Journal of Ethnobiology VOLUME |, NUMBER 1 MAY 1981 JOURNAL ORGANIZATION Co-DIRECTORS: Steven D. Emslie and Steven A. Weber, Center for Western Studies, Inc., P.O. Box 1145, Flagstaff, Arizona, 86002. EDITOR: Steven D. Emslie NEWS AND COMMENTS EDITOR: Eugene Hunn, Department of Anthropology, DH-05, University of Washington, Seattle, Washington, 98195. EDITORIAL BOARD BRENT BERLIN, Language Behavior Research Laboratory, University of Calif- ornia, Berkeley; ethnotaxonomies, linguistics. ROBERT A. BYE, JR., Department of Environmental, Population and Organismic Biology, University of Colorado, Boulder; ethnobotany, cultural ecology. RICHARD S. FELGER, Senior Research Scientist, Arizona-Sonora Desert Museum, Tucson; arid land ethnobotany, desert ecology. RICHARD I. FORD, Director, Museum of Anthropology, University of Michigan, Ann Arbor; archaeobotany, cultural ecology. B. MILES GILBERT, Adjunct Research Associate, Division of Vertebrate Paleont- ology, University of Kansas, Lawrence; zooarchacology. TERENCE E. HAYS, Department of Anthropology and Geography, Rhode Island College, Providence; ethnobotany, ethnotaxonomies. RICHARD H. HEVLY, Department of Biological Sciences, Northern Arizona Univer- sity, Flagstaff; archaeobotany, palynology. EUGENE HUNN, Department of Anthropology, University of Washington, Seattle; ethnotaxonomies, zooarchaeology, cultural ecology. V. KUHNLEIN, Division of Human Nutrition, University of British Columbia, Vancouver; ethnonutrition. Gary P. NABHAN, Meals for Millions Foundation, Tucson; cultural ecology, plant domestication. A. Posey, Center for Latin American Studies, University of Pittsburgh; ethnoentomology, tropical cultural ecology. M. REA, Curator of Birds and Mammals, San Diego Museum of Natural Sisixy- ethnotaxonomies, zooarchaeology, cultural ecology. © Center for Western Studies, Inc. Journal of Ethnobiology VOLUME 1, NUMBER 1 MAY 1981 MissOUR! BOTANICA® ALFRED FRANK WHITING, 1912-1978 INTRODUCTION The Journal of Ethnobiolagy i isa professional journal devoted to the interdisciplinary study of anthropology and biology which will of original research by Smemignveaney and other specialists. The — will consist of papers covering a broad range of topic , ethnobotany, ethnozoology, vated ecology, plant aaa zooarchaeology, archaeobotany, palynology, dendrochronology, and ethnomedicine. This ae which contains articles dealing with nearly all these subject areas, consists of pa resented at the Second Annual Ethnobiology Conference held in Flagstaff, Arizona, 6-7 April 1979. This conference was appropriately held in honor of 2 recently deceased prominent persieneeteata Lyndon L. rere and Alfred F. Whiting. This first issue of the journal is ated t to the ee men ll 1 bibliog of Alfred Whiting by Katharine Bartlett of the Museum of Northern. Arizona, F ‘oi Biographies of Hargrave have been published by Dick and Schroeder (1968), Emslie (1979), and Taylor and Euler (1980). The second article by Richard I. Ford was presented as the keynote address at the conference. The next 7 articles by Grayson, Hevly, Yarnell, Pulliam, Rea, Kuhnlein, and Berlin et al. comprised a special symposium of uiviied speakers to honor EErRrave and Whiting. The remaining papers in this i The Museum of Northern Arizona is acknowledged for their efforts in organizing this conference and for their considerable help and cooperation in allowing these papers to be published as the first issue of the Journal of Ethnobiology. These excellent papers are an ideal collection of research to begin a new journal. Proceedings of fut gy conferences will blished i i f the journal. The next issue, Volume | Number 2, will contain selected wapis presented at the Fourth Annual Ethnobiology Conference held in Columbia, Missouri, 13-14 March 1981. The journal will begin soliciting papers on original — for = firsts issue of 1982 (Volume 2 Number 1) in September 1981. Authors are the inside cover of this issue. By 1983, the journal may expand to a quarterly release and include book reviews and news and comments sections. Finally, this journal would not have been possible without the support and cooperation of 12 notable ethnobiologists comprising the Editorial Board; their experience and knowledge are what is required to ensure a successful and high quality journal. Steven D. Emslie Co-Director and Acting Editor Steven A. Weber Co-Director Journal of Ethnobiology LITERATURE CITED Dick, HERBERT W., AND ALBERT H. SCHROEDER. — EMSLIE, S.D. 1979. Introduction. Kiva 44 (2-3):77- Pp. 1-8, im Collected Papers in Honor of TayLor, WALTER W., AND ROBERT C. EULER. Lyndon Lane Hargrave (Albert H. Schroeder 1980. Lyndon Lane i 1896-1978. Amer. ed.). Papers Arch. Soc. New Mexico 1, hin. Antiquity 45(3):477-4 New Mexico Press, Santa Fe. J. Ethnobiol. 1 (1): 1-5 May 1981 ALFRED F. WHITING, 1912-1978 peer ee m of Nort Route 7 ‘hos 720, oa picts 86001 Alfred Frank Whiting was born in Burlington, Vermont, in 1912. After attending public schools, he went to the University of Vermont, located in Burlington, tite in 1933 with a Bachelor of Science degree.. He at once enrolled in the Graduate School at the University of Michigan and the following spring | received an M. A. in Taxonomic ee That summer he was included in a University Botanical E Potosi, Mexico, which may well have been responsible for arousing his interest in ethnobotany, the focal point of his career. In the summer of 1935, Whiting was appointed Curator of Biology at bey Museum of Northern Arizona, where he spent the first few month herbarium. In September he was joined by Dr. Volney H. Jones, ‘also from Michigan, and together they began a survey of Hopi Indian crop plants for the Michigan Ethnobotanical Laboratory. When the harvest was over, Jones returned to Ann Arbor, but Whiting stayed in Flagstaff to record with Edmund Nequatewa, a Hopi man on the Museum staff, the names and uses of cultivated and wild plants he and Jones had collected on the Hopi mesas. Al, whose title at the Museum had been ogee atc to Curator of Botany, continued: to collect and work on the wild plants of F At that time he entolled i in the University of Chicago to begin work on a Ph.D. in the combined fields of botany and anthropology. Whiting returned to Flagstaff in the summer Ethnobotany of the Hopi published as Bulletin 15 of the Museum of Northern Arizona in 1939. The school years of 1938-39 and 1939-40 were spent in Chicago working on his Ph.D. Here he married Dorothy J. West, whom he had met at International House at the ia ak ipine they both cl fo In September 1940, they came to F —— si ther next = £102aQ ad Ihe th on hee ethnobiology. He also continued to serve as Curator of eeany at the Museum. In the Se summer of 1941, Al was g y, unusual for that day and age, to be anthropological sehbicg to te " educational film psig rwon sl Coronet prtdbiclciee of Chicago, which , Navajo, Havasupai and Apache atari He assisted i in arrangements with the tribes for ‘Coronet to make the films and accompanied the photographers to the various reservations. This brought him in contact with the Indian people, their tribal governments, and Bureau of Indian Affairs personnel, altogether an enriching experience. The Coronet films completed, Whiting undertook a 6 month project, sponsored by the Indian Arts and Crafts Board of the U.S. Department of Interior and the Museum of Northern Arizona, to make an intensive survey of production and marketing problems of Hopi Indian arts and crafts. Although the films and the crafts survey were monetarily successful for the student with a wife and child, unfortunately, they diverted him from his Ph.D. dissertation, which he might have completed while still in Flagstaff. Due to World War II and other unforeseen circumstances, he never pee its Kgipewes or oe degree. In July 1942, the Whitings returned to the Mi work at Chicago until the fall of 1944 when he diced an Assistant Professorship at the University of Oregon, substituting for a member of the Anthropology Faculty who was serving in the armed forces in World War II. While at Oregon, most of Whiting’s time appears to have been occupied with teaching his first college classes and curatorial work at the Oregon State Museum. He published an article in American Anthropologist, ‘The Origin of Corn, an Evaluation of Fact and Theory,” based in part on his M.A. thesis at Michigan. a JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 The winter climate in Oregon did not agree with either of the Whitings or their 2 young sons and they were frequently sick. ne this time - pane ae separated and she and the 2 boys returned to Chicago where her p followed. Al wrote to a friend that he longed to get back to the edry Southwest, and so, in the spring of 1947 when his teaching term was up, he moved to Tucson and Tumacacori in the southern Arizona desert to spend the next several years. At the University of Arizona, in association with the Arizona State Museum where he found many old friends, Whiting settled down to do what he most enjoyed in life: research. Based on the results of some historical studies, he wrote ‘““A Kino Triptych” and “The ae cigp i Census of 1796,” the latter published in The Kiva of the Arizona Archaeological Society. The cover of this issue of The Kiva was illustrated with a water color painting by sa (reproduced in black and white) of a reconstruction of the mission church of Tumacacori. During his years in Tucson he painted a number of watercolors of other missions and scenes in the area. A stay at Tumacacori inspired 2 delightful stories based on personal exp with some Spanish-Americans of that vicinity, ‘“The Happy Cemetery” and “Miracle in the Living Room.” A trip to visit the Seri Indians, on the west coast of Sonora and Tiburon Island, to collect and study ihe. plants they used, renee in his reviewing all recorded data on that tribe. A s Seri work, he collected several loose leaf binders of data. In 1950- 51, apparently feeling the a of some income, he took a position as a master in the Santa Cruz Valley School. In the summer of 1951, Al was a member of the Cornell University Cultural Seminar, initiated by Dr. Alexander H. Leighton, which was well described by Bunker and Adair in their book, The First Look at Strangers. His friends, John Adair of Cornell University and Edward H. Spicer of the University of Arizona, were field directors of the 5 week summer course, then in its third year. Adair and Spicer may nave enlisted PWhitng to introduce the international group of students to the Hopis, among y s. The students spent a week each among the Papagos, Navajos and Hopis of Arizona one ie Rio Grande Pueblos and Spanish Americans of Truchas, New Mexico. This was an unusual an especially } the imaginative and skillful techniques Baba ee by Adair and Spicer in introducing fee to the peoples of the Southwest. In learning how to communicate with American Indians with whose language and culture they were unfamiliar, the students of the Seminar were preparing themselves to exchange ideas and information with native peoples anywhere. Participation in the Cornell Crnee- Coluret eccamaips rem to have marked a turning point in Whiting’s career, for by th g p g ngt Chicago to complete his Ph.D. Early in 19521 lied f d da2 tas District Anthropologist for Ponape, Eastern Carolines, U.S. Trust Territory ‘of the Pacific Islands. While he was in Washington, D.C. being interviewed for this position, he renewed his acquaintance with Marjorie Grant, a nutritionist in the U.S. Public Health Service, who had been a member of the Cross-Cultural Seminar the previous summer. Within a few weeks they were married, and soon left for the South Pacific. As far as we know Whiting did not publish anything about his work on Ponape, where he was the third anthropologist for sc Gisiract after brhisbe we) II. A general description of the duties and headaches of such an by J.L. Fischer (1979). A brief statement indicates that Whiting. as wellas others, was often at odds with the policies of the district administrator. In spite of whatever problems he may have had, Whiting was intensely interested in every aspect of his work with the nave people of the island, as attested by his collection of field notes, papers, | hich he presented in 1975 to the National Anthropological Archives of the Smithsonian Pit te where they occupy 7 linear feet of shelf space. Included are diaries recording daily events, correspondence and official reports, answers to questions of persons in the government of the Trust Territory aries 0! economic conditions of the islands; a card file of notes arranged by subject covering forA dist t lands, May 1981 BARTLETT 3 everything from history through material culture, social and legal problems to census ae archaeology, maps, linguistics, negatives and prints; and a collection of books Micronesia in Japanese. A Ponapean language sia an was begun by Whiting wai’ on the island, and several game are in the Sones When their tour of duty on Ponape d Al and Marjorie decided toG for a year, Al to teach and Marjorie to make : a survey for the U.S. Public Health Service. Leaving Marjorie on Guam, Al returned to the States, visiting Saipan and Japan en route. He planned a brief visit in Vermont with his mother and then to return to the South Pacific. However, he — that his nenee wife was s hospitalized and he went to Denver to nes after their 2 young so and his sons went East and he sought a teaching position, which he found in rae High School at Rockport, Massachusetts. Marj d in August, and they rented a large old farmhouse near the school and close enough to Boston for Marjorie to work on her Ph.D. Whiting had at last found a position well suited to his interest and his talents, a museum where he could work with college students and where intellectual curiosity was highly When Whiting arrived at Dartmouth College Museum he found a vast quantity of material pertaining to his department in storage, and many of the specimens lacking documentation. He spent most of his time the first years in reorganizing the storage collections, researching the origins of the specimens, and preparing new exhibits. Soon he offered to guide the museum tours for beginning Sociology classes to introduce the students to physical and cultural anthropology. This led to a weekly lab course in the museum to expose students to the various sub-disciplines in anthropology and to teach museology. He enlisted students and others as volunteer curators, who not only helped in organizing and researching the museum collections, me prepared exhibits and major: shows. In 1961 Al was promoted to th but retained his utle, Curator of Anthropology; 5 years later he became Adjunct Assistant Professor in the Department of Anthropology in addition to continuing as Curator in the Museum. He taught a course ae year, usually an advanced seminar, on a variety of subjects: Museum Methods, Africa Ethnography, Southeast Asia and the Pacific Islands, Primitive Art (which included = development of jazz and blues in America), and Primitive Technology. He supervised students with individual projects covering cultures from the Andes to the Arctic. In all these courses each student was required to prepare an exhibit case, selecting the material, eee * ae _ sila labels, to final polishing of the glass. Whiting never seemed too busy to discuss advice whether on the preparation of an exhibit ora ppeurne matter. His enthusiasm for the work at hand, innovative methods and imaginative techniques of = Sa combined with his kindly manner and his wit, endeared him to students and collea During his years in Hanover, Whiting published a number of book reviews, articles on museology, and articles on Hopi life. Occasionally, he vigaseigien to igs aonwent to spenda summer or a brief vacation to renew old friendships. He made a survey of material culture in the pueblos of Taos and Tesuque for the Museum of New Mexico, purchasing specimens and recording copious notes. Sometimes one or both of his sons accompanied him, but not his wife for they had separated soon after he went to Dartmouth. When the summer of 1974 arrived Al was ready to retire after 19 very demanding years, and left Dartmouth for Arizona. He spent several months in Flagstaff where he renewed his association with the Museum of Northern Arizona where he was appointed Research Ethnobotanist. He purchased a small house with one acre of land at Cornville, in the Verde Valley, Arizona, where the winters are mild and the syeamaces bmi He especially wished to escape the Flagstaff winters, f g g in New England. 4 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 He commuted to Flagstaff and the Museum once a week to work on a revision of Ethnobotany of the Hopi, the third printing having been sold out. During the summers of 1975-1977, he grew experimental plots of corn, beans, squash, and devil’s claw, with seeds he obtained from the Hopi, Havasupai, Apache, and Papago Indians of Arizona. He was studying the genetics of ene vane Bastian os to determine their Felationships: rom Dartmouth, Al brought capacity with notes and photographs containing the results of a considerable part of his research for the previous 45 years. He hoped to prepare for publication, during the pleasant Cornville winters, the numerous manuscripts he had accumulated. All his hopes and plans suddenly came to an end when he was taken il] in the late fall of 1977 and died a few months later. The publications of Whiting that appear in his bibliography represent only a small fraction of his interests. Before his death he arranged to leave his notebooks t to Dr. r. David Seaman of Northern Arizona University, with r the future. Al Whiting had a brithvant, active and creative mind, always far ahead of his manual dexterity. \ was written to his satisfaction, his agile mind leaped ahead to the next project. He simply could not endure the tedium of writing and ieee a text to suit an a ema r and so most of bis major research has never appeared 1 in print. I number of editors but all have given up in despair upon viewing the ‘ ‘completed” work. Al was constantly being lured away from one interesting project to another, although he was very thorough in conducting his studies and collected vast amounts of extremely valuable information which he planned to publish. Other notes were used for teaching and as the basis for museum exhibits. He hada Neneele a senate; which combined with a delightful sense of humor and a talent for words, | and contributed greatly to his success as a peoleasor and museum curator. He also wrote many scientific papers on a variety of subjec The anthropological profession lost a erie and versatile colleague when Alfred F. Whiting died. It is fitting that the Second Annual Conference on Ethnobiology was dedicated to him as well as to his former associate at the Museum of Northern Arizona, Lyndon L. Hargrave. ACKNOWLEDGMENTS We are grateful to Dr. P. David Seaman of Flagstaff who obtained a copy of “Finding Aid to the A-F. Whiting Collection,’’ National Anthropological Archives, Smithsonian Institution, Washington, D.C. April, 1976 Acknowledgment is made to Miss Mary E. Wesbrook, Administrative Assistant, Department of Anthropology, Dartmouth College, with thanks for her assistance in preparing this brief account of Whiting’s years in Hanover, N.H. LITERATURE CITED BUNKER, ROBERT AND JOHN ADAIR 1959. The First 252, in The Uses of Anthropology (Walter Look at Strangers. Rutgers Univ. Press, New Goldschmidt, se Amer. Anthropol. Brunswick, N.]. Spec. Publ. No. FISCHER, J.L. 1979. Government Anthropologists | WHITING, ALFRED ; eoneacuatinil In Harold in the Trust Territory of Micronesia. Pp. 238- S. Colton Papers, Mus. N. Arizona, MS 207-1. BIBLIOGRAPHY OF ALFRED F. WHITING gees Ponape Indian agriculture: I, background. collection, Mus. N. Arizona. Mus. Notes, 10 (2): . Arizona, Mus. Notes 8 (10): 51-53. 5-8. Prose The small herbarium. The preparation, 1937b. Hopi Indian agriculture: II, seed source organization and use of a small botanical and distribution. Mus. N. Arizona, Mus. Notes May 1981 10 (5): 13-16. 1939. Ethnobotany of the Hopi, Mus. N. Arizona, Bull. 15. (Reprinted 1950 and 1967). 1942a. The bearing of junipers on the Espejo expedition. Plateau 15 (2): 21-23 1942b. Junipers of the Flagstaff region. Plateau 15 (2): 23-31. 1944. The origin of corn: an evaluation of fact and theory. Amer. Anthropol. 46 (4): 500-515. 1945. Review: Burbank among the Indians, as told by Ernest Royce and edited by F.J. Taylor. Pacific Northwest Quart. 36 (2): 177-179. 1948. John D. Lee and the Havasupai. Plateau 21 (1): 12-16 195la. ctw : Hopi kachina dolls and their eal by Harold S. Colton. Arizona eee ): 91-92 951b. “The apps cemetery. Arizona Quart. 7 ial 1951c. Mahe in ey living room. Arizona Game 7 (4): 311-31 952. A Kino ape Arizona Quart. 8 (3): 238- Bp 1953a. Nan-Madol en Madol-en-ihmw, mimeo Dept. Education, Ponape. 1953b. The Tumacacori census of 1796. Kiva 19 (1); 1-12. 1955. Review: Indian corn in old America, by Paul Weatherwax. Indiana Mag. Hist. June: 169-170 1958. Havasupai cieue « in the onina. Plateau 30 (3): 5 Review: The Upper retail Indians, by Robert A. McKennan. Dartmouth Alumni Mag., November: ib. Themuseum, ine student and the library. BARTLETT 5 Dartmouth College Library Bull. 3 (1): 1-8. 1959c. Letter to the editor: Abenaki celebration of Roger’s raid. Darmouth Alumni Mag., December: 2. 1960. Letter on Identification. Pract. Anthrop. 7 (5): 240. 1962. To see ourselves: Word from Me anthro- pologist. Pract. Anthrop. 7: 138-14 1964a. Hopi kachinas. ae 37 (1): 1964b. Havasupai. Jn Encyclopedia Britannica. (This was reprinted — the new Britannica format was adopted in 965a. Hopi nocturne. sige 37 (3): 99-105. 1965b. The bride wore white. Plateau 37 (4): 128- 130. 1966a. Catalogues: Damn ‘em -an inter-museum office memo. Curator 9 (1): 85-87. 1966b. The present status of ethnobotany in the Southwest. Econ. Botany 20 (3): 316-325. 1967. Voiceless music in the Dartmouth College Museum. Dartmouth College Library Bull. 7 (2): 58-59. 1970. The current problem of the American Indian and Dartmouth’s...attempt to help. Letter to the editor, Dartmouth Review (7): 16. aby Leaves from a Hopi doctor’s casebook. Bull. New York Acad. Med. 47 (2): 125-146. 1971. Father Porras at Awatovi and the flying nun. Plateau 44 (2): 60-66. 1974. Professor Child’s collection of cider m Dartmouth College Library Bull. 14. (2):54-61. 1977. Hopi textiles. Pp. 413-419, im Ethno- graphic Textiles of the Western Hemisphere: 1976 Proceedings of the Irene Emery Round- table on Museum Textiles, The Textile Museum, Wash. D.C. J. Ethnobiol. 1 (1): 6-27 May 1981 GARDENING AND FARMING BEFORE A.D. 1000: PATTERNS OF PREHISTORIC CULTIVATION NORTH OF MEXICO RICHARD I. FORD University of Michigan, Ethnobotanical Laboratory, Museum of Anthropology, Ann Arbor, Michigan, 48109 ABSTRA\ The ss cept of dchiierate plant fenanacad north of Mexico were not single oc. in the Eastern Instead there were several periods of contact with Mesoamerica which resulted in the diffusion of specific plants into these areas, and they can be grouped into agricultural complexes. The first was the Early Eastern Mexican Agricultural Complex (Gourd Agriculture Complex) arriving in the East before 3000 B.C., resulting in gardens of bottle gourds and pepo gourds. The Eastern Agricultural Complex developed before 1000 B.C. and consists of 2 domesticated plants outside their modern range of diswibution. The Upper Sonoran Agricultural Complex pind in the higher elevations re the Sou thwest arosind 1000. B. Cc. ba esis BOUNCE, Be slightly later with beans. C the East where they were added to gardens and corn fields respectively. The Lower Sonoran Agricultural Complex is found in the more arid regions of the Southwest by A.D. 500 and although it includes several species of beans and squashes, cotton, and amaranth, only Cucurbita mixta and cotton became important outside areas where irrigation was almost mandatory. By A.D. 1000 pacman | aseqanenee valcca? Mexico Fesuleed in ata Late Eastern Mexican Agricultural Complex henopod. In contrast to the East, the indigenous Southwest Agricultural Complex ‘ds after Spanish contact and to date only the devil’s claw is recognized. The Hispanic Agricultural Complex hen Spanish i Scr a transported native tropical domesticates throughout their empire. Chili, tobacc 1 wheat and many garden crops soon were grown a their pre-Columbian 1 range. Each complex was grown initially in different ecological situations and had differential impacts on recipient cultures and subsequent cultural developments. INTRODUCTION In 1944, when Al Whiting published ‘The Origin of Corn: An Evaluation of he = Theory” (Whiting 1944), the archaeobotanical record was i 1 theory. Whiting assessed - sie ie ideas s for the botanical and cultural beginnings P of verification than others, none was sufficient without archaeological plant evident At that time the recovery - prehistoric plant remains was mostly happenstance. In the Southwest, for example, , Jemez Cave by Volney Jones (1935), had been published only in summary form, and Edgar Anderson had just begun to systematize archaeological maize and ethnographic exampies in the Pueblo area (cf. Anderson and Blanchard 1942). Despite th beginnings, sufficient botanical evid as accumulating from archaeological contexts, at least i in the Southwest, that Carter (1945) was able 4 one hypotheses, too, th h ical d. Sites rae as Bat Cave ( NEE and Smith 1949) ee the maize that botanists required from archaeologists to test their ideas. In the ensuing 35 years the recovery of cata remains has become both sophisticated and commonplace. Whiting correctly em that the archaeological record is the supreme measure of theories of the ee of plants and as a consequence hypotheses proposed by Carter and others continue to be subjected to re-evaluation. i 3 May 1981 FORD 7 in place of an emphasi icul which typified the era when Whiting began his | field snaties, attention has turned 0 the intricate crop history of North America north of h a particular subsistence pattern. The intent of this paper, then, is to delineate ‘the crop complexes of the prehistoric United States based upon the ev record and to discuss the implications of their addition to sical. economies. DISCUSSION Prehistoric A “be hieaiee rig gravel + mf hh or crop f an apparent common geographic pein and a mutual association within particular environmental parameters in which the complex developed, although afterward an individual species may experience a separate geograhical distribution and history. The idea for geographical-based complexes originated with Linton (1924), but received continental application by Carter (1945), e Southwest, the Gila-Colorado and Plateau, each with separate origins and routes of diffusion. In addition he distinguished an Eastern Mexican Corridor as a source of agriculture in the East, which diffused to the Plateau, and a West Mexican Cormidor (Carter 1945:12). Although the importance of each area relative to Linton’s and Carter’s theories has changed, nevertheless their insights are apparent in the agricultural complexes previously identified by Ford (1973) and expanded and elaborated upon in this paper. 1) Early Eastern Mexican Agricultural Complex (Gourd Agricultural Complex). Present evidence suggests that the first neers ote F Sets in me United States originated in eastern Mexico, probably diffused across Tex d the major river systems of the Midwest. This complex consists a Lagenaria siceraria, Cucurbita pepo, and perhaps Cucurbita pepo var. ovifera 2) Eastern

A i | 2 ae al mad ele oe \ ee e Chenopodium berlandieri var. nuttalliae t Nicotiana rustica ® Phaseolus vulgaris @ Cucurbita maxima x Cucurbita moschata @ Mexican Dent corn Fic. 6.—Late Eastern Mexican Agricultural Complex: sites in the east with beans, ca. A.D. 1000; Spanish introduced crops. May 1981 FORD 21 A Southwestern spain memeies L } ae | L No indigenous southwestern plant sy domestication in prehistory. Despite the possible domestication of the grain amaranth and tepary bean in the Sonoran Desert region, all archaeobotanical remains are fully domesticated and no antecedent developmental sages ave been excavated. A possible candidate is the devil’s claw cultivated or had its range extended by prehisvoric people. Provosesdce parviflora Woot. and Standl. Rcotie’ a daw pods yield edible seed and y Papago and Moencopi villagers, but Canetwe note Bell (1942:113) argue that its domestication is a historic response by the Pima and Papago to a commercial demand for baskets. Yarnell (1977:872) is less certain because he feels a minimum of several centuries is required to selectively breed plants with larger pods and white seeds. In this instance, no archaeobiological data are available to support or refute either position, but it remains an interesting possibility. Helianthus annuus L. The H for dye and food. Although sunflower seeds have been found in several archaeological contexts, none exceeds an uncultivated, Native Helianthus in size. In the the absence ot contradictory evidence, th I I was brought to the Hopi in historic times. Agave parryi Engelm. Minnis and Plog (1976) have noted the disjunct distribution of agave north of its natural range is correlated with the presence of a nearby archaeological site. They suspect that prehistoric people may have extended its range intentionally or accidently. Archaeological evidence for its utilization at these sites has not been forthcoming, but this does not negate the potentially active ae Southwestern Indians had in spreading this and other species beyond their modern rai Recently, ee wh has eounserated several native southwestern sone species, includin which z 1 Wao tl} = ] 1, | eo > | nee | VW + 4 J P | oe! Ww aait ak may have been encouraged through cae and even jot bea by them. No evidence demonstrates the genetic changes and human dependency associated with plant At present aS h Agricultural Complex has not been demonstrated beyond the devil’s claw. However, additional field research combined with botanical analysis may contribute additional species Hispanic Introductions Field, garden, and orchard crops derive their origin in parts of the United States from Spanish contacts. Early Spanish traders, missionaries, and colonists brought several domesticates native to the New World to regions where they were not grown in precontact times. They also brought many European plants to the Southwest in the sixteenth and seventeenth centuries. Wheat, barley, peaches, apricots, plums, walnut, peas, chick peas, and melons are but a few of the crops adopted by the Indians. Considering the contacts prehistoric Southwestern cultures reportedly had with Central Mexican cultures, it is surprising that the chili pepper, Capsicum annuum L., and tobacco, Nicotiana rustica, were unknown here until Hispanic times. No evidence of chili peppers has ever been found in unambiguous precontact contexts, not even at Casas Grandes. The history of tobacco in the Southwest is more complicated. The native western tobacco, Nicotiana attenuata, pioneers disturbed habitats, arid Pueblo people still collect and smoke it on ceremonial occasions. Archaeologists have shown that it had a number of prehistoric usages, and plant parts an and seeds were collected and stored (Yarnell 1977:871), but morphological analy di The Spanish brought Nicotiana rustica 2 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 early in the historic period. An important archaeological specimen has been identified as N. rustica from the post-1680 reoccupation of the Bandelier cliff dwellings (Volney H. Jones, personal communication). On the basis of this find it appears that rusticg was an early crop accepted from the Spanish and that it continued to be grown and Pueblo Rebellion. It still is planted and used corcmoninlls: in complementary relationship to attenuata. Prior to contact the tribes of the Missouri River drainage and the eastern United States grew only Cucurbita pepo. However, shortly after contact their complement of cucurbits was completed with the additions of C. moschata and C. maxima. A peduncle of Cucurbita moschata was recovered in the early 1700s Historic Period at the Fatherland site in Mississippi (Cutler and Blake 1973:37) and at the Historic Hidatsa Rock Village site in North Dakota (Cutler and Blake 1973:53) (Fig. 6). Cucurbita maxima originated in South America, and is assumed to have been an Hispanic introduction into Mexico, the American Southwest, and the East where the only authenticated find is from Fort Berthold Village, A.D. 1845-1874, in North Dakota (Cutler and Blake 1973:53). The Spanish were also responsible for the introduction of new corn types from Mexico. The large eared Cristalina de Chihuahua, which apparently evolved in northern Mexico (Cutler 1960), was recovered from a probable historic context at Casas Grandes (Cutler and Blake 1974) and northward in the historic Pueblos. Mexican Dent, an ancestor of the modern Corn Belt dent, also first appears in Spanish contact situations at Picuris in the Post-Rebellion deposits (Cutler me Blake ied Mexican Dent had a profound pmpat on the maize of the Rio Grande Pueblos g are a result. The Hispanic Agricultural Complex achieved widespread distribution and was continued during and after the Pueblo Rebellion. New maize types increased productivity and the great array of new annual and orchard crops intensified Pueblo use of arable land and brought relief from failure of prehistoric cultivars. o CONCLUSION n assessment of crop caspase guides of crop association, and the geographical distribution of d I isbandry an future research activities. Th td f the ind lent i luction f Mexico of the first cultivars in the East and the Southwest is less i important than their aod oak into the prehistoric economies. In the East squash and gourd were grown n gardens and supplemented gathered foodstuffs from the forest. In the major river ys starchy annual seeds were collected, and the exogenous origin of agriculture led to the cee eiel of a and sunflower at least. ue the West, into nie and squash and later gourd and common beans Perhaps 1000 years passed before corn became an economic . stable. Even with the establishment of sedentary communities in the Midwest and the spread of corn from the Southwest, an agricultural field system did not evolve for many centuries. Again, there is no evidence that any cultivated species or new race of corn immediately changed the cultural patterns where they were introduced. The sedentary villages of southern Arizona received a number of crops from Mexico, but these were merely added to an established agricultural pattern which had diffused from mountainous areas. What strategies were used for growing these crops and how they system to be explained. To heed Whiting’s appeal to the muerte record, evidence must be obtained to answer these and similar questions. e importance of changing cultural adapations for understanding plant breeding in prehistory i is conspicuous in North America. The achenes of sunflower and sumpweed, for example, increased in size heli after they were oe domesticated, and they may have under ergone their grea atest 1 g the b 54 f field agriculture, 2000 o May 1981 FORD 23 years after their domestication began. Corn demonstrates a similar pattern in both the Southwest and the East. The genetic variability and its adaptive potential was not appreciated by the casual horticultural aoe of hunters and gatherers or even by the Midwestern Woodland cultures with their large gardens. However, as cultural pressures changed, the productivity and adaptability of maize was realized and new varieties were developed in both areas. -D. 1000, with the pone exception of the devil’ s claw, all | prehistoric agricultural crops and complexes were in t were well-established. It was not until the arrival of Europeans that new crops were introduced and aboriginal economies underwent substantial change. LITERATURE CITED Apbovasio, J.M., J.D. GUNN, J. DONAHUE, AND R. United ein ea 41:81-86. STUCKENRATH. 1978. Meadowcroft Rockshelter, BOHRER, VOR . 1962. Nature and 1977: an overview. Amer. Antiquity 43(4): 632- foi eiea _ an nobotanical Materials 651. from Tonto National Monument. Pp. 75-114, ANDERSON, EpGarR. 1945. The Maize Collections in Archaeological Studies at Tonto National from Painted Cave. Pp. 77-85, in Painted Cave Monument, Arizona (Louis R. Caywood, ed.). Northeastern Arizona (by Emil W. Haury). Southwestern Monuments Assoc. Tech. Ser. 2. Amerind Found. Publ 970. Ethnobotanical aspects of 1 Racial identity of the corn from Snaketown, a Hohokam Village in southern Castle Park, Dinosaur National Monument. Arizona. Amer. Antiquity 35 (4):413-430. Univ. Colorado Stud. Ser. Anthrop. 2: 91-92. In Press. Cotton. To appear in Handbook ANDERSON, EDGAR AND F.D. BLANCHARD 1942. of North American =a Vol. 3. Prehistoric maize from Canyon del Muerto. si VorsILA L. AND GLORIA J. FENNER. 1974. Amer. J. Botany 29 (10):832-835. Cotto’ Squashes, Gourds, Uncultivated APPLEGARTH, J.D. 1977. Macrofloral Remains. Poms Pp. 314-317, in Casas Grandes, A Fallen Pp. 40-48, in Excavations at Sparks Rockshelter Trading Center of the Gran Chichimeca, Vol.8 (15 JO 19), Johnson County, Kentucky (by P.T. (by Charles C. DiPeso, John B. Rinaldo and Fitzgibbons, J.M. Adovasio, and J. Donahue). — - Fenner) ee Press Pennsylvania Archaeol. 47 (5). Flags AVID L. AND NANCY B. ASCH 1977. tic gg Te L., HuGH C. CUTLER, AND pod as cultigen: a re-evaluation of some JONATHAN D. SAvER. 1969. Carbonized _ Prehistoric collections from eastern North remains from two Hohokam sites, Arizon America. Midcont. J. Arch. 2(1):3-45. BB:13:41 and Arizona BB:13:50. Kiva 35 (1): "f 0 10. BRAUN, DAvID P. 1977. Middle ee Late and _ Social in e ts Lower Illinois Valley. Pp. 300-341, in The Prehistoric skin has US. Univ. Nature and Status of Ethnobotany (Richard I. Ford, ed.). Mus. Anthrop. Univ. Michigan, Microfilms, Ann Arbor. Anthropol. Paper 6 BROOKS, RICHARD H., LAWRENCE KAPLAN, HUGH 1979. Paper Piceenied at the Annu. C. CUTLER AND THOMAS W. WHITAKER. 1962 Meeting Soc. Amer. Auth Vancouver, Canada. Plant material from a cave on the Rio Zape, ASCH, NANCY, RICHARD I. FoRD, AND DaviD L. Durango, Mexico. Amer. Antiquity 27(3):356- 369 ASCH. 1972. ere dia ny of the Koster Site, the Archaic Horizons. Illinois State Mus. Rep. Invest. 24. BENN, Davip W. 1974. Seed analysis and its auwieals = ¢ Pes Ee eS ee re BROWN, WILLIAM L. 1974. The History of Maize. Review of: Corn, its origin, evolution, and improvement, by Paul C. Mangelsdorf. Science 5: 687-688. BROWN, WILLIAM L. AND EDGAR ANDERSON. 1947. The Northern Flint Corns. Ann. Missouri BLACk, MEREDITH 1963. The distribution an Bot. Garden 34(1):1-28 CARTER, GEORGE F. 1945. Plant Geography and in South Dakota. Plains Anthropol. 19(63):55- 72. archaeological significance of Elder, is annua L. Papers Michigan Acad. - ie Culture History in the American Southwest. Arts and Letters 48:541-547. Viking Fund Publ. Anthrop. 5. CASTETTER, EDWARD F. AND WILLIS H. BELL. 1949. BLAKE, S.F. 1939. A new variety of Iva ciliata from I Pima and Papago Indian Agriculture. Univ. ndian rock shelters in the south-central 24 JOURNAL OF ETHNOBIOLGOY New Mexico Press, Esco CHAPMAN, JEFFERSON, ROBERT B. STEWART, AND RICHARD A. YARNELL ss iii cia evidence for Precolumbian introduction of Portulaca oleracea and Mollugo verticillata into eastern North America. Econ. Botany 28(4):411-412. CHOMKO, STEPHEN A. AND GARY W. CRAWFORD 1978. Plant husbandry in prehistoric eastern ais America: new evidence for its develop- ent. Amer. Antiquity 43(3):405-408 Pacer C.B. 1947. Caves of the Weer Giisand Hueco Areas in New Mexico and Texas. Papers Peabody Mus. Amer. Arch. Ethn., Harvard et 34 (2):1-181. , C. WESLEY 1978. Prehistoric Use a siaibasice of Maygrass in Ea America: Cultural and Ph of Ethnobotany (Richard I. Ford, ed.). Mus. Anthropol., Univ. Michigan Anthropol. Paper 67. Cuter, HuGH C. 1952. A Preliminary Survey of Plant Remains. Pp. 461-479, in Mogollon Cultural Continuity and Change, the strati- graphic Analysis of Tularosa and Cordova Caves (by Paul S. Martin et al.) Fieldiana- Anthrop. 40 .1956. Vegetable Material from the site of San Cayetano. Pp. 459-462, in Upper Pima of San Cayetano del Tumacacori toy Charles C. DiPeso). Amerind Found. Publ. 1960. Cultivated plant remains from belo an: Chihuahua. Amer. Antiquity 26(2):27 git raat P. 46, in Basket Maker HI Sites near Durango, Colorado (by Roy L. Carlson). Univ. Colorado Stud. Ser. Anthrop. 