G-' CENTRAL CIRCULATION BOOKSTACKS The person charging this material is re- sponsible for its renewal or its return to the library from which it was borrowed on or before the latest Date stamped below. The Minimum Fee for each Lost Book is $50.00. for dl-lpllnory ortlon and may rwult in dl««lM«l «rom fhe Onlvewhy. TO RENEW CAU TELEPHONE CENTER, 3M-8400 UNIVERSITY OF ItllNOIS UBRARY AT UMANA-CHAMPAIGN ^ NOV 2 0 1S9A MAY 3o When renewing by phone, write new due date below previous due date. L.05 5. FIELDIANA Anthropology Published by Field Museum of Natural History Volume 63, No. 1 March 10, 1972 Paleoecology of the Hay Hollow Site, Arizona VORSILA L. BOHRER Assistant Professor, Biology University of Massachusetts FOREWORD Paul S. Martin Chairman Emeritus, Anthropology In order to make the researches of Dr. Bohrer more meaningful and to place them briefly and roughly in an archaeological setting, I shall attempt to give a general description of the site and of some of the data obtained. My statements are based on information given me by John M. Fritz, who headed up the research. A full treatment of the details I will give will be presented in his doctoral dissertation for the Department of Anthropology, University of Chicago. This is in preparation. Both Mr. Fritz and Dr. Bohrer have read and corrected the state- ments herein. Introduction Hay Hollow site in Hay Hollow Valley, located about 10 miles east of the contemporary town, Snowflake, Arizona, on the ranch of James Carter, is about 5,750 ft. (pocket altimeter) above sea level and is at 109°55' W. longitude and 34°31' N. latitude. It is situated on an erosional bench about 16 ft. above the flood plain of Hay Hol- low Wash. Excavations were started there in the summer of 1965 and research continued through two more summers until the autumn of 1968. The research was supported by the Field Museum of Natural History and grants from the National Science Foundation ^(GS-245, GS-984, GS-1910, and GS-2381). Grateful acknowledge- ment is made to the Museum and to the Foundation for this gener- ous support. The site was chosen for investigation because (1) it appeared to have been occupied before or near the beginning of the introduction Library of Congress Catalog Card Number: 73-179170 Publication 1144 1 2 FIELDIANA: ANTHROPOLOGY, VOLUME 63 of pottery-making and agriculture in the area; (2) it was unlike any other excavated site in the area except the County Road Site (about 1000 B.C.-350 B.C., GX-0274 and GX-0272, Geochron Laborato- ries) partly dug under the direction of Dr. James N. Hill but not written up for publication; and (3) it provided an opportunity to measure and analyze prehistoric subsistence systems at a time during which crucial changes were taking place — that is, at a time when the people of the area were responding to pressures that caused them to shift from their hunting-gathering subsistence adaptation to one of farming. The site (houses and pits) represents the structured remains of a hunting-gathering adaptation. It can be referred to as a Desert Culture Site — a term that now signifies a special adaptation of the Indians who occupied Danger Cave (Utah) during a time of gradual desiccation of part of the Great Basin. In reality, it may be that the "Desert Culture" represents not one but many differing adaptations, the shadowy origins of which suggest a genesis around or near pluvial lakes, forests, grassy uplands, and perhaps in less favourable areas — a genesis that may have started about 10,000-12,000 years ago in the Great Basin area. The common denominator of all of these adapta- tions was fishing, hunting animals (both large and small, extinct and existing), and the gathering of native, wild plants — herbs, roots, seeds, nuts, berries, and fruits. Agriculture was not known or prac- tised until sometime after 3,500 B.C. Hay Hollow Site may also be referred to as "early Mogollon cul- ture," a "cultural descendant" of the Desert Culture. If one prefers that sort of taxonomic classification, then one may properly call the site "late Desert Culture" or "early Mogollon Culture." The investigator, John M. Fritz; Dr. Vorsila Bohrer, our palynol- ogist; and I prefer not to fix the site in that kind of substructure or scaffolding, but to emphasize the subsistence system and adaptation of the people of that time and place and to explain (derive laws) that system in terms of attributes, resources, technology, ideology, and social organization. Brief Description of the Hay Hollow Site The site as outlined by Mr. Fritz occupied an area of about 6,000 sq. m. Excavations were done by means of a phased sampling pro- gram which insured a collection of data that reflected the attributes of the whole site without digging the whole site. More explicitly, a grid of 720 squares, 3 m. square, was superposed on the site. Some BOHRER: HAY HOLLOW SITE 3 areas (areas that were blackened from charcoal (?) and houses) were separately structured in the sample. After a topographic map was drawn, all surface cultural debris (lithic waste, blades, manos, metates, fire-cracked rock, and sand- stone slabs [house walls]) was gathered, sacked by squares, and analyzed. Then excavations were begun with a crew of six people with Mr. Fritz in charge, assisted by Dr. James N. Hill. Sixty per cent of the entire site was excavated and 90 per cent of all features. Trans- lated into numbers this meant that all nine houses were completely dug; and 250 pits out of a probable total of 265. Sixty-eight thou- sand artifacts and 13,000 "cooking" stones were recovered. Out of the 68,000 artifacts, we found that about 47,000 (2/3) were unuti- lized flakes; 18,000 were flakes modified by use and 3,000 were "tools," that is, stones modified over large areas of their surfaces, and that these had been employed for cutting, scraping, and piercing. In addition, we recovered 170 whole or broken milling stones; 460 stones that were used in the manufacture of stone tools, and 30 pot- sherds, most of which probably came from one pot — a "seed" bowl shape. The pottery is grayish-brownish and blackish, is friable, and not well-fired. It is among the earliest pottery of the area (about 300 B.C.). Attributes of all artifacts, houses, and pits were classified and tabulated by Fritz and their frequencies placed on IBM punch cards for use in a computer analysis. Factor analyses have been run. As of 1968, patterns of co-variation and association had been worked out for artifacts found in houses, pits, hearths, roasting pits, and storage pits. Complete analyses, descriptions, associations among artifacts, typological analyses, test implications of his models, and explana- tions