Plants by the Indians of the Missouri River Region. Reprinted. University of   Nebraska Press, Lincoln. Originally published 1919, Bureau of American Ethnology, Washington, D.C.

Harrington, H. D.
-1967 Edible Native Plants of the Rocky Mountains. University of New Mexico Press, Albuquerque, New Mexico.

Harrington, James F.
-1972 Seed Storage and Longevity. In Seed Biology, vol. III, edited by T. T. Kozlowski, pp.   145-240. Academic Press, New York.

Justice, Oren L., and Louis N. Bass
-1978 Principles and Practices of Seed Storage. Agriculture Handbook No. 506. U.S.Department of Agriculture, Washington, D.C.

Kirk, Donald R.
-1975 Wild Edible Plants of Western North America. Naturegraph Publishers, Happy Camp,   California.

Krochmal, Arnold, and Connie Krochmal
-1973 A Guide to the Medicinal Plants of the United States. Quadrangle, New York Times   Book Co., New York.

Martin, Alexander C.
-1971 Weeds. Golden Press, Western Publishing Company, New York.

Martin, Alexander C., and William D. Barkley
-1973 Seed Identification Manual. University of California Press, Berkeley, California.

Matthews, Meredith H.
-1979 Soil Sample Analysis of 5MT2148; Dominguez Ruin, Dolores, Colorado. In TheDominguez Ruin: A McElmo Phase Pueblo in Southwestern Colorado, by Alan D. Reed, Appendix B. Cultural Resource Series No. 7. Bureau of Land Management, Denver.

Minnis, Paul E.
-1981 Seeds in Archaeological Sites: Sources and Some Interpretive Problems. AmericanAntiquity 46(1):143-152.

Moerman, Daniel E.
-1986 Medicinal Plants of Native America. Technical Reports No. 19, Vols. 1-2. Museum   of Anthropology, University of Michigan, Ann Arbor.

Muenscher, Walter Conrad
-1987 Weeds. 2nd ed. Comstock Publishing Associates, Cornell University Press, Ithaca.

Musil, Albina F.
-1978 Identification of Crop and Weed Seeds. Agricultural Handbook No. 219. U.S.Department of Agriculture, Washington, D.C.

Quick, Clarence R.
-1961 How Long Can a Seed Remain Alive. In Seeds, the Yearbook of Agriculture, edited   by A. Stefferud, pp. 94-99. U.S. Government Printing Office, Washington, D.C.

Rogers, Dilwyn
-1980 Edible, Medicinal, Useful, and Poisonous Wild Plants of the Northern GreatPlains-South Dakota Region. Biology Department, Augustana College, Sioux Falls, South Dakota.

Schopmeyer, C. S.
-1974 Seeds of Woody Plants in the Unites States. Agricultural Handbook No. 450. USDA   Forest Service, Washington, D.C.

Tetra Tech, Inc.
-1987 Southeastern Wyoming Prehistory, vol. 1. Peacekeeper Program Cultural ResourcesTechnical Report No. 3. Tetra Tech, San Bernardino, California.

Weiner, Michael A.
-1972 Earth Medicine-Earth Foods. Collier Books, New York