8. . 1965. Cultivated Plants. pp. 107-109, in The McGraw Site, A Study in Hopewellian Dynamics (by Olaf H. Prufer). Sci. Publ. Cleveland Mus. Nat. Hist., New Ser. 4(1). . 1966. Corn, Cucurbits and Cotton from Glen Canyon. Univ. Utah Anthropol. Papers 80: 1-62 CUTLER, HUGH C. AND GEORGE A. AGOGINO. 1960. Analysis of maize from the Four Bear Site and two other Airkara locations in South Dak Southwestern J. Anthrop. 16(3):312-316. AND LEONARD W 1969a. Analysis of Corn from San Juan Capiatnitio. Pp. 107-109, haeology of Mission Vol. 2 (by Mardith K. Schuetz). State of Texas, State ager. Commission, Archeol. Prog Rep. | eatin op salle Corn from Cahokia Sites. Pp. 121- 136, in Explorations in Cahokia Archaeology Vol. 1, No. 1 (Melvin L. Fowler ed.). Illinois Archaeol. Survey Bull. 7. 1973. Plants from Archaeological Sites East of the Rockies. Missouri Bot. Garden, St. Louis. . 1974. Corn from Casas Grandes. Pp. 308- 314. in Casas Grandes, A Fallen Trading Center of the Gran rer (by Charles C. DiPeso, John B. naldo, and Gloria J. Fenner). Northland Sis Inc., Flagstaff. . 1976. Corn from Snaketown. Pp. 365-366, in The Hohokam ed Emil W. Haury). Univ. Arizona Press, Tuc CUTLER, HuGH C. AND WINTON MEYER. 1965 Corn and Gacurtss from Wetherill Mesa. bea: 136-152, in C i Archeological Project reg is ‘i! Douglas Osbourne CUTLER, BiugH c. Ps Taowas w. ‘Wrarranen 1961. H Cubiahits 3 in the Americas. Amer. Antiquity 26(4):469-48 DICK, saves? W. 1965. Bat Cave. School Amer. Res. Monogr. 27. DiPEso, CHARLES C. 1953. The Sobaipuri Indians t an; River Valley, pene Arizona. Amerind Found. Publ. peat Gee BRYAN J. 1977. A report on the floral remains from two archaeological sites in western Ohio. Paper presented at the 53rd Annu. Meeting Central States Anthropol. Soc. Cincinnati, Ohio. FERGUSON, LELAND G. 1978. Southeast. Pp. 756- ; mas P. Myers, Forp, RICHARD I. 1973. Origins of Agriculture in Native North America. Actes du VIII Congres International des Sciences Prehistoriques ¢t Protohistoriques Actes HI: 406-407. 74. Northeastern archaeology: past and future directions. Annu. Rev. Anthrop. 3:385- 413 5. Re-excavation of Jemez Cave, New Mexico. Awanyu 3(3):13-27. 977a. Evolutionary Ecology and the Evolution of Human Ecosystems: A Case Study From the Midwestern U.S.A. Pp. 153-184, in A Explanation of Prehistoric Change (James N. Hill, ed. ie Univ. New Mexico Press, Albuquerq ; 197. hichaedtivibiy of the Fort Bliss Maneuver Area II, Texas. Pp. de yar A in Settlement Patterns of the Bolson (by Michael E. Whalen). Centenial sank Anthropol. Paper 4. . 1980. Gathering and Gardening: Trends and Consequences of Hopewell Subsistence Strategies. Hopewell Conference, Chillicothe, May 1981 Ohio, 1978. Kent State University Press. GALINAT, WALTON C. AND JAMES H. GUNNERSON. i o r historic Southwest. ape Leaflets, Harvard Univ. Rc 117- Mus. par NAT, W.C. AND R.]J. stig PPE. 1961. seth archaeological evidence on the effects Teosinte introgression in the evolution ” ern maize. Bot. Mus. Leaflets, Harvar Univ. a 163-181. NGELSDORF, AND L. PIERSON. 1956. Esimats of Teosinte introgression in arch- gical maize. Bot. Mus. Leaflets, Harvard ous — 101-124. D THEODORE R. REINHART, AND Teen R ae. Lai digs ). Early cight-rowed 4. VV. Valley, New maize from th Mexico. GASSER, ROBERT E. 1976. Hohokam Subsistence: A 2,000 Year Continuum in the Indigenous ee of the Lower Sonoran Desert. Archeol. Report USDA akaitiy Service Southwestern Region — GILMORE, MELVIN R. 1981. Vegetable remains of the Ozark Bluff Dweller Culture. Papers Michigan sige Sci., Arts and Letters 14:83-102. HANSON, LEE Jr. 1966. The Hardin Villes Site. i Kentucky Press Stud. Anthrop. 4. Haury, EMI W. 1957. An alluvial site on the San Carlos Indian Sesreneoety Arizona. Amer. Antiquity 23(1):2-2 1976. The se Univ. Arizona Press, Tucson HEISER, CHARLES B. AND DAviD C. NELSON. 1974. On the Origin of the Cultivated Chenopods (Chenopodium). Genetics 78:503-505. IRWIN, H.J. AND C.C. IRWIN. 1959. Excavations at e LoKaisKa Site in the Denver, Colorado, Area. Proc. Denver Mus. Nat. Hist. 8. IRWIN- wee ~ AND S ree KINS. 1968. Excavations a at En Medio Shelter, New Mexico. City Hopewell Subsistence-Settlement System. “pill 7-15, in Hopewellian Archaeology in =| ens Jones, eae H. 1935. Pp. 61-64, in Report on the Excavation of Jemez Cave, New Mexico (by Hubert G. Alexander and Paul Reiter). Monogr. School Amer. Res. . 1936. The Vegetable Remains of Newt Kash Hollow Shelter. Pp. 147-165, im in Menifee County, Kentucky (by WwW Funkhouser). Univ. — — Arch. and Anthrop. 3. 942. Pp. 31-33, in Pima and Papago FORD 25 mee (by Edward F. Castetter and Willis H. Bell) Univ. New Mexico Press, Albuquerque. « WS) Plant River, Aaneiiin’ Found. Publ. Material from _ the utheastern Arizona. 1968. McKees Rocks Village Site. Pennsylvania hadaicisk $8(1-4):81- Bs i VOLNEY H. AND ROBERT L. FONNER. 1954. Maker II Sites near Durango, Colorado (by Ear] A Morris and Robert F. Burgh). Carnegie Inst. Washington Publ. 604. KAPLAN, LAWRENCE. 1956. The cultivated bean of the prehistoric Southwest. Ann. Missouri Bot. Garden. 43:189-251. lg Hone of Cordova Cave. Econ. Botany 17(4):350 .1963b. Beans. P. 47, in cae Maker III Sites near ce Colorado (by Roy a Univ. i ean Ser. - Anthro ara pene (beans). Econ. Botany 19:358- Beans of Wetherill Mesa. Pp. 153- 155, in Contributions of the Wetherill Mesa Archaeological Project (assembled ai Douglas Osborne). Mem. Soc. Amer. Arch . 1970. Plant Remains from rts hin Site. and Orrin C. Shane III). Kent State Univ. Press, Kent. . 1971. Phaseolus: pores agy and a of Origin. Pp. 416-427, (C.L. Riley, J. Charles i. C.W. vee an d R.C. Rands, eds.). Univ. Texas Press, Kay, MARVIN. 1979. Phillips Springs, Missouri: Report of the 1978 Investigations. Report of the S y Corps of Engineers Kansas City District. KING, FRANCES B. AND R. BRUCE MCMILLAN. 1975. Plant Remains from a Woodland Storage Pit, Boney Spring, Missouri. Plains Anthropol. 1 LINTON, RALPH. 1924. North American maize culture. Amer. nae 26:345-359. archaeological investigations in the Sierra de i Amer. Philosoph. Soc., n.s. 48(6):1-210. 26 —___. . 1971. Archaeological Synthesis of the Stevia: Pp. 573-581, in Archaeology of Northern Mesoamerica, part 2, (Gordon F. Ekholm Ignacio Be Handbook of Middle American Indians 11. MAGERS, PAMELA C. 1975. The cotton industry at Antelope House. Kiva 41(1):39-47. MANGELSDORF, PAUL C. 1974. Corn, Its he Evolution and Improvement. Harvard U Press, Cambridge. D WILTON GALINAT. 1966. “ibesiripaide: of Burned Corncobs. Pp. 43-46, in Excavation at Rainbow House, Bandelier National Monument, New Mexico (by Louis R. Cayw Natl. Park Service, Southwest Archaeol. Center, Globe. OBERT H. LISTER. 1956 ~ Aicacologca eeaonaserh on = nn of Maize ico Bot. Mus Leaflets a. oh 178, DE; E SMITH, JR. 1949. New Archaeo steiea Biden on Evolution in Maize, Harvard Univ. Bot. Mus. Leaflets a HERBERT ne DICK, AND JULIAN CAMERA- HERNANDEZ. 1967. Bat Cave Revisted. Harvard Univ. Bot. Mus. bee 22:1-31. D S. MACNEISH, AND GORDON R. ~ WILLEY. 1964. Origins of Agriculture in Middle Am Pp. 427-445, in Environments and Early Cultures (Robert C. West, ed.). Handbook of Middle American Indians 1. MARQUARDT, WILLIAM H. AND PATTY JO WATSON. 1977. Current State Research: Kentucky Shell- South- Arizona’s oldest 132(3418): 33-34. MINNIS, PAUL E. AND STEPHEN E. PLOG. 1976. A study of the site specific distribution of Agave mA rryi in east central Arizona. Kiva 41(3-4): 299-308. cornfield. Science MURPHY, JAMES L. 1971. Maize from an Adena oun thens County, Ohio. Science 171(3974):897-898. NABHAN, GARY PAUL. 1977. Viable seeds from rape eine Archaeobotanical remains Southwestern folklore. Kiva 43(2):143-159. ieee penton: HARD S. FELGER. 1978. oo in Southwestern North America—a biog: graphical and_ ethnohistorical study an Phaseolus acutifolius. Econ. Botany 32(1):2- 19 OzkeR, DorEEN B.V. 1977. An Early Woodland Community at the Schultz Site in the Saginaw Valley and the Nature of Early Woodland Adaptation in the Great Lakes Region. JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 Unpubl. Ph.D. Dissert. Michigan, Ann Arbor PAYNE, WILLARD W. AND VOLNEY H. JONES. 1962. The aniggien a and Archaeological ed from (Anthrop.) Univ. Signif of Giant Ra caine Bluff Sickie in the Ozark Plateau. Papers Michigan Acad. Sci., Arts and Letters 47:147-163. ROBINSON, SARAH. 1976. An Analysis of Charred Site in Central Missouri. Pp. 100-109, im Hopewellian Archaeology in the Lower Missouri River Valley ioe E. Johnson, ed.). Univ. Kansas Publ. Anthro SAUER, Reerccontyh D. 1969. eitiy of archae- ological grain Amaranths from the Valley of Tehuacan, Puebla, Mexico. Amer. Antiquity 34(1):80-81 AND LAWRENCE KAPLAN. 1969. Cana ns in at 3 prehistory. Amer. pare set 424, SHANE, “ie 1967. The Leimbach Site. Pp. 98- 120, in Studies in as Archaeology (Olaf H. Prufer and Douglas H. McKenzie, eds.). Press Western Reserve Univ., Cleveland. SMITH, C. EARLE, JR. 1950. Prehistoric Plant Remains from Cave. Harvard Univ. Bot. Mus. Leaflets 14:157-180. STAINS, DONNA. 1972. Seed Analysis: The Brewster Site (13 CK 15), Western Iowa. Unpubl. Thesis. (Anthrop.), Univ. Wisconsin, chee on. STRUEVER, STUART AND KENT D. VICKERY. 1973. The beginnings of reddele in the video riverine area of the United States. Amer. Anthropol. 75(5):1197- te Watson, PATTy Jo (EDITOR). 1974. Archaeology the Mammoth Cave Area. Academic Press, New York. 1969. The Prehistory of Salts Cave, ~ Kentucky. Tilinois State Mus., Reports Invest. 16. WEDEL, WALDO R. 1943. Archaeological Invest- gations in Platte and Clay Counties, Missoum!. Bull. U.S. Natl. Mus. 183 Laeger ALFRED F. 1944. The origin of corn: evaluation of fact and theory. Amer. elec 1. 46(4):500-515 WIMBERLY, Mark L. AND PATER L. EIDENBACH. 1972. Tularosa Valley Project, 1972 Excavations at Fresnal Shelter. Training Bull. Human Systems Res. Inc., Albuquerque, New Mexico. WINTER, JosePH C. 1973. The distribution and mfg ro of Fremont maize agriculture: some preliminary interpretations. Amer hata oe 439-451. RIGHT, HENRY T. 1964. A transitional Archaic porn te at Green Point (20SA1). Michigan Archaeol. 10(1):17 May 1981 ARNELL, RICHARD ASA. 1964. Aboriginal Relationships between Culture and Plant Life in the Upper Great Lakes Region. Anthropol. Papers, Mus. Anthrop., Univ. Michigan 23. 1972. Iva annua var. macrocarpa: ren aiecaas cultigen? Amer. Anthropol. 74:3 341. . 1976. Early Plant Husbandry in Eastern North America. Pp. 265-273, in Culture Change and Continuity (Charles E. Cleland, ed.). FORD 27 Academic Press, New York. .1977. Native Plant Husbandry North of ition. Pp. 863-875, im The Origins of Agri- culture (Charles A. Reed, ed.). Mouton, The Hague. . 1978. Domestication of Sunflower and Michigan Anthropol. Paper 67 J. Ethnobiol. 1 (1): 28-38 May 1981 A CRITICAL VIEW OF THE USE OF ARCHAEOLOGICAL VERTEBRATES IN PALEOENVIRONMENTAL RECONSTRUCTION ALD K. GRAYSON University of nai goal Department of Anthropology Seattle, Washington 98195 ALDI LI BRALL. bd cake J lente ere c W relied either up pie p or have oe - oe 3 c J C. J hL T e Ss a “2 © eee, a. 1 1 th sy nt< of é ae & o , p | 1, + = = | c. L 2. and oOo | 4 Le KR A. < 1 ye | > 1 £. 1 & a PIEKwMcostys t collection, h ll 1 . l L l Sr hae se Se t b 1 7 hd ~~ 2 L LL t if J 4 LK | Me an nd 2) we ora if ever, have any notion of the relationship between the quantitative structure of ite inferences, aA th .. 1 eee. | = Poe ae fae , eo eae er | ly studi rs hich treat taxa as attributes, and not as cabbies can routinely be treated as valid. Other major difficulties presented by archaeological vertebrates in paleoecological reconstruction are reviewed, with special emphasis placed upon hazards encountered in presence/absence studies. INTRODUCTION During the past few decades, b ins f haeological sites have b an increasingly popular source of information about past environments. It is easy to demon- strate this i ICC aRIIN popularity. Assuming mina the sample of paleoclimatic | literature, i Grayson (197. paleoclimatic ‘literature are representative of trends within the Sea anit literature as a whole, this bibliography may be used to assess the changing role of archaeological vertebrates in the analysis of past environments in North America. The bibliography lists no archaeological vertebrate faunal studies conducted for paleoclimatic purposes prior to the 1931- 1940 decade. Between 1931 and 1940, 2; 8% of the published paleo- climatic studies mad to 1.0% between 1941 and 1950, then increased to 4.0% between 1951 and 1960, increased again to 4.3% between 1961 and 1970, then i y years of the 1970’s to 10.7% (Table 1). While these data are North American, paleoclimatic literature from other parts of the world seems to exhibit the same trends, although the use of archaeological vertebrates in paleoenvironmental reconstruction began much earlier in the Old World. Given the increasing interest in the use of archaeological vertebrates as a source of information about past environments, it is interesting to note that the critical literature concerning ones studies i is quite small. While vertebrate paleontology has had a critical literature on j for well over a century (e.g., Dawkins 1869: Owen 1846), and this literature is rapidly becoming quite large (e.g., Behrensmeyer 1975; Shotwell 1955, 1958, 1963; Voorhies 1969; Munthe and McLeod 1975 and references contained therein), examination of the principles and processes of serene reconstruction g see Findley 1964). The lack of such a critical literature might suggest paleoenvironmental reconstruction using archaeological vertebrates i is aeonenenery straightforward, and can be conducted with little concern for y g the truth, as this paper demonstrates. oS 4 May 1981 GRAYSON 29 DISCUSSION Basic Approaches Of the several approaches to paleoenvironmental reconstruction which have been employed using archaeological vertebrates, 2 characterize the vast majority of the literature. In the bcd of _ bes aysouteneee the toa present in =n archaeological fauna are identified, and th I environments of th i th he f Guilday and Adam 1967; Guilday and Parmalee } 1972; Parmalee and Oesch 1972). In the eae approach, each taxon is treated not as an attribute which can be either present or absent, but as a variable whose abundance can vary discretely. In studies which treat taxa as variables, some measure of taxonomic abundance is employed to derive quantitative statements about the relative abundances of all taxa present (e.g., Bate 1937; Butler 1972; Grayson 1976, 1977b; Harris 1963). These 2 approaches are examined in detail here. Taxa as Variables It is not hard to see that treating taxa as variables holds a greater potential for providing paleoenvironmental information than does treating them as attributes. Let us say, for instance, that we are studying ore history of : Move owired ecosystem heise includes — 2 mammals, taxon A and taxon B. T b and only with fluctuations in temperature: when it gets hotter, A increases while B decreases, and vice versa. Let us assume we have a fauna which contains a sample of A and B which i is } Fepre- sentative of th g g the past 1,000 years. Analysis of this fauna shows both A and B have been p during thi entire period of time (Table 2a). All that can be inferred f his ob ion is th ature minima and maxima have not exceeded the tolerances of either taxon during “the period represented. Further analysis, however, shows the abundances of taxa A and B have fluctuated widely through time. Because the sample is representative of the environment when the sample was accumulating, and because abundances of these animals vary with temperature fluctuations, some fairly detailed statements can be made about temperature in the sampled area during the past : ,000 years — for instance, time periods 4 through 8 were much warmer than th li d (Table 2b). Clearly, treating taxa as variables holds the promise of providing much more detailed information on past ale ieacoe ieee than treating taxa as attributes, for the simple reason that presence/absence “i ] c . L a ri é 3 Sa hd 4q } anominal scale. When , only basis of paleoenvironmental inference. When taxa are treated as variables, fluctuations in abundances of each taxon, or of groups of taxa, become an additional target of aay. It can be argued, therefore, tl taxonomic abundance are preferred over those which treat taxa as presence/ absence attributes. Unfortunately, one can argue even more forcefully that paleoenvironmental Studies based upon counts of taxonomic abundance are not likely to provide demonstrably valid data about past environments (and here I use the term valid in its statistical sense: are we measuring what we think we are measuring?). There are 2 reasons for this: the nature of counts of taxonomic abundance, and, the nature of the faunal sample itself. There are only 2 measures available for quantifying the abundances of taxa represented within an perhecnnen: site: counts # mcaaaceia Specimens per taxon (NISP; in earlier publicati taxon (MNI; see Casteel 1978 and pt 1979 for a discussion of meat weights as abundance measure). I have treated these units at length elsewhere (e.g., Grayson 1973, 1978a, and esp. 1979), and will not repeat those discussions here. Ati is, however, necessary to to point out that NISP and MNI are similar in an i number provided by either measure and the actual number of animals which contributed 30 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 1.—Numbers of paleoclimatic studies in North America using vertebrate remains from archaeological sites (data from Grayson 1975) UDIES EMPLOYING DECADE ALL STUDIES (a) ARCHAEOLOGICAL VERTEBRATES (b) b(100)/a 1881 - 1890 2 0 0.0 1891 - 1900 ] 0 0.0 1901 - 1910 2 0 0.0 1911 - 1920 7 0 0.0 1921 - 1930 $2 0 0.0 1931 - 1940 144 4 2.8 1941 - 1950 200 2 1.0 1951 - 1960 302 12 4.0 1961 - 1970 559 26 47 1971 -.(*) 149 16 10.7 TOTALS 1398 60 (*): while the | fe in Gray (1975) dated 1974 g bl i decad tends only to 1973. TABLE 2.—An example of an archaeological faunal analysis in which taxa are treated both as attributes (Table 2a) and as variables (Table 2b). See text for explanation a: TAXA AS ATTRIBUTES b: TAXA AS VARIABLES (numbers represent absolute abundances) TAXON TAXON A B A B 1 x x 1 10 90 Z x x 2 10 90 =) 3 x x 3 10 80 g 4 x x 4 40 30 | 5 x x 5 60 20 6 x x 6 80 10 = 7 x x 7 50 05 ram 8 x x 8 40 05 9 x x 9 10 60 10 x x 10 10 80 (x - taxon recorded as present) G7 eT ee bones to the collection neem study i is, in all but trivial instances, unknown. That is, the meaning of both th y clouded because the relationship between estimated and actual abundances must always be unknown. For example, time period 4 of the site presented in Table 2b eee abundances of “40” for taxon A and “30” for taxon B without mention of wh e is being used. This is clarified by noting that th ini f individuals defined from 450 identified specimens ory A and 550 identified specimens of B. Knowing this, it is no longer clear what the actual abundances of A and B really are. It is possible to consult the voluminous literature on this point (see Grayson 1979 for areview), and recount the ——— for and against NESE and MINT as abundance measures. Instead, " can be counts and the number of animals which originally contributed to the collection is unknown and unknowable. Unfortunately, it is the original number which is the target of our estimates: May 1981 GRAYSON 3] There is, for i a 08 reason ragies the er number of animals deposited i in our. site could not have been B.C “60” and “80” are an accurate soso of the abundances of these taxa in the environment surrounding the site, andt ,it is not hard to see how misleading are the NISP values (450 for taxon A and 550 for taxon B), or the MNI values (40 for taxon A, 30 for taxon B). Because there is no way of working back from an excavated collection of bones to when it was deposited, and because there is no way of relating counts of identified specimens or minimum numbers of individuals to the number of animals which contributed to the faunal sarin neither NISP nor MNI isa reasonable quantifier of taxonomic abundance in t & case. We simply « do not oe and cannot know, what the counts they provide mean in t nce there are no other ways available for counting abundance in this setting, it is clear ore one of the bases for analyzing the taxonomic abundances of archaeological vertebrates is very weak indeed But there is another, even more damaging, problem involved in the paleoenvironmenta! use of archaeological vertebrates. In the example above, it has been assumed the sample of animals deposited in the site was representative of what was living in the area at the time the sample was accumulating. Unfortunately, the relationship between the archaeological collection and the actual population — the set of animals living in the area 6 chen time the archaeological sample was being deposited (the Boyais population”) — is own, except that the animals in the collection probably ca efrom tha kane: As with ‘ihe relationship rains as anal MNI and acta abundances, the rela p between which population is usually unknown and unknowable. This presents an insurmountable " difficulty for using ind Gang abundances of taxa within an archaeological site as a key to past environm The problem seems an obvious one. To continue with the example etal. I shall drop the unrealistic assumption that the collection under study accurately represents the abundances of taxa A and B in the surrounding environment at the time the fauna was accumulating, and note instead that the num bers g the site probably had more to do with the mechanism of accumulation than with the actual abundances of those taxa in the sampled area. The abundances of taxa A and B may have been in the ratio of 100 to 1 in the sampled environment, bese if the acc comnslatieiee tnechanisen sampled taxon B almost to the exclusion of taxon A, then taxon A and 80 individuals for taxon B (as noted above) are entirely possible. Is it unreasonable to emphasize that the risesiicuren ee the target population and the archaeological fauna is unknown? Can that that it becomes inappropriate in most cases to derive paleoenvironmental information from taxonomic abundances? A simple example serves to demonstrate the problem is, in fact, a severe one. People are just one in a set of mechanisms which accumulate vertebrate remains in pete tay sites. Other organic accumulation mechanisms include a variety of non- an predators and scavengers (see, for instance, Butler 1972; Guilday and Parmalee 1972; pratt 1960; Mellet 1974). It = obvious that predators and es, including people, cannot be relied on to environment. The saabnc introduced as a result of these varied accumulation mechanisms may be seen by examining the behavior of nin a class of predators whose peaepere a behavior seems simple compared to that of human e predation patterns of owls have been particularly well rida Maser et al. (1970), for instance, studied the food habits of 3 aires ot owls i <4 — Oregon eho = Horned ) Owl (Bubo virginianus), Short-earedO ong-ear 1 These authors gathered and analyzed 24 sets of ont pellets from these Species between February and July, 1969. With the e om areas adjacent to o springs (one each from B. virginianus and A. otus), ll coll were tats “similar in all areas’’ (Ibid. 1970:4). 32 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 TaBLe 3.—Mammalian contents of pellets from 3 species of owls (from Maser et al. 1970). BUBO VIRGINIANUS ASIO OTUS ASIO FLAMMEUS MNI % MNI % MNI % Peromyscus maniculatus 43 32 20 20 65 19 Microtus montanus 25 19 4 4 37 1] Thomomys talpoides 21 16 6 6 94 27 Perognathus parvus 19 14 64 63 101 29 Reithrodontomys megalotis 10 8 3 3 22 6 gurus curtatus 7 4 4 14 4 Neotoma ciner 3 F 4 0 0 0 0 Dipodomys ordii 3 2 1 1 16 5 Spermophilus beecheyi 1 1 0 0 0 0 S. townsendii 1 1 0 0 0 0 Totals ‘ 133 100 349 99 102 101 TABLE 4.—Modern ow! pellets as archaeological strata. Data from Table 3; see text for explanation. STRATUM 1 2 3 Microtus montanus and Thomomys talpoides 46 10 13] Perognathus parvus and Dipodomys ordii 22 65 117 Stratum 1 = Bubo virginianus (from Table 3) Stratum 2 = Asio otus (from Table 3 Stratum 3 = Asio flammeus (from Table 3) x2 values: | - 2: 44.15 (p 01) 2 - 3: 36.20 (p .O1) Some of the data from these 24 pellet collectidns are presented in Table 3, which displays the number lian individuals f h owl speci Iculated identified skulls and mandibles (C. Maser, personal communication). Rather than treating these data as synchroni i las havi lated time, the lets f B. virgimianus accumulating first, then those from A.otus, and finally situation is not far-fetched; several species of owls can be f y given pa (Bent 1938; Marti 1974), and shifts in the use of roosts by owls can be readily observed today. A stratified faunal collection from the pellets of 3 species of owls, all of which hunt the same habitat, has now been constructed. Using this collection of pellets as the basis of ; ues fp ‘ ee ee inferencec p those from A. flammeus. Such a di : h of habitat following the lead of the archaeological literature (Butler 1972; Grayson 1977b; Harper and Alder 1970), and by considering the Montane Vole (Microtus montanus) and the Norther? Pocket Gopher (Thomomys talpoides) as indicators of mesic environments, and the Great Basin Pocket Mouse (Perognathus parvus) and Ord’s Kangaroo Rat (Dipodomys ordii) aS indicators of xeric environments. } ‘ - a 4 q ‘4 May 1981 GRAYSON 33 [es | pO Hake fe ee t ] i h .. q ra é 4 H r) Lh sg: y between strata, with greater numbers of xeric rodents in stratum 2 than can n be accounted for by chance, a fewer number of mesic rodents than any hypothesis of randomness would allow in that stratum (Table 4). The conclusion of such an analysis would be: stratum 1 accumulated during a time of relatively high effective fies qononnere! producing a greater abundance of mesic habitats, while stratum nz abundance of xeric ¢ habitats. Stratum ree in turn, saw a return to conditions eatin those of stratum | times. Yet, all that has happened is that I have constructed a fauna using modern data under the reasonable assumption that different species of owl -_ ——— time, use the same roost. That is, it is assumed th hange through time. What this analysis has detected is not environmental change through time, but different predation eis “9 a set of sympatric owls. It is perhaps, one of the reasons why these owls can be sympatr The relationship between owl pellet forced one. It i of vertebrates from ow! pellets. But more important is the fact that owls, and other non- human predators and scavengers, represent one of a myriad of accumulation mechanisms which account for the deposition of bones in archaeological sites, and that changes in the accumulation mechanisms lead to changes in composition of the fauna which faunal auahynte iienately aay Clearly, any prlonememommenae analysis of archaeological the taxonomic abundances wees characterized ieee living community at } the time a fauna accumulated, and: the Tt .. 1 I faunas is not a and f. eee RE upon the accumulation mechanisms. Because those mechanisms are rarely known, ihe relationship between the population of animals in th be known with any precision. This is true even when problems relating to \oagargsem preservation of deposited materials are set aside. As a result, the validity of an paleoenvironmental reconstruction based on counts of abundance must always be in question. There are 2 reasons, then, why paleoenvironmental reconstruction based u quantification of azardous. First, the units available for counting eenenrsanig ee — ~ understanding of € processes w transform a d that the numbers provided by those units have much relationshi h ber of animals in the original pile. Second, we rarely, if ever, have any notion of the relationship between the quantitative structure of the target population, from which the sample was drawn and about which we are trying to make inferences, and the archaeological sample. Because of these problems, it is rare that counts of taxonomic abundance can tell us anything about known environmental parameters. If this is the case, taxa should be treated as attributes which can be either present or absent, rather than trying Pp eee ; . as if they necessarily p ning r This position is very similar to that taken by Sir Richard O ( ) y ago: Th Receiasle cit me ng dee ey, Pe eo as ok e eee ity of hi ‘+. + em 3 1 = a mel een —_ from the quantity of human bones in its oo RK ensmeyer Wolff 1973). ci si accept i as a working eae that the solution oe snesscoher 34 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 problems is logically prior to the use of archaeological and paleontological vertebrates for the extraction of paleoenvironmental information It is difficult to disagree with this principle. Vinee it would be optimistic indeed to think that taphonomic approaches will ipa ane ‘sufficiently refined to allow the easy Returning to the modern owls, and the fauna which they provided, makes this point more forcefully. As Maser et al. (1970) point out, fresh owl pellets from their study area were easily recognized as they were whole, held together by shiny mucous covering. After about a month, the pellets were rapidly disintegrated, in part because of the activities of a tineid moth which feeds on — sonar dante: on the pellets, and what pines is not = owl pellet, but th pe archaeological sites, there are 2 immediate difficulties: firstisa a need to recognize bones that were once part of an owl pellet; second, in order to establish continuity in accumulation mechanisms, there is a need to recognize that all owl pellets did, or did not, come from the same species of owl. [f these difficulties could be solved and bones could be separated from pellets of different and known species of owls, there would still be no usable information about taxonomic abundances except that a given taxon was present. This is true because owls take non-random samples of what is in the environment (Errington et al. 1940; Marti 1974). In the study examined here, Maser et al. (1970: 5-6) note that “although the deer mouse is generally the most in central Oregon « « -. the owls s caught farn — pocket mice than deer mice.” In other words, iti Maser et al. (1970) to the abundances of the captured mammals i in the hunted environment. Since that is the case, if would clearly be impossible to do so _ the archaeological eer Indeed, it may even be difficult to use th to the number of individuals which were eaten to produce that pellet (Raczynski aa Ruprecht 1974). The only reliable information about the local mammalian population in either the modern or the archaeological setting is the simple observation that since a taxon was present in the faunal collection, it was Ce present in the immediate. vicinity. Owls have been used are well-studied. The behavior of other raptors (Craighead and Craighead 1956), and of carnivores, wood rats, streams, and other mechanisms — including people — which accumulate faunas are no less complex. “4 matter how become, th W hich th ycan no answer: what i is the relationship between the abundances of taxa in an accumulating fauna and the abundances of those taxa in the surrounding environment? This is no criticism of the taphonomic amu as taphonomists do not have this goal in mind. It is, however, 4 pain din ae ia f. om fe PE : L ie : 1 1 * 1 faiinas which requires something be known of the relationship between those abundances and the abundances of the animals in the area from which those faunas were derived. Taxa as Attributes Since the paleoenvironmental meaning of taxonomic abundances from single archaeological faunas can never be known, presence/absence studies become the only arson approach to the paleoenvinoliniental analysis of those faunas. studies are actually quite simple; in fact, it is this simplicity which accounts for aoe of their value. In presence/absence faunal studies, one simply identifies what is present in a fauna and interprets the paleoenvironmental meaning. Even if abundances ar calculated, as they usually are, they are not interpreted (e.g., Guilday and Adam 1967; Parmalee and Oesch 1972). Instead, the attributes of the represented animals are used as the basis for statements about ition at the time of deposition. Guilday and Adam (1967) provide a good exmple of Ae | a study. After noting the presence of the collared lemming, Dicrostonyx, at Jaguar Cave, southern Idaho, they note this animal is “an obligatory tundra form with a long evolutionary association witha — May 198] GRAYSON 35 boreal climate’’ (1967:29), whose presence in the Pleistocene deposits of Jaguar Cave * indicative of a former tundra biome’”’ (1967:29). It would be hard to disagree with Guilday and Adam's statement. In fact, presence/absence studies (which are asymmetrical in that the interpretive emphasis is usually placed on presences) are usually quite sound. However, these studies are not trouble free. There are hazards in conducting presence/absence paleoenvironmental analyses of archaeological vertebrates, most of which are shared with approches that treat taxa as variables. Among these hazards are: 1) Assuming that the present ecology of specific mammals is the same as the ecology of those mammals in the past. It is extremely difficult to reconstruct the ecology of ancient mammals, though there have been attempts (e.g., Shotwell 1955, 1958, 1963; but, see also Grayson 1978b). If faunal analysts had to demonstrate the present ecologies of mammals were the same in the past for each time and place they conduct a pac study, they would not get very far. The ecologies f I this proble m can in part b d if suites veh taxa, which ne the same relationship dire ectly knowable, but tra today, parameter in the past. While habitat preferences of a single taxon might change through time, it is less likely that all members of a suite of taxa would change, and that all would change in the same direction. Findley (1964) has discussed this issue as well. 2) Assuming that ecological relationships remain stable across space and competitive settings. However, these relationships are not stable. In Oregon, for instance, the White-tailed Antelope Squirrel (Ammospermophilus leucurus) is an inhabitant of “the open, barren valleys far from timber, but usually where tufts of greasewood, sagebrush, and low desert shrubs furnish cover, protection, and food” (Bailey 1936:142). Not far to the south in central Nevada, they are seen in a different eT pinyon-juniper woodland well above valley floors (Hall 1946). To infer an area “‘treeless”’ or “‘treed”’ from these squirrels would be lous. Such adaptational plasticity may a be due to changing competitive relationships. As Cody (1974: 131) noted, “often no compelling innate-genetic or physio- but rather i ‘its position is flexible, and is determined by the restraints of its competitors.” Thus, on Bear Island, Iceland, Brunnich’s Guillemot (Uria lomvia) nests on cliff ledges, while the Common Guillemot (U. aalge) rests on flat ground. To the south, in Europe, where only U. aalge is present as a breeder, this species nests on both cliff ledges and flat ground (Lack 1968). Such examples of competitive release are common and well known. Much of what animals do is determined by competitive relations; and this lability in adaptation must be recognized in paleoenvironmental studies. Again, the danger of error from this source decreases as the number of animals used as the basis of inferences concerning some should be used as the basis of any paleoenvironmental argument. 3) Stratigraphic mixture. Any study based upon analysis of presence/absence data from archaeological sites must be conducted with the realization that such studies are extraordin- arily prone to error as a result of stratigraphic mixture; i.e., a single element identified for a given taxon carries as much weight as a thousand of those elements. Although it is appropriate to point out that only careful excavation can prevent such difficulties, it is also true that many sites are so stratigraphically complex that even the most careful excavations may not be able to detect all instances of mixture. Once again, the use of suites of taxa can help avoid errors due to this source. Ahi aoe 3 £ ed Poor stratigraphic resolution. g m originally sep stratigraphic resolution. I liffi h made which require finer stratigraphic resolution than was present or documented. This issue is becoming more important and increased attention is paid to the argument that P climates were more equable than Holocene climates (Axelrod 1967; Dalquest 36 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 1965; Hibbard 1970; Slaughter 1967), since one of the arguments used to support this hypothesis is that currently allopatric animals were sympatric during the Pleistocene. For example, Pleistocene sympatry of boreal and deciduous forest species has been argued to support the hypothesis of Pleistocene climatic equability (Graham 1976; Graham and Semken 1976). Yet, it is extremely difficult to argue that “‘stratigraphically sympatric”’ taxa in a single stratum of an archaeological or paleontological site were truly sympatric animals, and not actually allopatric or allochronic. A convincing demonstration! of sympatry in this setting would take remarkably fine is rarely encountered in either archaeological or ‘paleontological ates. 5) Long distance transport of skeletal remains. T b ds with examples of animals transported by humans to areas outside of their natural range. Other forms of long distance transport are also possible: movement by water provides one obvious mechanism. Although some instances of transport are readily detected (as the transport of Haliotis from the Pacific coast of North America to interior localities as Arizona [Haury 1976]), others will not be. The use of a — of taxa, all of which inform on a single enviraa mental variable, provides one means his difficulty, application of the principle of parsimony. For instance, it is far simpler to suggest that Guilday and Adam’s Dicrostonyx was a local resident than it is to argue that long distance transport was involved. In many cases, however, the local presence of a taxon whose historic range was not far removed from that area might be of concern ea jaan biogeograhic reasons (e.g., Grayson 1977a). Here, it might b more It to convincingly argue that long distance transport did not play a role in bringing the animal to an area in which it would not otherwise have occurred (see, for instance, the discussion by Alcorn [1940] of the introduction by man of Spermophilus townsendii into areas outside of its natural range). Clearly, these and other potential difficulties demonstrate that paleoenvironmental analyses of archaeological faunas which depend only upon presence/absence data are not trouble free. Nonetheless, the hazards associated with these studies are of a lesser magnitude than those which necessarily accompany studies which proceed by quantifying taxonomic ataundances, atike the latter, the memo unit with which analyst deals i in aaas abern studies Ity Stucics 7 CONCLUSIONS Two approaches to the paleoenvironmental analysis of vertebrate faunas from archaeological sites are in common use. In one of these approaches, the abundances of. taxa in the fauna are quantified using either counts of identified specimens or minimum numbers of individuals, and changing abundances through time are examined for paleoenvironmental meaning. This approach, while seeming to offer great precision in paleoenvironmental analyses, has 2 debilitating attributes: 1) The relationship between the abundance measure (NISP or MNI) and the actual numbers of animals which contributed skeletal remains to the collection is always unknown; asa result, the meaning of , with trivial exceptions, always unknown; 2) The relationship between t ic abund <— oh . ding the site at the time of fauna accumulated and the abundances of the animals present in an arch- aeological fauna is always unknown. As a result, even if the relationship between NISP oF MNI and the number of animals originally deposited in a site were known, the meaning of changes in these ahauaances through time would not be interpretable. This i is because che rarely } i gy in the surrounding environment, or if these abundances are reflecting changes in which are unrelated to th g alyst is attempting to monitor. May 1981 GRAYSON 37 Because of these sh i it is diffi have faith in the validity of th dies. As a result, analyses which 1 depend only ‘upon the t taxa recorded as p ithina f be preferred. While e by ll iated with all paleoenvironmental studies Sune ‘with ene or F utiboesil remains. Most importantly, presence/absence studies are not characterized by the 2 major and seemingly insurmountable shortcomings associated with vertebrate faunal studies which treat taxa as variables. Until wl ill be be possible to impro proceed simply on the fedis. of treating vertebrates as belonging to taxa which o can neither he present or absent, but whose abundance cannot be meaningfully counted. > 1 etindiec af h ACKNOWLEDGMENTS Th . rR L r Ty. =e) . L - a srayson, R. Lee I yman, is gratefully acknowledged. LITERATURE CITED ALCORN, J.R. 1940. Life history notes on the Piute ground squirrel. J. Mamm. 21:160-170. AXELROD, D.I. 1967. Quaternary extinctions of large mammals. Univ. California Publ. Geol. 74:1-42. BaliLey, V. 1936. The mammals and life zones of N. Amer. Fauna 55. . Paleontology: The fossil Excavations at the Wady El-Mughara. Vol. 1 (by D.A.E. Garrod and D.M.A. Bate), Clarendon Press, Oxford. BEHRENSMEYER, A.K. 1975. Taphonomy and leoecology of og tieinptae pana assemblages east o ard Mus. Comp. bag Sail 146004475. 