will be set forth in detail in a subsequent report. In addition to gathering and excavating of cultural items and debitage, Fritz collected about 250 sediment samples for palaeo- botanical analysis. Of this number, about 40 per cent or 100 speci- mens were turned over to Dr. Vorsila Bohrer — palynologist, Univer- sity of Massachusetts at Boston, for pollen analysis. Her research, presented in this report, gives us basic knowledge of past and present botanic environments and clues as to past climates and available moisture. I will not attempt to summarize her work since it is clearly presented hereafter. I should like to point out the signifi- 4 FIELDIANA: ANTHROPOLOGY, VOLUME 63 cance of one important facet of her research : the ability to distinguish between pinyon and other pine pollen — the first time this has been done. Another aspect of Fritz's research was to fix the site in time. A total of 22 carbon- 14 determinations was run on charcoal and human bone (Geochron nos. GX-0539, 0540; 0578-0582; 0727; 0796-0799; 0800-0809). This large a series probably reflects the total time range of occu- pation of the site. The radiocarbon dates span seven centuries: 470 B.C. ±115 to A.D. 305 ± 110. The mean of the series is A.D. 37.03 ± 95.6. Fritz dates the site at about 300 B.C. to A.D. 300. Analysis of the data suggests that the houses and pits fell into several separate clusters of houses surrounded by many pits. The earliest group consists of three houses and dozens of pits, the mean age of which is 30.7 B.C. ±97.5 years. The next younger group — perhaps two to five houses and asso- ciated pits — yields a mean date of A.D. 89.6 ±94.17 years. The most striking feature of the site was the large number of pits — some 265, all told. Two hundred fifty were excavated and de- scribed. They range in size from 20 cm. to 5 m. in diameter, and in depth from 10 cm. to 1 m. The pits may have been used for storage, caching foods or mate- rials, roasting or boiling, and thermal treatment of chert. The houses closely resemble the one well-preserved structure at County Road site (supra). The walls, sloped inward, were of jacal, with posts placed about the periphery or rim of a shallow, bowl- shaped depression, the diameters of which were about 5 m. Each house had been provided with a tunnel entrance that extended east- ward about 2 m. Interior house features consisted of at least one small hearth, several pits, and a kind of deflector composed of a series of upright, sandstone slabs about 30 cm. high set end-to-end to form a low partition wall between the entry and the hearth. The floor, the center of which was depressed below the present surface about 10- 25 cm., yielded milling stones and tools of chipped stone. One house smaller than the others may have been functionally different from the others. Given these data and developing models based in part upon eth- nographically known hunting and gathering cultures, Fritz has for- mulated a series of testable propositions: BOHRER: HAY HOLLOW SITE 6 1. Technology was generalized — that is, tools for food procure- ment and processing were not elaborated relative to those of more specialized agricultural adaptations which followed in the Southwest. 2. The population consisted of small groups that were dispersed over the area through which resources were distributed. 3. In times of famine individuals or families might have tem- porarily combined with groups in areas more richly endowed. 4. The nuclear family was the basic unit of social organization. However, aggregation of familial units, or bands, formed periodically when resources concentration occurred. 5. At Hay Hollow site, cooking, butchering, roasting, and broil- ing were done downwind and at some distance from the houses. 6. At this site food preparation such as milling was done in or adjacent to houses. 7. Hay Hollow site was a base camp which was occupied most intensively during the winter months and from which groups moved to procure resources throughout the area — particularly at higher elevations. 8. The resources used for food included maize. Both cobs and pollen of Zea have been found at the site. In general, then, Fritz employs the data from the Hay Hollow site to reconstruct the subsistence system and other sociocultural phenomena of the people at this site from about 300 B.C. to A.D. 300. April, 1971 Field Museum of Natural History, Chicago Paleoecology of the Hay Hollow Site, Arizona ACKNOWLEDGEMENTS Dr. Richard H. Hevly first aroused my curiosity about the pollen record of economic plants with his research (1964) in the Hay Hollow Valley. Dr. Paul S. Martin, Director of the Southwestern Archaeo- logical Expedition of Field Museum of Natural History, in co-op- eration with Dr. Paul S. Martin, Chief Scientist, Laboratory of Paleoenvironmental Studies of the Geochronology Department of the University of Arizona, provided an opportunity to satisfy my 6 FIELDIANA: ANTHROPOLOGY, VOLUME 63 curiosity. For the verification of Typha pollen and for matters con- cerning methodology and extraction I gratefully acknowledge the help of Peter J. Mehringer, Jr., Palynologist in the Department of Geochronology. John M. Fritz, University of Chicago, exhibited a sense of purpose and closely correlated sampling plan that greatly enhanced the value of the results. The viewpoint of much of my research reflects the influence of Professor Volney H. Jones, Curator of the Ethnobotanical Laboratory of the University of Michigan. None of the research could move forward without the financial assistance received under National Science Foundation Grant GS 245, GS 984, GS 1910, and GS 2381, and the Wenner-Gren Foundation Grant 2186 to Dr. Paul S. Martin of Field Museum of Natural His- tory. Assistance from a predoctoral fellowship awarded by the Uni- versity of Arizona and a National Science Foundation Grant GB 4694 to Dr. Paul S. Martin of the University of Arizona were also essential for the completion of this research. INTRODUCTION Plant macrofossils, such as leaves, seeds, or wood were infrequent in the Hay Hollow excavation. The pollen, as a record of vegetation, consequently increased in importance. Pollen spectra could be pro- cured from apparent replicates of cooking pit types, from similar sections of house floors, and from grinding tools. Pollen investiga- tions focused on revealing the prehistoric utilization of plants, both cultivated and wild, and discerning the nature of the plant com- munities characteristic of the Hay Hollow Valley during the occupa- tion of the site. I (1968) have presented detailed results which are summarized in this publication. POLLEN ANALYTICAL PROCEDURES The