578. 1978. Taphonomic and_ ecologi ishepenatices from bone weathering. Pa sl biology 4:150-162 PENT. are 1988. | Life e histories of North American Strigiformes. USS. Natl. Mus. Bull. 170. BRAIN, C.K. 1969. The contribution of Namib ottentots to an understanding of ee ee bone accumulations. . Papers Namib Desert Res. Sta. 39:13-22. BUTLER, B.R. 1972. The Holocene or postglacial ecological crisis on the eastern Snake Plain. Tebiwa 15(1):49-63 “ee 978. Saunt assemblages and the “weigemethode” or weight method. J. Field Arch § 5:71-77. CLaRK, J., AND Population dynamics of Leptomeryx. Fie Princeton Univ. Pees: Princeton T.E. Guenssurc. 1970. tructure Ba Biol. 7. CRAIGHEAD, J.J., AND F.C. CRAIGHEAD, JR. 1956. Hawks, owls and wildlife. Stackpole, Harrisburg, and Wildlife Mgmt. Inst., Washington. DALQUEST, W.W 1965. New Pleistocene nn Mech. Arts, Entomol. Econ. Zool. Section, Res. Bull. 277. Finney, J.S. 1964. Paleoecologic reconstruction: vertebrate limitations. Pp. 23-25, in The reconstruction of past environments (J.J. d J. Schoenwetter, eds.), Fort Burgwin 1977. recent human occupation site in Kenya. Quat. Res. 8:245-266. GRAHAM, R.W. 1976. oir goal mammalian faunas and environme ts of the eastern ery States. fclahiiogs 9.343. 250. H.A. SEMKEN. 1976. Paleoecological ee of the short-tailed shrew (Blarina) with a discussion of Blarina methodology of faunal analysis. Amer. Antiquity 38:432-439. ___.. 1975. A bibliography of the literature on 38 JOURNAL OF ETHNOBIOLOGY North American aed of og past 13,000 years. agen Publ., New Yor 6. The ee — avifauna and “alithermal Pp. 74-102, in Holocene Elston, ed.), Nevada Archaeol. Survey Res. Reports 6. . 1977a. On the Holocene history of some Northern Great Basin lagomorphs. J. Mamm. 58:507-513. . 1977b. Paleoclimatic implications of the Dirty Shame Rockshelter mammalian fauna. Tebiwa: Misc. Papers Idaho State Univ. Mus. 9. Minimum numbers and sample size in vertebrate faunal analysis. Amer. ame 53-65. "19 a a munities: a Race of Shotwell’ s method of paleoecological analysis. Paleobiology 4:77- 81. . 1979. On the quantification of vertebrate archaeofaunas. Pp. (M.B. Schiffer, ed.), Academic Press, New York. GuILpay, J.E., AND E.K 967. Small mammal remains from Jaguar Cave, Lemhi County, Idaho. Tebiwa 10(1):26-37. dinate PARMALEE. 1972. Quaternary periglacial records of voles of the genus Phenacomys Merriam (Cricetidae:Rodentia). Quat. Res. 2:170-175. HALL, E.R. 1946. Mammals of Nevada. Univ. Cabticnia Seige at and Los Angeles. HARPER, K.T., AND G.M. ALDER. 1970. The macroscopic pane remains of the deposits of Hogup Cave, Utah, and their paleoclimatic implications. Pp. 215-240, in Hogup Cave (by C.M. Aikens), Univ. Utah Anthrop. Papers 93. Harris, A.H. 1963. Vertebrate remains and past environmental reconstruction in the Navajo. Reservoir face’ Mus. Papers Anthrop. 11. Haury, E. eh 1976 The coereneange desert farmers n. Univ. na Press, Tucson. HIBBARD, C 1 1970. Fiohagiae mammalian local f New Mexico Press ments of the Central Great Plains (W. Dort, Jr., and J.K. Jones, Jr., Ast Dept. Geol. Univ. Kansas Spec. Publ. Lack, D. 1968. one adaptations for breeding in birds. Chapman and Hall, London Vol. 1, No. | LUNDELIUS, E., JR. 1960. Post Pleistocene faunal succession in western Australia and its climatic interpretation. Proc. Internat]. Geol. Congress 21: IV, 142-153. Marti, C.D. 1974. Feeding ecology of four sympatric owls. Condor 76:45-61. Maser, C., E.W. HAMMER, AND S.H. ANDERSON. 1970. Comparative food habits of three owl species in central Oregon. The Murrelet 51- 3:29-3 53:29-33. MELLET, J.S. 1974. Scatological origin of micro- rlgtonans fossil accumulations. Science 185:3 eee K:, AND S.A. McLEOD.. 1975. Collection of taphonomic information from fossil and recent vertebrate specimens with a selected bibliography. PaleoBios 19 NoE-NYGAARD, N. 1977. Butchering and marrow fracturing as a taphonomic factor in arch- aeological ae Paleobiology 3:218-237. OwEN, R. istory of British fossil mammals, eer birds. J. Van Voorst, London. WwW PARMALEE, .W., AND R.D. OkscnH. 1972. Pleistocene and recent faunas from the Brynjulfson Caves, Missouri. Illinois State Mus. Reports Inves ; CZYNSKI, J., AND A.L. RUPRECHT. 1974. The effect of digestion on osteological composition of owl pellets. Acta Ornithologica 14(2):25-28. SHOTWELL, J.A. 1955. An approach to the paleo- ecology of mammals. Ecology 36:327-337. . 1958. Inter-community relationships in Hembhiltian (mid-Pliocene) mammals. mage 2 271-282. 3. The Juntura Basin: studies in earth aay and paleoecology. Amer. Philosopb- Soc. Trans. 53(1). SLAUGHTER, B.H. 1967. Animal ranges asa clue to late-Pleistocene extinctions. Pp. 155-168, Pleistocene extinctions: the search for a caus€ (P.S. Martin mea E. Wright, eds.), Yale Univ. Press, New a Hav “MR. "1968. ie page and eTte- ocene ¥ brate fauna, ‘Knox County, Nebraska: Univ. 13:91-101. J. Ethnobiol. 1 (1): 39-54 May 1981 POLLEN PRODUCTION, TRANSPORT AND PRESERVATION: OTENTIALS AND LIMITATIONS IN ARCHAEOLOGICAL PALYNOLOGY RICHARD H. HEVLY Northern Arizona University, Department of Biological Sciences, Flagstaff, AZ 86011 ABSTRACT.—Within the past quarter century palynology has become an increasingly tm re t = L 1 . 1 } A 7° * L 4 ae ee PP | ee f site and room functions, ceremonial and medicinal practices, prehistoric diet and food preparation, correlative construction and chronologies, human modification of the local part ticularly as related to human demography and subsistence strategies. Apprehension concerning the nature and magnitude of palynological bias related to human activities, particularly as reflected by the sources of pollen hanvonatesd employed in nach stuaes, is ) justified but remained saptiid unexplored. E €a hat once the probability and magnitude of limitations me assessed, the, tney p } PI ens 5) ws ae INTRODUCTION Fossil pollen oe in carly studies of paleoecology was usually obtained from lacustrine sediments because ation of pollen in such environments. It was through such studies that me potential of palynology to yield paleoecological and paleoethno- botanical data was recognized (Clark 1954; Deevy 1944; Dimbleby 1955; Faegri 1944; Godwin 1956; Iversen 1949; Jessen 1935, 1949; Sears 1937, 1952; Troels-Smith 1956, 1960). ~ Non-lacustrine sediments were considered unsuitable for palynology due to low pollen concentration and poor pollen Paes despite the demonstration by Sears (1937) of their actual potential in the Am n Southwest (Dimbleby 1957, 1961). Modified extraction procedures finally vines a number of palynologists to recover deine preserved pollen in suitable quantities from aeolian and al application of palynology into archaeological sites where often mnure oon a control was available than in non 955; Leopold et al. 1963; Martin 1963; Martin and Byers 1965; Schoenwetter 1960, 1962; ea 1952, 1961; Sears and Roosma 1961). Within the last 20 years, palynology ha has become an ical research. A lucidation ot site and room functions (Berlin et al. 1978; Hevly MSa, 6; Hill and Hevl 1968), ceremonial and medicinal Practices (Hevly 1964; MSa; Hill and Hevly 1968), prehist 1964; Kelso 1970, 1976; Martin and Sharrock 1964; Ward 1975), correlative construction tod chronologies (Hill and Hevly 1968; Ward 1975), human modification of the local environment (Martin and Byers 1965; Wyckoff 1977), and the nature, magnitude and duration of climatic perturbations particularly as related to demography and subsistence strategies (Bohrer 1972; Dickey 1971; Euler et al. 1979; Hevly et al. 1979; Schoenwetter and Dittert 1968; Schoenwetter and Eddy 1964; Ward 1975; Weber 1981; Zubrow 1971). Increasing concern has developed about the potentials and limitations of these new applications in paleoethnobotany, particularly in regard to the production, transport and preservation of pollen (Bradfield 1973; Hevly 1964, 1968a; Bohrer 1972; Kelso 1976; Lyttle- Web 1978; Potter 1967; Solomon 1976). This concern is justified because the nature and magnitude of palynological bias related to human activities, particularly as reflected by the sources of pollen commonly employed in such studies, (e.g., room and ramada floor human coprolites, trash deposits, burials and artifacts) has remained largely unexplored. ‘: “ a ee i ] 40 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 MATERIALS AND METHODS In an attempt to provide at least partial answers to some of the expressed concerns, data from a number of sites in Arizona, Sein and New Mexico have been re-examined. The samples providing these data w d from ground tifacts, mummies, human coprolites, various pits and cists, ceramic bowls, floors and midden deposits as well as modern and prehistoric soils outside the archaeological structures. The pollen data from these sites were obtained by standard extraction procedures (Gray 1965). Pollen was identified using standard illustrations, keys and a small reference collection (Faegri and Iversen 1975; Erdtman 1952; McAndrews et al. 1973; Kapp 1969; Martin and Drew 1969, 1970). When possible a sgaeive was made of the first 200 pollen grains encountered while lly scanning th Records were also obtained of the number of grains per aliquot of pollen rich residue scanned, pine pollen preservation and ratios of AP/NAP (tree pollen/non-tree pollen), pine/juniper, and large/small pine (mostly referable to ponderosa and pinyon pines respectively). The fossil data were compared to modern pollen samples obtained from the plant community in which the site was located. DISCUSSION Pollen Production.—Pollen is produced in vastly different numbers by different kinds of plants. Anemophilous (wind pollinated) plants usually produce large numbers (e.g. 500 million per shoot of Cannabis; 350 million per 10-year branch system of Pinus) of generally small, smooth non-sticky pollen, while zoophilous (animal pollinated) plants usually produce low numbers (100s to 1000s per year per inflorescence) of generally large, rough, sticky pollen (Faegri and Iversen 1975). Production of pollen is influenced by both climatic, edaphic and biotic factors and consequently varies from stand to stand and from year to year within a stand (Faegri and Iversen 1975). Cone (both male and female) production i m cemilerty for exainpre, cer? parallels climate influenced growth, and even after th f cones both seed and pollen can be aborted by climatic factors such as relative available moisture (Daubenmire 1960; Leiberg et al. 1904; Lester 1967; Roeser 1942; Shoulders 1967). Many of these factors are manifest as individual plant variations rather than stand characteristics and hence are not reflected by pollen data from soil samples which incorporate the accumulative pollen of many years or several decades (Faegri and Iversen 1975). To be manifest in the pollen record, the effect must extend toa major portion of i stand and the effect must be either frequent or f long , g density of the p hi he ak f flowers Ecological factors which meet these criteria may y be limited to fire, climatic change, edaphic modification (e.g. volcanism and altered drainage patterns or water tables) and biotic exploitation, including disturbance by man and grazing by livestock or insects. Presented below are some data which provide evidence that these effects can be detected in the pollen records of archaeological sites. Citadel Sink is located in Wupatki National ene on a. tbe poner edge of the Sunset Crater ash fall area. S pollen analysis perm arious envir onmental known toh 1 tl eleventh veolcibain eruption, an eleventh-thirteenth century rise and fall of prehistoric agricultural population and the twentieth century grazing and juniper chaining (Fig. 1). The aboreal pollen (AP) 1s composed of 2 principal types, Pinus and Juniperus. Pine does not occur locally and its pollen is therefore a long distance transport type. High proportion of pine pollen therefore reflect poor local pollen production, while low pine proportions reflect good local pollen production (Solomon 1976). Pine proportions (relative to locally occurring juniper) do increase twice in the pollen record (during the eleventh and twentieth centuries) at times atypical for such phenomena in the Colorado Plateau pollen chronology (Euler et al. 1979)- Disruption of the local juniper population (which results in higher pine proportions in ara May 1981 HEVLY 41 = § i : + 3 > aes 3 z 32 y H Ze i al [ i Perse | 28) PCIE a oa 2358 He H hi % 50% 50% 50% 20% by ike 3s be 8 ¢ 3 3 ; § ara eeyeoe | © t and pollen from Citadel Sink, Wupatki National Monument, reflecting the aating sat Seblex, anal mere MS). The abrupt i increase of cinder at a ‘depth of 31.75 cm resulted ter about A.D 1066 (Colton 1962). The cinder contains little pollen, few stem and leaf fragments, but wt Said ea indicating that it is an original air fall rather than being secondarily deposited. ate deposition of pela initially favored the grow of desert shrubs, but increasing f The change i in pollen preservation, i 8 d relati hee [Cc 4h me ee ‘aid oo sine bly refi historic agricul The e changes in pine-juniper, pine-grass and j uniper-grass ra ratios in the upper ‘¢ 35 cm probably reflect twentieth century floristic modifications esta with chaining of juniper and grazing of livestock. pine-juniper ratios) in the twentieth century is probably due to local chaining operations. The eleventh century disruption of juniper is not likely to have been the direct result of mage from volcanic eruption considering the 22.5 km distance to the crater and the relatively shallow deposit of ash in the study area. Instead, the disruption of juniper is more ely to reflect the cutting of juniper for construction purposes by the prehistoric inhabitants whose local | I fr } 7 ais oO Lm 2 £ } 7 their Ormer by lava and cinder Ciesla « et al. 1978; Hevly et al. 1979; Pilles 1977). While j Juniper rerowered from this coeupGon, other plant types did not fair ” well due to the pe ft = | tyUyeicu Tt nen f this site Cheno- Am group ites Araerehehee) diminished in relative abundance as grasses € more abundant. A second modification of the floristic composition of this site (increased proportions of Compositae) appears to occur coincident with cultivation and probably reflects disturbance not unlike that detected at the nearby prehistoric cornfield where also the effect of man’s activity has persisted to the present (Berlin et al. 1978). The diminished proportions of Gramineae pollen in the latest twentieth century (surface) level could likewise reflect man’s activity, in this case grazing by domestic livestock which has resulted in a local deterioration of rangelan While the above examples appear to reflect charter of pollen production due to generally Persistant ae of the floristic community resulting from —— eruption or one by man, it nged pollen production with very little, if any, change of the local plant comenonity. For example, the proportion of pine in a pine- Juniper ratio (where both pine and juniper co-occur) parallels as expected the average rai Biwisti-- composition involving establishment and growth of trees to sufficient maturity for cone production (Fig. 2a The significance of the local Givirontoett relative to pollen production and transport is also manifest in the arboreal pollen proportions, particularly those of pine composition. 42 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 When local growing conditions were favorable, locally produced pinyon pine pollen exhibits high proportions in the fossil pine data (Fig. 2b). However, when local growing conditions were not favorable, a larger proportion of yellow pine pollen transported a pe eS of km Bee nearby paE PeCHIneS ai C eaatde scape pine RS Likewise, the i weeay succes- sonal diferent taxa predominating from year to year in response to seasonal changes of temperature and moisture availability (Hevly and Renner In Press; McDougall 1967; Solomon id Hays 1972). Persistant changes of their proportions over many ieee like those of pine-juniper ratio, probably reflect altered climatic conditions, such as seasonal distribution of moisture or succession within the local plant community coincident with abandonment and altered edaphic conditions. Incidentally it should be noted that the species composition of the flora of some areas of northern Arizona as well as the phenology of these plants is such that the seasonal indicator roles of Cheno-Ams and Compositae as described by Schoenwetter (1962), Martin (1963), Solomon and Hayes (1972) for southern Arizona are reversed in northern Arizona (Hevly and Renner In Press). In final analysis, changes in pollen production do occur in response to physical (e.g. edaphic or climatic) and biotic modification of the environment. Such alterations may be essentially permanent (e.g., deposition of volcanic cinder or depletion of soil minerals chreniets cropping); eee most are — t term resulting from such phenomena as climatic p and b Distinction of the particular ecological factor(s) ‘responsible ithe observed nature, magnitude and duration of palynological é f 8 3 is : 3 : 5 g ee ip eeeraes rol oe a ¥ E Eso Eas i sig 5 3 \ Pb \ 254 \ Sa T oF os ear aT ay T T T rt g 3 25 a LA A Nall J so" of a im fl Il /\ Email wa wv yee rv Wr z2| 6 ce 8& % ' dia O Hav Hollow fe} 5 7 le eg Bother if E27 Site Code w OUT R SOm GOP OK Ot oO O 7; NNT vaaare” J YX GOFG YXGGC FiG..2.—Comparisons of the d hi d and poll is f m Hay Hollow Valley eer 1972; Dickey 1971; Hevly 1964; Ward 197 1975) with tr ds f h ) and the White Mountains of California Senne 1974). Each site code letter reflects a pene site — leseets reflect diferent rooms as mee same ae ) Bahl» reflect — pollen production of these 2 genera locally, pine proportions declining during drought episodes as demonstrated in a study of historic allen (Hevly et al. 1980). Bg eh ah ae lel May 1981 HEVLY 43 a ie {> 18 g 3 g 8 Fl 3 i 5 POPULATION DENSITY * vi ahs ceded oe T | | | T U T q qT T eo ef us E54 zo Of 0 XN 5: Le SN . Egy | 2 aS. le £\, an at <3 = VW ba us PO eae -—— we 4 ~ See Se ole £ = Saad <<« we Se > crive dillea on oral U T T T T T SRSA, 5 3 tures Pine Pollen Depar ® o - < « u = < w So « < a $ = os =< = a Re waoP OK OL 0 OL J NNNI MNHIWKLI'J J YX GFG YxGGC f lie! ML . 4 A 2b. Departures of small pine proportions from th Be pine} Study area probably reflect local versus regional production of pine pollen and long-term trends of effective moisture. change(s) in the fossil pollen record can often be accomplished by evaluation of additional biological, geological, and archaeological data. Such analysis | i pal environmental reconstruction in which the relative effects of climatic perturbations, fire, biotic impacts by man and insects, as well as volcanism can be derived (Hevly et al. 1979). Pollen Transport and Deposition.—The majority of pollen does not travel far from the plant producing it. Even in wind pollinated taxa most of the pollen falls immediately beneath the canopy (Silen 1962; Wright 1953). Nevertheless, wind transported pollen travels further (10s-100s km vs. 10s of cm to 100s of m)than i ported pollen which app in very low concentration in soil samples from open situations, being recovered most frequently in close proximity to the plant producing it or where it may on occasion have pped by insects transporting it. In archaeological sites or caves entomophilous Pollen can be more abundant than in modern soils (Fig. 3a) and is most likely to have been introduced there by man or rodents (Briuer 1977; Hevly 1970; Hevly et al. 1979; Kelso 1970, 1976; Lyttle-Web 1978). _ The majority of the pollen found in features wit icted openings (€.g., caves shelters, fissures and man-made structures) is transported into such features primarily by wind but also by man and rodents. Pollen and sediment transport into such features should be slower in open sites, but pollen will enter more freely than the larger and heavier inorganic sediment. Hence, if sediment accumulation is slow and a given sample reflects the accumulation of many decades or centuries, its pollen concentration should be high compared with that observed in open sites. These trends appear to be observable in the Poller concentration data available from Southwestern archaeological sites (Table 1). Continuously open sites such as rock fissures, caves and shelters have high pollen concentrations for individual samples, while man-made structures such as pithouses and 44 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 ~ — insect pollinated types i - ——-— Population ia -_—— _- -_— 3 ig ec ~ Ee Se ~~. Relative Population Estimates _~— ~— _-— ——_— Cultivated or Native Economic Pollen Percentages 800-4 fae 400- 5 » eens wi A.D. 800 900 tr &l pi CI Gramineae 1600-4 CC i posit ei Population o J J : ; E z 21200 4 ° : r4 $ § a é a 2 “ban : 3 : o w Ke 2 = 400- T T AD. 800 900 ann a proportion of non-arbo a cultivated plartts and llen lin of ech plants acti reflects the relative agricultural success of the local human community, some change of environment is suggeste 3b. Change in melating propartions of ‘pollen: from other largely annual non-arboreal taxa whose germination, largely controlled by the seasonal distribution of moisture and emnperanue The proportion of late-spring and early summer flowering Cheno-Am and Low-spine Compe tae pollen in the ae non- arboreal poles 4 sum is shown by the long, open bars. € proportu { Gramineae p igh-spin e Compositae and Artemesia (late-summer and early-autumn wanaata is ae by the shorter bars and shorter scale at right. The data suggest replacement of Cheno-Ams and Compositae by Gramineae and could reflect secondary hess = faced site, recaps ati ees abandonment and sre cent nal however distribution of moistur re. The latter i interpretation could | provide at least a partial explanation for the putative diminishment of agriculture inferred from pollen data of cultivated plants shown abov May 1981 HEVLY 45 pueblos with limited duration as a space within which to trap sediment have floor samples with more limited pollen content. Objects such as a small storage jar with a small hole open for centuries has accumulated a high concentration of pollen. Objects such as storage pits d cists open only for limited times during the ia aed of the site and sealed for centuries by burial have low pollen concentrations. Com soils in an open situation with the pollen content of floor sediment ofa pithouse or pueblo room and of storage pits or cists manifests This phenomena is called qvenyacon, and Probably 5 reflects progressively smaller target openings for pollen f content of soils on or in which man- Naud Heels structures are o buile appears to be diminished by oxidation and mechanical breakage during construction (i.e. fossil soil, floors, cists, pitand subfloors in Table 1). Since wind transported pollen can travel so far it might be anticipated that pe sae would enter structures with restricted apetures en sali facility; however, contrary t expectations, different wind transported types appear to be transported differentially (Currier and Kapp 1974; Hevly 1970; Hevly et al. 1978. Tauber 1977). When the sediments o rock fissures are compared with outside soils, not only does the concentration of pollen change, but there is also an increase in the proportion of some wind pollinated types and decline in others (Fig. 4). The increase in pine pollen and decrease of other types is most noticeable in the Grand Falls Fissure, which has remained open for centuries collecting pollen. The magnitude of differential transport appears less in the structures occupied by man where pollen collected for more brief periods of time. Pine pollen 3 is slightly under represented, while the NAP types, which were g Pp TABLE 1.—Pollen concentration and Preservation in Archaeological Sites. Preservation is expressed as the percentage of entire pine pollen. Concentration is expressed as mean numbers of grains per aliquot of pollen rich residue, numbers of grains per gram of exfencien es end & sth carat ed rains counted while counting 150 grains from p ypt Concentration Pollen/ 150 Preservation Spores/ Provenience aliquot gram Eucalyptus aliquot Fecies (Sheep) | 1094 41,572 547 98% 210 Fecies (Human) 2 1050 53,200 315 ie - Mummy Alimentary Canal 8 1128-2198 43,050-55,650 itn 100% — Grand Falls: 4 Rock Fissure Floor 10,000 11,000 oe 84% 17 Storage Jar 20,000 i — 100% 518 Rock seine Shelter 5 500-1500 506-1520 pees 70-80% = Modern Soi 1061 1075 135 87% 275 Fossil Soil 7 2250 =e 75% a Pithouse Floor 7 332 336 —— OR sich Pueblo Floor 7 658 666 nas 66% see Hogan Floor | 212 214 eee 84% 300 Metate 1,7 297 BH srs Pe aes ee a Mono 1 127 oer nets =e ae Fire Pit 8 207 210 —— 87% -—— Storage Pit 8 116 17 cen 93% ~— Cist 8 50 50 eae, 90% is Subfloor 3 47 47 — 65% me eerie I, ag etal. 1980; 2. Hevly and Hudgens MS; 3. Hevly MSa; 4. Hevly 1970; 5. Hevly et al. 1979; 6. Hevly unpubl. data; 7. Hevly 1964 and Unpul ta; 8. Hevly MSb-d Vol. 1, No. 1 JUEDHTUs HS Sutpasoxa saimiedaq ‘s[ios snoauezoduiajuoo ut uat[od jo suonsodoid ay) woyy I I d [tssoj jo sainuiedaq—'f ‘olf \ for o o\dweg OL 10014 O19eNd JOURNAL OF ETHNOBIOLOGY SS « $100)4 asnoH iid s10014 181104S pue ered mm ewe eae ee ae ae ia ae ia a ee a i oe ae a a a i ae ee -|- sro ew OR sap 881015 Dm ee me eee eae ae fe ee ae ee ia i ie ew = eS seats OR | 46 2 = ' + al - Lad ' + ” ° e e ' + 2 ° “ - ' + = ° bail a ' + 2 ) o ' JON3INIAOME vow e110, wae evcseiny swy ouey) snsediune snug Na110d seetines ereoeeiey 1a snojjudoweuy HEVLY May 1981 BL61 “TY 19 UTP °L1 56261 NOS “OL “GSW MIQIEH G1 SW AIA9H 'b1 TL6T A2421C “81 GL61 PAE M 21 SL61 49440g * 11 O61 A140} “OI (SIN ABBY pur Ajaagy “6 togyy Ajaagy @ “2510 A14aH4 *L LL61 B2NUg “OSI POH LPH OSW ALAAH b SGR961 AIAPH'S “6L61 TE 9 AjA0H 2 ‘OL61 AIAPH I ve T | T r ’ YJ ; ] ' CA is | . oe ; ST ; 40) eon perodns br ! | ! ' \ ! ! ! suites ee emi g A OH8R vew ' | | i 0 1 1 | ! ' pie ve . cam ' I i i ‘ 1 ! 1 i | ey Curie eno} 8A \ \ ' ' 1 \ ' \ : “4 = ig i ' : ' | ‘ , | L. 1 Mo . | my i c , \ seiioimos : ‘ j 1 | ue tein a 1 Lo Lal : — i eal i f L 1 " a a = on l i ' ! ! ' _ - { \ | | { ! ; un = Ee | ieee 1 1 1 2 eh % T ; on ) ' ; 661 1 an -_ ' 661 OL = | ! I ' 61 O18 :s 1 1 l 1 , ' (A? OF a | ' } } ngs we oe ! ed . i] 1 | i] (Ww) st i . 2% —— ~ ! a ! | | Pe [tom | ieee J Feel . \ 1 1 | if | | at on : i Ls a : 4 (a) t ps i | I ! ' ri am 1 ! I I ' ' vi ' ' l ' 1 1 | q cal a, I ' (i cd ‘ u ! q REE Di ; ; | eames cee) ee me 4 Se ; Sid edei01g eae ‘seeseaee | ! ! | , A | om -_ ' \ ! n ' ' i, L " RK | | ! ! I ; 4 - ! | | ( ! | « | | ! ! | ! = 1 rT 1 4 4. 48 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 frequently over-represented in the structures occupied by man, particularly when comparisons are made with modem rather than ancient soils (Fig. 4). Such over- representation reflects natural or human (Fig. 5) disturbance favoring growth of pioneer species at least in part (Diggs 1979; Gish 1979; Halbirt MSa; Scott 1979). Macroscopic evidence from human feces and food storage and preparation features suggests that such pollen over-representation may also reflect the introduction of wild (or even encouraged or semi-cultivated) plant parts bearing pollen into the occupation areas (Bohrer 1972; Cutler 1964; Hevly MSc; Hill and Hevly 1968). Recognition that the pollen record of archaeological sites is not identical to that of con- temporaneous soils due to differential transport of pollen might seem a critical if not fatal. blow to all attempts of However, selection of indicator species least affected by such differential transport and using samples from proveniences exhibiting minimal transport bias such as small sites may obviate such problems. Furthermore, when the general patterns of transport capability are recognized, various statistical procedures may be used to correct for over or under representation of particular types (Mosimann 1963). Comparison of various man-occupied structures and non-archaeological sites permits recognition of general patterns in pollen transport and deposition, departures from which can be interpreted as changed modes of transport (Fig. 6). In archaeological contexts such changed modes of transport, which often are cflected by: significant over-representation of particular types (Fig. 4), provides i g g activity (e.g., storage, food preparation and medicinal or ceremonial). Pollen Preservation.—Different pollen ty t equally d, being subject to chemical, mechanical, and biological degradation (Having on It has been suggested that juniper pollen is less well d th i llen (Bradfield 1973; Potter 1967). This is contrary to experimental studies of relative pollen. preservation and has not been substantiated in modern pollen studies of soils (Havinga 1971; Hevly 1968a). The explanation for tiese Boeiny results may relate to the different seasons of pol dispersal in pine (ear the depositional Liaiewei If j juniper calle were to lie on an pai Se soil surface for several months prior to burial by wind mixing of sediments or secondary transport into a cistern with the onset of summer rains shortly after the pollination of pine, it would be expected that differences of preservation might be manifest. If, on the other hand, junipet and pine pollen are both buried shortly after their wind transport and deposition, preservation would be about equal as found in the experimental studies. The problem is worthy of much further examination since no one has checked the relative preservation of pine and juniper in diff Preliminary studies would suggest that differences might occur since pine pollen is not equally well preserved in wet vs. dry sediments of different plant communities (Fig. 7 Preservation of pollen could be an juan factor in fossil pollen studies of archaeological sites, since relative abundance of a pollen type such as pine appears to negatively correlated with the percentage of broken grains in both modern and we samples (Fig. 7). Fortunately, th samples is about equal (except i in | samples from a bumed structures) despite the jt lower pollen pared with that of modern soils from open environments (Table 1; Fig. 7). The variability of pollen preservation might also seem a fatal blow to environmental reconstruction; however, the types critical for environmental reconstruction appear to be about equally well preserved. Preservation of pollen in different depositional situations is also variable, but archaeological sites, particularly in grassland or woodland situations, seem to provide best preservation. In fact pollen often provides the only record of plants whose macroscopic record is totally lacking in archaeological context having bee? posed by bacteria or fungi, eaten by animals or destroyed by fire (Bohrer 1972; Schoenwetter 1962; Hevly 1964, 1968b, MSc; Martin and Byers 1965). NAP Ratios May 198] HEVLY 49 2545 O ® 2.0— a 1.55 104 O 3 8) 1@) e 0.5= re) a 3 o 3 1°) e O e o ° s =e ee cette 0 y i 1 50 100 150 Rooms/Site ® Graminae/ Cheno-Am O High Spine/ Low Spine Compositae {data after Halbirt 1978) Pig Gis : ’ . rede A comparison of non-arboreal pollen ratios with the size of archaeological sites (rooms/site). sai 1 Sites are - : sé y = igh prop : —— . ae ligh spine Compositae pollen, eu € larger sites are characterized by higher proportions of Cheno-Am and Low-spine Compositae n. Fin : The latter plants are characteristic pioneer plants favored by disturbance. 50 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 Pollen Pollen AP Departures % Pine Departures % Alluvium Arc ical Alluvium Archaeological hasolog 4p) ~ -30 - + 1 >. PIS -15 0). 4155 3+ py spe | -F, +P |e AAA MAA waa aS ie a 4 4 NE Fic: 6.—Comparisons of the arboreal pollen dat ll d arch in the same locality (Black Mesa, Hevly, unpubl. patie acai departures of arboreal pollen from the an of the study area are negatively correlated between alluvial and archaeological sites, sefiesiee sentin ne differences of mode of pollen transport into such environments. Departures of pine in pine-juniper fatios sites the modern ap big the pay area exhibit generally parallel i Le iy ~ © ols ff. Afic tCW Gir su ae pe pine-juniper ratios can reasonably be substituted for AP/ NAP ratios in pine Hittin reconstructions if the latter ratio is biased, for example by NAP over-representation. Forest % Pine Pollen Pinyon- Juniper Woodland ® 204 e 104 T T , > cee T ° 1 2 3 6 ? 8 9 © science % Broken Pine Pollen ® aquatic site FiG. 7.—Preservation of pine pollen (data from Dickey 1971; Hevly 1964; Ward 1975). 7a. A comparison of the preservation of pine pollen with th n different depositional environments within mttovens plent communities. Generally, the penne of pine increases as expected in the higher elevation con owever, pine also er and cor cease? low pollen prides ds high oe production (compare savanna) conifer forests. Within the Pi inyon- Juniper Woodland and Grassland or Savanna pce preservation was generally better than in the pine and mixed conifer forests except in aquatic sites. May 198] HEVLY 51 rr a ry z $ Ss e © = 50- 2 : & 3 40- 5 2 ; , < $ 30- 4 . 8 a o § eo ol a ie 204, ¢«e ) e H £ - re J e ° e a / ® : see j « Rae, SMa ST ec Bae sf * a, ee ; * * ¥ ) 80 100 140 200 300 400+ Pollen Concentration Pollen Preservation ( of pollen/aliquot) (%pine breakage) ® fossil ® fossi! pine % modern *% modern pine “ss A comparison of pine pollen preservation and total pollen concentration. Pollen concentration is ower than in modern soil samples, however, the range of preservation is about the same except in burned sites where more than 50% of the pine pollen is broken. preg een f.3. eee Rh 4 oe 1 2 eee G lly the relative abundance of pine pollen diminishes as the breakage of pollen increases. CONCLUSION The pollen record contained in archaeological sites, like that of any other depositional h W ee | = 4 1 3 Ld 1 In: ° Dy ‘ ry pollen record of archaeological sites is not identical to that of contemporaneous soils due to differences of mode of pollen transport into different depositional envi d €ven of differential capability of pollen to enter archaeological sites due to location, size and Seasonality of openings. Comparisons of pollen taxa with other previously demonstrated sensitive environmental indicators reveals that some pollen types are useful for purposes of Paleoecological reconstruction. Comparisons of pollen taxa within and without archaeological sites indicates that other pollen types are probably more useful as indicators of human behavior. Different pollen types are not equally well preserved and the preservation of pollen in the . ological record may not be assumed to be similar for all time periods. Experimental Studies are badly needed in the American Southwest, but data which are at hand suggest the 52 JOURNAL OF ETHNOBIOLOGY i= Vol. 1, No. 1 polle nt h utility for p ] foetal ‘aliinit equally well preserved. Thus, if preserved by incorporation in soil soon after dispersal, such pollen types should retain their utility as paleoecological indicators. Thus, in final analysis, the potential limitations posed by improving understanding of pollen production, dispersal and preservation are real but not so limiting as to pre reasonable reconstructions. eclude inferences of human behavior and attempts of paleoenvironmental LITERATURE CITED gees R.Y. 1955. Pollen analysis, a research 1 for the apd of cave deposits. Amer. , R. HEVLY AND G. ralicaiot of a Sina n/plant analysis and oo. ogy. Ame. Antiquity 42: 588-600. BOHRER, si aleoecology of the Hay Hollow Site, Pata Fieldiana-Anthrop. 63: 1- 30. BRAGHELD, M. 1973. Rock cut cisterns and pollen * in the vicinity of Old Oraibi, Arizona. ars 46: 68-71. BRIvER, F.L. 1977. Plant and animal remains from caves and rockshelter of Chevelon Canyon Arizona - methods for isolation cultural depositional processes. Unpubl. Ph.D. Dissert. (Anthrop.) Univ. California, Los Angeles. CLARK, J.G.D. 1954. Excavation of Star Carr. Cambridge COLTON, HS. 1962. Archaeology of the Flagstaff area. Pp. 171-172, in New Mexico Geol. Soc. thirteenth Field Conf. Guide book of the Mogollon Rim Region. CURRIER, APs Sain R. Kapp. 1974. leeaiaree 0 Montcalm County, Michigan. patos Soe 7. 1-225. CUTLER, H.C. 1964. Plant remains from the Carter Ranch Site. Pp. 227-234, in Chapters in the prehistory of eastern Arizona II —e P.S. Martin et al.). Fieldiana-Anthrop. 55:1-2 DAUBENMIRE, R.F. 1960. A seven-year study of cone IBM Re as related to xylem layers and temperature in Pinus ponderosa. Midland v0 64: 187-193. DEEVEY, E.S. ‘aia Pollen analysis and Mexican archaeo an attempt to apply the method. Amer. Austicpiien 10: 135-149. Dickey, A.M. 1971. Palynology of Hay Hollow Valley. Unpubl. a thesis (Biol.) Northern Arizona Univ., Flags DicGs, R. 1979. Ten achaiad years of land use at the 1 Ranch locality near Springerville, Unpubl. M.A. thesis (Anthrop.), n Arizona Univ., Flagstaff. Amer. DIMBLEBY, G.W. 1955. Pollen Analysis as an aid to the dating of prehistoric monuments. Proc Prehist. Soc. 20: 231-236. 1957. Pollen analysis of terrestrial soils. New Phytol. 56: 12. —__— 1961. Soil pollen analysis. J. Soil Sci. 12: 1-11. ERDTMAN, 1952. Pollen Morphology and Plant Taxonomy - Almquist and Wilksell, Stockholm. EuLER, R., G. GUMERMAN, T. KARLSTROM, J. DEAN AND R. HEVLY. 1979, Cultural dyna and pal Science 205: 1089- 1101. FagcRI, K. 1944. On the introduction of agricul- ture in Western Norway. Geol. foren Stockh. forh. aie 449-462. ND J. IVERSEN. 1975. Textbook of Pollen cine 3rd ed. Hafner, New York. Fritts, H.C. 1965. Tree-ring evidence for climatic changes in western North America. Mon. Weath. Rev. 93: 421-443. GisH, J.W. 1979. simdapannestg results of the Coronado Project, Second Annu. nf. Ethnobiol., Flagstaff, Arizona. April 6-7, 1979. Gopwin, H. 1956. The History of the British Flora. Cambridge. Gray, J. 1965. Extraction Techniques. In stone functions. Unpubl. (Anthrop.), Northern Arizona Univ., Flagstaff. HAvinGa, H. 1971. A experimental investigauion into the decay of pollen and spores in various soil types: In Sporopollenin. (Brooks et al. eds.). Academic Press, New Yor HEvLy, R.H. 1964. Pollen analysis of Quaternary archaeological and lacustrine sediments from the Colorado Plateau. Unpubl. Ph.D. Dissett. (Botany) Arizona, Tucson. 1968a. Studies of the modern pollen rain in northern Arizona. J. Arizona Acad. Sci. ¥ 108-127. coe May 1981 1968b. Sand Dune Cave pollen studies. Pp. 393-397 (Appendix VI), in Survey and Excavations north and east of Navajo Mountain, Utah, 1959-1962 (by Lindsay et al.) Mus. N. Arizona Bull. 45, Flagstaff. 1970. Botanical studies of a sealed stora: jar cached near Grand Falls, Arizona. Plateau 42: 150-156. ——_— MSa. Palynology of Basketmaker sites on Cedar Mesa and mummies from Grand Gulch, U MSb. Palynology of a Gallena Phase on from northwestern New Mex (submitted to Mus. New Mexico). MSc. Pollen studies of three pithouses, Coronoado-Dinosaur Transmission Corridor St. Johns to seers: Arizona (submitted pe Mus. N. Arizo —_—— d. Sie palynology at LA 11843, a Pueblo II-III site in cea arid New Mexico 4 eau to Mus. New Mexi . Macroscopic plant ie a Fea s Cave, Arizona. (File copy, Mus. N. Arizona). HALBIRT AND C. JEFFERS. 1980. Palynology Chap. 12 in Vol. 4 of Prehistory and e Ojo Amarillo (David T. Kirkpatrick, ed.). Report #276 Cultural Resource Mgmt. Div. Dept. Soc. Anthrop., New Mexico State Univ., Las Cruces ——_, M.L. Hever AND S.J. “OLSEN. yea Paleoecological reconstructions upland Patayan rockshelter, 4 @ y eens Acad. Sci. * 67- hen . MS. Paleoenviron- ments of he nie He o and their prehistoric exploitatio , G. ANDERSON AND OLSEN. 