excavators of the Hay Hollow Site gathered the pollen data in the form of sediment samples from the floor and fill of about 40 per cent of the pits and from all definite houses. Samples were re- moved with a trowel and placed in clean plastic bags tagged with provenience and date. About 250 samples were collected, 234 of which have been catalogued by the Geochronology Department, University of Arizona. Of the 100 samples extracted, 30 samples had sufficient pollen for a 200-grain count. About half came from the floors of outdoor pits and half from house floors. BOHRER: HAY HOLLOW SITE 7 The interpretation of fossil pollen counts from the Hay Hollow Site was enhanced by studying the modern plant communities in 25 locations together with the pollen spectra derived from soil surface samples. I adapted the quarter method (Curtis, 1950, p. 25) for sampling the modern soil surface among pines. First a topographic map was used to establish a compass direction for a transect that would run at right angles to the prevailing slope. A pair of numbers from a random numbers table determined the paces to each sampling point of four reference trees. Forty double-pinches (one pinch is ca. .7 ml.) of surface soil were accumulated in a single plastic bag and assigned a geochronology catalog number. Except for omitting the four trees as reference points, the same technique was usually followed in collecting soil surface samples among junipers and in grassland. All samples were collected between January 30 and April 9, 1967. I extracted the pollen following the HCI-HNO3 schedule of Meh- ringer (1968). In addition, I inspected screenings for seeds, aceto- lysed the sample just prior to the treatment with potassium hydrox- ide, and dehydrated the residue in 95 per cent and 100 per cent ethyl alcohol followed by thiophene-free benzene prior to mounting in sili- con oil (Dow Corning 200 fluid, viscosity 12,500 centistokes). Pollen taxonomic categories conformed to those used by Hevly et al. (1965) with a few modifications. Short-spine was substituted for low-spine and long-spine for high-spine in the Compositae. Opun- tia grains were divided into the Cylindropuntia or cholla type and the Platyopuntia or prickly-pear type, on the basis of puncti-bacu- late ektexine in the former and reticulate ektexine in the latter (Kurtz, 1948; Tsukada, 1964). Juniper was omitted from all pollen sums to minimize the effects of differential pollen preservation and the impact of juniper destruction by ranchers on the natural pollen rain. Pollen measurements from 26 ponderosa pines (Pinus ponderosa) and 14 pinyon pines (P. edulis) in the Snowflake area were used to evaluate the pine pollen record in conjunction with surface samples. Modern and fossil pine saccus breadth frequencies were based on 100 observations. The relation of Indian plant uses to pollen deposition was ex- plored through experiments and by the use of the 95 per cent con- fidence interval of the binomial distribution (Herskowitz, 1965; Fryer, 1966). 8 FIELDIANA: ANTHROPOLOGY, VOLUME 63 MODERN VEGETATION The escarpment known as the Mogollon Rim forms a rugged arc from Flagstaff, Arizona, southeast through central Arizona into New Mexico. It marks the southern edge of the Colorado Plateau, which extends through adjoining portions of New Mexico, Colorado, and Utah. At 7,500 ft. the crest of the rim supports a broad sweep of ponderosa pine forest with occasional admixture of Gambel oak (Quercus gamhelii), Douglas fir (Pseudotsuga menziesii), or aspen {Populus tremuloides) . The rim also cradles a less stately forest, known as a pinyon-juniper woodland, on its lower slopes that merge with the Colorado Plateau. The woodland interdigitates with the grassland vegetation. Islands of pinyon-juniper woodland cover mesas and eminences; a grassland predominates in the valleys, some- times as a juniper savanna in the sense of Dansereau (1957) (see fig. 1). The Hay Hollow Site vegetation probably represents only a relict of a pinyon-juniper woodland. Piny on pine grows on portions of north-exposed mesa slopes and intermixes with some Utah juniper (Juniperus osteosperma) in the rocky, dry channel east of the site. Few pinyon live on the mesa top and none occupies the Hay Hollow site, located on a terrace below a northeastward extension of the sandstone mesa. Broadly spaced one-seed juniper (Juniperus mono- sperma) are conspicuous on the site terrace (fig. 2); the junipers gradually decline in density as they mix with the alkali-sacaton (Sporobolus airoides) grassland at slightly lower elevations to the north. Since the site terrace was devoid of annuals for three years, shrubs and other perennials were especially obvious. They included : squaw- bush (Rhus trilobata), adelia (Forestiera neomexicana) , banana yucca (Yucca haccata), wild buckwheat (Eriogonum mearnsii), sagebrush (Artemisia bigelovii), and heavily grazed Mormon tea (Ephedra tor- reyana) . This sparse cover of primarily entomophilous shrubs helped to account for the high mean pine pollen frequency (36 per cent). Such a high pine percentage in the modern pollen rain would not be expected in the desert grassland of southern Arizona where grasses and annual anemophilous herbs are abundant and where the local pollen rain is presumably heavier (Martin, 1963). Fig. 1. (opposite). Vegetational transect from the Mogollon Rim to the Hay Hollow Site. 10 FIELDIANA: ANTHROPOLOGY, VOLUME 63 In the summer of 1967 the excavation backfill on the terrace pro- vided a habitat for pioneer species. Nevertheless, despite a moist summer, only three species were observed to colonize the disturbed area. They represented disseminules of nearby perennials such as grama grass (Bouteloua sp.), alkali-sacaton grass, and blue gilia (Gilia longiflora). The weedy elements one might expect on dis- turbed ground were missing, perhaps because there was no local seed. Even if the seed source were closer, weed establishment would be doubtful. One weedy species, Chenopodium fremontii, formed sparse rings beneath junipers only in rainy summers in the valley. Popu- lations of Chenopodium fremontii might be restricted to such locations because the cattle grazed the inflorescence in spite of otherwise abun-3 dant vegetation. On the other hand, another potential weed, pros- strate pigweed (Amaranthus graecizans), exhibited ungrazed shoot tips but failed to spread beyond the relatively restricted habitats along intermediate-sized drainways. Climatic or edaphic factors, rather than grazing pressure, seemed to influence its distribution. Many plants of potential economic value are not productive under the present climate. Although banana yucca produced