1979. peat effects of climatic change, cultural impact and volcanism in the Paleoecology of Flagstaff, Arizona A.D. 900- 1300. In Volcanism and Human Habitation (P.D. Sheets and D.C. Grayson, eds.) Academic Press, New York. AND L. RENNER. In Press. Pollen, spores and other atmospheric particulates. Res. Reports Mus. N. Arizona 18. ———— AND G. Rocce. MS. Pollen studies at the Webb Tank Site, a Mogollon Pithouse village in “alten Arizona. Report on file, Mus. N. ates RS. SCHLEY AND R.W. BERRY. MS. Sediments, Pollen, and Paleoecology: Sunset Crater, Sink, Wupatki National Monument, Arizona. (submitted J. Arizona- N Acad. Sci HILL, J.N. AND R.H: HEvLy. 1968. Pollen = at K Pueblo: some new interpretation Amer, Antiquity 33: 220-210. HEVLY 53 PERSE: J. 1949. The influence of prehistoric man vegetation. Danm. Geol. Unders. 4 (3):, 6: 1- 25, JESSEN, J. 1935. Archaeological dating in the history of North ae Vegetation. Acta Archaeol. 5: | 1949. saat in late Quaternary deposits and flora-history of Ireland. Proc. Royal Ir. Acad. 52 B: 85-290. Kapp, R. 1969. How to Know Pollen and Spores. Wm. C. Brown Co. Publ., Dubuque. KELso, G. 1970. Hogup Cave, ( Utah: ‘Comparative Pollen Anal fill. Append. IV, in Hogup Cave (by M.C. ae Univ. Utah Anthropol. Papers 93: 1- 1976. ALR 1 WW £. J to the interpretation of human activities in Northern Arizona. Unpubl. Ph.D. Dissert. Ppa ietagt Univ. Arizona. LaMARCH, V.C., JR. 1974. Paleoclimatic inferences from long tree-ring records. Science 183: 1043- 1048. LEIBERG, J.B., T.F. sonics evi DODWELL. 1904. Forest condition Francisco Mountain Forest pein etait U.S.G.S. Prof. Paper No, 22: 1-95. LEOPOLD, E., L. LEOPOLD AND F. WENDORF. 1963. in Ne exico. Pp. 265-170, in UNESCO and World Meterological Organiza- tion Symposium on ge of Climate with special reference to the arid zone. Rome, 2-7, Oct. 1971. (Arid Zone Res. Ser. No. 20) UNESCO. Paris. LESTER, D.T. 1967. Variation in cone production of red pine in relation to weather. Canadian J. 1978. The raeagny y of Miami Wash, Gila County, Arizona. A.D. 1100 to 1400. Unpubl. Ph.D. Dissert. (Anchrop.) Univ. ona. McANDREWS, J.H., A.A. BERTI AND G. NorRIs. 1973. Key to the Quaternary Pollen and Spores of the Great Lakes Region. Misc. Publ. Royal Ontario Mus., Toronto. McDOUGALL, W.B. 1967. Botany of the Museum and Coltin Ranch Area. IV. Vegetation changing in field five. Plateau 39: 134-142. pain P.S. 1963. The Last 10,000 Years. Univ. zona aay Tucso . BYERS. “1965. Pollen and archae- elias at at Wetherill Mesa. Amer. Antiquity 19: 122-135. AND C.M. Drew. 1969. Scanning electron ning eleciscnt photomicrographs of 54 western pollen grains. J. Arizona Acad. Sci. 6: gah pieieanee ay SHARROCK. 1964. Pollen Aaiec of Prehistoric ree fecies: a new approach to mer. Antiquity 30: 68- 1 MOsIMANN, J.E. 1963. On the compound mult- nominal] distribution, the multivariate B distribution and correlation among proportions. Biometrika 49: 65-82. PiLLEs, P. 1977. The field house and Sinaqua Demography. In Limited Activity and Occupa- tion Sites (A.E, Ward, ed.) Contrib. Anthropol. Stud. I. Center for Anthropol. Stud., Albuquerque. Potter, L.D. 1967. Differential pollen accumla- uo ii watcl J Ecology 48: 1041-1043. ROESER, J., JR. 1942. The influence of climate on seed production in Douglas-Fir. J. Forestry 40: 304-307. SCHOENWETTER, J. 1960. Pollen analysis of sedi- ments from Matty Wash. Unpubl. M.S. thesis papers yeh Arizona, Tucson . The po len ainalyais of eighteen pant sites in Arizona and New Mexico. Fieldiana-Anthrop. 53: 168-209. AND E RT. 1968. An ecological iiherpretaxion of Anasazi settlement patterns (A.D. 200-1600). Pp. 41-46, in Anthropological Archaeology in the American Southwest. (eds). Anthrop. Soc. Wash. AND F. Eppy. 1964. Alluvial and via ne reconstruction, Navajo Reser- District. Mus Mexico Papers mniheeles 13:1-155. San SCOTT, =E..J. 1979. Pollen. analysis in selected rooms at Walpi and along modern transects first mesa. Second Annu. Conf. Ethnobiol., Flagstaff, Arizona. April 6-7, 1979. SEARS, P.B. 1937. Pollen analysis as an aid in dating cultural deposits in the United States. 61-66, in Early Man (G.G. MacCurdy, ed.). Lippincott, Philadelphia. 5 alynology in southern North America. I. Archaeological horizons in the basins of Mexico. Geol. Soc. Amer. Bull. 63: 241- 1961. Palynology and the climatic record JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 of the Southwest. Pp. 632-641, in Solar Variations, Climatic Change and Related Geo- physical Problems. Ann. New York Acad. Sci. 95: 1-740 cote AND DA. RoosMa 1961. A climatic sequence from two Nevada caves. Amer. J. Sci. 259:669- 678. SHOULDERS, E. 1967. Fertilizer application, inherent fruitfulness, and rainfall affect flowering of longlife pine. For. Sci. 13: 376-383. SILEN, R.R. 1962. Pollen dispersion considera- tions for Douglas Fir. J. Forestry 60: 790-795. SOLOMON, A.M. 1976. Pollen analysis of alluvial imen implications of experimental evidence from sedimentary and acne pollen oh reece AMQUA abstr. 159. H.D. Hayes. 1972. ah pollen pro- ‘in F qualitative influence of moisture. J. Arizona Acad. Head. Sci. 7: 52-74. Tauser, H. 1967. Investigations of the mode of pollen transfer in forested areas. Rev. Paleobot. Palyn. 3: 277-286 1977. Investigations of aerial ed transport in a forested area. Dansk. Bot. Ark. 1-12) TROELS-SMITH, J. 1956. Neolithic period in Switzerland and Denmark. Science 124: 876- 879 1960. Ivy, mistletoe and elm, climate indicators-fodder plants. Danm. Geol. Unders Series 4, 4: 4- Warp, J. 1975. nical subsistence pace Joint Site as function of paleoclimate. Unpubl. M.S. thesis (Biol.) Northern Arizona Univ., Flagstaff. Weer, S.A. 1981. Palynological interpretation of the Sinagua and their changing environment. Unpubl. M.A. thesis (Anthrop.) Northern Arizona Univ., Flagstaff. WRIGHT, J.W. 1953. Pollen-dispersion studies: some practical applications. J. Forestry 51: 114- 118. Wyckorr, D.G. 1977. Secondary forest succession following abandonment of Mesa Verde. Kiva 42: 215-231. : ZuBRow, I. 1971. Carrying capacity and dynamic equalibrium in the prehistoric Southwest Amer. Antiquity 36: 127-135. (st ke eee J. Ethnobiol. 1 (1): 55-60 May 1981 INFERRED DATING OF OZARK BLUFF DWELLER OCCUPATIONS BASED ON ACHENE SIZE OF SUNFLOWER AND SUMPWEED RICHARD A. YARNELL University of North Carolina, Department of Anthropology Chapel Hill, North Carolina 27514 ABSTRACT.—Samples of cultigen sunflower and sumpweed achenes recovered from archaeological sites in eastern North America gradually increase in average size during the lact 2h , es Ee Alsth ae | > L W = D3 be * ¢ oe re fr the same general time period fall within a relatively restricted mean size range. Achene samples of both sunflower and sumpweed have been recovered from several Ozark Bluff Shelters, but the dating is highly problematical. The pweed ples fall into 2 disti size categories which indicate that they derive from 2 separate time periods, one during Mississippian ti d the other during early Late Woodland times. The sunflower samples display a more continuous mean size variation and seem less reliable as chronological indicators. INTRODUCTION Seeds and achenes of cultigen sunflower (Helianthus annuus var. macrocarpus Ckll.) and sumpweed (Iva annua var. macrocarpa Jackson) have been recovered from many archae- ological sites in eastern North America ranging in age from about 1500 B.C. to late prehistoric times. Sunflower husbandry has continued to the present, but there are no reports of historic sumpweed husbandry. The prehistoric record of these plants has been extensively reported by Asch and Asch (1979), Black (1963), Heiser (1951, 1955), Struever and Vickery (1973), Yarnell (1972, 1979), and others. What is of concern here is the gradual sumpweed. DISCUSSION Sunflower achenes apparently increased in mean size from approximately 6 x 3mm up to about 12 x 7 mm, while sumpweed achenes increased from 3.5 x 2.5 mm up to 7.5x 5mm. Taking into account an apparent doubling of thickness, the overall increase in sumpweed achene size was approximately eight-fold, while sunflower achene size increased twice that much. Overall increase in sumpweed achene size from the wild progenitor appears to have been approximately twelve-fold, which again is only half the comparable increase for sunflower, These increases can be interpreted as having been more or less regular and continuous through time even though the available data are stil] less abundant than preferred, even though there are exceptions to the expectations. The summary data portrayed in Table | Present a preliminary indication of the patterns of size increase of sunflower and sumpweed achenes. (See Yarnell 1972 and 1979 for more detailed data and sources.) It shows that the average of means of achene length times width gets progressively larger from Terminal Archaic through Early Woodland, Middle Woodland, and early Late Woodland to Mississippian times. unflower and sumpweed achenes from the Ozark Bluff Dwellings and from Newt Kash Hollow shelter in eastern Kentucky are not clearly placed chronologically. Neither are the sunflower achenes from the Mammoth Cave vestibule or the sumpweed achenes from Cloudsplitter and Hooton Hollow shelters in eastern Kentucky. In addition, sample size is (00 small to be reliable for 8 sites with sunflower and 4 sites with sumpweed. This leaves 17 sumpweed samples (N = 1] to 879) and 11 sunflower samples (N =9 to 1000) which were used AL OF ETHNOBIOLOGY Vol. 1, No. 1 d JOURN ‘(8S61 19S19H) ..(90L1 A) AV - HBLPS1).008Z1-922 “ON “ET UeBIY>I Jo Ansusatuy,, se parynuapr pur ‘19s19}4 Aq paunseaus “1aBur Jaq. Aq par2a[]09 sauaY>e g S>PNIIUT SIYLs 2086 - SUSpy SE= TSX GL Gh (991-#-Z§) PAY 9S = SPX FL 096 S01 - Pe-Ty G§= 8h X SL 16 69¢ - PPOppeID 9F = OGXTL 9¢ 9S 96=TS8X*6TT 8L - NI UMoig | =SLX FI 8h - L9 YopperD Ost 69=€9X*X60I b8E - 99 YOppeID of 49=¢9* LOI 981 - 99 YOppeID LI 19=6¢9%* §Ol O8€ - 99 YOPpeID 16=6E* 9 00€ «ol 0¢ =39 X16 90LI - SUsPy 12=6E* GG 006 squdtpe o7z18 evar Saude jo ‘ou ueoul uebau jo ‘ou ra 68=LLX9 II sauayre J981e] 96 =6RX EL Fb ra 89= 9X 8O0I s1eT paula} ra GG=PFSXTOI SoUdYIe Joes 16=6E* GG é ssayays {fn} q 44070 (¢) (08 = L9* 611) (IW ‘P2OjTEM Noy) b OL=19* FI wuerddississim,, 98 = 6 EX ZL g (3) (Gg=FEXO'8 ) (IW ‘pAog inoyitm) § Ig =OFX SL pue[poom ae] Apeq pue a[ppry 16= 8 X9°¢ 8 b b=SEXEL puel[poom Ajreq-oreypry jeuruna y, Sl=cS8 * oF 9 poSe9ae suvout uu ut pouag aunty uu ut podei2ae suvou ‘dod jo ‘ou suvaul JO aZe12ae SUBIU JO aBe12A8 ‘dod jo ‘ou WaMOTANOAS aqqaMdWns ‘sauay av sayjays f{njg y4vzQ 01 pasvduso2 sv spotsad aust} juasaf{{tp wosf sauayrv samojfuns puv paamduns fo azs a8vizay—'| TIAVI, | oe a, ee ge oat ae ie fc Le a a Dime ee gL UR Sheba 3 Oe Aa ac oa la el nN dd gee mL oF Ra = Ot Tes a ee LS LS co eae ee ee es ee Boe Nd Sy eM ak A Mk Seo a yt May 1981 YARNELL 57 in order to derive an average achene size for each of 3 broad perhistoric periods: Terminal Archaic and Early Woodland, Middle Woodland and early Late Woodland, and “Mississippian” (including Fort Ancient). The number of usable sarepiee is rege but the results are generally supported by data from the smalle sch and Asch (1979) and Andrea B. Shea. In addition, Gaver are indications ant the reconstruction factors for estimating original achene size from carbonized sumpweed seed and achene size tend to underestimate the mean size of larger achenes (Asch and Asch 1979; A.B. Shea, personal communication). It is suspected that the same is true for sunflower. The product of length and width in mm is taken to be a reliable indication of achene size for purposes of comparison. For sumpweed these figures are 13, 21, and 35 for the 8 broad periods from earliest to latest. The comparable figures = sunflower — are 24, 31, and 70. Two sunflower samples from the Yazoo Basin in western Mississipp small for their age. Achenes from the wai Site with an coins date ee A. D. 500 Ub Connaway, personal communication) average th same site (see Table 2), and achenes from dei Miskiaisiplon period Wilford Site have a mean length times width of only 41. It appears that these sunflowers, grown at the southern margin of the prehistoric sunflower belt, perhaps in damp soil, produced smaller achenes than those produced elsewhere at the same time. If we delete these 2 samples, the size progression for sunflower becomes 24, 35, ds. This seems nearer to the reality of prehistoric evolution of sunflower schene size under domestication. It alsoisa better indication of the vast increase in sunflower achene size during Late Woodland and Mississippian times. If we compare the sizes of sumpweed and sunflower achenes from Newt Kash Hollow (21 and 29; see Table 2) to the size progression portrayed in Table 1, they seem to fit best into Middle Woodland to early Late Woodland times; but they may have a mixed composition. S age seems about right for the Hooton Hollow sumpweed also, but the initial Cloudsplitter sumpweed collection fits well into the Early Woodland size category. The Mammoth Cave Vestibule sunflower, collected by Nelson and measured by Heiser, is much too large for an Early Woodland assignment but accords well with a Late Woodland designation. On the basis of a limited series of measurements, I had assumed until recently that all of the sunflower and sumpweed from the Ozark Bluff shelters were Mississippian in age, Probably not earlier than A.D. 1100 to 1200. Early in 1978 the University of Arkansas useum graciously allowed access to collections there in order to select samples of Arkansas TABLE 2.—Sumpweed and sunflower from the same source. SUMPWEED SUNFLOWER f no. of eics mean —_s ee size achenes 309 3.7x2.7=10 Salts Cave J IV: 4-11 ee - er 80 40 4.0 x 3.1 =12 Mammoth Cave cadaver ses sata . ay 1000 879. 42x32- 13 Salts Cave feces 7.4 . : hs nt 744 5.5x39=21 Newt Kash Hollow, KY pes 10 20. 6.1x42=26 Boyd, MI eee 36 4 13 5.7 x 3.9 = 22 Hooton Hollow shelter, KY od ae ry, 2 74 6.0 x 4.2 = 25 Haystack shelter, KY o - ny " - " 8.6 x 4.1 =; 19 6.2 x 4.2 = 96 Rogers shelters, KY 7 2 300 55x39=21 Edens - | 7 ewes 17 300 -5.5%3.9=21 Craddock 66 - 380 Se res 19 70x 4.5=32 — Paul’McCulloch, MO 12.6 x 6.6 = 58 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 Bluff Dweller sunflower and sumpweed achenes for study. With the generous assistance of museum personnel I was able to locate 6 collections of sunflower achenes and 6 of sumpweed, only 2 of which contained both species. Measurement of the sumpweed achenes revealed ior the sien die fall distinctly into rks size categories. Four samples from Craddock, Alred and E 3x 4.9mm = 36. This is approximately the size expected for all of the Ozark Bluff Dweller r samples. Howeie: 2 samples from Edens and Craddock 66, both of which have mean sizes of 5.5 x 3.9 mm = 21, are clearly outside of the expected Mississippian period size range. In fact, they are the same size as the Newt Kash Hollow sumpweed and fall between the Middle Woodland and early Late Woodland expected sizes. This suggests that they should date to around the fifth century A.D. Yet, data included in Tables | and 2 indicate that the mean sizes of sunflower achenes from these collections occupy intermediate positions between ae arhy Late Woodland size, on the one hand, and the f good Mississippian period collections and the other Ozark collections, on the other hand. Thus it would appear that the 2 Ozark samples with smaller sumpweed and sunflower achenes should date to approximately the seventh century A.D. This assumes that there was no contamination of the sample by larger sunflower achenes from a later deposition. Two collections of sunflower ek ot ~~ Craddock 66 ahelaes whued AVCTARS 10.8 x 6.3 = 68 may be early Mississippian in till larger achenes ae Craddock 67 and Brown Bluff a should date to well within the Mississippian period. In fact there is a radiocarbon date of A.D. 1110 + 110 (M-1711) on materials with the same collection number (BR - 78) as the Brown Bluff sunflower (Crane and Griffin 1968: 92). These achenes have a mean size of 11.9 x 8.1 mm = 96 which is the largest of any prehistoric collection on record. Mean thickness of achenes in the Ozark sunflower collections 1s consistent with mean size as determined by length and width. There is one additional sample of 10 sumpweed achenes, measured by Richard I. Ford (personal communication), with a mean size of 7.0 x 5.2 mm = 36. This is from the Proether shelter in southern Missouri and clearly falls in the Mississippian period size category. Heiser (1953) has measured 2 samples of sunflower achenes from unidentified Ozark shelters. A sample of 9 achenes recovered by! M. R. Harrington (Heye Museum No. 11/7265) has a mean size of 9.3 x 4.8 mm = 45 the early Late Woodland size. Another sample of 10 achenes (University of Michigan, Museum of Anthropology No. 13250) witha mean - 11.4 * F fi 0 7 Rael is sale hcl in ‘Size. Th ¢ 1 ™m nureRd achene size in the Ozark shelter collections are that carly Late Woodland sleet Mississippian occupations are represented, possibly with an early Mississippian component as well, occuring within an inferred time range of the seventh century A.D. or earlier to the twelfth century A.D. or later. These estimates are based exclusively on sunflower and sumpweed achene sizes and are presented with somewhat limited confidence. They were derived independently of the available radiocarbon dates which can be interpreted as providing a chronology that differs in some respects from the chronology based on achene size. Crane and Griffin (1968: 88-93) have xpportess 17 dates from 8 Ozark — ranging from 40 B.C. to A.D. 1950. Since there i the 3 latest dates of A.D. 1670, A.D. 1810, and A.D. 1950 should be of no eolicern, certainly not the last Zz Also it seems unlikely that the date of 40 B.C. (M-1694) from Red Rock shelter is relevant to the sunflower and sumpweed remains. Except for an Edens shelter date of A.D. 630, the remaining dates form 2 clusters. The earlier cluster includes dates of A.D. 200, A.D. 360, and A.D. 370 from Edens, ap es hota —_ men Bluff cues The latter 2 dates might be seen to indicate the age of the coll Edens and Craddock shelters were it not for the size of the associated sunflower achenes: The later cluster includes 9 dates 1080 to A.D. 1350. It is likely that the collections with large sumpweed achenes and thos€ RI seme Ta RRO eRe: May 1981 YARNELL 59 with large sunflower achenes date from this period. In fact, as noted earlier, the Brown’s Bluff collection containing large sunflower achenes has been dated to this peri The single early Late Woodland date of A.D. 630 + 120 (M-1703 A) seems most likely to represent the age of the samples of smaller sumpweed achenes from Edens and Craddock shelters, but their actual age may be between this date and the e date of A.D. 1080 + 110 (M- cere pich wa was determined on other materials from th (Burial E- TOM Sie and Newbridge i in the lower Illinois Valley and some of the vodkihelieri i in eastern entuc CONCLUSIONS To summarize, conjecture, and conclude, I suggest that in general we can reasonably expect that the product of mean length times mean width in mm for sizeable collections of sumpweed and sunflower achenes to be approximately: Sumpweed Sunflower 8tol2 and 20to24 for Terminal Archaic samples 12tol16 and 22to26 for Early Woodland samples 16t020 and 251035 for Middle Woodland samples 20t0 26 =and 35to60 for early Late Woodland samples, and 25to40 and 50to100 for Mississippian period samples These estimates are densi on the earaee ares and only aero take into account the expectation that a be encountered because of laa of a variety of influencing factors. Currently a data indicate that ante an be menecied for sunflower, which might further indicate that greater varietal diversity had developed in sunflower. In any case, it appears that continuing increase in size of achenes took place more uniformly sumpweed and that this species should be a better chronological indicator than sunflow sit ante - accuracy of the eee size ranges of achenes for each period will ah Reale etermined when we have an adequate set of measurements for additional aay of archaeological sumpweed and sunflower achenes which have been reliably ACKNOWLEDGMENTS Valnah 1 | iby the k of Nancy and David Asch i in Illinois, by eer i among their sellin s. LITERATURE CITED CRANE, H.-R. AND JAMES B. GRIFFIN. 1968. Asc, Davip L., AND NANCY B. ASCH. 1979. The University of Mic sep radiocarbon dates, XII. €conomic ‘iced! of Iva annua and its aaa importance in om eae Illinois Radiocarbon 10: 6 alley. Pp. 301-341, in The Natureand Statusof HEISER, CHARLES - . 1951. The sunflower Ethnobotany (Richard I. Ford, ed.), Univ. among the North American Indians. Proc. Michigan, Mus. Anthrop., odiecpol. Papers Amer. Philosoph. Soc. 95: 432 No. 67. 1953. The archaeological record of the cultivated sual ouiek with remarks concerning the origin of Indian agriculture in eastern North America, MS, files of the author. B eae MEREDITH. 1963. Fhe distribution and ological significance of the marsh elder, va annua L. Papers Michigan Acad. Sci., Arts and Letters 48: 541-547. 60 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 1955. The origin and development of the cultivated sunflower. Amer. Biol. Teacher 17: 161- 167. STRUEVER, STUART, AND KENT D. VICKERY. 1973. e beginnings of cultivation in the Midwest- Riverine area of the United States. Amer. Anthropol. 75: 1197-1220. YARNELL, RICHARDA. 1972. I; macro- carpa: extinct ey edaae felt aig Amer. Anthropol. 74: 335- 1979. Gerald of sunflower and apiece I. Ford, ed.), Univ. Michigan, Mus. Anthrop., Anthropol. Papers No. 67. J. Ethnobiol. 1 (1): 61-68 May 1981 ON PREDICTING HUMAN DIETS H. RONALD PULLIAM State University of New York, Department of Biological Sciences, 1400 Washington Avenue, Albany, New York 12222 ABSTRACT.—Optimal Foraging Theory (OFT) has helped animal ecologists understand the dietary preferences of animals. This paper addresses the question ‘Can OFT shed light on the food choices of humans?” In its simplest form, OFT predicts the diet which maximizes Lh f, T = } = fh vd } net energy gain tot g y be forced t nd hg , ae bd 1 ; ie L he LO TMAGAALILTLIZE COelr Cnersy ’ ’ I y Seer dab iti lly bal d diet th b gy-rich one, OFT theory can predict nutrient-contrained diets but the theory b plicated and the requi information on the nutrient contents of wild foods is usually lacking. Nonetheless, some evidence does suggest that humans often choose food so as to meet their nutrient requirements. The area of OFT known as Central Place Foraging Theory may be directly applicable to Stationary, hunter-gatherer societies. The theory predicts that people should be food generalist when hunting and gathering near home but should become progressively more specialized in foods they choose to bring home when they forage farther and farther afield. The evidence for and against this prediction is discussed. Finally, the evolutionary mechanisms which might result in optimal human foraging behavior are discussed in some detail. INTRODUCTION In recent years, animal ecologists have become increasingly int the criteria which animals use to select their diets. Many ecologists believe that animals do not select food at random, but rather select food according to criteria that have evolved by natural selection (Pyke et al. 1977; Schoener 1971). More precisely, ecologists argue that the neural] and sensorimotor mechanisms which, in large part, determine food choices, have evolved by natural selection to maximize Darwinian fitness. This viewpoint has led to a body of ideas known collectively as Optimal Foraging Theory. The purpose of this paper is to ask whether Or not this theory might be useful to human ecologists studying human diets. ; The usefulness of Optimal Foraging Theory to animal ecologists has largely been that it has allowed them to organize their thoughts and to ask new questions. Ecologists have long argued that food selection is, in some ill-defined way, adaptive. Optimal Foraging Theory has focused the adaptation picture by forcing i ig t precisely what they een by adaptive behavior. The result is that, for tl ul logi f lating testa hypotheses about food selection. These hypotheses are being tested in laboratory and field a DISCUSSION Mechanisms of Evolution Even in the case of non-human behavior, ecologists have usually avoided the difficult Questions of mechanisms. Ecologists have argued that adaptive foraging behavior results om natural selection working on heritable variation, but they have avoided the questions 62 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 of how, in learning animals, behaviors are transmitted from generation to generation. Pyke et al. (1977) state, without further elaboration, that ‘‘the selection that has so far been considered, either implicitly or explicitly, is Darwinian natural selection coupled with genetic inheritance, but the evolution could also be cultural and yet gaverned by selection.” In the present paper, I consider 3 possible mechanisms of evolutionary change. These are: 1) Genetical evolution of learning mechanisms, 2) Cultural retention of individually adaptive behavior, and 3) Cognitive evaluation and retention of beneficial customs. The first 2 mechanisms are variations on the old theme of natural selection. The third requires no differential fitness yet pees adaptive behavior. a er present my personal view of the relative importance f human diets. Genetical selection is the change in gene frequencies accompanying the differential survival and reproduction of individuals. A coherent viewpoint of the genetical evolution of behavior has been developed by ethologists and, more recently, by sociobiologists (see Wilson 1975 and Tinbergen 1951). This viewpoint, in my opinion, reveals little about the behavior of learning organisms, particularly humans. a basic tenet of sociobiology has been that genes may cause animals to behave in certain ways which propagate those genes in future generations. Human genes do not cause humans to behave; however, genes do influence how humans learn. No doubt, some ways of learning are more adaptive than others. The real problem is to specify how genes influence learning (see Pulliam and Dunford 1980). The likelihood that an animal repeats a particular behavior depends, in part, on the reinforcement it experiences. The sensory and neural mechanisms of reinforcement may have evolved by genetical selection in 2 ways (Pulliam and Dunford 1980). Ens, genes specify certain primary reinforcers which guide early learning. Second, g y certain learning programs which control how experience with reinforcers i is integrate ed. For example, the taste of mother’s milk is a positive primary reinforcer. That is, the genes specify certain connections between the sensory elements which detect this taste and the central nervous system mechanisms which evaluate experiences as positive. A child learns that certain behaviors increase its likelihood of access to this reinforcer, and that the other be stimuli consistently sabctasc with the primary reinforcers shall become secondary reinforcers. Thus, the visual image of the child’s mother become a secondary reinforcer by association with mother’s milk. Once this association is perceived, the child will work for the reward of access to secondary reinforcers just as it will work for access to primary reinforcers. The relevance of primary reinf food selection is tk i inf ide the learning of food habits. Certain tastes, such as from low concentrations of salt and inter- mediate concentrations of sugar, positively reinforce the eating of certain foods. Similarly, the taste of ae compounds negatively | reinforces eating — ane Circumstantial evidence for from a variety of studies. For RES C.M. Davis (1928, 1939) found that human infants grew normally and maintained good health on a diet they selected for themselves. More carefully controlled experiments with naive rats have shown similar results. (For a revieW, see Nachman and Cole 1971). If ov mechanisms of reinforcement have evolved b Iselecti daptive g, then the reinforcing properties of food should change as the nutritive needs ¢ of the individuals change. The t evidence in favor of this view of abits comes from studies of specific hunger in laboratory animals. Numerous animal coal support the idea that = with a particular nutritional need “learn to develow . preference or aversion t (Nachman and Cole 171: 340). For example, rats given a diet deficient in thyamine ae a preference for any food containing thyamine. Similarly, after several days © salt a — -_— May 1981 PULLIAM 63 deprivation, sheep aocept previously rejected, salt-rich foods. The evidence for in humans ial. For example, a 3.5 year old boy with a severe adrenal deficiency maintained himself by eating salt by the handful (Nachman and Cole 1971). Th tly after hi by physicians who thought that such high salt intake could not be good for the child. The specific cravings of pregnant women have also: been interpreted as specific hungers for required ae All inall, very little is food selec “Although Li and | specific hungers are surely important to human food selection, their social learning. Parents not only can control what ch their children hee available to eat, but they also use many forms of persuasion to influence their children’s eating habits. P by the selective presentation of reinforcers. This process of strong parental influence leads toa second ane of evolutionary change: selective retention of adaptive behavior. in recognized the ¢ importance of social rancid to human behavior when he wrote that facial Pita ‘serve as t ther and her infant; she smiles approval and this encourages her child on the right path, or frowns disapproval” (Darwin 1896: 364). Darwin’s viewpoint is echoed by the modern theory of social exchange. According to George Homans, an eminent sociologist, human values are “learned by being linked with an action that is successful in obtaining a more primordial value” (Homans 1974: 27). Homans (Ibid:27) goes on to say: uppose a mother ofteri hugs her child and getting hugged is probably an innate value - in sie aay in which the child has behaved ere wien other children, and, as _ er says, ‘better.’ The ‘behaving better’ than othe to become, as we say, ‘rewarding in itself.’ By va a processes of — men may learn ane maintain long chains of behavior leading to some ultimate reward. The important point that I wish to draw from Homans is ie parents can control what their children learn by manipulating their social experience. Some of the ‘values’ acquired by social learning are what I call ‘ideas’ about behavior. An idea about behavior is what a Person perceives as the relationship between behavior and access to rewards. Children learn ideas about behavior during the socialization process because adults, particularly their Parents, control their access to positive “we nenatiee: reintorcets. What does social | lear ning have to do ior? Parents tend to teach their children the same ideas that they once 4 from their own parents. Thus, acquired ideas may be passed from generation to generation. Since ideas are perceptions about the relationships between behavior and rewards, ideas motivate behaviors (Pulliam and Dunford 1980). Some behaviors affect individual fitness, i.e., they affect an individual’s chances to survive and reproduce. Ideas that motivate behaviors which affect individual fitness ie affect the likelihood that those same ideas will be acquired by the next 8eneratio As iebicesonty stated, a basic tenet of sociobiology has been that genes can cause people to have in ways that increase or decrease the likelihood that those genes are replicated in future generations (Dawkins 1976). I argue that ideas can motivate people to behave in ways that affect the likelihood that those ideas are replicated in subsequent generations. Furthermore, I argue that such cultural selection of i evolution of human behavior during the past million years than has genetic selection. umans have many ideas about what to eat and what not vi eat. Decisions to avoid cating certain is or to eat beans w ith ri rice, ea to eat insects yO decisions which p | poter atially aff tion. The ideas which motivate tions of the a between feeding behavior and iii _ o> welfare. Some i ideas about food selection! have no doubt been “‘weeded out” of hu Other ideas have been selectively + retained, i.e., replicated: in subsequent generations, because they motivated adaptive behavior (see Durham 1976 and Ruyle 1973). 64 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 ening: enough, — ee ideas which evolve by selective retention are percep ween behavior and welfare, they may be misconceptions and yet lead to adaptive aa ome For example, an herb may have been added to the pa” to appease the gods. The idea to do this may have b 1 from generation t because the herb contained a rare vitamin i d inf, hip Parents who defied the gods were unlikely to have children who survived long enough to learn their parents’ non-orthodox beliefs. The point is that cultural retention of an adaptive behavior may be based on misconceptions about physical reauity. p= fs} Lildren of parents who had maladaptive ideas. Cultural evolution need not be such a blind If people correctly perceive the relationships between their behavior and their own welfare, adaptive behavior may evolve without any selective deaths. This is what I refer to as the cognitive evaluation and retention of beneficial customs. Cognitive evaluation and retention of new ideas (i.e., of the relationships between new behaviors and welfare) depends on both individual perception of the welfare of others who have already evaluated the new ideas and individual perception of the consistency of new ideas with those already personally evaluated. The way in which these 2 factors influence food habits can be better understood by briefly considering the sociological theory of cognitive balance. How I decid kind of food? The evaluation of the new food can be analyzed using the P - On X cycle of balance theory (Davis, C.M. 1939 and Heider 1958). P is the person making the evaluation and X is the new food being ve (more precisely, X is the idea that eating the new food is beneficial). O is a si sed in making the evaluation. In this case, O is another person, institution or idea. orThe evaluation of the new food (X) depends on the person’s previous evaluation of O and the person’s perception of the relationship between O and X. For example, if a person's respected friend accepts a new food, then the person P is more likely to try it too. On the other hand, if a new food is sanctioned by the church O, then the person will try it if the person evaluates the church positively, and may reject it otherwise. he purpose of bringing up a cognitive balance theory is that I think it sheds some new light on the question of when human behavior will be adaptive. I have argued that humans have genetically-inherited learning mechanisms which also play a role in the evaluation of new foods. If cognitive balance leads a person to try a new food, the food may be reevaluated according to its taste. A person will only accept a new food that “‘tastes bad” if the person b: evaluation of the significant other (e. g., church or friend) outweighs the negative reinforcement of the taste. For example, the behavior of eating a noxious-tasting medicine when ill, because the doctor recommends it, will only persist if evaluations of doctors = very pasate’: Furthermore, the pe enim - “doing what the doctor says is best” will b leads, on average, t0 an increase in individual fitness, or, at least, toa perception of increased welfare. I am proposing that the initial evaluation of new ideas depends on their own consistency with older, more established ideas, but that new ideas must also lly pass the tests of individual reinforcement and cultural retention. Primary reinforcers are genetically- inherited guides to behavior that slow down the acceptance of potentially dangerous aie ideas even if the new mesiee are ‘Consistent with established ideas. Furthermore, seas even once totally accepted, may b y ptive behaviors. This is lividuals who have learned maladaptive ideas are less ms es! pass their ideas on to children of ie next sie tence: Asa result, many more or less neu ideas may be retained from generation to generation but those which truly lower the Darwinian fitness of the idea bearers will be gradually “weeded out” (Durham 1976). The atlective sieve of cultural retention can aa be expected to operate efficiently phe _ Pp i den ile Oo riwree ye 1, ° Cc ’ g ideas mail be more saad more likely to ee May 1961 PULLIAM 65 h s = rape = 3. = taut = ae | A society whose ideology is well adapted can quickly adapt to minor changes in conditions. rman it conditions 7 t ] i ee lad 1 h “4 1 ee | ‘ s a é s _ “¥ * ry 52% 1 kh 4 iaening yet other ideas. For is only in stable cultures of people that have inhabited the same region for many genera eratio ons. Optimal Foraging Theory may help human ecologists predict what foraging behavior is adaptive in a particular environment. The question of how well behavior is adapted to an environment can then be resolved empirically. u Citic, - Optimal Foraging seit Optimal Foraging Theory consists of a set of i which can be made precise and which are, therefore, testable. For example, it seems reasonable that energy- Stressed animals might make foraging decisions so as to maximize their net rate of energy intake while foraging. An optimization model turns this general supposition into a set of precise predictions hes can be vinteg or eee his igiene to real data. Energy not mea much as they can, whenever they can. As applied to humans, the as states that when energy-stressed, 1) people choose food so as to harvest as much energy as possible during the time they devote to hunting or gathering, or 2) people hunt and gather in sucha manner ast inimize the ti juired tth energetic requirements. f course, minimizing foraging time also maximizes the time available for other activities (Smith 1979). The theory makes both q itati d ive predictions. I expect that, in general, the quantitative predictions will oaks be veel for human societies during periods of €xtreme food shortage. Nonetheless, the qualitative predictions of Optimal Foraging Theory are likely to predict trends in the foraging behavior of all hunters and gatherers. No doubt, during some periods food is abundant and people are not so much energy- limited as they are nuntrient-timited. During such periods, models of protein maximization Or energy maximization with nutrient constraints (Pulliam ngs might be more appropriate. Of course, the theory is only as g its nd for any particular study must be based on uid field studies. aaaie most of the qualitative predictions are the same for energy maximization models and nutrient Maximization models. It is these robust, qualitative predictions that are most likely to Predict trends in human foraging behavior. In the case of energy-maximization, the value to a forager ofa particular prey item is defined as the ratio of its energy content to its handling time. Handling time is the average total time required to pursue and capture a food item ane it has been encountered. - the case of human foragers, handling time would ist mostly of g I g for a hunter, and digging and picking time for a gatherer. A forager that maximizes net rate of energy intake, pursues prey with the highest value €very time they are encountered. The decisions to pursue or ignore prey of lower value depend only on how frequently the higher-value prey are encountered. One of the most robust predictions of the theory is that when high-value prey are rare, the diet expands to include more low-value prey, and, conversely, when high-value prey are common the diet contracts. A number of animal studies have tested qualitative and quantitative predictions of Optimal Foraging Theory. For example, Werner and Hall (1974) offered bheee —— a choice of prey of 3 different sizes. They n “= ae ne available in the aquarium and siitionel when the fish should eat and when t ey shou ignore the 2 kinds of lower-value prey. They found that not only did the fish diets expand 66 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 when the highest-value prey were rare, but hat the fish b hel value prey at the abundance of highest-value peey tha was predicted by the e theory. Krebs etal. an presented Great: g 1] ] They varied th the prey a bundanc by putt belt h d prey by the des at sdjtiaeoe saan The ask as ta that the nace to eat or ignore the small mealworms was independent of the abundance of small prey and dependent only on the abundance of large prey. As expected, when mealworms were common, the tits ate only large ones, but when mealworms were rare, they ate both large and small ones. However, the theory predicted a specific abundance of large prey at which the tits should quite suddenly change their behavior and eat every mealworm presented. Instead, Krebs et al. (1974) found that the small prey were only gradually added to the diet. The results of this experiment support the qualitatiave predictions of the theory but not the quantitative predictions. A few investigators have tried to test Optimal Foraging Theory under more or less natural conditions in the field. For example, Goss-Custard (1977) studied Redshanks foraging on invertebrates on a natural beach. He found that whether or not these birds accepted small prey depended only on the abundance of large prey and not on the abundance of the small prey themselves In a study of Fee ee Sparrows in a natural oak woodland (Pulliam 1980), I found that these sparrows preferred seeds of high energy value and expanded their diets when high- energy prey were less abundant. However, I also found that the prey were not eaten in the same frequencies as predicted by theory. So again, the qualitative but not the quantitative predictions of the theory were supported. How close animals come to matching the quantitative predictions of the theory seems to depend on how energy-stressed they are. This is shown dramatically by the experiments of Caraco etal. (1980) on risk aversion by Yellow-eyed Juncos. The theory of risk aversion is an extension of classical optimization problems to the situation where an animals must choose between food of high-value and high-risk and food of lower-value and lower-risk. The theory predicts that animals will be more likely to maximize energy intake even if a risk 1s involved when they are energetically stressed. Caraco and his coworkers gave selow ers] Jun hoi ne side ofa ther side to receive 7 seeds half of the time, but no seeds the other half. The expected reward was then 3 seeds on hcg cow reward a vane Ci side and 3.5 seeds on the “high reward, high risk”’ side. As predi gh- risk strategies when ‘energetically stressed and low-risk strategies ‘when not cereal stressed. I PAC Cases We bese vv ill Spry tv SaLGALLY CASALL humans A final 1i hich 1 eo ae ee £ shat iS known to ecclogists as Central Place ee Theory, Many foragers start from a central place such as a nest, a den, or a village to which they return with food. The theory predicts how prey choices will vary as a function of prey abundance and how far the foragers will go from the central place. Optimization models of central place foraging predict that as prey abundance declines and foragers go farther from the central place, they prey they choose to bring back. This is because once the forager has travelled far from a central place, any extra time required to select a better prey may be small compared to transit time to and from the central place. This prediction holds regardless of whether the foragers are eneTgy- maximizers or nutrient-maximizers. CONCLUSION During much of the history of mankind, leh i by | d gathering During periods of food shortage, individual survivorship has, ne doubt, often “depended critically on individual decisions about which foods to hunt and gather. These decisions - —_ May 1981 PULLIAM 67 Asés YY 1 ] kK s ee! I 2 have been made, in part, by reference tot techniques, favored hunting grounds, alternative foods, e Traditional knowledge about hunting and ee consists of culturally-inherited ideas about the relationship between Ranging behavior and individual welfare. If ideas leading to adaptive behavior have been more likely to be culturally retained as part of traditional knowledge, then nara trends of foraging behavior should be predictable by Optimal Foraging iota ory Among the litati ici f Optimal Foraging Theory as applied to Stationary human hunters wits gatherers who have inhabited the same region for many generations are the following: 1) Human foragers should become more selective in their prey choices as the abundance of preferred prey increases; 2)Decisions to Aeagiaa or generalize the diet should be independent of the abundance of less preferred p 3) During hl ne food shortage, prey pref hould roughly be ordered ding the ratios of energy content to handling time; 4) Human foragers solanaine be more willing to take risks for high energy gains during periods of food shortage 5) People should be food generalists — hunting and ae near home (a central] place) and become p g = they forage farther ‘afield. . Probably, human foragers never exactly maximize their rate of energy gain while foraging. Nonetheless, their behavior is probably much closer to this ideal than it is to well the behavior of hunters and gatherers is adpated to their energetic requirements. If the qualitative predictions listed above are supported, the quantitative predictions should also be tested LITERATURE CITED Caraco, T., S.P. MARTINDALE AND T-S. Brace, es vich. WHITTAM. 