flowers, neither it nor wolfberry (Lycium pallidum) bore fruits. In poor years pinyon pine seed production was limited to a few trees in the drainways near the site. Shrubs like squawbush grew only 3 ft. tall and lacked vigor. The low density of weedy annuals mentioned above would not sustain human life even in good years. It is hard to imagine successful maize cultivation at this site under the above conditions. PREHISTORIC CHANGES IN PLANT COVER The Hay Hollow Site received higher effective moisture during the time it was occupied than today. Apparently enough moisture was available when house 17 was occupied to support a pinyon- juniper woodland. The nearest woodland today is 12 miles distant. I do not know if the former woodland was isolated from the present woodland or continuous with it. The alkali-sacaton grassland probably grew near the site, as it does today. Although salt-bush may have been less abundant than today, the herbaceous Chenopodiaceae and Amaranthus (table 1) were certainly more abundant. Populations of other herbs were prob- ably more widespread than today. Among them would be tansy- mustard (Descurainia obtusa), bee plant (Cleome serrulata), and 11 Table 1. — Flora of the Hay Hollow Valley and nearby areas with a list of University of Arizona Herbarium accession numbers. Vouchers are identified from Kearney and Peebles, 1960. PINACEAE. Pine Family Pinus edulis pmyon pme 158252 Juniperus monosperma J. osteosperma one-seed juniper Utah juniper 158249 158253 EPHEDRACEAE. Joint-fir Family Ephedra viridis E. torreyana Mormon tea Mormon tea 163897 158191 GRAMINEAE. Grass Family Bouteloua gracilis Hilaria jamesii Lycurus phleoides* Muhlenbergia torreyi Munroa squarrosa Oryzopsis hymenoides Sporobolus airoides S. contractus Tragus sp. blue grama galleta grass wolf-tail grass ring muhly grass false buffalo grass Indian rice grass alkalai-sacaton grass 164318 161745 163246 164330 164329 163408 164317 LILIACEAE. Lily Family Nolina microcarpa Yucca baccata Y. baileyi beargrass banana yucca 158255 163909 163899 SALICACEAE. Willow Family Populus spp.* Salix sp.* Cottonwood willow POLYGONACEAE. Buckwheat Family Eriogonum alatum E. jamesii E. mearnsii* winged eriogonum antelope sage wild buckwheat 161738 161777 CHENOPODIACEAE. Goose-foot Family Atriplex canescens Chenopodium fremonti C. graveolens C. watsoni* Eurotia lanata Salsola kali four-winged saltbush goose-foot 158217 164340 164333 161417 164328 AMARANTHACEAE. Amaranth Family Amaranthus albus* A. graedzans A. powellii* A. torreyi prostrate pigweed 161760 164335 163210 164332 NYCTAGINACEAE. Four-o'clock Family Mirabilis multiflora Oxybaphus linearis four-o'clock 164324 164321 * Collected between Pinedale and Taylor, Arizona. 12 Table 1. — ^Flora of the Hay Hollow Valley and nearby areas with a list of University of Arizona Herbarium accession numbers. — Continued PORTULACACEAE. Portulaca oleracea P. parvula P. rettisa Portulaca Family purslane purslane purslane CARYOPHYLLACEAE. Pink Family Drymaria fendleri* BERBERIDACEAE. Barberry Family Berberis haematocarpa red mahonia CAPPARIDACEAE. Caper Family Cleome serrulate Rocky Moimtain bee-plant CRUCIFERAE. Mustard Family Descurainia dbtusa* Dithyrea wislizeni Lepidium montanum Stanleya pinnata SAXIFRAGACEAE. Ribes sp.* ROSACEAE. Rose Family Chamaebatiaria millefolium* Cowania mexicana LEGUMINOSAE. Pea Family Amorpha fruticosa* Astralagv^ spp. Parryella filifolia tansy-mustard spectacle-pod pepper-grass desert-plume Saxifrage Family goose-berry fern-bush cliff-rose false-indigo loco-weed POLYGALACEAE. Polygala alba* EUPHORBIACEAE. Croton texensis Euphorbia revolutia ANACARDIACEAE, Rhv^ radicans* R. trilobata Milk-wort Family milk-wort Spurge Family Cashew Family VITACEAE. Grape Family Vitis arizonica* MALVACEAE. Mallow Family Sphaeralcea sp. LOASACEAE. Loasa Family Mentzelia pumila* CACTACEAE. Cactus Family Echinocereus sp. Opuntia whipplei O. sp. poison-ivy squaw-bush canyon grape stick-leaf hedgehog cactus Whipple choUa prickly-pear * Collected between Pinedale and Taylor, Arizona. 13 164331 164508 164316 163240 158213 163896 164326 164325 158192 158195 161742 164336 164314 158256 14 FIELDIANA: ANTHROPOLOGY, VOLUME 63 Table 1. — Flora of the Hay Hollow Valley and nearby areas with a list of University of Arizona Herbarium accession numbers. — Continued ONAGRACEAE. Evening-primrose Family Oenothera runcinata evening-primrose OLE ACE AE. Olive Family Forestiera neomexicana adelia Phlox Family blue gilia Borage Family POLEMONIACEAE. Gilia longiflora G. multiflora BORAGINACEAE. Cryptantha flava LABIATAE. Mint Family Salvia reflexa* SOLANACEAE. Potato Family Chamaesaracha coronopus Lycium pallidum Physalis fendleri Rocky Mountain sage wolf-berry ground cherry SCROPHULARIACEAE. Cordylanthus wrightii* Penstemon linarioides Figwort Family club-flower beadtongue COMPOSITAE. Sunflower Family Artemisia bigelovii A. filifolia Aster arenosus Brickellia brachyphylla Chrysothammis sp. Franseria acanthicarpa Gaillardia sp. Gutierrezia lucida Helianthus sp.* Pedis papposa Senecio longilobus Thelesperma sp. Verbesina encelioides Xanthium sp. Zinnia grandiflora sagebrush sand sagebrush brickell-bush rabbit-brush bur ragweed blanket-flower snake-weed sunflower chinchweed thread-leaf groundsel crown-beard cocklebur 161818 158214 164315 163901 161759 161902 164320 161744 161741 164327 158210 158218 164323 164399 164322 164337 164397 164334 164319 * Collected between Pinedale and Taylor, Arizona. ground cherries (Physalis fendleri and Chamaesaracha coronopus). All of the foregoing plants were traditionally gathered for food by the Hopi (Whiting, 1939) and Zurii (Stevenson, 1915). The use of beeweed and tansy-mustard extends into the prehistoric plant record (Jones and Fonner, 1954; Schoenwetter and Eddy, 1964, p. 74; Martin and Byers, 1965). The structure of the alkali-sacaton grassland during the occupa- tion of the Hay Hollow Site changed according to the available mois- BOHRER: HAY HOLLOW SITE 15 ture (fig. 3). Moist summers provided an abundance of alkali-saca- ton grass and dry summers allowed expansion of the herbaceous Chenopodiaceae and Amaranthus populations. In summers of abun- dant rainfall large stores of alkali-sacaton grass seed could be set aside with relative ease. In dry summers the prostrate pigweed, an amaranth, would spread in the open spaces among the grass and furnish an alternate harvest. Prostrate pigweed could easily invade a cultivated field. Other species of Chenopodium and Amaranthus would furnish seed in intermediate years. No similar wild harvests in drought years are available today except in limited and widely scattered habitats. With higher available