1980. A empirical demonstration of | KREBS ascend AND ER; CHARNOV. 1974. risk sensitive foraging preferences. Anim. Hunting b 5 A Behavior 28:820-8 tudy of patch use by chickadees. Anim. Davis, C.M. 1928. Self shlection’ of diets by newly ain Lae weaned infants. Amer.J. Diseases of Children © MAYNARD ean een Optimization theory in 36:65 1-679, evolution. — Rev. Ecol. Syst. 9:31-56. 1939. Results of the self-selection of diets | NACHMAN, M. AND LP. COLE. ia Role of taste ri young children. Canadian Med. Assoc. J. in specific hunger Handbook of Sensory 2257. Physiology 4:337- rive J.A. 1963. Structural balance, mechanical PULLIAM, H.R. ‘ee Diet optimization with solidarity and interpersonal relations. Amer. J. nutrient constraints. Amer. Natur. 109:765-768. Soc. 68:444-462. 1980. Do chipping sparrows forage Darwin, C. 1896. The Expansion of Emotions in optinaally? Ardca 68:75-82. Man and Animals. D. Appleton and Co. AND C. DUNFORD. 1980. Programmed to Cai, R. 1976. The Selfish Gene. Oxford Learn: An Essay on the Evolution of Culture., Univ. Press. Columbia Univ. Press. DurnAM, W.M. 1976. The adaptive significance Pyke, G-H., H.R. PULLIAM AND E.L. CHARNOV. of cultural behavior. Human Ecol. 4:89-121. 1977. Optimal foraging: a selective review of Goss-Custarp, J.D. 1977. Optimal foraging and theory and tests. Quart. Rev. Biol. 52:137-154. € size selection of worms by redshanks RvyYLe, E.E. 1973. Genetic and cultural pools: (Tringa totanus). Anim yeas 25:10-29. some suggestions for a unified theory of bio- HEIDER, F. 1958. The Psychology of Inter- cultural evolution. Human Ecol. 1:201-215. Personal Relations. John Wiley and Sons. SCHOENER, J.W. 1971. Theory of feeding Homans, G.C. 1974. Social Behavior: Its strategies. Annu. Rev. Ecol. Syst. 11:369-404. Elementary Forms. Revised ed. Harcourt, SmirH, E.A. 1979. Human adaptation and 68 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 energetic efficiency. Human Ecol. 7:53-74. WILSON, E.O. 1975. Sociobiology: The New TINBERGEN, N. 1951. The Study of Instinct. Synthesis. Belknap Press of Harvard Univ. Oxford Univ. Press. WINTERHALDER, B.P. 1977. Foraging strategy WERNER, E.E. AND D.J. HALL. 1974. Optimal adaptations of the boreal forest Cree: an foraging and size selection of prey by the evaluation of theory and models from evolu- bluegill sunfish (Lepomis mochrochirus). tionary ecology., Unpubl. Ph.D. Dissert. Ecology 55:1042-1052. Cornell Univ., Ithaca, New York. — ae a 2 ARR rere J. Ethnobiol. 1 (1): 69-83 May 1981 RESOURCE UTILIZATION AND FOOD TABOOS OF SONORAN DESERT PEOPLES AMADEO M. REA Curator of Birds and Mammals, San Diego Museum of Natural History, San Diego, California 92112 ABSTRACT.—Resource utilization and food taboos of 8 Sonoran Desert cultures (Riverine Pima, Papago, Sand Papago, Pima Bajo, Seri, Colorado River Yumans, Maricopa and Western Apache) are compared. Taboo (or dietary prohibition) herein i ding l (species banned to the entire community) rather than specific sense (species banned to a particular age and/or sex class at specific times). The purpose is to compare nutritive resources (plant and animal species) available to 2 or more cultural groups exploiting the moran desert. Western Apache, with the greatest number of taboos, had access to more ecotone resources and to more than one major life zone. Within the Pima-Papago cultural complex there is a probable underlying adaptive (ecological) basis for the fact that most restrictions were found with Riverine Pima (resource-rich ecotone habitat) and the fewest restrictions with Sand Papago (most harsh habitat of groups considered). There is a probable i. Pie re. Gast Bes, jp } * £ a: PRES rey ry i ma and no. agricultural resources). Speakers of mutally intelligible languages, even though disjunct geographically, tended toot I imal taboos if t — impoverished. Plant use was cross-cultural. Tab y & INTRODUCTION A number of factors determine the dietary items any heterotrophic organism utilizes as food, including its own anatomical mechanism for obtaining the food, its physiological ability to assimilate the food, and the availability of the prey items themselves. All these factors play a major role in the dietary of man, an omnivorous animal. But we cannot stop there. Several factors radically alter the dietary categorization of man‘as an “omnivorous animal” and these are culture and language. It is almost an anthropological maxim that man’s diet is not simply determined by his anatomical and physiological ability to handle prey items (both plant and animal) that happen to be available in the environment. All humans that we know live in a cultural context, speak at least one language, and practice dietary selectivity. (Our own culture provides examples of rigorously observed but unwritten, perhaps even unconscious, rules specifying dietary selectivity; see appendix.) In addition to culture and language, man, especially in “archaic” societies, differs from other i inadi i ight call a sense of the sacred (Eliade 1959; Rappaport 1971). All 3 modify diets. Dietary selectivity has been discussed for a number of areas of the world, but to my knowledge there has been no intercultural comparison made of the aboriginal peoples living in the Sonoran Desert of the American Southwest (Fig. 1). ee A number of questions came to mind when I decided to look at the variation in dietary resource utilization and taboos in desert peoples: 1) Did the utilization of plants as well as animals differ from one group to another? 2) Insofar as there were shared resources, were there dietary differences between groups Speaking closely related (even mutually intelligible) languages, as between the Riverine ima, Papago, Pima Bajo, and Sand Papago or between the Maricopa and the Colorado River Yumans? 3) Does agriculture-modify the range of wild foods hunted and gathered? ) f 4) Do dietary restrictions arise because of ecological determinism or do they arise and unction as symbols of group identity? age és relatively greater Besse of a ate subsistence data is available on Amerindians living in tropical areas (Carneiro 1968; Ross 1978; Chagnon and Hames 1979; others) even 70 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 q Approximate Ranges . 1 Colorado River Yumans 2 Maricopa 3 Riverine Pima 4 Western Apache 5 Sand Papago 6 Papago 7 Seri 8 Lowland Pima Bajo ee Sonoran Desert Fic. 1.—Localities (ca. 1850) of 8 groups discussed. (After Rea 1979a). May 1981 REA 71 though considerable investigation remains (cf. respondents to Ross 1978). Tropical situations are characterized by high species diversity but low population numbers (i.e., there are more species but fewer individuals per species than in temporate forest communities). Relatively little is known of resource utilization in New World deserts where human cultures might be thought of as marginal. More is known of subsistence in Old World deserts (e.g., Kalahari, Australia). Even though absolute quantitative intercultural data from the American Southwest are now no longer retrievable, the comparative qualitative data presented here are probably largely valid and useful. (We must realize in any comtemporary study that the Sonoran Desert today is an artifact; community structure since the Pleistocence has been radically altered due to | f megaf d iated animals([e.g., Rancholabrean birds] and, more immediately, a century of disastrous overgrazing by Eurasian herbivores. No matter how hard we try to close our eyes to the facts, Southwestern deserts have been radically altered by the last century of abuse!) Some philosophical debate surrounds the topic of the origin and function of food taboos. That taboos are cultural inventions is incontestable. But do they function as cosmological symbols in a culture, maintaining as orderliness and structural indentity? Are they understandable as daily, visible projections of values emanating from a common metaphysics? Or can they be reduced to some evolutionary fitness factor, fulfilling (unknown even to the adherents) a sanitary or hygenic function, or perhaps an ecological function of sustained yield or predator strategy theory? Can why man eats what he eats be understood in terms of nutrition and calories? Can what people hunt, gather, or grow be reduced to terms of cost-benefit? The dietary restricitons of Sonoran Desert cultures might Suggest some answers. METHODS eh eee . id dh For comparative purposes onl p sins ; more cultural groups had access (see Table 1). Hence, Prairie Dogs (Cynomys spp.) ’ Western Apach ttaken int here t =o Comparisons are not to be taken in an absolute sense because (1) relative’ abundances of various prey items aboriginally are unknown, and there i devid eee ae of plants or animals are now decimated or locall irp t iinet 7 the desert habitat (Hastings and Turner 1965; Rea 1977); (2) the relative importances of Specific taboos are unknown; (3) the coverage of even absolute taboos weiss il ca of the data (specifically for Westen Apache and Yumans) are from the literature at st been verified. Literature citations are sometimes limited by the investigator's incomplete understanding of his own Linnaean taxonomy and the ethnotaxonomies of his native informants, and (far too often!) incomplete interrogation. Also, Senna ened — have had varying, usually long periods of contact with Europeans. This — “— . technologically dominant culture has resulted in ag labandon aE ways. Of the groups considered here, only the Seri preserve a viable native subsistence pattern (though all the others preserve parts). Data for the Riverine Pima, Papago, and Pima Bajo are based on my own field work. Table 2 gives the life zones and ecotones that the 8 8toups exploited. oe t te, DISCUSSION Kinds of Taboos—Food taboos are equivalent to dietary prohibitions. Food taboos Considered here are only general taboos, that is, those imposed on the enure € thnic group at all times. Excluded from discussion are specific taboos that restrict a particular food species for a specific age/sex class of the population at a particular time. Such prohibitions may affect, for instance, only men while hunting, or women while nee tiie lig RE menstruating. (A study of these taboos here now would take us too far ree this Practical restriction is not to imply the lack of importance, either symbolically or ecologically, of specific or temporal taboos.) 72 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 TABLE 1.—Resource utilization by desert tribes. RIVERINE PIMA PAPAGO SAND SERI MARICOPA YUMAN WESTERN PIMA BAJO PAPAGO SPEAKERS APACHE Grasshoppers ag ? dy ? — 49 T ? Acrididae + ? ? ? eee ? ? ? idae Caterpillars + + + + + + * * Celerio lineata Mullusks 0 0 0 + + 0 + 0 Mull Fish a > 0 - + + - Mud Turtle + T T + aoe ? z ? mos Desert Tortoise + + + + + + 4% +? h agassizi Gila Monster gt ? i ? ~ ? ? ? Helod pectum t Ig T 0 + wv ? + 0 Chuckwalla 0 0 #8 + + 0 + +7 uromalus obesus Small lizards T gi +. bs ty ? + T Iguanidae, Teidae Snakes (all) tr T 8 ? +/T* T E T Colubridae, Crotalidae ks & Geese Zs ? 0 ? + + + + iformes wks & Eagles ui ? T i + ? ? zt Accipitridae Vultures a _ zy i i ed ? ? T Turkey 0 + 2 0 0 3 0 3 Meleagris gallopavo Quai + + + + s s: . + Odontophorinae Herons & Egrets T ? 0 ? + ? ? T Ardeidae + + + + * ry + sd Colum Roadrunner 7 ? T 2 T ? ? ? fen life Great H 1 + T? T ? + ? ? T Bubo virginianus Is ry ? ? ? + ? ? T Strigiformes ? ? £3 ? T ? ? t Corvus spp. Other Songbirds + ? ? 2 + + + i Passeriformes Birds’ eggs + + H is ? +* T* + +* Cottontail & Jackrabbit + + + + + + + i lvilagus, Lepus Ground Squirrel - + + ? - ag + 7 Otosperma, Citellus, Ammospermophilus Gopher _ + ? 0 T 2 + T Small mice + + ay + eee ? + % Perognathus, Peromyscus t T 0 T ? 7 4 ? + . Di: Beaver + + 0 0 0 + - £ Castor canadensis May 1981 REA 73 | TABLE | Continued i RIVERINE PIMA PAPAGO SAND SERI MARICOPA YUMAN WESTERN . PIMA BAJO PAPAGO SPEAKERS APACHE Muskrat. + + 0 + 0 ? + ; Ondatra zibethicus pine rT 0 0 ? baie! T Fi + Erethizon dorsatum le a § > i a ? i ad a +? T? Canis latrans i -_ s ? ‘i ¥ Es sy ‘anis familiaris Kit & Gray Fox E ? T ? + vg + i Vulpes macrotis, 7 Deore M Black Bear + 0 T ? 0 y iy ? 1% Ursus americanus Raccoon + + 0 0 + + + + q Procyon lotor : Badger +? + T ? + T + + Taxidea taxus 7 ; Skunk F + 5 ? t yi yy + Mephitis spp Puma + * + + + = ‘ + | Felis concolor i Bobca: ¥ +? + + + z ss + Lynx ru “pd + + + ? + + ¥ Y ‘% ’ Mule & White-tailed Deer + + + + + - . bi j hemionus, O. virginianus . + Fs < ’ + + + + : Antilocapra americana Bighorn + 0 + 7 + - + 7 Ovis canadensis Cauail + ? 0? 0 eee + + Typha spp ‘ P P Grasses + ? + + ' Palin 0 i 0 ~ 0 + 0 vm 0 ? + + + + + ; Yucca baccata, > Y. arizonica A A . Century Plant + + + + . Agave cn Spp. oe 0 0 vo) + q Quercus spp. > > Mistletoe ¥ ? + + + ; Phoradendron californicum . ter Greens + +? ~ 0 ? . 7 Monolepis, Atriplex tri, Pic > > + + + 0 Allenrolfea occidentalis ‘ + + > > * Amaranthus, Trianthema, etc P . > Velvev/ Honey Mesquite - + + i ‘Osopis velutina © P. juliflora 0 Screwbean + 0 0 0 = 7 Prosopis pubescens ? . 0 Palo Verdes + + + ™ Cercidium > Hog-Potato send + 0 . 0 Ai r Hoffmanseggia oo + + + ? + + + 74 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 TABLE 1 Continued RIVERINE PIMA PAPAGO SAND SERI MARICOPA YUMAN WESTERN PIMA BAJO PAPAGO APACHE PEAKERS Graythorn + + + + Ps + 0 Condalia lyctoides” Saguaro + + + rs + “e + Carnegia gigantea Organpipe cactus 0 ‘ + + 6 0 0 Lemaireocereus Barrel cactus + + + + + ? 0 Ferocactus spp Prickley Pear + + - * ep + + Opuntia spp Cholla bu + + + + + + + Opuntia spp food 0 0/+ . 0 0 + 0 Ammobroma sonorae lfberry + 0 + + + + 0 Lycium spp. Chiltepin +o + + +98 ? 0 + Capsicum annuum Broom-Ra # 0 + + + ? 0 Orobanche spp. Wild gourds * Be + 0 ¥ + T+ tee: bot, ft, Osas: . & C. digitata Sunflower tal 0 ae 0 ? + * Helianthus spp. KEY TO SYMBOLS: 0 ism d i z explained in text ? no information trade item + utilized T absolute taboo low a tili conflict in literature ie Se a field FE, 1OTA-U preparation techniques for edible plants Taboos are a luxury. I think there are eaaieas caotsineces ser relatively ah ss peoples with agriculture (Pima Bajo, Riverine Pima, Yu a good number of animal species, whereas those from the more sea areas (Sand Papell and perhaps Seri) had to be more generalized in their dietary selections. For instance, the Sand Papago survived because they relaxed a strong reptile taboo of their ecologically richer eastern relatives. A strictly ecological approach (cost/benefits, optimum yield) fails to explain the function and maintainence of taboos, but I think does explain their limitations. Simoons (1967) has ile “‘biol ee s for the major meat taboos of the Old World. I would agree with Thurton (1978): ‘Food tabooes will show a certain minimum opi o fit with ecologeal: ise technical ‘realities Iti is, of pepe: h babl a major key Sa ? useful to have s oe = Ghemeehves:’ tth I I ic writers would like to make them out to be. They have no qualms about exploiting a species to extinction and undoubtledly have in many instances in the past (cf. Paul Martin’s Pleistocene overkill theory). The Zuni hunt flickers and orioles extensively to obtain feathers for ceremonial prayer sticks, I am sure with little regard to whether they are decimating the local population. I find it impossible to accept the ne (contra Ress 1978) that an Indian culture invents a taboo in order t When the lasts sea or is harpooned in the Gulf of California, the Seri will enjoy the feast j j predecessors for cow VS. dees, ) An arbicrany defined category 0 of “‘game”’ mountain lion, rodents, coyote, bese de never F peaches an American table. 5) Muscle tissue is acceptable; organs are = eed (e.g., roasts and other meat cuts vs. intestines, pancreas, gonads, brain.) Liver is the standard ex xception. Note that in the mnie Aen eS folk As a result of these rules, strictly tabooed for the vast majority of U.S. Ansericane are: dog, cat, horse, bare, rodents, crayfish, bear, , songbirds, insect larvae, snakes, lizards and many so-called Raper — these cultural taboos. LITERATURE CITED BAHR, DONALD M., JUAN GREGORIO, Davip I. BUSKIRK, WINFRED. 1949. Western Apache LOPEZ AND ALBERT ALVAREZ. 1974. Piman subsistence economy. Unpubl. Ph.D. dissert., shamanism and staying sickness (Ka:cim Univ. New Mexico, Albuquerque. Mumkidag). Univ. Arizona Press, Tucson. Povervetsi RoBERT. 1968. Slash-and-burn Basso, ELLEN. 1973. The Kalapalo dietary system. cultivation among the Kuikura and its Atti del XL Congresso Internazionale degli implications for cultural development in the ire 2:629 Amazo in. Pp 145, in Man in BOHRER, VorsILA L. 1975. The prehistoric and Adaptation: The Cultural Present (Yehudi TEES role of the cool-season gras e Cohen, ed.), Aldine Press, Chica t. Econ. Botany 29:199-207. TTER, EDWARD F., AND WILLIS H. BELL. Brown, Davin E., AND CHARLES H. Lowe. 1978. 1942. Pima and Papago Indian Agriculture. Biotic communities of the Southwest. U.S. Inter-Americana Studies, Vol. 1. Univ. New pt. Agric. Forest Service, Gen. Tech. Report. Mexico Press, Albuquerque. 1. LE: OT i ae =, ey ee a ae May 1981 OE 51. Yuman Indian agriculture: Siauitive agg on the Lower Colorado and Gila Rivers. Univ. New Mexico Press, Albuquerque CHAGNON, N AND RAYMOND B. HAM 9. Protein deficiency and triba in Amaz new lence 203:910-91 sicoins eho sacred and the profane: the nat arcourt Brace Jananovich, New York and pce FONTANA, BERNARD L. 1974. Man in arid lands: the Piman Indians of the Sonoran Desert. Pp. 489-528, in Desert Biology, Vol. 2 (George W. Brown, Jr., ed.). Academic Press, New York. S, JAMES R., AND RAYMOND M. TURNER. 1965. The chanc; mf irre este REA 83 avifauna of the Gila River Indian Reservation, central Arizona. tnt Ph.D. dissert. (Zool.), Univ. Arizona, T .1979a. The yon of Pima fields. Nemes Southwest 484:8-1 b. Hunting lexemic categories of the ue preg Rive 44:113-119. Indians of Arizona. ypuene deposited: Paleoenviron- acbtal Lab., Dept. Geosci., Univ. Arizona, Tucson. Ross, ERIC BARRY. 1978. Food taboos, diet, and hunting strategy: the adaptation to animals in rere Ripe ecology. Current Anthrop. 19(1):1- RUSSELL, aaa 1908. The Pima Indians. Twenty-sixth annual report of the Bureau o American Ethn ology. Reprinted 1975 with introdi iction, vegetation change with time in the lower mile of an arid and semiarid region. Univ. Arizona Press, Tucson. HRDLICKA, ALES. 1908. Physiological and medical observations among the Indians of south- western United States and northern Mexico. Bur. Amer. Ethn. Bull. No. 34. Linares, OLGA F. 1976. “Garden hunting” in the American nt Human Ecol. 4:331-349. LEE, RicHarp B 1968. What hunters do for a living, or, eae to make out on scarce resources. Pp. 30-48, in Man the Hunter (Richard B. Lee Paace Irven De Vore, eds.), Aldine Publ. Co., NARHA, ( Gary P. 1978. Tepari bean domestica- Kio acutifolius evolution. Unpubl. M. 5 Thesis, Univ. Arizona, Tucson. Rappaport, Roy A. 1971. The sacred in human evolution. Annu. Rev. Ecol. Syst. 2:23-44. » AMADEO M. 1977. Historic changes in the by Bernard L. Fontana. Univ. Arizona Press, Tucson. sone FREDERICK J. 1967. Eat not this flesh: food avoidance in the Old World. Univ. Wisconsin Press (Madison, Milwaukee and Lon SPIER, ant) 1933. Yuman tribes ik the Gila River. Univ. Chicago Press, Chicag' Taukz0%, Davib. 1978. iain a ‘Eric Ross’ ood s, diet and hunting strategy). ent cambio 19(1):26-27. UNDERHILL, sae” aan Univ. Press, New 976. Singing for power: the song magic e Papago Indians of southern Arizona. enna 1938 edition]. Univ. California Press, tkeley, Los Angeles, London nbc James. 1968. An introduction to Hadza ecology. Pp. 49-55, in Man the Hunter (Richard B. Lee and Irven De Vore, eds.), Aldine Publ. Co., Chicago. Papago I “a Yor J. Ethnobiol. 1 (1): 84-94 May 1981 DIETARY MINERAL ECOLOGY OF THE HOPI HARRIET V. KUHNLEIN University of British Columbia, Division of Human Nutrition, Va ncouver, B.C., Canada V6T 1W5 ne nists are meeceted one ses seas s oldest Native Indian groups who 1] continue to li ands they have relinquished many of their pubes sulearal practices, including ciel related to agriculture and diet. Pood cOneMEES gi Pp beans, and squash, aug r bey SF ® = tc } A iL Hoe. us survey i 420 Hopi d children showed d that | less nan 25% sasperacat one Leyes d th t the varie Clty dramatically — that described i in the early pty die Bass changing aes 1 status. \ high levels of all “nutritionally essential elements. Additionally, Hopi cultural practices reinforced the use of unusual eancral- rich P wes erugte and salts ee were a saad beens, Leu careste id tian the diet ry I 1 If 1 - 1 whale caicium and phosphorus of | ie a *¢: ait the latter group had | higher levels of ai zinc, copper and lead. INTRODUCTION The Hopi are considered one of America’s oldest Native Indian groups who continue to live on traditional lands. Hopi villages on the mesas of northeastern Arizona (Fig. 1) have been continuously inhabited since A.D. 1150-1417, and although the Spanish and Anglo- American acquisition of Indian lands considerably reduced the Hopi food land base in the late nineteenth and twentieth centuries, cultivation of the corn-bean-squash foods conunued to furnish the major components of the diet until recent times. Previously, wild plants, animals and salt were gathered from areas as far away as the San Francisco Peaks (near present day Flagstaff) and the Grand Canyon. While the introduction of domestic livestock by the Spanish brought new protein resources, the eventual reduction of the fragile desert vegetative cover by grazing has resulted in erosion and exacerbation of moisture conditions for native and cultivated vegatation (Thornthwaite 1942). Dem ographic influences on Hopi food culture and preparation include an increasing practice of maize meal grinding by the Hopi women. Using stone mano and metate, corm grinding would take 3-4 h daily for the ge eae of enough food for one day pei the family which usually included 7-9 member. of fields using traditional agricultural pee is also recognized as a labor intensive occupation for the Hopi man from mid-April to October (Hack 1942; Forde 1931). When all Hopi foods were supplied entirely by their native environment (except for the small amounts obtained by trading with neighboring tribes), Hack (1942) and Bradfield (1971) independently estimated the required farmland per capita to be 0.8 ha in corn and 0.2 ha in other vegetables. Thus, approximately 4.2 hl of corn were consumed per person pet year (about 316 kg of cornmeal) and an equal amount was stored for lean years (Stephen 1936). In 1893 it was — that there were 1458 a in fOrn, ate he in cultivated vegetables, primarily beans and squash, and 405 ha in 2000 Hopi (Donaldson 1893). Peaches were BELSIF i! by the Spanish in the aa century and were a popular food and trade item. May 1981 KUHNLEIN 85 F HOPI/NAVAJO. JOINT USE AREA Pd = BOUNDARY OF HOPI DISTRICT ADS i“ @ INHABITED VILLAGES @ AWATOVI, ARCHAEOLOGICAL a SITE “, MESAS SCALE OF MILES BRae eR , ARIZONA FIG: l.-— The Hopi area in Arizona (boundaries in the Joint Use Area are being negotiated). For various social and geographical reasons, much of the traditional farmland has fallen out of use. As an example, the Oraibi (a Third Mesa village) Hopi cultivated about 972 ha in the 1890’s; in the 1960’s Bradfield (1971) recorded only 373 ha under cultivation by this village. In 1974, 2573 ha of “annually or intermittently” cultivated farmland were noted in the Hopi area (BIA 1974). For a population of 7500 Hopi, this averages about 0.3 ha per person. It appears, then, that the use of traditional agriculture and harvested foods has declined. tis the intent of this report to describe the ct ging itional envi ftheH with regard to selected dietary minerals. A description of general dietary change is followed by data on selected minerals contributed to several indigen pi foods t those minerals of particular interest. Finally, mineral data from seventeenth century and contemporary Hopi deciduous teeth permit comparison of human accumulation of certain minerals in preindustrial and current settings. DISCUSSION Dietary Change A dietary study of 420 Hopi women and children in 1974-75 has confirmed the above 8tographical and social trends and shown an infrequent use of traditional Hopi food (Kuhnlein and Calloway 1977a). During the survey period, less than 25% of the individual daily records contained one item of traditionally prepared foods. The composition of the contemporary diet is primarily of items distributed in commercial food networks throughout America. The work of early anthropologists which was thoroughly compiled and expanded by Whiting (1939) in Ethnobotany of the Hopi permits a summary of foods used in pre-Spanish times. This is compared to the foods of the contemporary Hopi family in Table 1. The most prevalent items in terms of q itati pu ted. Itcan 86 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 be seen that the total variety of food plants was much greater in the earlier period, and that the composition of the diet in all food categories has dramatically changed. While the diet was previously composed primarily of corn, beans, squash, seeds, wild greens, and fruits and water, today the diet is primarily beef, mutton, eggs, wheat, potatoes, some canned vegetables, fruits, and fruit jucies, lard and other fats, coffee, tea, milk, and several commercial pastries and sweets. Although the contemporary diet is not strikingly income populations (Kuhnlein and Calloway 1977a), there is a preponderance of refined cereals, animal fats, and sucrose-rich foods, all of which were not used previously. These A a . é ne ee : ee : q a, hh problems which have been documented in contemporary Southwest Indians (Sievers 1966; Sievers and Cannon 1974). TABLE 1.—Indigenous and Contemporary Hopi Family Foods® Food Category Indigenous Contemporary Fruit agave uniper, melon, peach, apple, grapes (seasonal), eeenaisiics alia, Scchles pear, range, banana, Sommer preserved squawbush, currant, rosehips, yucca AD and juices (seasonal, some preservation) High protein beans*: tepary, lima, kidney, scarlet —_ Beef*, pork, mutton*, eggs*, poultry, runner (ca. 20 var.), pinyon nuts, fish, small amt. wild game, peanut eeds* (many), wild game butter, beans*, nuts Vegetable fresh sweet corn*, wild eet corn*, snap beans*, tomato, potat pumpkin*, squash*, fresh rari wild aes chili*, squash, spina greens*; (ca. 39 sp.) (seasonal, some potatoes* Seominnesiaal and ea! grown, preservation) some preservation) Grain corn* (12-14 var., ca. 70 methods of _ wheat*: yeast bread, fry bread, quick preparation, often with culinary ash) sania tortilla, pasta, etc. (mostly com- Indian rice grass, millet ercial) corn: mush, grits, occasional old Hooi dishes (piki, nokquivi, etc.), Tice, various breakfast cereals* Separated fat lard*, margarine*, oil*, butter Dairy products milk*, cocoa, cheese, preserved milk* Other beverages _—_—water*, herb tea coffee*, tea*, pop*, sweetened juice drinks, powdered sweet drinks* Sweets and snacks sweet roots: 11 species (seasonal) milk pudding*, candy*, jello*, chips* and other deg and salted snacks, pastry*, his, biyvet) | 4Adapted from Kuhnlein and Calloway (1977a). bPre-Spa sh “Foods identified in a survey of 420 women’s and children’s diet records in 1974-1975. May 1981 KUHNLEIN 87 TABLE 2.—Approximate Percentage of Desired Daily Intake“ Provided by 200 g Portions of U.S.D.A. Commodity and Hopi Cereals. Element Calcium Magnesium Manganese Iron Zinc (need) (800 mg) (350 mg) (7 mg) (15 mg) (15 mg) Cereal (% of above contained in 200 g)> Commodity m Meal ] 23 7 36 8 White rice ] 15 34 41 16 White flour 5 15 17 35 7 Rolled Wheat 9 (es 91 45 35 Hopi Corn meal 4 80 18 43 35 Piki bread 40 102 56 121 55 Bivilviki 142 159 57 224 49 @Desired intakes are arbitrary ‘ ” re NRC all 1973 Product ined 8 dry solids, i.e. “‘as is’’ dry product basis. From Calloway et al. (1974). TABLE 3.—Composition of Hopi Chamisa Ash* g/kg mg/kg Na 1.57 Mn 710.00 K 174.00 Fe 3840.00 Ca 125.00 Cu 96. Mg 52.00 7n 169.00 P 10.80 Se 8.00 Rb 189.00 Sr 1060.00 Pb 11.00 oa he ee eee @From Calloway et al. (1974). Minerals in the Hopi Nutritional Environment Beaglehole (1937) and Nequatewa (1943) recorded that in addition to the usual —. and kfast corn foods, piki and bivilviki, the Hopi commonly prepared at least additional dishes of blue or pink cornmeal with a culinary ash. The preferred ash seni = different dishes, cooks and villages, but the most commonly used ashes are pre ng the green plant leaves and stems of cham: canescens) or dry bean pods and vines (Phaseolus sp.) isa (four-wing salt bush, Atnplex The rich mineral content of Hopi 88 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 chamisa ash is noted in Table 3. Culinary ash is also prepared from juniper branches, corn cobs or sheep dung. Today, some cooks substitute baking soda for ash, as the a produces a similar color change in the anthocyanin pigment within the cornmeal. alts.—Trace elements were also contributed to the earlier diet by the use of crude - which were gathered by the Hopi from several geological deposits in the area. Sites in the Grand Canyon and Zuni Lake were most often frequented (Hunter 1940). Mineral levels in .I1 salts from oe oe which are mite aa to have been used by Native People are given in Table 4 ple. The indigenous salts contained iron, arsenic, bromine, rubidium and strontium in substantial levels. If the average Grand Canyon sample contained about 500 ppm iron, and 5 g of salt were consumed on a daily basis, this would provide 2.5 mg of iron, which is one-fourth to one- eighth the normal human requirement (Food and Nutrition Board 1974). Absorbability, however, is not known. The effects of these quantities of arsenic, bromine, rubidium an strontium are also not known. The small amounts of manganese, copper, zinc, nickle and selenium probably had only minor influence on improving mineral adequacy of the diet. Sherds.—Ceramic sherds from the Hopi area were identified at the Harold F. Colton Research Center i in Flagstaff and analyzed by x-Tay INTERNE: The data given in Table 5 have b Calloway 1979a). These fragmenta ceramic vessels used in the Hopi area were found to contai f all trace elements analyzed. It is possible that if these were used for food vessels, some of the minerals would leach out _ foods, especially if the food pH was low. Today the Hopi rarely use ceramic container Calcitum.—This sieeath is especially interesting for study in traditional Hopi foods because milk was not used in the culture except when infants were breast-fed. It was found that there were several excellent sources of calcium in Ho s which were clearly important in pregnancy and childhood when skeletal growth a rapid. TABLE 4.—Elements in Indigenous Salts From the Hopi Area Compared to Commerical Brand. Cu eo in ee Ae a Se ik (ppm) petinn ie ON Siero aaa Zuni Lake Domestic <2 5+) «9 1345 2+1 3442 <1 442 = Zuni Lake G3-411 3+2 2+1 <8 3645 2+] 141 <1 432 . Grand Canyon G3-4195 3+2 3+] <9 15+5 12+] 420+21 =. 224+11 «2 a Grand Canyon E2345€ 6+2 11+2 36+7 1530+70 414420 1060450 19449 4512 638 Grand Canyon E2346 341 541 <11 505425 6943 554427 10845 50:3 ee Grand Canyon E2347¢ 3+1 +1 10t6 =. 360418 = -12746 = 372+18 130+6 1542 " Grand C G3.790 2+1 3+1 <8 5745 1 -288+14 85+4 542 (4 Camp Verde G3.230 8+2 3+1 <8 13+5 ra) 18+1 41 <2 +" Camp Verde G3.69 2+1 <2 <8 18+5 <1 241 <2 241 “ Medicine Cave NA863.1© 342 4+] €8 436421 <1 <1 «1 6 f° Wupatki 405M.44 “3 2+1 10+7 —-265+13 41 4+l a 542 Co-Op lodized <3 2+) <8 5t4 <) 6743 <2 oy ” eee a eee SC Except where noted, undetected elements were: Cr 12 ppm; Ni < 3 ppm;Se < 2 ppm. Zuni Lake Domestic Salt was obtained from 4 Hopi kitch d Co-Op Iodized } Sin folk wager; pales 1 ‘ a) See ey on of Norther™ Le Pp Arizona ] ber. izona 4 g € Ni = 1242 | May 1981 KUHNLEIN 89 TABLE 5.—Elements in Seven Pot Sherds from the Hopi Area. 206 Fe Mn Cu Zn Hg As Sr Pb (ppm) Tusayan 1.8% 83430 4542 49+2 643 742 19548 3247 Leupp black-on-white ar 320+70 4442 6242 643 10#2 14246 3128 San Bernardo polychrome?! 1.8% 180+50 3542 90+4 613 1342 9544 2248 Tusayan black-on-white 1.6% 110430 = 82425042 548 5. 259 «558 Tusayan corrugated 2% 130+40 95+2 68+3 748 5+2 179+8 46+8 Tusayan corrugated> 3% 120440 3644 7043 <9 642 14946 3438 Contemporary First Mesa redware 6% 50+20 2443 20948 1544 943 43+5 82411 4 Typical style made in Awatovi. b Found in Awatovi. In Table 2 it can be seen that if only 200 g (dry weight) of cornmeal-ash foods were nutritional needs, even if total availability were in question. Other good are water (Dutt and McCreary 1970) and seasonally collected green plants (Kuhnlein and Calloway 1979b) and probably other locally-grown plant foods as well. Ton, zinc and phytate.—Iron i I itional|} ial mineral of thic area nf A interect in Hon highly colored with iron-containing This, in turn, may have caused the skeletal deformities (El-Najjar and Robertson mci oe Najjar et al. 1976). Zinc is also known to form insoluabl l with pl ( tal. 1976). c and metate grinding and during mechanical grinding is shown in Ta . . 1 metate grinding in considerable quanuty, oa L 1} AAan in a accumulated in the meal prep witt Probably as minute particles of rock. Manganese, calci | phos} the traditional procedure. The electric grinder with stone rubbing surfaces did not make significant additions of any of the minerals analyzed, even though the resultant meal can also be classified as “stone ground.” Zinc was not added in either procedure. te analysis was performed on modern Hopi corn for which mineral data were available, and these results are given in Table 7 (Kuhnlein and Calloway 1979b). The payin Tanged from a low value of 0.4% for lyophilized fresh sweet corn to 2.2% a ee ye — com normally used for hominy. When chamisa ash was used to make the bivilviki, the ania of calcium, iron and zinc were increased and the molar ratios of phytate/mineral were reduced. Molar ratios of phytate/minerals were lower for the sweet corn than for so egren corn. The high level of calcium in the bivilviki might accentuate the formation of an insoluble phytate complex. o 90 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 TABLE 6.—Elements in Hopi Blue Maize Cornmeal Prepared with Mano and Metate vs Mechanical Stone Grinder® Sample Treatment Fe Mn Ca 5 (ppm) (ppm) (mg/g) (mg/g) Al After ing } Is fi b by hand 30+2 10+2 0.05 1.63 A2 After washing with stored water 33+2 10+2 0.06 1.52 AS A 8 8 323+16 10+2 0.07 2.40 A4 After ing in i keul 372+18 8+2 0.08 2.50 A5 After fine grinding 496+24 18+2 0.73 223 A6 After sifting with wire sifter 576+28 18+2 0.67 1.42 Mechanically st g Ao BI After removing kernels from cob mechanically 31+2 9+2 0.04 1.83 B2 After washing 27+2 8+2 0.04 1.30 B3 After coarse — with “meat genial 47+2 9+2 0.05 2.43 B4 (No. B2) After fine g 29+2 8+2 0.04 1.19 B5 (No. A2) After fine grind “s machineb 26+1 9+2 0.05 1.44 Laboratory controls CA6 After microwave cooking and lyophilizing 502+25 20+2 na& n.a. CB4 After microwave cooking and lyophilizing 29+2 8+2 n.a. n.a. CB5 After microwave cooking and lyophilizing 33+2 10+2 n.a. ae: LJi ga). ri —_ wird Rlsctsic Gebesdes: “All Grain Co., Tremonton, UT S/N 1442 c Also given are mineral values for corn excavated from Antelope House in Canyon de ea (dated about A.D. 1200) where porotic hyperostosis was identified in skeletal ains. The iron and zinc sualiaes are similar to the contemporary Hopi corn, but calcium is substantially higher, Unfortunately, ib phage waraple for phytate analysis was not obtain binding « of ner to phytate i in foods, the vane Posie formed between the many diff 1 phytate bil- ity in the gut have still to be Pate (Oberleas et al. 1966). However, it is nisl to — that the use of culinary plant ash emerged as an experiential response to limited amounts of animal foods ees ba ype of absorbable qeaental mineral nutrients (particlany iron, zinc and 1 p cereal-legume See —High levels of strontium in the Colorado River and in northern Arizona soils and plants indicate that this area of Arizona is rich in natural strontium (Kopp and Broner 1967). This is shown in elevated levels in some native salts reported here, and also in water, water-extracts of soils, native leafy green plants and other Hopi foods (Kuhnlein and Calloway 1979a). The culinary ash used by the Hopi in blue cornmeal foods contains large amounts of strontium as well as calcium and was very likely a major contributor of both elements in the indigenous diet. Strontium and calcium are closely imerelated biologically, and it is thought that strontium can substitute for calcium in the apa complex of skeletal tissue (Likens et al. 1961 Analyses of Hopi harvested foods reveal 10- 30 Mg Sr/g Caand these ratios are considerably higher than the ratios of the contemporary Hopi diets (about 5 mg/g Ca) (Kuhnlein pie This is not unusual considering that relatively small amounts of cornmeal-ash foods 4 included in modern Hopi diets. eee OE eee NN TTR LT ee ( May 1981 KUHNLEIN 91 TABLE 7.—Phytate and Minerals in Hopi Corn* Phytate/mineral ratio Fe Phytate Cc Zn Fe Zn gm/kg gm/kg mg/kg mg/kg mole/mole mole/mole Blue Cornmeal 18.7 0.09 32 26 50 7 Bivilviki (from the above cornmeal) 14.8 5.67 168 37 8 37 Hominy Corn 21.9 0.07 31 38 59 55 Roasted Sweet Corn 6.5 0.12 64 60 9 i Antelope House Corn (yellow kernels) n.a.b 0.43 23 26 na site: From Kuhnlein and Calloway (1979b). Not analyzed. TABLE 8.—Minerals in Deciduous Tooth Dentin* ME recy 17th Century Contemporary Contemporary ny i Hopi Hopi Californian Value Value n=10 n=16 n=12 lumn Column Element (1) (2) (3) 1 vs 2 2 vs 3 eee te : Adg/g dentin - 7.0+3.8b 27.6+15.2 16,645.4 <.001 O15 = 478.0+86.4 97.7423.3 87.3¢13.7 001 260 Zn 134.0431.1 178.0452.9 151.0430.5 013 099 ee 10.1+2.0 22.9424.4 15.8+16.5 055 371 Hg 4,041.7 5.64.7 5345.4 237 908 g/100 g dentin = 26.8+1.0 29.443.1 25.1415 .005 001 : 11.6415 10.5+3.0 11.3+0.6 244 _ eobeRirbti ric nk BP/gCa 0.42+0.05 0.36+0.11 0.45+0.01 026 004 mg St/g Ca 1.78+0.29 0:34+0.09 0.35+0.09 ¢.001 700 Be Ne) en Se I eet a : om Kuhnlein and Calloway (1977b). Mean + S.D. : : ORL Ot oe eeespaeeneny ay Wee Bhrdenceg er Lead.—Lead is naturally present in lead and t ne was not abundant in the indigenous diet because geological conditions, especially that of high soil pH, render lead insoluble and plant physiology excludes lead. Although it has determined that lead does not enter locally-grown Hopi foods i the — Setting (Kuhnlein and Calloway 1979a), some Hopi corn foods were founc tain lead in levels high enough for concern if those foods were to comprise the major part of the diet (Calloway et al. 1974). Lead was not present in large amounts inthe native salts reported here, but it was found in chamisa ash?, and ceramic sherds from the Hopi area (Kuhnlein and Calloway 1979a). | ; is one mineral element generally thought to have increased in the modern environ- ment due to combustion of fossil fuels which release lead into the atmosphere and to the use of many lead-containing products of technology. For instance, lead in canned milk = Canned infant formulas (which have become the preferred substitutes of human breast mi 'n this area) was greater than 0.5 ppm during the late 1960's (Lamm et al. ai Plumbing, caoki geq if i¢. 3 es Pi h the Hopi environment. j and pro 5 92 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 Minerals in Seventeenth Century and C y Deciduous Teeth Comparison of the composition of seventeenth century and contemporary Hopi deciduous tooth dentin was made to assess the change in the human burden of several elements resulting from the change in diet and general environment since the preindustrial period (Kuhnlein and Calloway 1977b). Deciduous teeth from 10 individual skeletons dated from the seventeenth century were obtained from the Hopi collection of the Peabody Museum of Harvard University’. The excavations were made in the late 1930s from the Franciscan compound in the pueblo of Awatovi, which is located on the eastern section of the current reservation (Fig. 1). The burials were placed within the church after the Pueblo Revolt in A.D. 1680 when the monks were expelled, and presumably before the abandonment of the pueblo around A.D. 1700 (Brew 1949). Spree exfoliated uineciims posegi Sass wee samples" were randomly obtained in 1974 direc .Athird group of samples, sie cauactie exfoliated, were donated through 2 dental offices i in 1974- 75. These were from Renny, otieatsaas asap oe children in the same age range living in relatively ind hern California. Dentin was removed from the teeth and analyzed for lead, strontium, zinc, copper, and mercury by x-ray fluorescence and for calcium by atomic absorption. Phosphorus was measured with the Fiske-Subbarow method (Kuhnlein and Calloway pies Seen and peg were # Pp i bak a . Pp ee a popes! S = rn ] | are ard tissues i and are detectable with the same method. Calcium and phosphorus are major - elements in teeth and have been used as indicators of skeletal integrity. The results in Table 8 show significant differences between the seventeenth century and ae Hopi teeth. mien sine and copper are higher i in “abrieaniend teeth, reflecting tha perce natural background levels in the Hopi area; rather, it ei the introduction of many technological products in the last 50 years. Mercury was present in similar concentrations in all groups, indicating that the amount deposited in hard tissues has not changed; however, it is not known how well dentin reflects dietary content of mercury. Recently, lead and zinc have been shown to increase with urbanization and industrial- ization in Norwegian populations. Deciduous teeth from Medieval Bergen and modern urban and rural communities were analyzed and results gave the same trends as those reported for the Hopi (Fosse and Justesen 1978a, 1978b). Strontium is more than 4-fold higher in the seventeenth century teeth when compared to the 2 contemporary groups. The Ca/Sr ratio in the seventeenth century teeth is higher than any here-to-fore reported, although Steadman et al. (1958) noted similarly high levels from Tonga and Texas. Since this element is not considered a product of modern technology, its presence in tooth structures can be viewed as a natural consequence of the local geological contribution to the nutritional environment. Calcium and phosphorus levels in all groups are within accepted limits, but it is interesting that the ani aaa: Hopi levels differ from the other groups. This is reflected in their higher Ca/P ratios It appears, then, that the Hopi have b i i ingly posed to lead, zinc, and copper since the preindustrial period, and that their exp trontium has decreased. In fact, it seems that uptake of these minerals from the Hopi nutritional i lay is similar to that of northern California where there is obviously greater industrial development. Bs cet aereaet These data can offer only a few pera tives ond clear that native grown and prepared foods. It naturally follows that exposure to other nutrients has dietary mi ecology of the Hopi. ey ' 6. toys ly en ey ee May 1981 KUHNLEIN 93 ] h f, a changed as well, with declining use of , Native salts, and other uniquely Hopi food practices, in favor of refined commercial foods, there has been a concommitant decline in intakes of calcium, eens iron, site other sae and also phytate. On the other hand, the d oot erall increase has occurred in intakes of zinc, copper. in lead since the preindustrial penta Within the same time, these data confirm a decrease in strontium exposure. Given the variety of minerals which are important in human physiology, the complexity of assessing mineral availability in foods and the difficulties in estimating human mineral utilization, there is all likelihood that the total picture of change in Hopi dietary minerals from early to contemporary times will never be completely understood. However, these studies do give some understanding of the modern historical change in diet and mineral status. It is also very clear that many traditional Hopi foods were important sources of minerals, and that these are no longer being used to full advantage. LITERATURE CITED asain P. 1937. Notes on Hopi Economic Yale Univ. Publ. rege No. 15, Yale Uni v. Press, New Have roel R.M. 1971. The changing pattern of Hopi agriculture. J. Royal Anthropol. Inst. Occ. Paper No. 31. BREw, J.O. 1949. Franciscan Awatovi. Part I. The History of Awatovi. Part II. The Excavation of . Papers Peabody Mus. Arch. Ethn. H Univ. 36: 3. BUREAU OF INDIAN AFFAIRS. 1974. Soil and Range Inventory of the Hopi District Six Area. Arizona, U.S. Dept. Interior, Branch of Land Oper., Hopi Agency. CaLLoway, D.H. GIAUQUE, R.D., Costa, F.P. 1974. The superior mineral content of some Indian foods in comparison to federally donated counterpart commodities. Ecol. Food asa oP said OF NATIONAL HEALTH AND Nutrition Canada. National 1893. hss hudiwns of Arizona Oo “eictinas ns ew Mexico. U.S. Seog Office oompnelity eons Extra Census ull UIT, GR. AND McCreary, T.W. 1970. The quality of Arizona’s domestic, agricultural = industrial waters. sea t 256. Agric. Exp. » Univ. Arizona, Tucso sien M.Y. AN , JR. 1976. oye bones in prehistoric pen Science nate satge ae AN, DI. Turner, C.G., aa DR OBERTSON, ne States. at - Phys. Anthrop. 44 (3): Fo. "i AND NUTRITION BOARD, NATIONAL ESEARCH COUNCIL. 1 1974. Recommended Dietary Allowances, 8th ed. Natl. Acad. Sci., Washington, D.C. Forpe, C.D. 1931. Hopi agriculture and owner- ship. J. Royal errr tee Inst. 61:357. Fosse, Pyesd AMD ao Nes Pere ie Jee. a I 3 Agch: Enviton. Hebi $8:968. | 1978b. Zi ic ind of Norwegian children. Internatl. J. Environ. Stud. 13:19. Hack, J.T. 1942. The changing physical envir- onment of the Hopi Indians of Arizona. Papers Peabody Mus. Arch. Ethn. Harvard Univ. 35 (1):1. Hunter, H.V. 1940. The ethnography of salt in A. aboriginal North America. Unpubl. Thesis (Anthrop.), Univ. Pennsylvania, i hia P, J.F. AND , FRONER, R.C. 1967. Trace po Div. Pollution Surveillance. Cincinnati, Ohio. KUHNLEIN, H.V. 1976. Stronuum and lead in the Hopi nutritional environment and teeth. Unpubl. , Dissert., Univ. California, Berkeley. , AND CALLoway, D.H. 1977a. Contem- pokary Hopi food intake patterns. Ecol. Food Nutr. 6:159. piece 1977b. Minerals in human teeth: differences between preindustrial and con- temporary Hopi Indians. Amer. J. Clin. Nuw. 30:883. _ 1979a. Adventitious mineral elements in Hopi Indian diets. J. Food Sci. 44(1):2 _ 1979b. Composition of traditional Hopi ‘oods. J. Amer. Dietet. Assoc. 75: cae S., COLE, G., GLYNNE, K., . ULLMAN 1973. Lead content of milks fed to aioe 1971- 72. New England J. Med. 289 (11):574. 94 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 LIKINS, R.C., POSNER, A.S., PARETZKIN, B. AND vation of A lati he geochemical A.P. Frost. 1961. Effect of crystal growth on the environment. In Trace Substances in Environ- comparative fixation of Sr§? and Ca*® by mental Health VII. (D.D. Hemphill, ed.). Univ. calcified tissues. J. Biol. Chem. 236:2840. issouri, Columbi NEQUATEWA, E. 1943. Some Hopi recipes for the | STEADMAN, L.T., BRUDEVOLD, F. AND SMITH, F.A. preparation of shes ene foods. Plateau 16:18. 1958. Distribution of strontium in teeth from OBERLEAS, D., MUHRER, M.E., AND O’DELL, B.L. different geograhic areas. J. Amer. Dent. 1966. The eaitauetliy of zinc from foodstuffs. Assoc. 57: 340. In Zinc Metabolism (A. Prasd, ed.). C.C. wreeibic A.M. 1936. Hopi Journal. (E.C. Thomas, Springfield. Parsons, ed.), Columbia Univ. Press, New REINHOLD, J.G., FARADJI, RB., oP AN York. ISMAIL-BEIGI, F. 1976. Dec — es “ THORNTHWAITE, C.W. 1942. Climate and acceler- calcium, magnesium, eg and phosphorus by ated erosion in the arid and semi-arid South- humans due to increased fiber and phosphorus west and special reference to the Polacca Wash consumption as wheat bread. J. Nutr. aed Drainage Basin, Arizona. U.S. Dept. Agric. SiEVERS, M.L. 1966. Disease patterns am Tech. Bull. 808 Southwestern Indians. Public Health ten U.S. PuBLic HEALTH SERVICE. 1973. Ten State 81:1075. Nutrition Survey. Washington, D.C. ., AND CANNON, H. 1974. Disease patterns = WHITING, A.F. 1939. Ethnobotany of the Hop of Pima Indians of the Gila River Indian Reser- Mus. N. meine Bull. 15, Flagstaff, Avital NOTES 'Courteously provided by Dr. Christry Turner III, eee level for plant foods, ca. | i it would Arizona State University. centrate about 10-fold upon a “The 1] ppm reported in Table 3 is not unusually These were Kindle supplied by Mr. Al cara high. If the fresh weight plant contained a 5 ge J. Ethnobiol. 1 (1): 95-108 May 1981 THE PERCEPTUAL BASES OF ETHNOBIOLOGICAL CLASSIFICATION: EVIDENCE FROM AGUARUNA JIVARO ORNITHOLOGY BRENT BERLIN, JAMES SHILTS BOSTER, AND Department of Anthropology, University of California, 720 Berkeley, California 94 JOHN P. O’NEILL Museum of Natural Sciences, Louisiana State University, Baton Rouge, Louisiana 7080. _AssTRACT.— Preliminary experiments conducted to explore the principles of Aguaruna Jivaro bird classification of 164 commonly occurring species reveal that classification is primarily determined by the perceptual salience of each species. Those species of birds rated independently by a western ornithologist as perceptually highly salient are shown to have stable, codable Aguaruna names. Birds of low perceptual sali have | dability. Birds of all levels of perceptual salience are consistently more codable for males than for females, indicating a marked division of ornithological knowledge by sex. Methodological and theoretical implications of these findings for further kin mi hnobiological h are discussed. INTRODUCTION Ethnobiology is moving slowly toward an understanding of the principles underlying Native systems of biological classification. Work over the last several years has explored the formal structural f f ethnobiological taxonomies. A major conclusion of this work is that native systems of biological classification are structurally quite similar. (Berlin 1972; Berlin et al. 1973; Bulmer 1974; Hays 1974; Hunn 1977). Considerably less research has been devoted to determining the substantive nature of ethnobiological classification. Ethnobiologists are not yet able to predict which groups of plants and animals will be given conceptual recognition by the local society nor predict how the recognized groups will be perceived to be related. This is perhaps the most serious weakness in our understanding of how ethnobiological systems of classification actually work. In this paper, we will explore some of the substantive features of ethnobiological ificati s : classification by describing a portion of our research on Aguaruna Jivaro ornithology’. We ssification. First, we will try to we will examine why some named species are perceived to than to other species. Answers to these questions req correspondence of native names to individual biologica correspondence employs 2 methods, each leading to complementary conclusions. The first es. Here we determine which species uire an understanding of the 1 species. Our analysis of this Patterns, we believe, help reveal the internal structure 0 ; Contribute to an answer of our second question, why certain biological species are perceived to be similar to one another. We believe that our findings will contribute toa better understanding of the structure and function of ethnobiological systems of classification. 96 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 DISCUSSION Part I: The Psychological Salience of rentals Ewen’ The fundamental plant and animal taxa in been called folk generic categories (Berlin et al. 1973, 1974; Berlin 1972, 1976, 1977). While folk generic categories have been defined in differing ways by several authors (Bartlett 1940; Bulmer 1970; Cain 1956; Conklin 1954; Rosch et al. 1976), we believe that underlying these various formulations is a recognition of essentially the same conceptual and biological unit?, Work in ethnobiological classification has up until now made the assumption that folk generic categories of plants and animals differ little from one another in their perceptual distinctiveness, however this might be measured. This assumption is probably wrong. Several years of field work in ethnobiology have led us to the belief that biological species differ considerably in their overall distinctiveness from one another, and that this differential distinctiveness leads to the formation of folk generic categories of differing degrees of perceptual importance. Following traditional perehcena! terminology, we will use the term salience to refer to t or distinctiveness of any specified species of bird. As will be seen below, “perceptual salience can be measured indirectly. The Linguistic Codability of Ethnobiological Categories.—An examination of the ways local Dae een apply names - species of pane ane aniinals xeveais that people agree sely o n the name of other species. Following the usage of Brown and Lenneberg (1954) i in their ail study of color naming, we will use the term codability to refer to the degree of variability in naming responses for any specified stimulus, in this case, for any particular species. The greater the variability in naming, the less codable the species. One should expect species with high salience ratings to be highly codable while species of lower salience ratings should be less codable. If eich a a icduetic could be ate to hold true, it would suggest that a society’s linguistic codification of the biological universe is constrained by the character of the natural discontinuities in a particular habitat. The Naming Experiments.—In order to collect information to test the relationship of salience and codability, we undertook a series of naming experiments in a small village of Aguaruna Jivaro i in the Upper Maranon River Valley of northern Peru. Specimens of 157 species of birds were collected and Lpacparet wor use as seicuciionsy in 8 naming oe _~ species used in our experiments d found in the region inhabited by the Aguaruna, an issih that is treed t to include some 500 species. We are confident, however, that they represented the most common and regularly occurring birds in the immediate locality of our field site. As far as possible, sexually dimorphic species were represented by specimens of both sexes in the naming experiments. A total of 260 specimens were ultimately employed as stimuli in our experiments. Because of the large number of specimens, 3 separate naming experiments were conducted. Specimers were placed on long, cane work tables against backgrounds of tancolored botanical corrugates. The placement of specimens along tables was arbitrary, with th ption that speci fth i d directly adjacent to one ayac f ag b Pec | ies io Cale ee ee h imen ‘“‘wajim patya, f “What is it called?” and the response was recorded in a notebook. Each subject required So 15-20 minutes to complete each experiment. All participants were paid a minal sum at the completion of each tas pennant informants (18 male, 10 female) participated in the first experiment, 25 (21 ec I al ——— Se —caiaiieeeas ee seal ae May 1981 BERLIN ET AL. 97 males, 4 females) in the second, and 27 informants (21 males, 6 females) in the third experiment. Practical ideration de it impossible for all informants to participate in each of the 3 experiments. Furthermore, women were hesitant to participate as subjects in the study, leading to a sample biased toward male informants. The Measurement of Perceptual Salience.—There are no established criteria for d into 4 major salience oe (ao 4 \ categories — from very high (116 peci ), high (94 specimens), r (32 sp ) to low (18 specimens). Male and female birds of the same species showing strong sexual dimorphism were usually given separate salience ratings. All of the saliency judgments were, of course, carried out ind pendently of +f ion for the linguistic codability of the individual species. The criteri li blishing perceptual sali ill be di d below. Measurement of Codability.—A codability measure was needed that would characterize the degree of agreement among subjects for each specimen. A number of possibilities were considered. The most obvious was the size of the mode, the number of people offering the Most common name for the specimen. Another would be the number of distinct names offered, comparable to the use of the number of different species in an area as a measure of Species diversity. The Brown and Lenneberg (1954) codability score used a combination of the above measures. While all of the measures considered gave similar orderings of specimens by their codability, we decided to determine codability with a variant of the Shannon measure of uncertainty (Clifford and Stephanson 1975; Garner 1962), namely: s N(loggN)- £ nj(loggni) i=l _ N(loggN) where N is the number of naming responses (i.e., subjects), s is the total number of names, and nj; is the frequency with which the ith name was given as a response. . This measure ranges from one, in the case of total agreement, to zero, in the case of car disagreement. W, h 2 a *.: Jj age : £-1 C. } all of the names applied to a given specimen rather than just the most common name Or names. Findings.—Table 1 shows the relationship found between linguistic codability and Perceptual salience. Whether measured with Peason’s or Spearman's r, the correlation tween salience and codability is statistically significant at the .001 level. Interestingly, Most of this effect seems attributable to the contrast of the ‘very high’ group with the lower Salience levels since the correlation coefficients remain very nearly the same when the lower 3 levels are collapsed into a single group, see Table 2. ie a Currently, efforts are underway to codify the ornithological criteria used as the basts for quantifiable? As predicted by Hunn for ethnobiological classification in general (Hunn 7:74), size is quite strongly correlated with the codability of a bird, as 98 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 ABLE 1.—Correlation of Salience with TABLE 2.—Correlation of Collapsed Sali ives Codability Pearson Spearman Pearson Spearman : -R Prob R Prob N R Prob R Prob N 478 .001 492 001 252 467 .001 472 001 260 TABLE 3.—Correlation of Size with Codability. Pearson Spearman R Prob R Prob N Codability with Size 396 001 369 001 274 Salience with Size 203, 001 9 033 248 TABLE 4.—Multiple Regression of Codability on Salience and Size. Variable Fto Prob Multiple R R Square Simple R N Enter R Square Change Salience 69.4 001 469 .220 .220 469 248 i 34.2 001 562 315 095 397 248 TABLE 5.—Comparison of Men with Women by Codability. Mean S d N Deviation Men .700 244 289 Women 408 350 289 Difference 292 272 289 T =9.49 ep € € 001 The correlation of codability and size is significant at the .001 level. However, since salience ratings were in part based on size, MES is ako a bigs bu: statistically aignihae correlation of size with salience ratings. To show level of salience, the results of a multiple regression of codability on salience and size e are pinatar ce m Table ig It can be seen that, while, salience: i is the better of the z predicts . codabilit peed in codability. The apparent independence of size and salience may be ane attributable to the many small birds which were highly salient and codable due to their striking plumage color. Linguistic codability not only varies according to the characteristics of what is perceived but also according to the characteristics of the human perceiver. One would expect that ornithological knowledge will increase with age, and our initial studies support this intuition. Earlier work on inter-informant variation in Aguaruna plant identification SE. © een ee al Ree i ee i ee May 1981 BERLIN ET AL. 99 (Boster 1977) supe Steet that we might also expect differenices by in of subject. Fable 5 indicates that the bi omen. A T test was performed treating the men and wome en ’s codabilities of each token as matched samples. The T value was 9.49 with 288 degrees of freedom, p An ahead s skill in the classification of oe the outcome of ‘bite experience, role expectation and some component of natural talent. sexual division of labor. The third does not. In oe of the fact that women have : ample opportunity to observe birds on a casual basis, — roles = = train them to become ined — ' Aguaruna — on a other han age g birds When hunting, they have many opportunities to see the birds just as they apes in our bird sania: tasks: dead, chose at hand, with feathers ae Even Ain a woman h pp } | is nf f : Bl 1 | ae ey Bho here task woman for puahiiece On the other hand, in the classification of manioc varieties, it is the mature women who are the experts, reflecting their role and experience as manioc cultivators. These findings suggest that the Aguaruna division of intellectual labor mirrors the division of physical labor. Part II: The Internal Structure of Ethnobiological Categories ‘Our focus to this point has been to examine the et of the biological species to the various linguistic expression(s) used to refer to that speci We now take the name as a starting point, and ask in what ways a single name maps onto biological species. Just as one can measure degree of agreement on the name of a given Species, it is also possible to measure the range of species referred to by a particular name. € examination of the biological range of a term affords an understanding of the internal Structure of ethnobiological categories. As mentioned in the discussion on the naming experiments, there may be more than 500 Species of birds in the immediate region of tropical rainforest inhabited by the Aguaruna. It is an area of ornithological diversity exceeded by few other regions of the world. There is little reason to expect that the human population in the area will develop a system of classification that provides a separate name for each of these 500 or more species. It is Notable, nonetheless, that we have elicited, independent of the naming experiments, more an 300 distinct Aguaruna generic names for birds. Our ultimate goal is to specify the biological ranges ofeach of these terms, and to outline the ways in which these categories are conceptually organized into a coherent system of classification. Before discussing the experimental results that bear on the internal structure of ethnobiological categories, it is essential to present an overview of our current understanding of Aguaruna bird taxonomy. Aguaruna Bird Taxonomy.—The most general terms in Aguaruna ornithological Systematics are chighi and pishak. The former term is best glossed ‘large game bird’ and may Sat understanding of the organization of the 300 basic generic categories Is far from Ily store these 300 complete. However, it seems clear that the Aguaruna do not menta Categories in some arbitr trary list structure. On the contrary, sy stematic elicitation and informed dbeebyaiken reveal that subjects form “horizontal groupings” (Hunn 1977: 62) of Species of birds that are perceived to be closely related to one another on the basis of overall Similarit et Such groupings were first described for ethnobotanical systematics as “covert categories 100 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 (D’ Andrade MS; Berlin et al. 1968) and later as ‘‘covert complexes”’ (Berlin et al. 1974). Covert complexes have been discovered in several other ethnobotanical systems of plant classification (Berlin 1977; Hays 1974, 1976). For ethnozoology, these mid-level categories are described in detail by Hunn (1977), as well as by Bulmer (1974), Dwr (17D and Majnep and Bulmer (1977). While not all folk groupings, complexes include the vast majority of all Aguaruna birds. Finally, Aguaruna complexes are remarkably similar in content to well recognized ornithological taxa at the family and sub-family a of Siteiication: The Ag rh } A in Ee Be t ) om that at least some of these mid-level complexes are named (e.g., pinchu ‘hawk,’ yampis ‘dove,’ Shitk ‘puff bird’). The taxonomic structure for Aguaraun bird classification seems to be a shallow one comprised of approximately 300 basic generic classes, most of which are members of mid- level complexes. Most of these complexes are unnamed. Finally, some small number of generic taxa are sub-divided into what we have called “folk species.” Typically, a generic category will be segmented into 2 or 3 folk species, though we do have a few examples of 5 or more. This taxonomic structure is illustrated in Figure 1. The Naming Experiments, Part II.—Our sample of 157 species of birds elicited 275 distinct naming expressions in our experiments. Some of these names were applied by all informants to one or more specimens in the naming tasks but others were used much less frequently. Of the 275 expressions, less than half were judged to be ‘‘good names,” either by observing the frequency of their application to a given specimen or by the restriction of their application to a seagate mx uological range. The to the species in our sample provides a tentative picture of Aguaruna bird glamuiticatiog. Some 140 of the 157 species are found to be members of one of the mid-level complexes described above. Another 10 are seen as isolates, not participating as a member of any complex, and another 7 are not systematically paired with any Aguaruna name. Because only 157 of the approximately 500 birds in the area were examined in our experiments, none of the complexes is represented by their total membership. Nonetheless, the basic outlines of the structure of each are apparent. The Complexes.—Methods of establishing the membership of covert complexes in ethnobiological classificadins have been outlined i in Berlin et al. (1968), hloys on ee and Hunn (1977). T ] pishak ‘bird’ mid-level } complexes pinchu ‘hawk’ yampis ‘dove’ (woodpeckers) (tanagers) Dl oe (isolates) folk genera’ O = Oo 2 Otigeens OO Or a ale OM OO, igo Ly O50 Din DiGi On «28 folk species /| N = 300+ generic categories —— a 8 ge ee May 1981 BERLIN ET AL. 101 types of evidence. First, we systematically interviewed 3 knowledgeable Aguaruna males. The subjects were asked to form groups of all those generic bird names that they considered to be pataji ‘in the same family,’ or kumpaji, ‘closely similar.’ The third subject, quite knowledgeable but nonliterate, was interviewed over several days in lengthy, tape-recorded sessions. The ethnographer, working through an alphabetically arranged dictionary of all the potential bird names that had been collected from earlier interviews, asked the subject, “Brother, does such and such a bird have any relative?”’, recording the names of the birds on cards. Because of the number of bird names A Rare shee ethnographer found i am impossible to remember which names had already bee The subject, ses would emphatically state, “We have already named that one.” "The stability of h discovered fier he had. provided mutually exclusive complexes ‘of bird names with a exception of 2 generic names which had been assigned to more than a single group! second kind of evidence used for the recognition of ethnobiological complexes comes from observing the actual distribution of naming responses. Often, the biological ranges of Aguaruna names will systematically overlap in such a way as to suggest close perceptual similarity of the categories involved. As an illustrative example consider 2 Aguaruna bird names, A and B. These Seoressions refer to at least 3 species — 1, 2, and 3 — in the following fashion. Species 1 ; A, species 3 predominately receives name B, and species 2 receives name A by half of soe subjects and name B by the sence hall. Hf 20 individuals are involved in the naming task, w shown in Fj igure 2. The distribution of seatttiis responses allows one to infer that categories A and B are conceptually related in that their biological ranges overlap. When the generic categories of a covert complex elicited from knowledgeable subjects with those Seneric categories tied together on distributional grounds such as those just described, we have good reason to believe that we have discovered a stable, salient, grouping of birds. 20 20 eetet OK x XX B54 Cx xX B iS "er Name A 15 Name Z 15 S505 sina xxx 6 & SKS gg xX gx xx woe OOO). =e Lo. ©.@. AAA = 10 10 decks related biological species, the splitters will make a greater number 0! ++) *! distinctions in the set than the lumpers. As a result, the lumper’s name refers te * © ren biological range while the splitter’s name is more narrowly focused. Even in the-¢ ca 1B i TT, i TTT | a el TCT, cei — tT A ee ————— — May 1981 BERLIN ET AL. Des tie woodpecker species used in naming experiments reported upon in this paper. From ri o bottom: Phloeoceastas melanoleucos, P. rubricollis, Celeus spectabilis, C. elegans, C. grammicus, optilus punctigula, Veniliornis affinis, V. passerinus. 105 106 JOURNAL OF ETHNOBIOLOGY WESTERN BIOLOGICAL TAXONOMY OF WOODPECKERS Vol. 1, No. 1 svar Pa TAXONOMY OF WOODPECKERS inferred from naming tasks data) Phloeceastas melanoleucos Phloeceastas melanoleucos P. rubricollis af P. rubricollis | Celeus elegans Celeus elegans C. spectabilis C. spectabilis C. grammicus C. grammicus Chrysoptilus punctigula Chrysoptilus punctigul liornis p ; Veniliornis passerinus V. affinis V. affinis Fic. 5.—Comparison of western and Aguaruna classification of woodpeckers. which lumpers outnumber the splitters, the splitter’s names can be recognized as those pega! applied to a narrow biological range and rarely used outside of that range. ur hearts belong with the splitters. These subjects provide the most intricate and internally consistent categorizations. Lumpers’ categories are not only broader but sometimes almost random. For these reasons, among others, we equate splitting with ee In our ape e-Sjeis tend to be older men though a few of the younger men also irds. However, it should be understood that the lumpers and se were not absolutely fixed groups for sometimes even the most expert made ‘mistakes.’ The typical lumping strategy is to classify 2 or more closely related biological species by the same name while the typical strategy in splitting is to separate the species. Furthermore, the lumper’s name for the category will generally be that used by the splitter to refer only to the most salient of he related specs: ahs can be illustrated 1 in Figure 8. In the case of the woodpecker genus P to both P. melanoleucos and P. rubricollis, tatasham while splitters reserve jaiachasn for P: ggapeuaregbats a calling the less salient P. rubricollis, samamu. In t , lumpers refer to - species - peel (with some = cma on C. grammicus), while splitters an only C. am k with the name jz7p. In sum, lumpers and ines agree on the name of the most salient ae and differ only in their designation of the least salient. This amplifies the correlation of salience with codability. m 1Ccu 5S, Sawa €, CONCLUSIONS In this paper, we have examined the patterns of correspondences between Aguaruna names for birds and the ae species to which these names refer. The pasterns O as _ informant variation that ca components of the structure a process of ethnobiological classification. First, we yee that highly perceptually salient species are more codable than are less salient species. Salience was demonstrated to be determined in part by size of bird, though we believe that distinctiveness of plumage coloration, frequency of observation, and distinctiveness of vocalization, also contribute to overall salience. We found that female subjects disagreed more often on the names of birds than do males, and we suggested that this could be explained as an outcome of the sexual division of labor in Aguaruna society. Second, an examination of the distributional patterns of bird names over multiple biological species has allowed for a better understanding of the internal structure of ethnobiological categories. A detailed analysis of one covert grouping of birds, the woodpeckers, revealed that the prototypical members of the complex were perceptually May 1981 BERLIN ET AL. 107 LUMPERS Phloeceastas melanoleucos tatasham P. rubricollis (Woodpeckers) Celeus grammicus sawake C. elegans : C. spectabilts (other woodpeckers) SPLITTERS tatasham Phloeceastas melanoleucos samamu P. rubricollis (Woodpeckers) sawake Celeus grammicus — C. elegans jiip C. spectabilis (other woodpeckers) FiG. 6.—Aguaruna lumping and splitting strategies for the classification of woodpecker species. Furthermore, species that are closely related ornithologically were shown to be also closely related perceptually in the folk system. Both the west ientific and folk sy ere thus Seen to be quite similar in overall structure. Finally, we noted that the patterns of lex ref tegies of classificati disagreement in naming within th ected individual st biological categories nomenclaturally recognized for a particular culture d the SuDsequent and independent verification of the predictions by field investigations” (Hunn 1977:72). Our findings suggest that such a test may one day be feasable. Not only should we ereey be able to specify which species are likely to b di icular habitat, utfwech 373% ld ¢ 1 . 7. be related, those which are likely to be focal and peripheral to so .gory as well as those species that are likely to be totally unknown. The most logical place to make such a series of predictions would be in a society Tesiding in a habitat similar to that of the Aguaruna with a comparable tropical forest avifauna. While we are aware that the number of factors involved-make such an exercise difficult, the outcome of the experiment should bring us closer to an understanding of Precisely what constraints govern the formation of ethnobiological categories in general. LITERATURE CITED BARTLETT, H.H. 1940. History of the generic Chapter I, in Cognition and Categorization (E. concept in botany. Bull. Torrey Bot. Club 67: Rosch and B. Lloyd, eds.). Erlbaum Publ., 49-362. Hillsdale, New Jersey. BERLIN, BRENT. 1972. Speculations on the growth BERLIN, BRENT, DENNIS E. BREEDLOVE, AND PETER of ethnobotanical nomenclature. Language in : ; ert categories and folk Society 1:51-86. taxonomies. Amer. Anthropol. 70:290-299. —— - 1976. The concept of rank in ethno- 1973. General principles of classfication biological classification: some evidence from and nomenclature in folk biology. Amer. na folk botany. Amer. Ethnol. 3(3):381- Anthropol. 75:214-242. : . 1974. Principles of Tzeltal Plant Classi- ——— .- 1977. Ethnobiological classification. fication: An Introduction to the Botanical 108 JOURNAL OF ETHNOBIOLOGY Ethnography of a Mayan-Speaking People of aes Chiapas. Academic Press, New te JAMES SHILTs. 1977. Inter-informant ari. study of language and cognition. J. Abnormal Soc. Psych. 49:454-462. BULMER, RALPH N.H. 1968. Worms that croak and other mysteries of Karam natural history. Mankind, aa Anthropol. Soc. Australia 6(12):621-6: . 1970. eee came first, the chicken or the egghead? Pp. 1069-1091, im Echanges et Communications, —— tal a Claude evi-Stra’ Y anniversaire (Jean “Pouillon and Pierre , eds.). M 1974. Folk biology in the ‘ets Guinea Highlands. Soc. Sci. Information dial BULMER, RALPH N.H. AND M.]J. LER. 1968. Karam classification of frogs. J. aitisanns Soc. 77:333 ee CAIN, A.J. e esas a evolutionary taxonomy. pene Zool. 5(3):97 CLIFFoRD, H DW. se 1975. Introduction Numerical Classification. to Academic Press, New York. Har NKLIN, OLD. 1954. The Relation of Hanunoo Culture to the Plant World. Unpubl. Vol. 1, No. 1 Ph.D. dissert. (Anthrop.). Yale Univ., New Haven D’ANDRADE, Roy. MS. Report on oo ae Tzeltal Ethnobotany. Unpubl. sea PETER. 1977. An seat “a Rofaifo taxonomy. Amer. Ethnol. 3:425-445. FELD, ), STEVEN. 1979. Sap you pers are shige hale they ait and Sentiment), Unpu my Ph. D. dissert. New York . Mauna: Explorations in Ndumba Sihnobcuay: Unpubl. Ph.D. dissert., Univ. Washington 76. An onbirical method for the identi- ration of covert categories in ethnobiology. Amer. Ethnol. 2:489-507. HUNN, EuGENE S. 1977. Tzeltal Folk Zoology: The Classification of Discontinuities in Nature demic Press, New York MAJNEP, SAEM AND RALPH N.H. BuLMER. 