moisture, yucca, cholla, and pinyon would all bear fruit more heavily. Wolfberry and squawbush would un- doubtedly be more productive. Wild grapes might grow in the area and a different species of cholla, although the floristic composition would be essentially similar to the modern one. Today a relict-stand of pinyon in the midst of a juniper grassland is all that remains of the pinyon-juniper woodland which apparently grew nearby at the time of initial occupany. Herbaceous Chenopo- dium and Amaranthus populations also appear relict at present, and the water table is lower than in times past. While these are valid ecological contrasts, and the average available moisture seems to be low enough to limit Chenopodium, Amaranthus, and pinyon popula- tions, I do not know the extent to which other factors over the past 2,000 years have contributed to the modern plant ecology. EVIDENCE FOR HIGHER EFFECTIVE SOIL MOISTURE Herbaceous Pollen Cat-tail {Typha sp.) pollen in the prehistoric spectra implies higher ground water levels than today. The cat-tail pollen recov- ered from the floors of houses 17, 32, and in a cooking pit (68051815) probably originated locally. If the pollen were wind transported, they probably came from less than 7 miles distant (cf. Potter and Rowley, 1960). If man inadvertently introduced the pollen, then we may infer local availability. People who used cat-tail, such as the Acoma, Laguna, Pima, and Paiute, for everyday purposes (Rus- sell, 1908, p. 133; Castetter, 1935, p. 53; Stewart, 1941, p. 375) had the plants available locally. No cat-tail grows near the modern site today, not even in localized habitats. 16 FIELDIANA: ANTHROPOLOGY, VOLUME 63 A. During site occupancy alkali- sa c a t on gra ss pros t rate pigweed e ve n I n g prim ro s • fll/Ww^W alt- bus h LOW AVAILABLE MOISTURE \ / alka I i. so c at on grass ^ sa I t-bush HIGH AVAILABLE MOISTURE FiG. 3. a. Extreme fluctuations in plant cover (September aspect) adjacent to the Hay Hollow Site as postulated during the site occupation. The record of the Chenopodiaceae + Amaranthus pollen type, together with the seed that implied food utilization, pre-supposed an abundance of herbaceous plants. One would either need a jeep or helicopter today to exploit the limited and widely scattered herba- ceous Chenopodium and Amaranthus habitats in the area. During a modern drought year, there may be no herbaceous Chenopodium and Amaranthus whatsoever. More regularly available moisture in an average summer must have maintained the populations at a broader distribution during the site occupation to make the seed BOHRER: HAY HOLLOW SITE B. During i96o'$ 17 ^11. ^ salt-bush alkali' so c a ton gr ass LOW AVAILABLE MOISTURE N / ^^^^■^^■^■■^ prostrate salt-bush alkal i- pigweed sacaton grass HIGH AVAILABLE MOISTURE Fig. 3. b. Extreme fluctuations in plant cover (September aspect) adjacent to the Hay Hollow Site as observed in the 1960's. collection practical. The shifts in structure within the alkali-sacaton grassland are inferred from the modern ecology of a swale 2 miles west of Snowflake, Arizona. (See section on plants used by prehis- toric occupants for a more complete discussion of evidence.) The recovery of pollen from evening-primrose (Onagraceae) in prehistoric pollen spectra is also suggestive of slightly higher levels of available moisture. During my three seasons observing the flora in the Hay Hollow Valley, I found no evening-primrose. Oenothera runcinata, one member of the family, did grow in the alkali-sacaton 18 FIELDIANA: ANTHROPOLOGY, VOLUME 63 Fig. 4. breadth. Pine pollen grain with two sacci as seen in distal view. a=saccus swale 2 miles west of Snowflake, Arizona. The same species grew at Window Rock, Arizona, in a similar habitat or in areas that received extra run-off from the road shoulder. Assuming Oenothera is not present today, slightly higher moisture values would probably be necessary to maintain a population of evening-primrose in the Hay Hollow Valley. Arboreal Pollen As the artificial introduction of pine pollen into the site was un- likely and as only two naturally occurring pine species (Pinus pon- derosa and P. edulis) were expected, I endeavored to separate the two on the basis of bladder breadth (fig. 4) and to use their pollen frequencies as the estimate of the past position of the pinyon-juniper woodland border. The smaller pinyon pollen bladders formed a nor- mally distributed population that overlapped in its upper size ex- tremes with the larger ponderosa pine pollen (fig. 5). However, bladders ^ 30/i seemed to be almost entirely pinyon pollen. A series of modern soil surface samples were analyzed to determine the char- acteristic frequency of pine pollen bladders ^ 30/i in grassland, at the borders of and within the pinyon-juniper woodland, and in a ponderosa pine forest. By using up to 10 soil surface samples from different locations within a given vegetation type, it was possible BOHRER: HAY HOLLOW SITE 19 ponderosa from mesic habi f at (n s 200 ponderosa from xeric ha bi to t (n= 200 ) pi nyon (ns 280) 20 30 40 50 60>i Fig, 5. Graph of pollen saccus breadth frequencies of ponderosa and pinyon pine. to statistically discriminate between the modern archaeological site, a pinyon-juniper woodland, and a grassland on the basis of the fre- quency of pine pollen ^ 30 /x (see table 2). Pollen extractions from the prehistoric site were measured to ob- tain the frequency of pine bladder breadth equivalent to ^ 30/i in a sample of 100 pine pollen grains (table 3). The frequency of small 20 FIELDIANA: ANTHROPOLOGY, VOLUME 63 Table 2. — Summary of small pine pollen saccus breadth (^30^) frequencies with 95 per cent confidence interval for modern surface samples in major plant communities near Snowfiake, Arizona. No. of sub- samples from different Mean 95% confidence Plant community geographic locales frequency interval Ponderosa pine 4 18 14-23 (N=400) Pinyon-juniper 10 44.2 41-47 (N=1000) "Juniper-Grassland" of Hay Hollow Sitei 2 33.8 31-37 (N=1000) Grassland 10 29.4 26-32 (N=1000) 1 The modern Hay Hollow Site is characterized by junipers, an extremely low density of pinyon trees, and a lack of herbaceous cover. 0 pine pollen from one of the earlier houses (house 17) compared to an equivalent frequency from either the modern environment or from a denser pinyon-juniper woodland. I assumed the latter choice was correct because the previously described independent indicators sug- gested higher effective moisture values. The trend in C^^ dates associated with the three prehistoric pollen samples (table 3) analyzed for pine bladder size paralleled an inter- pretation of increased woodland decimation. The accompanying record of Typha pollen in both the oldest and intermediate samples (houses 17 and 32, respectively) seems incompatible with an inter- pretation of less favorable climatic conditions for the growth of pin- Table 3. — Interpretive summary of prehistoric small pine pollen saccus breadth (^27.5/i) frequencies with C" dates. Description House 17, X1754220 Frequency of small sacci 42.5 95% confidence limits 36-49 Plant community pinyon-juniper or modern site^ C" yrs B.P. 1950 2095±105 1995±100 1895±110 House 32 (com- 29 posite of X3251, X3253, X3254325 Pit 30051815 24 Pit 33853825 18 20-39 16-34 11-27 modern site* or grassland modern site* or grassland grassland 1030 ±80 1920 ±75 1720 ±90 * The modern Hay Hollow Site is characterized by junipers, an extremely low density of pinyon trees, and a lack of herbaceous cover. BOHRER: HAY HOLLOW SITE 21 yon. The persistence of Typha pollen is more harmonious with an interpretation of woodland decimation under generally favorable cli- matic conditions. I consider man as the destroyer of the woodland instead of climate. The contrast of 12.7 per cent pine pollen in house 17 with the 36 per cent pine pollen from the modern surface at Hay Hollow might seem to indicate an early retreat of the number of pine trees in pre- historic times (Schoenwetter and Dittert, 1968, p. 46). Modern Southwestern surface transects show that the pine percentages de- crease with elevation and with distance from pines (Hevly, 1964, 1968; Schoenwetter and Eddy, 1964, p. 68). Application of the gen- eralization to all archaeological sites may not be justified. The generalization implies that the herbaceous cover is constant and the arboreal pollen is the variable. Exceptions can be found in modern surface samples (table 4). Samples derived from Bull Hollow, a grassy valley adjacent to a vigorous pinyon-juniper woodland, showed a lower percentage of pine pollen (18.5 per cent) than the modern Table 4. — Percentage of pine in pollen spectra near Snowflake, Arizona, ranked by decreasing density of herbaceous cover. an! Geochron : Cat. No. Description %pine Distance to nearest pine and elevation in feet 1 9065 Alkali-sacaton swale 7 No closer than 5 mi. to pinyon; 5,600 1 9066 Meadow 2 }4-}^ ini. to ponderosa pine; 6,400 2 9141 Juniper grassland 25 No closer than 134 nii. to pinyon; 5,900 2 9142-2 Border of pinyon- juniper woodland 18.5 100 to 200 feet to pinyon; 5,900 2 9147 Grassland with Morman tea 19 No closer than 8 mi. to pinyon; 5,500 3 9063 Grassland with saltbush 12 No closer than 1 mi. to pinyon; 5,800 4 9148 Eroded grassland 28 No closer than 20 mi. 5,200 4 9061 Hay Hollow Site 34.5 100 to 200 feet to pinyon; 5,800 5 9142-1 Interior of pinyon- juniper stand 58 5,900 5 9057-6 Interior of pinyon- juniper stand 79 6,100 5 9146 Interior of pinyon- juniper stand 68 6,300 22 FIELDIANA: ANTHROPOLOGY, VOLUME 63 Hay Hollow Site with its relict pinyons. I assumed neither the herbaceous or arboreal pollen to be a constant in this study. My assertion that the cultural introduction of pine pollen is negligible involves the following assumptions: (1) No pine pollen residue adhered to pinyon seeds possibly used as food. This assump- tion has supporting experimental evidence (table 7) . (2) The limited Zuni practice of eating pinyon and ponderosa pine buds and shoots (Stevenson, 1915, p. 96) concerned an early enough developmental stage of pollen formation to be ignored as a source of pine pollen in- troduction. (3) No Pueblo Indians gathered pinyon and ponderosa pine pollen for ceremonial use as did the Navajo (Vestal, 1952, pp. 13-14). (4) Any possible pollen distortions resulting from spreading pine needles on the floor where maize was stored can be ignored at the Hay Hollow Site, although excavators at Point of Pines noted the practice at Ariz. W:10:50. (5) The concentrations of pine pollen in an infant burial at Broken K Pueblo near Snowflake, Arizona (Hill and Hevly, 1968, p. 206) resulted from interment on a day within the period of pine pollination. Pollen rain on any one day may be quite different from averages of pollen from living surfaces and may not necessarily indicate cultural introduction. Pine values reach as high as 78 per cent (n= 54) in weekly June extremes in the San Augustin Basin (Potter and Rowley, 1960, p. 16). Seed Macrofossils The recovery of a seed of Portulaca retusa from a prehistoric fire pit (table 5) is perhaps less significant as an indicator of increased available soil moisture, since Portulaca grows in the area today. However, no seeds were recovered from two soil surface pollen sam- pling transects at the site. If available moisture were greater, Portu- laca retusa would grow more abundantly and increase the probability of recovering the seeds in the modern soil surface. Considering the small amount of sediment used in pollen analysis (24-36 ml.), the widespread recovery of a minute charred, unknown seed (table 5) was intriguing. The seed was hexagonal in cross- section and oblong in longitudinal view. It measured about .5 mm. long and half as wide. The dark, non-reflective surface of the minute seed discouraged photographic attempts. Modern surface samples at the Hay Hollow Site did not contain the seed, but I recovered it in a soil surface sample in a mesic tribu- tary of Bull Hollow (12 miles distant). I redeemed the same seed at the Tule Springs Site in southern Nevada, where oxidized seed BOHRER: HAY HOLLOW SITE 23 Table 5. — Seeds screened from prehistoric sediment samples prior to pollen extraction. Geochron. Cat. No. Field Museum No. No. and type of seed; remarks 9068-2 X3252 y^ T*; house 32 floor 9069-1 X3253 Herbaceous cheno-am fragment; 90% cheno- am pollen; from under grinder in house 32 9070-2 X3255 IT*; house 32 entryway 9072-2 X2553 1 T*; house 25 floor 9076 16853815 1^ T*; firepit floor 9080 22852815 2 T*; firepit floor 9089-2 29450815 2 T*; firepit floor 9095-1 31253815 J^ T*; 1 Chenopodium graveolens, 82.5% cheno-am pollen from firepit floor. 9101-1 44753815 6 T*; 1 Portulaca retusa, herbaceous cheno- am fragments; 81% cheno-am pollen from firepit floor. 