1977. Birds of My Kalam Country. Univ. Auckland Press, aeesiveg RoscH, E., C.B. MERvis, W.D. Gray, D.W. OHNSON, AND T.D. BoyEs-BRAEM. 1976. Basic objects and natural categories. Cog. Psych. 8:382-439. SCHANERSES, REET seine: 1970. A Gate a. Li Co., Wynnewood, sath oto NOTES 1. This work has been supported by a generous grant from the National Science Foundation (BSN 7916746, “Field Research in Ethnobio- logical Anthropology,” Brent Berlin and ome L = tton, Co-Principal Investigators), an out with the collaboration of the ales de Historia cree “Javier Prado,” balancer snes ba de San Marcos, Lima cultura, Direccion de Forestal y de Fauna, Lima, Peru. We are appre- ciative of the comments and suggestions of numerous persons, but would especially like to thank Elois Ann Berlin, Barbara Cohn, Eugen Hammel, Eugene Hunn, Paul mee wile Kempton, Susan Niles, and Pilgrim. 2. The ethnobotanist H. * Bartlett, the first 2Ci LU refer categories, defined them as those categories which are ne or less consciously thought of ac th A S22UUCi li W name”’ (1940: 341). ‘For the | vee es refer to biological world that can be rec 3. Size me without close study’’ (1956:97). For the ethno- zoologist, Ralph Bulmer, these fundamental units (in his terminology, (Bulmer 1968, 1970; Bulmer and Tyler 1968). Rosch et al. (1976) use the term “basic level rege ts” to refer to these fundamental categor ies. For us, generic categories are definable terms of a series of Secelatie taxonomic, psych- ological, and biological criteria, (c.f., Bet lin 1977). Generic categories in ethnobiological systems * classification cryout to be named urements were taken from The Birds items involved in each ‘ between 248 and 274. Effect of this fluctuation on the correlation coefficients seems tO been mini 4. Comparably named suprageneric groupi ings have recently been described by Steven Feld for the Kaluli of New Guinea (Feld 1979). J. Ethnobiol. 1 (1): 109-123 May 1981 QUELITES — ETHNOECOLOGY OF EDIBLE GREENS — PAST, PRESENT, AND FUTURE ROBERT A. BYE, JR. University of Colorado, Department of Environmental, Population and Organismic Biology, Boulder, Colorado 80309 r W je, J 1 M hI rf I y MaSVUlI Ally alia Ulicy iS food } - © Ey otetmerte | = l 1 eee wu: eae SSE ABSTRACT.—Quelites are edible greens usually derived from young, tender annual herbs but they may also include flowers, infl , and stem tips of ial ca ese }. hy en | 1 f this r maintenance of this resource, 3) greens form a nutritionally important ponent of annual diets, 4) queli I prod f ecologically 1 agricultural practices and yields are based upon the multiple cropping model, 5) encouragement of this urce may have led to the domestication of such plants as Amaranthus, Brassica, and Chenopodium, and 6) these plants may be a valuable resource in future food production systems. INTRODUCTION Until recently, the significance of uncultivated edible greens in the traditional native American diet has not been appreciated. As the intensity and depth of botanical, ethnological and archaeological investigations increase, practical and theoretical concepts are being applied to the elucidation of the principles of resource exploitation by man. The employment of undomesticated greens — referred to as “quelites” in Mexico — as food Provides an opportunity to investigate the ethnoecology! of this poorly understood food resource. + Seater Pa ee | ee 1 1 } dt der vik eens as “ guiribS ” to which the Spanish term, “quelite,” is generally applicable. These plants are usually immature when consumed and are eaten raw (in a few cases) or lightly pects which are being man exploitation of 110 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 TABLE 1.—Some common edible greens or quelites of the Tarahumara. All of these species are commonly found in and along cultivated fields. arahumara Name sé Scientific Name (Arranged by Family) Mexican Name Season of Procurement AMARANTHACEAE Amaranthus retroflexus L. basori, wasori spring/summer quelite del agua CHENOPODIACEAE Chenopodium ambrosioides L. chu’a’ summer/fall epazote Chenopodium berlandiert Moq. chu’a spring/summer quelite de cenizo COMPOSITAE Bidens odorata Cav. sepe spring/summer Cosmos paraviflorus (Jacq.) HBK. _hu’ve spring/summer CRUCIFERAE Brassica campestris L. mekuasare spring/summer coles fall-cultivated Lepidium virginicum L. rochiwari winter/spring fall-cultivated MALVACEAE Anoda cristata (L.) Schlecht. rewe spring/summer PORTULACA Portulaca oleracea L. chamo summer/ fall verdulaga URTICACEAE Urtica dioica L. ra’uri, ra’oke spring/summer eee Species Richness Richness in the number of species and in the phenological types is an important parameter in evaluating the ecological potential of any resource system. The Tarahumal@ are known to employ over 120 species of quelites. Most of these plants are ingested in the form of immature leaves and stems of herbaceous dicots although a few plants have the edible portion represented by bulbous leaf bases (e.g., Pitcarnia palmeri), pseudobulbs (e-8- Gongora sp.), succulent stems (e.g., Opuntia spp.), and immature inflorescences (e.8+ Jacobinia candicans). Of these 120 plus edible species, only 10 are consistently consumed today in the sierras (Table 1) and all are found in anthropogenic communities (Fig. 1). These common species have been erroneously referred to as “wild greens” althougha few researchers recognized their relationship to h disturb (M 1972; Wilken wall 1 - tc adanl Biologically, these plants are weeds which luti y and co gical f Pp to survival in habitats disturbed by human activity. Without constant human interaction over thousands of years, these forms would not be present or in sufficient density to be an adequate food resource. Th quelites are land rep 3 major life forms which are important in the availability of culturally acceptable and seasonally distributed resources: 1) winter annuals (e.g., Lepidium), 2) spring-summer annuals (€.g., Amaranthus, May 1981] BYE 1]] Fic. 1.—Some edible weeds form an anth pogeni ity (maize fields and margins; May 197 Cusarare, Chihuahua). Top row (left to right): Amaranthus retroflexus*, Chenopodium berlandien*, Brassica campestris*, Lepidium virginicum*. Bottom row (left to right): Galinsoga semicalva, Simsia eurylepis, Bidens ordorata*, Cosmos parviflorus*, Ipomoea hirsutula, Dalea sp., Anoda cristata, Urtica dioica. An asterisk (*) denotes the preferred species. Scale equals 5 cm. Bidens), and 3) summer-fall annuals (e.g., Portulaca). It should be noted that there are only a few perennials and that the ecologically wild Species play a relatively minor role in the total diet. One notable exception to this statement would include certain species of prickly-pear cacti, Opuntia spp., found wild in the barrancas (although it is known to be a tolerated weed, encouraged weed, or even cultivated wild plant in some regions). Human Disturbance Human disturbance is an important factor in determining the presence and density of these common edible weeds. They are members of various anthropogenic communities® which are maintained by the Tarah ees le 1 fields, field-fence margins, dwelling Sites, corrals and trailsides. A general ethnoecological principle to be documented in the future states that the existence of large human populations depends on the net Productivity of the ecosystem which is available only in the early developmental stages of Succession. Based on Odum’s (1969) Ecosystem Development Model (Fig. 2), = «st €arly stages, certain resources can be manipulated directly (e.g., cultivated fields) or indirectly (e.g., weed communities) so as to concentrate th ploitable resources in ume and space. Recently, this principle has been illustrated in a restricted sense by the development of the “garden hunting” concept using a tropical ecosystem and animal resources (Linares 1976). The exploitation of quelites represents an analogous situation with plant resources. Interestingly, Bohrer’s (1977) speculations on the food habits in i12 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 ae co pe ai tu = lu B = biomass Po= gross productivity Pye net productivity R = respiration TIME tie DEVELOPMENTAL STAGES MATURE STAGES (including OU nents communities) Fic. 2.—Generalized E Devel Model (after Odum 1969). Note that Net Productivity is available in the development stages and not in the mature stages. an ik 5 < soe lnited hominid evolution suggest that plants of the early ] were Exp as soe eaitets Eas members a ie € more mature communities. Man ] f Ecosystem Devel nt(Odum 1969) ar are beneficial to human exploitation of cuniceniend resources. These Yacen al include: 1) low Raa ace 2) low biomass, 3) linear food chains, 4) grazing food ch ains, and 5) short lived organisms with simple life cycles (e.g., annual plants). The attributes of low species Maile a low biomass may seem contradictory until one assesses the quality of the species and the manent. = general, species richness and ngicomlenee! increases with ecosystem development but of herbs to w early stages (Fig. 3) Le ttle ‘The moe sae cig 1S ech siemens low due to the nature of herbaceous annual pla woody perennials of later stages. The presence and density of edibl depend on several f which are only poorly known today. Many weed seeds have evolved 1 mechanisms for long distance apes - order to colonize distant habitats wh increase the seed bank for maintenance of local population) (Baker 1974; Harper ort Disturbance (by digging, plowing, etc.) of the upper layer of the soil is critical to the germination of weed seeds so that seeds near the surface and light germinate and emerge faster than if they were deeper in the soil (Fig. 4) (e.g., Dawson and Burns 1962; Wiese and Davis 1967). The ecological importance of disturbance to light flash and seed germination has been discussed by Sauer maar Struik (1964). Denisty of a species in = stages © succession tends to be related t Davis and Cantlon ( 1969) found that Amaranthus retroflexus tends to increase in density as the open area increases during the first year of experimental secondary successional studies in New. Jersey- It is posable} thas agricultural Practions originated, - n part, in response to human preference for Partially domesticated plants (1.€-, apiaticaiie altered from undomesticated progenitors) may have been encouraged, sown and subsequently selected in the manipulated habitats which auras into agricultural and garden habitats rather refuse mounds to pepiiaed fields and subsequently selected. aman 3 May 1981 BYE 113 herbaceous perennials woody perennials annuals DIVERSITY (number of species) — y 6-1] 11-15 16-20 21-25 ¥ LJ Oe eee TIME-SINCE DISTURBANCE (years since last plowing) Ac. 3.—Generalized model of relative change of annuals, herbaceous perennials and woody peren- nials in the early Stages Of succession in abandoned agricultural fields (after Beckwith 1954). / NUMBER OF SEEDLINGS SOIL DEPTH FIG. 4.—Generalized relationship of seed germination to seed depth in the soil. 114 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. | Productivity Quelites are an ee aia Vectans act eles of the capeare ecosystem sh rt i the Tarahum g be coon in apace ways. A few considerations are outlined below. Being subsistence agriculturists, the Tarahumara depend on an annual diet cycle based upon maize, bean, cucurbit and chile which are consumed from fresh plants in August through October and from stored, dried forms in October through May. Often times the stored cultivated food supplies are limiting from April through July. During this latter period, the diet is augmented by hunted and gathered resources such as _ walt ererag roots, bulbs, or ‘hearts”’ of maguey (Agave spp.). Itis during this per riod th t ds dominate the diet. May-June period also marks the end of the dry season and the a els of the rainy period and the start of the annual growing season. The seeds of weeds as well as planted maize emerge in the fields in mid to late May in response to the increased temperature and moisture. The coincidence of the marked change to warm moist regime with the germination and emergence of edible weed seedlings with the depleted food reserves is critical to the survival of the Tarahumara populations in the sierras. The weeds can also provide food after the initial growing period. July and August may be frequented by severe hail storms which destroy the young maize plants. Also, animal pests such as crows and insects can destroy portions of the maize crop at different stages. The tender apices of the older weed plants as well as the late emerging aint can be collected and consumed. The quelites represent a living emerging food reserv en considering primary productivity in ethnobotanical terms, one must account for not only quantity in time but also quality. Although studies are in progress, preliminary data indicate that in the sierran cultivated maize fields, 100 g of edible seedlings of Amaranthus retroflexus (Fig. 5) can be harvested in May and early ae from a plot varying rom 1-4 m®. Regeneration of another 100 g of edible weed seedlings can occur during this period in about a week. A daily serving of A. retroflexus consists of about 100 g per adult individual and is prepared by slightly cooking itin warm water and rinsing it in cold water or 3 times and then eating it with a little salt along with tortillas or pinole. FiG. 5a.—Tarahumara woman collecting Amaranthus retroflexus (Bye 8532; 30 May 1978; San Ignacio Arareco, Chihuahua). May 1981 BYE 115 Fic. 5b—Seedlings of A. retroflexus at the early developmental stage when they are consumed as quelites (Bye 8510; 28 May 1978; Cusarare, Chihuahua). The quality of quelites can be measured in several ways. One system involves cultural Preference based upon beliefs and cross-cultural comparisons. For example, some Mexicanized Tarahumara no longer eat certain quelites because the dominating Mexican pharmacology and flavoring. i he nutritional requirements of the Tarahumara and the value of their present diets _ not known at this time. A preliminary evaluation of the Tarahumara maize-bean-cucurbit diet indicates that the following items are deficient: protein, calcium, vitamin A, thiamine, ribo lavin, and vitamin C. The first 3 components are only present at about a quarter of the minimum Recommended Dietary Allowance (RDA) for an adult (National Academy of Sciences 1974) while the latter 3 components are marginally dificient. An addition of 100 8 of quelites (e.g., Amaranthus, Brassica, and Chenopodium; see Table 2) has only a slight impact on the protein yet provides sufficient calcium, vitamin A, thiamine, riboflavin and vitamin Cto meet the RDA standard for the United States. It should be noted that nutritional loss by traditional Tarahumara preparation techniques using warm (not boiling) water Is Probably minimal based upon knowledge of loss of ascorbic acid through various cooking methods (Caldwell and Gim-Sai 1973). Other preparation techniques such as sun wilting and mineral additions may enrich the value of quelites as well. Toxic materials may be removed from food plants through selective breeding and genetic manipulation of domesticated plants or through gathering and preparation techniques applied to non-domesticated plants. The Tarahumara collect only the young, tender leaves which tend to accumulate in the older, senescent leaves which are not gathered. Aqueous Cooking and leaching (rinsing) practices can also reduce the amount of these substances. 116 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 TABLE 2.—Nutritional value of some weedy greens (per 100 g edible portion) (Leung 1961). Ca Vit. A Thiamine Riboflavin Ascorbic Plants (mg) (IU) (mg) (mg) Acid (mg) | Amaranthus spp. 313 1600 0.05 0.24 65 Brassica campestris 252 1335 0.12 0.29 118 Chenopodium berlandieri 156 2765 0.17 0.47 109 Average 240 1907 0.11 0.33 97 Palatability is another factor which affects the edibility of quelites. In general, only the young leaves and stem tips are consumed. These tender structures are relatively unlignified compared to mature tissue. Chemical constituents of certain edible weeds may provide additional values due to flavoring and pharmacological activity. Chenopodium ambrosioides, a common weed along margins of fields and fences, is often added to beans and meat dishes. It imparts 4 distinctive flavor to the food. Also, the leaves contain ascoridole as part of the Oil of Chenopodium which is known to be an effective anthelminthic medicine (Guenther 1948-1952; Santos 1925). The Tarahumara often collect edible weed seedlings from week 2 to week 6 after germination. After this time the plants are often too large and lignified for consumption (although the stem apices and terminal leaves can be consumed in times of emergency OF famine). Recent study on the nutritional value of leaf protein in Africa included species of Amaranthus, Solanum, and groundnuts (Oke 1973). The extractable protein nitrogen, 4 measure of leaf protein, was found to peak during week 5 to 6 and was followed by rapid decline in nutritional value in later weeks (Fig. 6). Itappears that the Tarahumara gathering of palatable leaves occurs when the potential extractable nutrient value reaches its peak. 6607 nt naa Fa aan a total N above ground © - < 5005 es 3 er rd Te — 406, ai ~ YIELD N extracted protein N extracted ee ae T T T T Noa See be 5 7 9 11 TIME (weeks after planting) FIG. 6. Ch 4 f i io 3 Ef 7 . ‘UL zy L | groundnut leaves; from Oke 1973). : ; May 1981 BYE 117 Ecological Benefits eel the Tarahumara practice of leaving the weeds in the field for extended periods (Fig. 7) may appear uneconomical, this strategy may be ecologically sound. Unconscious dispersal of weed seeds by Tarahumara movements while harvesting maize during the previous year and the soil for pl les the weed seed bank to build up in the soil and to be closer to the surface to insure high rate of germination. When the weeds emerge, they are not weeded out until 6-8 weeks later. Subsequent weeding of cultivated fields at similar intervals allows for the establishment of new weed populations which provide emergency food reserves. This system allows weeds to be the first crop with the second crop, maize, being available later. This double crop system allows for the harvest of reliable yields of one type of net productivity in an environment where maximum yields of one crop systems are not abe due to poor soil fertility, limited moisture and unpredictable pests and weather. Only recently have the saaitete aspects of multiple cropping systems been considered in applied techniques and theoretical terms (Papendick et al. 1976). The essence of the multiple cropping is the complementary use of growth resources by different components of the system. The rate of exploitation of each resource by each component is separated by space and/or in time. Hence, the shallow rooted amaranth weeds should be extracting water and nutrients in the upper soil surface above the deeper planted maize seeds. After a certain period of growth the roots of both species would be competing for the same resources in the same space and time, to the detriment of each species. Future research will investigate the hypothesis that the Tarahumara remove weeds when they begin to mio with maize for the same resources. Before that time (6-8 weeks) the weeds do not c te with maize and therefore should not negatively affect the maize yield. Net nid of reliable yield therefore has 2 temporal peaks — early in the growing season with weed seedlings or quelites and late in the growing season with the harvested maiz Tentative eae i sa AE can be seen in experimental work carried out at Chapingo, M d Hernandez X. 1972). Plots of maize were treated with —A field consisting of two crops: 1) edible weeds (Amaranthus, Chenopodium, Bidens and Fic. Cosmos) and 2) maize. (June 1973; San Ignacio Arareco, Chihuahua). 118 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 different weeding practices. It was found that the weeds left in the fields for days 1 to 30 and for days | to 62 after ae had no. effect on the maize yield compared to the control pia free plots). Maize y f weeds were left in the fields after these periods (Fig. e 2 weeds used in this experiment were Amaranthus and Simsia, 2 Tarahumara sath ied The Tarahumara concept of multiple, reliable yields appears to illustrate mle cropping ecological theory. Weeds may also provide other ecologocial benefits such as soil protectors, dispersion of food resources for various predators, and other factors which merit urther investigation. Domestication The exploitation of weeds may represent one pathway to domestication and subsequent agriculture. Weeds and domesticates represent end products of genetic and ecological alterations mediated by human activities (Fig. 9). Domesticates appear to be the result of human directed evolutionary changes in ae in order to increase and saunas genetically the valued plant parts. These plants produce valued yields in a s, on the other hand, are not directed by conscious human selection but are evolutionary responses to human disturbed habitats which vary in time and space. As we know more about domestication, the more important weeds become in understanding this evolutionary process (De Wet and Harlan 1975). This domestication process recognizes weeds as one type of progenitor which was suggested by Vavilov (1951) with respect to priors enh of an of crop exis (e. e ” in cultivation in northern Europe). People’s response to edible resources aiid in human disturbed ar pelea maeeed ‘(trigger — sowing and selection of weed seeds. opo ved from — een inee: (Fig. 10) (Sauer otis Wilson and} Sart oa in both ew northern and sout f the Western misphere. This North- South pattern — waa be pea sola Perera a Rapa? Fee Lepidium ) erican >| 8015 kg/ha 1 | SINIAWLY Sul 2 — el | LJ i io) T T 7 | era | T ; pan w S —) co) MAIZE YIELD (%) Fic. 8.—Competition pens igh maize e oad weots (Amaranthus and saniiggeny (based upon data from Blanco and Hernandez X. 19 maize free from W' days 1-30 (after planting); hs maize free from weeds days 31-62; D, maize free from weeds days 63 94; E, maize and weeds together during total growing season; F, weeds alone ae .* - = ee ee a enmememee ae eee ee | ON RN pen May 1981 BYE 119 WEED A natural selection unconscious human selection WILD PLANT -» WEED (colonizer) ” (colonizer i man- : in human-altered disturbed conscious human habitat: habitat) selection (including WEED-CROP) v DOMESTICATE Fic. 9.—A generalized pathway of domestication involving weeds and domesticates. Amaranthus — grain amaranths A. powellii §—————_____________» A. hypochondriacus (sw US; nw & c Mex.) A, hybridus ——p A. cruentus (s Mex. & Cent. Amer.) A. quitensis p A. caudatus (S. Amer. Andes) Chenopodium — grain chenopods C. berlandieri —> C. nuttalliae (c & s Mexico) C. hircinum $$ ______—_» C. quinoa (S. Amer. Andes) FIG. 10.—Weed progenitors of domesticated food plants. Bolivia and Peru (Gade 1976; Leon 1964). Although no native Lepidium is known to be domesticated in North America, the Tarahumara plant seeds of Lepidium virginicum asa weedcrop in cultivated fields (Fig. 11). Perhaps the domestication of Lepidium in the northern latitudes is proceeding slower. An “experiment” to examine this domestication ypothesis started nearly 300 years ago and is still in progress. Brassica campestris, a weedy mustard introduced by the Spaniards, has been considered a potential candidate for domestication in South America although there has | gand select Future Resources As we begin to understand the evolution, ecology and nutritional values of edible weeds, Plants can become more beneficial in the future. Strategies of germplasm conservation hk Oe ee li f ] lau oO é Pe / of conomically important plants t i 120 JOURNAL OF ETHNOBIOLOGY FiG. 1la.—Cultivated plot of Lepidium virginicum, a weed-crop. a a, RN oie SET Oct. 1975; east of Cusarare, Chi huah d-crop, L. virginicum, from plot in Fig. la. (Bye 7040; 71 ua). Vol. 1, No. 1 19; 10, 15 May 1981 _ BYE 121 well as wild spauesie Domesticates, weeds progenitors ane weed byproducts result from ongoing plant-m t t. These interactions involve degrees oe pate tae as well as synergism between plants and man with changes in response to various biological, ecological and cultural factors For our agroeconomic societies, quelites should provide new stimuli for evaluating productivity, cultural Se api and value systems. A few grams of certain edible weedy greens in low energy input ecosystems may be more nutritious and cheaper than cultivated vegetables from high energy input industrialized ecosystems. Despite negative cultural pressure, some edible weeds (e.g., Portulaca, Chenopodium) are still available in Mexican open air markets and i supermarkets (Fig. 12). Perhaps our young, modern civiliza- tion has a lotto] y from older civilizations which have survived thousands of years by eating weeds as one component of their subsistence. CONCLUSIONS Survival of the agricultural Tarahumara populations is dependent upon edible weedy greens from cultivated fields. The diversity of plants and the ecological and evolutionary bases of their exploitation of quelites suggest that certain generalities could be drawn and +, an ¢ he rt on tN 1G. 12aand b.—W Id in Chihuah a, Che — berlandieri (Bye 9322; March 1979); b, Portulaca oleracea (Bye 9100, 9101, 9102; 2 Desi 1978 30 122 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 applied to the development of ethnoecological principles. One principle appears to be that disturbance of the ecosystem in order to push the ecosystem back to the early developmental stages and to maintain the communities at these stages is important to the biological existence of human populations. Net productivity is available for exploitation in the early stages of succession and is subjected to variation in quantity and quality depending on human activities. are able to study the processes of plant-man interactions today in order to elucidate ethnoecological principles. Edible weeds are consumed by the Tarahumara. This plant-man interaction appears to be based upon biological and ecological theory. The principle of this resource exploitation should apply to other present-day cultures as well. Because these processes are evolutionary in nature, we should expect evidence of weed food resource exploitation from archaeological studies in the forms of phytoliths and epidermal tissues from coprolites, field soils, and preparation implements. This principle should also apply to the future. Once it is understood and applied, we should expect a more realistic basis for developing relationships between human populations and their ambient vegetal environment. ACKNOWLEDGMENTS I wish to express my sincere appreciation to the Tarahumara Indians for their cooperation and assistance. Enthusiastic encouragement from C.W. Pennington, L.J. Verplancken (S.J.), J. Candler, J. k and R. Shuster is greatly appreciated. Financial support covering travel in Mexico during which certain data were obtained was extended by the Botanical Museum and Department of Biology of Harvard oo Department of Environmental, Population and Organismic Biology, U niversity Museum, and Council on Research and Creative Work of the University of Colorado; Nationa al Geographic Reciee! and National Science oc (GB-35047). ER pe not cited, voucher specimens for the Tarahumara work are deposited at ECON, CHAPA, ag O, GH, and — ei (abbreviations in accordance with Index OR camera The plants d with the pe of the Tarahumara Indians, Secretaria de Agricultura y Ganaderia de Mexico (Departamento de Forestales), Consejo Nacional de Ciencia y Tecnologia de Mexico, and Universidad Naci onal Autonoma de Mexico (Instituto de Biologia). LITERATURE CITED ALCALDE BLANCO, S., AND E. HERNANDEZ X. 1972. Estudio preliminar sobre la competencia we eiasiiasne: “ne arvenses ae el A osianen: oe sus acid content of Malaysian leaf vegetables. Ecol. Davis, R.M., AND J.E. CANTLON 1969. Effect of p. 94- size of area open to colonization on $ reso Latinoamericano composition in early old-field succession. Bull. Sociedad Botanical de Mexico, S.C. BAKER, H.G. 1974. The setae of weeds. Annu. |DAwso Rev. Ecol. Syst. 5:1-24. of barny: BECKWITH, S.L. 1954. Ecological succession on abandoned farm lands and its relationship to wildlife management. Ecolog. Monogr. 24:349- 376 96, Resumes, I Con ardgrass, green foxta foxtail sivas from various soil depths. Weeds 10: ae isd DE WET, J.J.M., AND and domesticates: evolution in the man R. HARLAN 1975. Weeds J -made BOHRER, V.L. 1977. West African dietary elements as relicts of hominid evolution. J. Anthropol. Res. 33:121-132 Bye, R.A. 1976. Ethnoecology of the Tarahumara of Chihuahua, Mexico. Unpubl. Ph.D. dissert. (Biol.), Harvard Univ Incipient domestication of mustards in northwestern Mexico. Kiva 44:237- 256 CALDWELL, M., AND Y. GiM-Sal 1973. The effect of cooking method and storage on the ascorbic habitat. Econ. Botany 29: 7. GADE, D. .W.. 1972. Setting the stage for donee tion: Proc. Assoc. Amer. Geogr. 4 1976. Personal es iwiication March 1976. GUENTHER, E. 1948-1952. The Essential Oils. Van Nostrand, New York . HARPER, J.L. 1977. uae Biology of Plant Academic Press, New Yor LEON, J. 1964. The‘ ate hngh EPR meyentt), 4 18 May 1981 BYE 123 little known food plant of Peru. Econ. Botany 2122-127, LEUNG, W-T.W. 1961. Food oepasien Table for Use in Latin America. I of Central America and Panama (Guatemala) and Interdepartmental Committee on Nutrition for National Defense (U. eee ). Natl. LinaREs, O.F. 1976. ‘‘Garden hunting” in the American tropics. Human Ecol. 4:331-349. MEssER, E. 1972. Patterns of “wild” plant con- sumption in Oaxaca, Mexico. Ecol. Food Nutr -332. NATIONAL ACADEMY OF SCIENCES 1974. Recom- The strategy of ecosystem Se ent Science 164: 262- 270. Oke, O.L. 197 review. satin Sci. 15:139-155. PAPENDICK, R.L., SANCHEZ, AND G.B. TRIPLETT. 1976. paar ae ga Amer. Agr. Spec. Pub 27. Madison, isconsin. SANTOS, J.K. 1925. A pharmacological study of Chenopodium ambrosioides L. from the Philippines. Philippine J. Sci. 28:529-547. SAUER, J.D. 1967. The grain amaranths and their relatives: a revised taxonomic and geographic survey. Ann. Missouri Bot. Garden 54:103- 7 SAUER, J. AND G. STRUIK 1964. A_ possible ecological relation between soil disturbance, light-flash and seed germination. Ecology 45:884-886. VaviLov, N.I. 1951. The origin, variation, immunity and breeding of cultivated plants. Chronica Botanica, vol. 1 WIESE, A.F. ANDR.G. Davis 1967. Weed emergence ILKEN, G.C. 1970. The ecology of gathering ina Mexican farming region. Econ. Botany D C.B. HEISER 1979. The origin and a relationships of ‘huauzontle’ (Chenopodium nuttalliae Safford), domesticated chenopod of Mexico. Amer. J. Botany 66:198-206 NOTES ‘Ethnoecology is the area of study es examines the ecological bases of huma interactions with and relationships to “ee mbient environment. latuncasea Seenity is generally sanreueeniin to neict For the purpose of this paper, the emphasis is upon richness, which can be defined as a variet given community. 5Anthropogenic community is a plant commun- ity initiated and maintained activities and represents an early hieaaaes Roemer communi ity. f which 5s elaeeaihant in the ecosystem over a period of time (usually on an annual basis). It can be defined by the difference between Gross Productivity and Respiration in a given community or ecosystem J. Ethnobiol. 1 (1): 124-134 May 1981 ON THE RELATIVE CONTRIBUTION OF MEN AND WOMEN TO SUBSISTENCE AMONG HUNTER-GATHERERS OF THE COLUMBIA PLATEAU: A COMPARISON WITH ETHNOGRAPHIC ATLAS SUMMARIES EUGENE S. HUNN University of Washington, Department of Anthropology, DH-05, Seattle, Washington 98195 ABSTRACT.—The subsist d d odes in Murdock’s Ethnog ic Atlas have been used to evaluate hypotheses: as 0 oe etaniie nen of men versus banat to — gatherer subsistence. Th summaries of the role - gathering, hunting, and fishing in the societies of the sample. The validity of these coded data is evaluated for repeaemabine Meats American nae cases by comparison with estimates of the caloric and protein contributions of the m: resource aypes based « on saipesienare tea aa “ | ethnobiological iiactaie: The - sat aia wie shown to b gathering feeamaiceic INTRODUCTION . Lee’s data on the relative caloric contribution of the products of hunting and gathering among !Kung Bushmen (Lee 1968:39) clearly demonstrate that the pervasive stereotype of men as providers, women as economically dependent childrearing specialists, does not apply to all foraging societies. In fact, comparable figures reported for other Bushman groups (Tanaka 1976) and for Australian Aborigines (Gould 1977) suggest that the female economic contribution as gatherer (measured as percent of caloric ‘amit provided) ranges between 60% and 80% generally among foragers in the arid tropics. Lee’s cross- cultural sample drawn from the Ethnographic Atlas (Murdock 1967) pabeaies that only above ca. 40° latitude does the male economic contribution through fishing and hunting meet the bulk of subsistence requirements (Lee 1968:4 However, the case is apparently not yet considered chelate: Ember has recently argued to the contrary that among the 181 Atlas hunter-gatherer societies—those rated zero on Murdock’s “subsistence erpenience” code for animal husbandry and sane not wo men, . ty pic subsistence” (1978:44 1). If we accept the ‘Atlas as at face value and sample * representative, Ember is conreeet “_ % O producers) of greater value than ‘gathering ‘“ cubsistencs. However, Picker auene to a the significance of the geographical bias in the Atlas. Fifty-seven percent of the hunter gatherer cases in the Atlas are from at or above 42° latitude compared to only 17% of the total cases at or above that paces ae erga | inane of the hunter-gatherer cases in the Atlas are from North America that continent (see Table 1). The statistics from Sigeds America do not differ significantly from worldwide figures, meeeree si ——— of latitude with ~ importance og gathering shows up one x societies s below a but among 98% of those at or above that latitude. gs. Ember equates Murdock’s scale with Jeng of caloric Fequirements met (1978:441, 445). Murdock made no such claim. Rathe gathering, hunung, fishing, animal husbandry, and oi rated 0 to 9 with respect to “the relative dependence of the society”’ on the factor in question (Murdock 1967:46). There is no mention of calories, nor is any operational definition of aunetnene dependence” offered. To interpret Murdock’s subsistence scale in terms of Pp ectivity a — Ne i ecm May 1981 HUNN 125 to the Atlas data. I hope to show here that in at least jor culture area, the Columbia- Fraser Seer mine Pi aan 13 his _ 181 hunter-gatherer cases in the Atlas— Murdock’s subsisten y biased in favor of the hunting- fishing contribution if interpreted in caloric terms. These 13 societies represent an area of 750,000 sq km drained by the Fraser and Columbia Rivers in what is now British Columbia, Washington, Idaho, and Oregon (see Fig. 1). The northern or British Columbia portion of the Plateau area is largely forested, while the southern portion is an arid sagebrush steppe and grassland surrounded by pine parkland and mountain forest. A diverse subsistence economy based on gathering, fishing, and hunting has supported continuous aboriginal populations for 10,000+ years (Cressman 1977). In no Plateau case does Murdock rate gathering as contributing more than 35% to “subsistence dependence.” Cz,' Chilcotin we 3 Shuswap =? as a p) S “WW, Lillooet ae - NX “4 i - H, utenai RIV a I ye dilbenent: European influences date to ca. 1730 when horses reached the area indirectly from Spanish sources in the Southwest (Haines 1938). Epidemics of smallpox dating to before mid oe MSa) apecadt from sand pre: or Great Plains contacts. Fur traders nt Indian wars followed close on the overland explorations of ae and Lewis and Clark, reat ssa the restriction of most of the native population to reservations by the end of the nineteenth century. Hydroelectric dam construction on the Columbia begun in 1931 has now all but eliminated any semblance of traditional subsistence fishing patterns (Pacific Northwest Regional Commission 1976). Thus estimates of the relative contribution of gathering, — a ashing, to Een’ colonic requirements ‘Test on limited ethnohistorical tion by ear informants ct by twentieth century ethnographers. These caer awoke quite detailed a he time and place of harvest— must now be interpreted i in ‘the light of scientific knowledge a the natural history and biochemistry of resource species. Previous attempts to characterize the ecological parameters of Plateau subsistence have focused almost exclusively on salmon (Hewes 1947, 1973; Kew MS; Palmer 1975a; Sn sitle 6 No atienapt has yet pee: made to quantify the vegetable input. This one-sided v the = = ie f native plant foods. For example, ogee er anatied that “the satisfaction of this demand [for ares must have been largely up to the isheries. ously low in ey since other me | 4 ie May 1981 HUNN 127 fuel value,’ among which he specifically includes bit d (Ibid.:151), 2 Plateau staples. Hewes’s estimation of the fuel values of native plants is simply wrong (e.g., for camas see Konlande and Robson 1972). Furthermore, recent ethnobiological and cultural ecological research in the area! clearly indicates a much more important role for vegetable products as sources of food energy than Hewes recognized. For example, French and I have d 1a folk classifi } y detail applied to a single taxonomically difficult genus, Lomatium, of the Umbelliferae. Fourteen basic folk taxa are named in Sahaptin, the language of the middle Columbia, at or below the scientific species level. Most of these speci imp native foods including the staples, Lomatium canbyi Coult. & Rose and L. cous (S. Wats.) Coult. & Rose. These along with bitterroot (Lewisia rediviva Pursh), camas (Camassia q h{Pursh]G ), and huckleberries (especially Vaccinium membranaceum Dougl. ex Hook.) account for the bulk of vegetable foods gathered in the southern half of the Plateau region. Preliminary studies of densities and harvest rates of these species suggest the feasibility of reliance on vegetable resources in this area for the bulk of the calorie requirement (Hunn and French MS). t a Estimates of Salmon Consumption Hewes’s estimates of salmon pti the Pp P for the region. However, his interpretation of the nutritional factors is misleading. He does not allow for the fact that the edible fraction of whole salmon is generally considered to be approximately 80% of the total weight (Martinsen, pers. comm.). Furthermore his caloric calculations are based on commercial samples. These are biased in 2 respects. They selectively represent the fattest species, Chinook (O hynchus tschawytscha) and Socke} (O. nerka), and they represent individuals taken at the beginning of the spawning migration. Yet Idler and Clemens (1959) have shown that migrating. salmon (Fraser River Sockeye) lose on average 75% of their caloric potential during this migration, as do Amur River Chum Salmon (O. keta) (Pentegov et al. 1928). Table 2 cites salmon samples on which the present argument rests. The 20 samples ee tad tn Asatte Tange of the group cited. For group y stre Y, lated the ratio as that of the distance from the Columbia River mouth to the mid-range of the group to the total distance to the limit of salmon migration on that tributary.’ This ratio is then multiplied by 0.75, the average caloric value lost by salmon in migration, and the result subtracted from one. I use Hewes’s value of 2000 kcal/person/day as the minimal daily requirement (MDR), in the absence of reliable estimates of body weight or population sttucture for pre-contact populations of the region. 1€ tabulated calculations clearly show that estimates of salmon consumption fall Consistently well short of the percentages of subsistence dependence cited by Murdock, with € exception of Thompson. The caloric contribution of salmon throughout the Plateau based on Hewes’s consumption figures averages 26% compared to the 44% average dependence on fishing cited by the Ethnographic Atlas. While other fish contributed to the total dependence on fishing (Hunn 1979), waste, loss to scavengers, and the use of salmon as fuel (Thwaites 1959:124) should tend to offset any increment from non-salmon fishing Sources, except among groups such as the Flathead with restricted access to salmon. 128 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 Tas_e 2.