9105 53052815 2 T*; firepit floor 9111-2 59251825 IT*; firepit floor 9116-2 62954825 HT*; firepit floor 9138 269 1 T*; burial * Tule Springs Site, Nevada, unknown. See section on seed macrofossils for description. clusters, derived from spring mound IV, stratigraphic unit E2s (Mehringer, 1965), were in association with Cyperaceae, Scirpus sp., grape (Vitis sp.), and ash (Fraxinus sp.) macrofossils. The ecology of the transect in the Bull Hollow tributary and the mesic species composition of the Tule Springs stratigraphic unit im- ply that this unknown seed is also characteristic of a damp or mesic habitat. Since the seed was not found in modern surface samples at the Hay Hollow Site, I inferred the site must have been more mesic during human occupation. PLANTS USED BY THE PREHISTORIC OCCUPANTS Dr. Hugh C. Cutler, Curator of Useful Plants, Missouri Botanical Garden, recovered maize macrofossils while examining organic matter from the fill of 60 pits. An almost spherical maize kernel (about 4.5 mm. diam.) probably came from a pop or very hard flint type. A cob fragment that seemed to come from the tip of an ear had 12 rows, grains 3.4 mm. thick, and a cupule width of 4.4 mm. The 24 FIELDIANA: ANTHROPOLOGY, VOLUME 63 'fragment resembled a popcorn, like Reventador, or a very small form of a hard flint (correspondence). Pollen evidence suggested that alkali-sacaton grass seed prob- ably was ground on a metate. Seeds from the pollen category of Chenopodiaceae + Amaranthus were harvested. The milling stone used in processing the seed as well as an associated baking pit was identified by pollen analysis. Two species of cholla (Cylindropuntia) were harvested and at least one cholla roasting pit was identified from the high concentrations of cholla pollen in the pit. High con- centrations of pollen from long-spine Compositae on a grinding stone also suggested plant utilization. The pollen spectral distortion of mormon-tea (Ephedra sp.), indicative of usage, was associated with one storage pit and possibly a fire pit. Dr. Hugh C. Cutler identified a portion of the carbonized wild seeds from the site. A few Chenopodiaceae or Amaranthaceae seeds and a partly burned seed of Opuntia sp. reinforce conclusions derived from pollen analysis. In addition, carbonized juniper seed, or prob- able fragments, were recovered from houses 13 and 17. The juniper berries might have been consumed by the inhabitants, as the Hopi do today (Whiting, 1939, p. 63; Nequatewa, 1943, p. 18). The key to the identification of certain pollen spectral distortions, indicative of plant utilization, lies in the high concentration of a given pollen type. If, for example, food supplies were replenished often enough in a given location, the associated pollen might have reached a concentration that could not be duplicated by the pollen rain sampled in the soil surface. An interpretation of pollen spectral distortion seems justified when a frequency is so high that it fails to come within the 95 per cent confidence interval (binomial distribu- tion) of the highest probability of occurrence in a known normal spec- trum. For example, Ephedra pollen reached a concentration under natural conditions of 26 per cent with 95 per cent confidence limits of 20 to 33. A fossil spectrum from a cultural context (X1754365) had a frequency of 44 per cent, with 95 per cent confidence interval of 38 to 51. The fossil sample might yet be matched by undiscovered or unanalyzed samples of the modern pollen rain. But a plant com- munity exhibits sufficient homogeneity that I do not expect very many new samples to exceed the confidence limits of the extreme sample used above as an illustration. One could not conclude that a fossil Ephedra sample statistically exceeding the confidence limits of the modern sample will never be matched by the natural pollen BOHRER: HAY HOLLOW SITE Table 6. — Pollen spectral distortion at the Hay Hollow Site. 25 Pollen t5rpe % Field Museum No. Provenience Cheno-Ami 80 X3251 NE quadrat, house 32 Cheno-Am 90 X3253 SW quadrat beneath hand grinder, house 32 Cheno-Am 79 X3253 SW quadrat, house 32 Cheno-Am 80 X325335 SW quadrat, floor of pit in house 32 Cheno-Am 79.5 22852815 Outdoor cooking pit type 2b from area 1 Cheno-Am 82.5 31253815 Outdoor cooking pit Cheno-Am 81 44753815 Outdoor cooking pit Cheno-Am 80 52951825 Outdoor cooking pit Cheno-Am 86 53052815 Outdoor cooking pit type 2b from area 2 Cheno-Am 79 57252825 Outdoor cooking pit Cylindropuntia'' 9 31351865 Outdoor cooking pit Ephedra tor- reyana type' 44.5 X1754365 Floor of pit in house 17 * Chenopodiaceae + Amaranthus. Modern maximum was 59% with 95% con- fidence interval from 52-65. Prehistoric mean of 79 % has 95% confidence interval of 72-84. ^ Unless sampling directly beneath cholla, the modern frequency is 0% with 95% confidence interval 0-2. Prehistoric mean of 9% has 95% confidence interval of 6-14. ' Modern maximum was 26% with 95% confidence interval 20-33. The pre- historic 95% confidence interval is 38-51. rain. One could conclude that this will rarely happen and that the distorted nature of the fossil Ephedra sample probably had a cul- tural cause. The identification of pollen spectral distortion (table 6) led to experiments to ascertain how pollen concentrations might accumu- late. When one considers that pollination precedes seed formation, the assertion that pollen spectral distortion may be created by the use of seeds seems paradoxical. However, experiments suggest that pollen adheres to the seeds, fruits, and husks in six of the seven spe- cies of plants tested (table 7). The nature of the blossom, or the collective blossoms (inflorescence), provides a possible explanation. In the case of maize, husks enclose the cob from pollination to maturity. Pollen accumulates on the husk exterior and may be transferred to the kerneled cob by the person doing the husking. Cholla {Opuntia whipplei) and sunflower {Helianthus sp.) have bios- 26 FIELDIANA: ANTHROPOLOGY, VOLUME 63 Table 7. — Relative abundance of pollen washed from modern seeds, fruits, and husks. Pollen Name of plant Plant part Common Abundant to rare Absent Pinyon (Pinus edulis) seed* X alkali sacaton grass {Sporobolus airoides) caryopsis* X maize {Zea mays) husk caryopsis X X goose-foot (Chenopodium dessicatum) seed* X Prostrate pigweed (Amaranthtis graecizans) seed* X Whipple choUa (Opuntia whipplei) fruit and stem X sunflower {Helianthus sp.) achene* X * Winnowed prior to treatment soms which shrivel and dry but persist at the apex of the ripening ovary or fruit (fig, 6) , Wind and moisture may transfer pollen from the old flower to the developing fruit beneath it. The pollen of flowers arranged in the indeterminate inflorescence as in grass, Cheno- podium and Amaranthus, was unintentionally collected in the same container as the mature seed while beating the inflorescence for seeds (fig. 7). The foregoing experiments raise the question of whether the use of seeds or flowers created pollen spectral distortion. The alterna- tives are difficult to weigh in the case of the long-spine Compositae pollen type. Sunflower {Helianthus sp.) blossoms might be ground for ceremonial purposes (Stevenson, 1915, p. 93) or the achene might be cracked on the milling stone to facilitate separation of the shell from the seed (Gushing, 1920, p. 252). Either situation might con- tribute to the long-spine Compositae pollen concentration. The use of many other genera in the same family would produce similar results. The recovery of small, black seed-fragments resembling Cheno- podium or Amaranthus in association with distorted frequencies of the Chenopodiaceae + Amaranthus pollen type indicated that the prehistoric pollen probably was originally introduced along with the seed (table 5). BOHRER: HAY HOLLOW SITE 27 old petals f I o w e r frui t Fig. 6. Flower with inferior ovary and developed fruit. Old stamens with pollen residue are retained with the old petals in choUa {Opuntia whipplei) and in sunflower {Helianthus sp.). In the case of grass, I infer that it was the seed ground on the stone, since I know of no modern use of blossoms of grass that entails grinding or crushing. Normally, prehistoric pollen of grass genera seldom shows consistent differences at magnifications of 400 X, ex- cept in size range. Because certain small grass pollen occurred in clumps, in suspiciously large amounts, and conformed in size to pollen of modern Sporoholus airoides (23^ ±1.6 n=50) which grows nearby, inference on the use of this species was postulated. It was assumed no other small grass pollen now absent (i.e., Leersia ory- zoides, Phragmites communis, Calamagrostis sp.) produced the pollen record in question. Although only one species of cholla (Opuntia whipplei) now grows in the Hay Hollow vicinity, knowledge of former utilization of an additional species is based on significant differences in the prehistoric cholla pollen diameters (58. 5^ ± 22.2 n= 24) from the diameters of modern Whipple cholla pollen (63/i ±6.0 n=24). It is a matter of speculation if people traveled long distances to collect the other spe- cies of cholla or even traded for it. Cholla joints are considered more of a starvation food than a delicacy (Castetter, 1935, p. 35). Cholla buds are favorably regarded (Nequatewa, 1943, p. 19), but I am 28 FIELDIANA: ANTHROPOLOGY, VOLUME 63 pollen-bea r i ng f lowe r s Fig. 7. Indeterminate inflorescence typical of Chenopodium, Amaranthus, and members of the grass family. acquainted with no modern records of extended travel or trade for cholla. This leads to speculation that cholla was locally exploited until only one species survived. POLLEN INDICATORS OF SEASONAL OCCUPANCY If pollen spectral distortion has been accurately assessed, the seasonal presence of the creator of the distortion, man, may be rec- ognized in certain cases. For example, I would assume that the people who lived in house 17 were present in the spring when Ephedra pollinates and in late summer when maize, alkali-sacaton grass, and prostrate pigweed, mature. Cholla pollen should probably be dis- regarded as evidence of seasonal occupation at the Hay Hollow Site since it is possible to gather the fruit or vegetative portions of 0. whipplei in winter, and still retain pollen (table 7) . Pollen spectral distortion will tell the investigator little or nothing in regard to winter occupancy. On the other hand, it may provide important clues to occupancy during certain portions of the growing season. REFERENCES BOHRER, VORSILA L. 1968. Paleoecology of an archaeological site near Snowflake, Arizona. Ph.D. dissertation, Univ. Arizona. BOHRER: HAY HOLLOW SITE 29 Castetter, Edward F. 1935. Ethnobiological studies in the American Southwest I. Uncultivated native plants used as sources of food. Univ. New Mex. Bull. Biol. Ser., 4, no. 1, pp. 7-62. Curtis, John T. 1950. Plant ecology work book. Burgess Co., Minneapolis. CusHiNG, Frank H, 1920. Zimi Breadstuff. Indian Notes, Monographs, 8. Dansereau, Pierre M. 1957. Biogeography; an ecological perspective. Ronald Press, New York. Fryer, H. C. 1966. Concepts and methods of experimental statistics. AUyn and Bacon, Inc., Boston. Herskowitz, Irwin H. 1965. Genetics, 2nd ed. Little Brown and Co., Boston. Hevly, Richard H. 1964. Pollen analysis of Quaternary archaeological and lacustrine sediments from the Colorado Plateau. Ph.D. dissertation, Univ. Arizona. 1968. Studies of the modern pollen rain in northern Arizona. Jour. Ariz. Acad. Sci., 5, no. 2, pp. 116-124. Hevly, Richard H., Peter J. Mehringer, and Harrison G. Yokum 1965. Modern pollen rain in the Sonoran Desert. Jour. Ariz. Acad. Sci., 3, pp. 123-135. Hill, James N. and Richard H. Hevly 1968. Pollen at Broken K. Pueblo : some new interpretations. Amer. Antiquity, 3, no. 2, pp. 200-210. Jones, Volney H. and Robert L. Fonner 1954. Appendix C, Plant materials from sites in the Durango and La Plata Areas, Colorado, pp. 93-115. In E. H. Morris and R. F. Burgh, Basket Maker II sites near Durango, Colorado, Carnegie Inst, pub, no. 604. Kurtz, Edwin B., Jr. 1948. Pollen grain characters of certain Cactaceae. Bull. Torrey Bot. Club, 75, pp. 516-522. Martin, Paul Schultz 1963. The last 10,000 years. Univ. Arizona Press, Tucson. Martin, Paul Schultz and William Byers 1965. Pollen and archaeology at Wetherill Mesa. Amer. Antiquity, 31, pt. 2, pp. 122-135. Mehringer, Peter J„ Jr. 1965. Late Pleistocene vegetation in the Mojave Desert of southern Nevada. Jour. Ariz. Acad. Sci., 3, pp. 172-188. 1968. Pollen analysis of the Tule Springs Site, Nevada. Ph.D. dissertation, Univ. Arizona. Nequatewa, Edmund 1943. Some Hopi recipes for the preparation of wild food plants. Plateau, 16, pp. 18-20. 30 FIELDIANA: ANTHROPOLOGY, VOLUME 63 Potter, Loren D. and John Rowley 1960. Pollen rain and vegetation, San Augnstin Plaines, New Mexico. Bot. Gazette, 122, pp. 1-25. Russell, Frank 1908. The Pima Indians. Bur. Amer. EthnoL, 26th Ann. Rept., 1904-1905, pp. 17-389. ScHOENWETTBR, James and F. W. Eddy 1964. Alluvial and palynological reconstruction of environments, Navajo Res- ervoir District. Mus. New Mex. Pap. Anthropol., no. 13. SCHOENWETTER, JAMES and ALFRED E. DiTTERT 1968. An ecological interpretation of Anasazi settlement patterns, pp. 41-66. Anthropological Archaeology in the Americas. Anthropological Society of Washington, Washington, D. C. Stevenson, Matilda C. ' 1915. Ethnobotany of the Zuni Indians. Bur. Amer. Ethnol., 30th Ann. Rept., 1908-1909, pp. 31-102. Stewart, Omer C. 1941. Culture element distributions: XIV Northern Paiute. Anthropol. Rec, 4, no. 3, pp. 361-445. TSUKADA, M. 1964. Pollen morphology and identification. II. Cactaceae. Pollen et Spores, 6, pp. 45-84. Vestal, Paul A. 1952. Ethnobotany of the Ramah Navajo. Pap. Peabody Mus. Amer. Ar- chaeol. Ethnol., Harvard University, 40, no. 4, pp. 3-94. Whiting, Alfred F. 1939. Ethnobotany of the Hopi. Mus. N. Ariz. Bull., no. 15.