—Salmon proximal analyses used, per 100 g. G WATER G PROTEIN G FAT KCAL* Rivera 1949: canned 11 samples, 6 species 66.95 22.17 8.61 rz Rivera 1949: fresh 2 samples, sockeye & steelhead 67.7 22.0 9.13 176 Watt and Merrill 1963: fresh 1 sample, Chinook 64.2 19.1 15.6 222 1 sample, pik 76.0 20.0 3.7 119 Watt and Merrill 1963: canned 1 sample, Chinook 64.4 19.6 14.0 210 1 sample, chum 70.8 21.5 5.2 139 1 sample, Coho 69.3 20.8 i 153 1 sample, pink 70.8 20.5 5.9 141 1 sample, sockeye 67.2 20.3 9.3 171 AVERAGES 67.7 215 8.7 170 *kcal for fish is calculated on the basis of 4.27 kcal/g of protein and 9.02 kcal/g of fat (Watt and Merril] 1963). TABLE 3.—Estimates of salmon consumption (pounds/person/year), caloric yields (kcal/ person’ day), and percents of estimated MDR (2000 kcal/person/ day). Annual Gross Calorie Net % Percent Con- Caloric Loss Caloric _— off Atlas _ Differ- SOCIETY sumption Yield Factor Yield MDR° Rating — ence en: ee WISHRAM 400 676 88 594 30 50 -20 TENINO 500 845 87 735 37 50 -13 UMATILLA 500 845 81 684 34 40 e 300 507 52 264 13 40 ‘ NEZ PERCE } 582 983 52 511 26 40 -14 SINKAIETK 500 845 67 566 28 40 “12 S IL 500 845 62 524 26 50 “24 COEUR D'ALENE 100 169 25 42 2 30 -28 FLATHE 100 169 25 42 2 40 -38 KUTENAI 300 507 25 127 5 40 -35 CHILCO 600 1014 64 649 32 50 -18 SHUSWAP 500 845 675 570 28 40 “12 LILLOOET 600 1014 11 4] 50 -9 THOMPSON 900 1521 81 1232 62 50 +12 (1967). Calorie loss for ~ Fraser River Eroups, Pies eas penny Lillooet, _ Thompson, are re from Kew deat 2 Gross caloric yiekds ar «= derived from eee is pee: mane as . em of = * Ogee ee Uae Se hji- RB; } fab a ie Columbia or, if th yee ib bv the di ae ee - h x § The resulta aoe s so O75 the £ cea +f 1.0. The net caloric sid is simply the gross does not Occur in the region od iscus A day of ceaaes digging often netted as high as a bushel of roots. . . The skins of roots. . slipped off as they were dug, or more commonly at camp in the evening (Ray 1933: 97- 98). th =r ey June i is aah main collection season, ... This root[Lomatium cous], ety, with dine winter use. A good digger gathered 50-75 = of /qamsit/ [L. cous] i ina single ps (speaking of the Nez Perce [Marshall 1977:52)). These different locations had camas med S cisgson quanorert at soca times; the lowest, warmest ones were exploited in t st [sic.] could be wo’ orked until September. . - Harbinger (1966) said that a good digger a gather 80- a4 pounds per day of hard labor, p y informants estimate that women gathered camas for two to ise ators (speaking of the Nez Perce, See 1977:55, 57). eople moving to the mountains for berries. They obtain at this season the large mountain ti eberry dass senior echo ae aanet are usually absent on these excursions —_ [away swin thei ‘olumbia —— foe four tos SIX wore duri which, each family lays i in, for winter use, f oo hia group, diary entry for August 19, 1843, of the missionary H. K. W. Per rkins [Boyd M The second ‘of : f bl les is f y preliminary ime- and-motion studies of contemporary Indian sO0t- digging. One U matilla woman, working ata normal pace and using the contemporary steel version of the traditional digging stick, 48 33 L. cous tubers/h, or 3.79 kg/h of peeled roots. I find that I can dig and “pocket” a L. Canbyi tuber in 7 s. Allowing 3 s to find the next plant, we have 6/min or 360/h, which at 2 0 g/tuber (N N=52) gives 3.96 kg/h. These estimates tend closely about a figure of 4 kg hora cop Ales 30 5am in 7.5 h, not an unreasonable day's S saad > = een accord betw ee ese ae are Besbdiinavined ti in Table 4. The low she =for the Kutenai bitterroot harvest cited is 130 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 perhaps due to the fact that the Kutenai are on the northern fringe of that species’ range (Daubenmire 1975), and the high value for the Flathead camas harvest is noted as a remarkable achievement (Geyer, in Hart 1976:16) Per capita caloric consumption is based on a producer/consumer ratio of 1:4 with kcal/100 g standards as in Table 5. The harvest periods of tubers in spring, camas in summer, and huckleberries in fall were largely distinct. Thus we may add the estimates for spring tubers, camas, and the berry harvest to arrive at a rough but conservative annual per capita consumption figure: SPRING (Lomatium spp. and bitterroot) 900 kcal EARLY SUMMER (camas) 400 kcal LATE SUMMER/FALL (huckleberries) 50 kcal ESTIMATED ANNUAL PER CAPITA CONSUMPTION 1350 kcal This figure is more than double the estimated contribution {rom salmon for this area and 67.5% of the estimated MDR. Compare this to the 30% “‘subsistence dependence” attributed by Murdock to the Wishram, Tenino, Umatilla, Nez Perce, Sinkaietk, Sanpoil, Coeur D'Alene, and Flathead, all known to have harvested several of these species in quantity. It might be argued that harvests were not continuous during the periods of resource maturity. It is certainly true that women were called upon to preserve both fish and game harvested by the men. pei length of harvest figures cited in Table 4 are generally conservative.* Trans of the harvest to the winter villages might also pose problems, given the quantities involved especially in prehorse times. However, spring roots an nd huckleberries were ore transport reducing their weight by over 50%. In addition, there were many ie fruits, berries, tubers, bulbs, and greens eaten on the spot which have not been included in this estimate. Thus 1350 kcal/person/day from gathering seems reasonably applicable throughout the southern Plateau. The more northerly groups generally lacked these staples, relying instead on a variety of liliaceous bulbs other than TABLE 4.—Estimates of plant food harvest rates (kg/woman/day), total harvests (kg/women/yeat), and caloric yields (kcal/ person/ day). Estimated Harvest Total Daily Period/ Annual kcal SPECIES Harvest Days Harvest Yield _—_ Locale SPRING: Lomatium canbyi 30 30-40 1050 800 Sanpoil! Lomatium cous 22.7-34.1 ca. 40 1136 988 Nez Perce? 33.3* ca. 30 999 Umatilla? Lewisia rediviva 30.3* ca. 60 1818 112] Umatilla’ 6.5 7 45 28 Kutenai‘ EARLY SUMMER: Camassia quamash 36.4-40.9 14.21 677 524 Nez Perce? 18.2-22.7 14-2] 358 277 Nez Perce? 2160 1672 Flathead? LATE SUMMER/FALL: Vaccinium spp. 98-42 63.9-80.2 31 Tenino- Wishram? 98 42 Umatilla’ Soret (1) eas! ci (2) Marshall biel (3) Hunn wed — MS, (4) Hart 1976, (5) Geyer 1845-46, Boyd MS6. ote: isha on 8 hour days. ~ OT pO.) mm I I, | pm SS a err cD Oe eek May 1981 HUNN 131 camas, such as Fritillaria spp., Erythronium grandiflorum Pursh, and Lilium columbianum Hanson in Baker, and to a more considerable extent upon hunting (Palmer 1975a). On Measuring Subsistence Dependence The data compiled here do not demonstrate that the Atlas subsistence scale is incorrect, only that those scales cannot be reliably interpreted in caloric terms. Murdock’s figures are based on ethnographic reports that are almost without exception mere impressions. For example, his rating of the Sanpoil as “32500” (i.e., 26-35% gathering, 16-25% hunting, 46-55% fishing, 0-5% animal husbandry and agriculture) is clearly in accord with Verne Ray’s characterization of Sanpoil subsistence emphases. Gathering, says Ray, the Sanpoil ethnographer of record, is but ‘‘a valuable supplement to the meat and fish that hold first place in the diet of the Sanpoil (1933:97).”” Ray devotes 20 pages each to fishing and hunting among the Sanpoil and but 9 to fruits and vegetable products. Yet Ray’s own statements on the spring root harvest (quoted above) proves the contrary. Clearly the Ethnographic Atlas reflects both the bias of the ethnographer and of his informants for the less predictably available foods (cf. Lee 1968:40, for a similar oe oe among the Bushmen), which seem most always to be the special task of men to pur In the final analysis, one eras cannot tbe reduced to calories. — calories are the body’s first and larg balance of nutrients over e foaled run. Salmon provided protein in 1 more than adequate amounts, a nutrient the region’s starchy staples largely. lack. —_ salmon is rich in Vitamins, especially A and D (Rivera 1949). G when other foods were in short supply. Fruits and berries, even lichens (Turner 1977), ‘contributed other vitamins and a variety of mineral nutrients, while “Indian celeries,” eagerly sought in ai winter and early spring after a winter on a diet of dried stores, are rich sources of Vitamin C.5 © single out one resource, one nutritional requirement, or one sex as the key to understanding the success of hunting-gathering adaptations is to miss the point entirely. Human foragers survived to colonize nearly the entire land surface of the earth by virtue of judicious selection of an ample and varied diet from an extensive, empirically sound folk biological inventory of the flora and fauna. To argue that either men or women were of Paramount importance in the evolutionary history of the human species is to ignore the most human ecological characteristic, familial economic cooperation. TABLE 5.—Plant food proximal analysis used, per 100 g. G eG c Carbo- SPECIES Water Protein Fat hydrate kcal Lomatium ab 6 dried root samples! 11.68 2.58 1.48 82.41 $52 same, en fr water content 71.9 0.9 0.47 26.22 112 ' sample, fresh 19 0.8 0.12 9 108 Lomatium cous 1 sample, fresh? 67.9 1.0 0.4 30.0 127 Lewisia rediviva ' sample, fresh? 76.6 0.7 0.1 21.6 90 Camassia quam ! sample, fresh? 70.0 0.7 0.23 27.1 113 Vaccinium s blueberries, raw’ 83.2 0.7 0.5 15.3 62 Sea "es: (1) Washington MS, (2) Benson et al. 1973, (3) Watt and Merrill 1963. 132 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 CONCLUSIONS I have summarized ree which demonstrates that the importance of vegetable resources gathered by w ushmen ye Australian Aborigines. Naw! 4s plant foods play an insignificant role everywhere above 40 latitude. Murdock’s Ethnographic A tlas pcoreeiorsd FEpOnene code summaries to the contrary, the food f the Co lumbia-Fraser —_ of northwestern North America (at ca. 45°-48°N 1 de) obtai I £70% of their ips energy needs a plant —_ harvested id women. The ives divers between th this region, raise serious doubts about the general validity of the Atlas subsistence codes. ACKNOWLEDGMENTS My research in Plateau aeTatgoee has been supported by NSF grant BNS 76-16914 and by grants from the Melville and Elizabeth Jacobs Research ply basse om Museum Foundation, and ea Graduate School, Donati of Washington. This wo fited f the collat assistance of the Kamiakin Research Institute of on bares Indian Nation, and the Botany and Nutritional Sciences departments of the University of Washington. LITERATURE CITED BENSON, E.M., J.M. PETERS, M.A. EDWARDS, AND L.A. HOGAN. 1973. Wild edible plants of the Pacific Northwest. J Amer. Dietetic Assoc. 62:143-147. Boyp, R. MSa. Small; gtheI f th Northwest Coast: 1765 to 1865. MS, Dept. Anthrop., Univ. Washington, Seattle, December 6, 1978. MSB. (The journal of H.K.W. Perkins with annotations). Unpubl Anthrop., Univ. Washington, Seattle, undated, untitled. CRESSMAN, L.S. 1977. Prehistory of the Far West: Homes of Vanished Peoples. Univ. Utah Press, Salt Lake ig! DAUBENMIRE, R. 5. An ecological life-history of Lewisia ne (Portulacaceae). Syesis 8:9- 23 EMBER, CE. 1978. Myths about hunter- gatherers. rigger 17:439-448, FAGOT, 1970. Recent Changes in the Numiers a Salmon (Oncorhynchus) and Steelhead (Salmo) that Return to. their Producing Areas in the Columbia River Basin. Unpubl. M.S. Thesis, Univ. Washington, Seattle. Geyer, C.A. 1845-1846. Notes on the vegetation and general characteristics of the Missouri and Oregon Territories, during the years 1843 and 1844. n J. Botany 4:479-492, 653-662; 5:22-41, 198-208, 285-310, 509-524. GouLD, R.A. 1977. Discovering the Australian Desert Culture. Pacific Discovery 30:1-11. HAINES, F. 1938. The northward spread of horses among the Plains Indians. Amer. Anthropol. 40:429-437. MS, Dept. | HARBINGER, L.J. MS. The Importance of Food Plants in the Maintenance of Nez Perce Cultural: Identity. Unpubl. M.A. Thesis, Washington State Univ., Pullman. Hart, J. 1974. Plant Taxonomy of Salish and Kootenai Indians of Western Montana. Unpu M.A. Thesis, Univ. Montana, Missoula Montana—Native Plants and Early Montana Histor. Soc., elena, HEWES, ce 1947. Aboriginal Use of Fishery Resources in Northwestern North America. Unpubl. Ph.D. dissert., Univ. California, Berkeley : 1978. Indian gone productivity in pre- Pp; almon ar hata Autiepl. Res. Notes 13: 1-19. A D D.H. FRENCH. MS. Lomatium: A Key R ce for Columbia Plateau Native Subsistence. MS, Dept. Anthrop., Univ. Washington, Seattle; Dept. Anthrop., Reed College, epee Aiagegi IDLER, D.R., AN CLEMENS. 1959. The _ energy ena at Fraser River Sockeye . Intern ral. Progr No. 6. New Westminster, fecicich Colum Kew, M. MS. sora Abundance, Techno logy Fraser River Watershed. MS, Dept. -Anthrop. and eee Univ. British Catoinbia: Vancouver, April 1976. egy nn amen ee Le ee, * “s - NL I LL ON ee ee, eee, May 1981 KONLANDE, J.E., nn J.R.K. Rosson. 1972. The nutritive value o ed camas as consumed by Flathead oe Ecol. Food Nutr. 2:193- 195. Ler, R.B. 1968. What Hunters Do fora ‘Living, or, How to Make Out on Scarce Resouces. Pp. 30- 54, in — the Hunter (R.B. Lee and I. DeVore, eds.). Aldine, Chicago. MARSHALL, A.G. 1977. Nez Perce Social say _ Ph. . 1979. Personal communication. Univ. Washing, Seattle. MURDOCK, 1967. Ethnographic Atlas. Univ. Pittsburgh Sa Pittsbur ACIFIC NORTHWEST Recwoial. opene ae 1976. Columbia Basin Salmon a eelhead Analysis. Pacific Chick irae? oO Summary Report, September 1, PALMER, G. 1975a. Cultural ecology in the Canadian Plateau: pre-contact to early PENTEGOV, B. P., Y.N. MENTOv, AND E.F. KURANAEV., 1928. Physico-Chemical a istics of Breeding Migration Fast of Keta. B cae Sci., Fisheries Res. Sta. (Vladivostok) 2 ane 1933. The Sanpoil and Nespelem: alishan Peoples of Northeastern Washington. Univ. Washington Publ. Anthrop. 5:1-237. ecu T. 1949. Diet of a Food-Gathering oe with Chemical Analysis of Salmon and _ atoons. Pp. 19-36, in Indians of the Urban Sages os Smith, ed.). Columbia Univ. ork. st, ay fe Of Salmon and Men: Investi- 0a n of Ecological Determinants and riginal Man in the Canadian Plateau. Pp. on — ho Petit a Man and Environments eau of Northwestern America (A. HUNN 133 Stryd and R. Smith, eds.). Student Press, gary, Alberta TANAKA, J. 1976. Subsistence Ecology of Central Kalahari San. Pp. 98-119, in Kalahari Hunter- Gatherers (R.B. Lee and I. DeVore, eds.). Harvard Univ., Cambridge, Massachusetts. TuwalTEs, R.G.,ED. 1904-1905. Original Journals of the Lewis and Clark Expedition, 1804-1806. se 3. Reprinted 1959. Antiquarian Press, New or pian N. 1973. Plant Taxonomic Systems and thnobotany of Three Contemporary Indian Groups of the Pacific Northwest (Haida, Bella Coola, and Lillooet). Unpubl. Ph.D. dissert., Univ. British Columbia, Vancouver. ___ . 1977. Economic importance of Black Tree Lichen (Bryoria fremontii) to the Indians of western North America. Econ. Botany 31:461-470 . MS. Thompson Indian Ethnobotany. ~ MS, ‘Powel Mus., Victoria, British Columbia Se BOUCHARD, AND D.I.D. KENNEDY. 1980. Ethnobotany of the Okanagan- -Colville Indians and of British Columbia and Washington. MS, British Columbia Provincial Mus. No. 21, Occ. Paper pastes British Columbia UNITED STATES HOusE OF REPRESENTATIVES. 1952. The Columbia ay and its Tributaries. 81st Congress, 2nd Session, House Document 531 (8 volumes). WALKER, D.E., JR. 1967. Mutual Cross-Utilization of Economic Resources in the Plateau: An presented to the 29th Annu. west Anthropol. Conf., April 10, 1976, Ellensburg, Washington. WatTT, B.K., AND A.L. MERRILL. 1963. Composition of Foods. U.S. Dept. Agric., Agric. Handbook No. 8. Washington, D.C. NOTES 1, rach Studies include research with Wishram mee enino by D. and K. French, Umatilla and ima by E. Hunn, Nez Perce by A. Marshall, anapum and Sinkaietk by N. Washington, unn and French MS, Marshall 1977; Palmer Bouchard, and Kennedy 1980; Washington MS ). 9. Kew’s calorie loss ratios are almost certainly overestimates since he states that, ‘Total caloric value of a sockeye measured at the river mouth will be reduced to nearly one half when z reaches the Upper Stuart spawning groun (MS:6). Idler and Clemens cite losses of 2. 1% for males and 79.8% for females at the time of death on the Stuart Lake migration path (1959: 18). 134 3. For the groups cited here, the Wishram mid- range is taken as the Dalles (Columbia River mile 190), the Tenino at the Deschutes River mouth (Columbia River mile 202), the Umatilla at the mouth of the river of that name (Columbia River mile 300), the Nez Perce at the confluence of the Clearwater and the Snake Rivers (Columbia River mile 324 + Snake wa mile. 140), the Sinkaietk at the mouth of th Okanogan River (Columbia oe mile aa, the Sanpoil at the San River mouth oe Columbia ae mile 615), the D’Alene at Spokane Falls, limit of migration on _ Clark a. eset 0’ — the River, and on either the Columbia River (Columbia a Lake) or the Kootenai River (below Kootenay Lake). JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 Limit of migration on the Snake River is at Upper Salmon Falls (approximate Snake River mile 400). Mileage figures abstracted from Fagot (1970:111-124) = United States House of Representatives (195 in . canbyi was arly could have been harvested as deadly as Febeiaii In 1979 L. canbyi and Lewisia rediviva were mmonly available up to m elevation ae April 1. In 1977 L. cous ee Lewisia rediviva were still being harvested by Umatilla Indians at 1400 m in the Blue Mountains of Oregon on June 22. Since camas may be harvested into September (Marshall 1977:57), a root harvest period of 100+ days is possible. . Benson et al. cite 66 mg/100 g ascorbic acid or the young growth of Lomatium nudicaule (Pursh) Coult. & Rose (1973:145), an important “Indian celery” of the region. ee ee ee, ae ae a ~— - — ~ —_—— aaieeetneliamtnttiy same a, ii a ge OS NE ll | “wn i - J. Ethnobiol. 1 (1): 135-164 May 1981 DEVIL’S CLAW DOMESTICATION: EVIDENCE FROM SOUTHWESTERN INDIAN FIELDS GARY NABHAN, University of Arizona, Office of Arid Lands Studies, Tucson, Arizona 85719 ALFRED WHITING, (deceased) HENRY DOBYNS, F lorida State University, Tallahassee, Florida 32306 RICHARD HEVLY, Northern Arizona University, Department of Biological Sciences, Flagstaff, Arizona 86011 ROBERT EULER, National Park Service, Grand Canyon National Park, Grand Canyon, Arizona 86023 \ } re } oe | 9 CiaWw \f b 7 / rf and grasslands, have been utilized for food and fiber by numerous Indian groups in south- western North America. A white seeded devil’s claw, with longer fruit providing more useful basketry fiber, has been cultivated by basketmakers in California, Nevada, Utah, Arizona and Sonora. Historically, this devil’s claw taxa has been poorly understood, and has often not been recognized as being genetically distinct from wild Proboscidea spp. in the region. , eer . ] ] . 1 A ch 7 ] ; : ‘ou = Ll D. - oe mt r gh with other FProvoscidea, in Lae ee Re 2... % os ed L the wild and under cultivation, “al g g PE white seeded ppears to b t closely related to typical Proboscidea parviflora (Woot.) Woot. & Sandl., and their differences are in those characteristics most often altered via domestication. It is suggested that cultural selection by basket-making native farmers in the Southwest, and natural selection in their field environments can account for the distinctive- ness of the white seeded devil’s claw. Additionally, ethnographic and linguistic information elucidate the white seeded race’s affinity with P. parviflora, yet also its distincti ive cultivated crop. Af luati these various data, itis concluded that P. parvifl lerg YP domestication, increasing its usefulness as a basketry fiber producer. It does not merit the status of a cultigen—or fully domesticated plant—since its survival from year to year is not entirely dependent on man’s intentional planting. Yet, white seeded devil’s claw is today highly associated with cultivation in a few Indian rancherias in Arizona. S L 1 1 . f INTRODUCTION To domesticate a plant literally means to bring it into the human household. The process of domestication involves cultural selection for economic characters, as well as natural Selection in the man-altered : where the pl grown. The intensity of these Selective pressures is not h hti through space. It varies with the demand for the economic product, the kind of horticulture or agriculture practiced by the people involved, and the degree of geographical or phenological isolation between the cultivated Plants and their wild relatives. Often, an incipient domesticate has not been recognized as such. This is because the Cultivated plant may still have the appearance of its wild relatives. Additionally, the early Stages of cultural adoption may not involve formal husbandry so much as simple seed selection, sowing and protection in an otherwise unmanaged environment, which looks wild” to observers from another culture. ae Given these conditions, it is not surprising that it took Europeans more than 2 centuries in ses thwestern North America before they questioned whether certain plants the Indians utilized 5 I rely wild crops. In the case of devil’s claw (Proboscidea), the use of the Plant for food and fiber ized d e its outright cultivation was noted (Fig. 1). Additional time passed before scientists first suggested the plant as a possible Sucate. 136 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 Tont Y ae Eg © i} \ Koma) awe rae Apache al nn Apache + ) ‘ H (\Tarahumor : Warihio —. j ‘ + = Used devil's claw in basketry @ = Cultivated devil's claw ? = Proboble Fic. 1. Locations of tribes growing or using devil’s claw (cartography by Alison Habel). arly researchers suggested the presence of “introduced” kinds of devil's claw ——e the ie However, Castetter and Bell (1942:113, 202) hae? almost agree dependent upon cultivation. They noted that a second kind of aes devil’s claw among the Pima and Papago was different from the wild kind in Arizona in respect to several characteristics. They claimed that this longer clawed, white seeded kind was ear only under intentional cultivation or as a volunteer in agricultural fields. has indicated (in Correll and Johnston 1970: os that strains of Prokosais jariiloia (WiSeL. ) Woot. and Standl. semi-cultivated b groups ar anomalous for Proboscidea in that they have white seeds. Recently, Yarnell (1977), without further data or analysis, concluded that Proboscidea parviflora was one of only Spee definitely domesticated north of Mexico. If one goes to the ‘Teservations of the Papago and Pima Indians today, one finds 2 somewhat than that described by Castetter and Bell ( 1942). Both black and white seeded devil’s claw are cultivated in fields and gardens; additionally white seeded devil’s claw can be found on roadsides and in arroyos within Papago ranchenia’; growing nearby the more common black seeded Proboscidea parviflora (Nabhan and Fritz 1977). Given this information alone, we feel that the data in Castetter and Bell’s work do not place devil's claw in enough of a cultural and botanical context to-convince the scien! community that domestication has actually occurred. | The proposed status of domesticate for Proboscidea parviflora has thus been ine unrecognized and untested. We will use the devil’s claw example as a case study 1n ho ific LS As ge 9, rg —— a, — ——< a . May 1981 NABHAN ET AL. 137 anthropologists and botanists methodologically determine when a plant has undergone cultural selection and domestication, over and above mere cultivation. Additionally, we discuss problems in the interpretation of historic specimens and ethnographic data, and eet some testable indicators of domestication. g how and wi hed i ight have occurred for Proboscidea, we wish to emphasize how much has yet to be learned. We! hope.to encourage further research of devil’s claw as well as of other little-known crops. Such research is urgently needed, since many minor crops have been abandoned within this century as modern monocultural agriculture has usurped the land and water formerly allotted to smaller scale mixed crops. Devil’s claw cultivation is a case in point. Today it is practiced in only a few “islands” within its former sii oe to the demise of traditional Saaheny and agriculture among several § : d Proboscidea germ plasm have eroded within the last half century. DISCUSSION Botanical Background and Historical Recognition Within the New World family Martyniaceae, the genus Proboscidea is divided into 2 sub- genera: Dissolphia, including 3 yellow-flowered perennial species; and Proboscidea, including 10 species, most of which are annual, with flowers of cream, pink or purplish hues (Van Eseltine 1929; Hevly in Correll and Johnston 1970; Hevly 1969a; 1970). We will be concerned with 3 annual species of the southwestern United States and adjacent Mexico: Proboscidea fragrans (Lindl.) Decne.; P. louisianica Miller (Thell.), and P. parviflora (Woot.) Woot. and Stand. Partially overlapping in range, (eg-» * in ‘Teeas); — species are nevertheless phenetically distinct and ble (Table 1). However, the 3 species ave been found to be experimentally cross-commpatible. First generation (F!) flowering and fruiting hybrids can easily be obtained, although F' fruits contain few seed (Anderson 1922: 141; Perry 1942; Hevly, unpubl. data). A more thorough treatment of the genetic and bio- geographic relationships of these species is currently being prepared by Peter Bretting at Indiana University, in a taxonomic revision of the genus Proboscider In the 1870s, Dr. Edward Palmer published 2 of the fi western devil’s claw. Palmer (1871:422) noted as “ Apache a cooked the immature fruit of Martynia violacea for food, and basketry. Additionally, Palmer (1875:112) descriked the preparation of fruit of Martenia Proboscides as a black basketry ornamentation, as it is used “‘by all the tribes of Arizona.” At the time that Palmer made these comments, only 2 annual species of devil’s claw were recognized in North America, both with large calyces; M. fragrans, for which M. violacea isa synonym—with purple flowers; and M. yng: for which Martenia proboscides is a misspelled synonym—with white to pink flow Neither article acknowledges if Palmer ened voucher specimens to substantiate these identifications: thus there is no way of checking | wi suggestion that 2 species were then utilized. The paucity of voucher specimens, a sisted into recent decades; “br still not clear if more than one devil’s as species has been utilized in Southwestern Ske During the decades that followed Palmer’s articles, it became apparent that the most pees kind of geval’ s coed in biome it New aycrel and adjacent Mexico is distinct from sal 1ifferently colored flowers. This species was named Martynia parviflora Wicacu is in , 1898, but was transferred along with M. louisianica and M. frangrans to the genus Proboscidea because their flower and fruit characteristics were incompatible with Martynia (Hevly 1969b). fC th Vol. 1. No. 1 JOURNAL OF ETHNOBIOLOGY 138 snaouos snqwog “dds vdov0}4x saaq Aq Buissory UDG"G-6X8-G°9 »peIq-uMoIg ajug-prey uD9z 01 91 08-0F uorod paaind Ajavenzue uo sejnpuels ‘uiod }uawYpENe MOjaq asojUswOL 4O snojjtA Ajasieds ‘snouqe|) “yiZuay aqni ainua 24) Siop ajdind-pas yrep % sayr0]q JOM ‘eoutjapind Mood 1yBug sayno(q adind yim uaijo quity ‘ary Ajazea ‘aydand ystppas 07 ajdand 1ajor, UIWIG9-G¢ wuz] si2eIq !WUGZ ‘aderjoy ay) Buisseding aseq 18 [e1areyenbe o1 aeps0) !wo¢z 01 dn yipmm ‘aioe sasnuts !paqoy f-¢ 01 ainua Apieau ‘aeao Ajpeoig 01 proijag 9piM WZ x []B) Wig’ OF ‘aIpeW ELIA ay? Jo 18e2 OoIxayW UsaIseay OU % Sexa.] s099q-UEL | SNVHOVESA ‘d “nddiiud) “.vnbi1qgo ae a wuinsse) 801s] ‘snpinsaf jroo “91Dt475 01240] 4903N p ‘,81)D1U2p1990 meneroyery* estunwmwUo? sapoen}7W wous}OApuD xafidso vdor0) 4X ‘uansouontsmag vdor0) kX *4ds saposstay ‘,01.snBuv pssopBou2y ‘wiafrau sidp vouzj}04puv xajidio vdo20)4y ‘snaouos snqwog ‘ipany v1piag ny 37S) (Buyjag) (aancayja %G1 Buryjas) (¢Burjjas aurog) s20q Aq Butssorr) saaq Aq Burssor) saaq Aq Burssorry saeq Aq Burssoiry %92-08 08-02 ‘%SS-06 %9°GS-6' 8S 49 Sb-SE %Ob-SE %E'8E-GE ME OS '6E uaaaun ‘padejag uaaaun “padetaq usaaun ‘pakeiag UdAa ‘EIPIWIUT WOW] uMolg-ye1g Aes8-99e1q Avad-42e1q Aes8-a11y UIWIQ-GXOI-L UIWIG-G' >XOT-B UWGSO' L39°X0L0"-b86" WWOFO'91S X60 896" LOSI 299-04 68S 31F9 64 yPe1qQ-uMoIg ypPe1G-uMOIg yeIQ-uMOIg RIG 197" 9ynq-puepy 2ug-peyy anuq-paepy aiqetd-ryog W206-6 upg PL 's1 WL 3F0 LI WORST EGS a | TBS GBs $9Z 08-OF 08-06 001-09 002-SL ‘uonsod ‘Wuiod juauyseNe MOjaq asoiuauIO) —-aaoge snoiqe|3 Inq 1uswyDeNe Jo — aaoqe snoiqe]8 nq yuaurypene JO aaoge snoiqess nq yusunypene jo JO snoj]tA Ajasieds ‘snouge|s) = utod s1ay) Mojaq 10 We Le[NpUETs) —UIOd say) MOjaq 10 18 azeynpuel) qurod syay) Mopaq JO 18 JeENpETD duasqe ‘aqM) ainua ayp punore Buryre] Ajyensn saynojq ‘aqni amo, Buryre] Ajpensn saynoyq !aqna Jamo] 14 yep yo yyey siop ajdind-pas 10 adues0 ‘sayno(q WIOIA ‘sauTpapInd aZues0-Mopjar Ul WY?) YIM palelosse slop par [Jes 2% ‘sautpapind moyjad 1ySug UT WAY? IPIM paleosse Op par [Jews xB ‘sautjapin mopjad 1yBug ur ‘womp iim pareizonse sop pat [Jews ¥ ‘saurjapind moppad 1ysug saqo] 1addn uo saqo, saddn uo saynoq adund yo] sadd ynorq saqo] 1addn uo 4O 1Uasqe sayno]q adind ‘ajdind dand yi yO quay ‘arty ystyurd yeymauwios 07 any [ING Ajszeou 07 yurd ‘ajdind-ystppay wugg-G¢ UID -GZ wiwg]-/ Si9e1q ‘UNWUIOZ-GT wUg-g S19BI1q ‘UWI ]-O| ‘aderjoy ayy Bursseding a¥eryoy ay) Aq passedins 10 Surjpenby ‘aseq 18 pe1aepMbaut oO) aveps09 ‘wo cz “a8eq 18 aIeps0d ‘WD O¢ o7 dn yiptm ‘aeynonuap ‘asmqo oi dn ria ‘a1enuts 01 ainua ‘a7eAO0 Sasnuls ‘paqo] [-§ O1 aNNUa ‘a1eAO Q[-g Ol aINUA ‘a1eA0-protjap (01 areao Ayreau o yurd ‘ajdand-ystppay ystppar Apanes tyurd on any Tue | WIULIb$-$§ UWI S-1E WUT [-p SI9BIq /WWIOZ-¢ | LI-8 q LS aSe1joy ay) a8e1o} ayn Aq passedins 10 3urjjpenby Surjpenbe Ajases 10 ‘Aq passeding “aseq 18 ph np nes ‘aseq 18 d jeiaepmbeaur 01 aepso0d ‘ogg on dn YIP ‘are1;nonuap ‘asmgo sasnuts saepnonuep Peqol a, s 2 os Dp Pim ule" x (Te) WZ OL oPIM Wig"] x [[e) Wig OF Spe gg eo ol “epeAaN Ul ‘arayMas]> = aed Ajawianxa ‘serusoptye ut ey “ENUOJI[ED Ul BANUDAPY “SEXIT “HIPPY BLaIg JO 189M OOTXIP YON “SEXI] SOIIG-SURLT 3B ‘OIKaW euozuy ‘operojoy 01 'g¢) wiaynog YOINVISINOT ‘d VUOMIAUVd “d Q'IIM FHL NI SUT aWOS ‘aneao- -1e1191q30-Qng opiM Wg x [JE WE] OF OoTmKaW MAN 3g euOZUY ut Ajpeuawtadxa % (Sagi may &) euoztiy BIOUOG ISBIYyLIOU BR “euOZLTy ‘eIUsOyI/eD) IseayINOS “Ye Wiaynos ‘epeaa_ WayInog VAOTAIAUVd ‘d *‘NOLLVALL1NO WaaNn dOVU GAdaaS-ALIHM (ALVOLLSANO G4.1S499N5) (aoreurtjod..) utarold 1a. Jag [!0 1ua2 Jag uoneumasy 40[0) azis Mina Jed saquing ‘agas 40/07) Auqiqettd suaosagnd juaweyty uoneuaueui0 jeusaiut eyfor07) aouarsao[juy sanex] aris Wed uonnginstp arydesB00n) wWaLOVaVHD ‘IsaMyINOS ay) ut eapissoqgoig fo uostundwoy— | T1AV.L. ee en 1 t May 1981 NABHAN ET AL. 139 Over the hundred years since Pahoer s introductory notes, the use of devil’s claw has been North America (Table 2). In addition to basketmakers’ use of fiber splints from the dried nes, aa s claw fruit and seed have been eaten, and used medicinally; the fruit have been ls and ornaments, and have been given supernatural significance. Again, because written references have seldom been accompanied by voucher specimens, and because obsolete nomenclature has often been utilized, we can only guess which devil’s claw species b=) seen After the turn of the leg ethnographers began to comment on the planting and protection of devil’s claw (Table 3). Russell (1908:133), os (1928:134), and Roberts (1929:141) imply that cultivation or lack of it was directly related to the abundance of wild TABLE 2.—Devil’s claw use in southwestern North America: Early ethnographic references. Basketry y Early Other Culture Group Use Use References Identifications References ueblo x Robbins et al, 1916:57 Martynia Jemez Pueblo x Castetter, 1939:notes Martynia hiti Pueblo X _ Lange, 1959:150 Zuni Pueblo xX Stevenson, 1909:46 M. louisiana this report Hopi Pucblo x X Hough, 1897:33-44 M. louisiana Whiting, 1939-92 Hane P Robbins, et al, 1916:57 Martynia packs (genesal) x X Palmer, | 2 M. violacea Palmer, 1875:112 Warm Springs Apac X Gifford, 1940:45 Martynia Mescalero Apache x Gifford, 1940:45 Martynia Chiricahua Apache xX Gifford, 1940:45 M. louisiana Castetter and Opler, 1936:45 Huachuca Apache x Gifford 1940:4 Martynia Ci Gifford, 1940:45 Martynia a Xx Gifford, 1940:45 Muxtynia Buskirk, 1949:164 acta po Mason, 1904:512 M. louisiana Rea, 1977: notes San Carlos Western x Hrdlicka, 1905:404 cat's claw Roberts, 1929:141 Apache mee Yavapai x X Corbusier, 1886:324 s cla Gifford, 1936:281 3 weern Yavapai x ie Los 281 Martynia —— x ra Martynia McKennan, 1935:80 oe x ar ag M. louisiena Spier, 1928:134 thern P; (general) aaa x x si mee noted in Bye, Virgin River and Moapa So. Pai x x this report puvwits So. Paiute x Stewart, 1942:340 M. proboscidea Drucker, 1941:110 a vam x Stewart, 1942:340 M. proboscidea Kelly, 1964:78, 80 — x Mason, 1904:519 Mar Stoffle and Evans, 1976:4 a x Merrill, 1923:7 M. pro aa idea Zigmond, 1978:202 — x Coville, 1892:358 M. proboscidea Merriam 1903:826 Shek” x Steward, 1941:338 M. proboscidea Jaeger, 1941:248 = etate x Steward, 1941:338 M. proboscidea tenes oe x Merrill, 1923:7 M. proboscidea lee — x Merrill, 1923:7 M. proboscidea Voegelin, 1938: Ki - Akwa? x Merrill, 1923:7 M. proboscidea Mari a nt Fai) x? rucker, 1941:110 Martynia Gita py x x Forde, 1931:124 Martynia Spier, 1946:129 ee x x Rs pace Martynia Russell, 1908:133 a x Watson, 1898:66 N Imen arihi X Paid & Moser, 1976:23 P. altheaefolia jee . Xx entry, 1963:92 M. annua, M. fragrans Onova? (Lowland x Pennington, 1980 arenari Rea, 1978:notes tac e, Sarna te eee I eee ee P. sinaloensis 140 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 TABLE 3.—Ethnographic references to devil’s claw cultivation. Culture Group Citation Pima Hrdlicka ]1906:43 Pima Russell 1908:133 Papago & Pima Kissell 1916:202 Pima Breazeale 1923:42 Papago Castetter & Underhill 1935:57 Pima & Papago Castetter & Bell 1942:202 Pima Curtin 1949:107 Papago Dobyns 1952:211 Havasupai Spier 1928:134-135 Havasupai Whiting 1942:378 Havasupai McKee, McKee & Herold 1975:13 Apache Roberts 1929:141 Tubatulabal Voegelin 1938:30 Shoshone hoshone & Steward 1941:338 Northern Paiute Shoshone Jaeger 1941:248 Shivwits & Kaibab Stewart 1942:340 Southern Paiute Kelly 1964:80 hern Panamint-Death Smith and Sinnpson Shoshone 1964:46 Quotation blag cays arsied A eae My. the Pima in their melon patches. + a oa Altes aor “oy for the piped basket. The supply i larg gh, anda f re ie nted each year ” ee basketmakers. Alth di fields, since they find the cultivated sing yields Leg with hooks of — igi finer grain and a better black. The g il I have never seen it growing out upon the f y cultiv ated fi es 1 wash. The indians erg Se stalks around sgl ‘houses, as the wild i basketry.; 1 1 s1 1 (M x ce y Now many sow ~~ = _- raise a ae crop. Bey Set | re = my of od black | ee 4 oh. £. : nl f. = te l They were P P planted in hills ... ee ae Iti sk f, a. | ee Ie} Gn ild lains and mesas. P. di i d Loih (ad aae co Ei Pe een \ i} Lh iiank | for basket desi a The second variety, with hooks 25 to 30 cm long, was introduced by ae a bs: Altho =~ the wild plants are also used, it is customary th corn same t TP. . en ae 7 a | 4 4 . “1 sar DrTOduUct |Pagaq) F whole field of the introduced y> & fF t W : ERS ER ee Te re apes eee YS WG 4 4 oh lant from the i. « 4 Meds, aT Bs 6 he Boni} while still i 5S another ‘gested “ Yavapai a as a source. eg : ] lrivated. The latter y is adeq rn Oy eae a ] f basketmakers gather the ent: wild form. le 1 1 a i +. £ a 1 Folin their country in its wild state. Yo, Fe L > i | a hl 1 Soff. 7 A f orn. a” Martynia proboscidea Glox ... which is classified as weed, grown occasionally in gardens ow, pods sometimes NP-F. pf Fish Springs, California Bishop}: 5 V. mI ray Se c ° re oe pub of Death vaueyh planted any s claw in n gardens. kh hy hh f Hungry M. parviflor. a 4 it from Saline . nf Big ae f king blac rims in their baskets 0 fiber of the fruits. He procured seeds and pean them in Joh the p! still flourishes there. Devi s claw (Mert per idea) S K [Southern Paiute, Kaibab} Use learned from SS Pai (Kelly, ms, states seeds i have been planted near pata: Arizona). ; cee St. George, Utah, planted locally. btained from a Mamie Button’s basket is woven bed willow, Joshua ip roots and fibers _— the fruit devil’s . The dark b h Mame 8 . SS ates ae re ane se ewe rae | 1 in their garden for use in basket-making. ne , devil’s claw in their area at the time. It is usually stated that devil’s claw is grown for its fiber used in basketry, although in certain cultures (e.g. the Papago), seeds were no doubt eaten also Although split devil’s claw fiber splints have been found in cave po agen — roughly ted A.D. 1150 in Arizona (Exhaus Cave - Hevly and Hudgens, MS) and New Mexico Tularosa Cave - Kaplan 1963), the antiquity of devil’s claw cultivation is an on ere Did cultivation for basketry fiber occur in previous centuries, Saige: e ignored by feet or did it begin this last century to keep pace with basketry sa Begi g with Spier (1928:134-135), there are sasements that a longer sai culties variety is apelin rather than being indigenous t (1916:202) implies that the wild devil’s ami in “a country is sno’ 3 in peer al — that cultivation produces longer clawed fruit with better qualities for basketry. ~~ May 1981 NABHAN ET AL. 141 From the 1930s onward, specimens accompanied by limited ethnographic data were deposited in museums and herbaria (Table 4). Associated f ambiguous. For instance, Percy Train’s note that at Moapa, Nevada, devil’s claw is “In Indian field,’ does not clarify whether or not he collected an intentionally planted crop, a self-seeded feral plant, or a wild ‘“‘weedy”’ volunteer. Particularly in terms of fruit size, we don’t know if collectors chose an sypicaly large fruit, a representative individual, or a conveniently small fruit that could be “‘pressed and mounted” easily. Botanists continued to abel their specimens with obsolete AE aE pat of course anthropologists were no more aware that finer taxonomic distinctions wete possible. Fig. 2 maps the sites of cultivation. Castetter and his ewan during their studies of Pima and Papago ethnobotany, amassed considerable information regarding devil’s claw cultivation. Yet even their information is ambi 1gU0US ON some ma jor points, and at sacs it is contradictory. Castetter and Underhill (1935:57) note that Mart wild in Papago country (sic), but ong ago, women began to protect fertile patches of it; later they began to sow its seeds. Castetter and Bell (1942: 113) identify as M. louisianica both the wild black seeded variety, and the white seeded, longer clawed kind that “never grew wild,” but is propagated by planting in holes. They doubted that Pimans who asserted that devil’s claw has been cultivated for a long time, and suggested that only wild Martynia was utilized before a commercial stimulus increased basketry production. 1S Tre) 105 ‘ t aa ‘ ; 7) Bishop p \. Stide?) i Raw! oe i! i yee j oKern River 35 : % Clarkdaie% Camp “J Qq fo Pre: scott” Pe Verde ? white! ) Fone a. . River AN} \ j Fort meine a e * ©® SAN Corios’ E. Le Se ' = Akchin | oy mai Blackwater j me ~—s o Kohatk } — is B olucson ] 4 } Pozo Verde rte! ieee, i 30 mins \~ A a es , vibe: | © = White seeded race in cultivation (specimen) ® = Black seeded p é=e P parviflora in cultivation (specimen) *nographic record of Cultivation (no specimen) FIG. 2. Locations of devil’s claw cultivation by Native Americans (cartography by Alison Habel). Vol. 1, No. 1 JOURNAL OF ETHNOBIOLOGY 1G 74 ((Wd BT Me] ‘paas aity M) “(wo 1% MEP ‘paas ity M) ,,Aneyseq Ul pase) yNy aq, qweu o8edeg™aydind 1431] s1aMozy,, ‘(Spaas a1tyM «° YTOU-S Pa[[E> St pss IGM VY Sud 1% MEP) 4 A JOS SUL, “YOOY,2 pareanjncy,, (WD ggg-G°gT SMETD) _lods aydind tir airy s19MOT] ‘paas a1Ty M,, , Suapses may) ur penuue (ats) papaas-yoe]q sty pareanyns suonng B SINUNFY IY T,, “WO G°gz SMETD ‘pas aITYM ‘puodaq 10 splay ur Jt ray JON “IINYy ameUWT UO WD £1] 01 dn SME] ‘paas Jy M, Anayseq UL pase) ‘(Buoy payooy) yynAryvy voy “spyaty UI palUe]d “pees ae aw? ¢ 38) Me]? oo PIM aup Ayqissod ‘aumyeutunt yins4) ,, cin yayseq Ul suerpuy Aq pasy) “spjayy UT,, (wo ZT x 6] 01 dn ‘adse] SPABAT SW Z'Zs “9 ‘LS ‘BZ SMEID “YEI() Ur ey) (aouasar0[jut Butssedins saaeay ‘paas aryM WD §°bZ B 1S SMETD) , TOS PTY uaprey,, (eperaany UI PIIM Ul aley ‘aoUadsa10]JUI Burssedins SOABIT] “OOD JAMO] YS!) ,,"Platj UEIPUT UT, (aaoge SB QUIRS JDUaIaJUT :S910U 19y10 ON) ,,pedeosy,, (19 OZ JO YIBZUaT METI YIIM ‘saTpaaN Woy psonpo.nur sem I s1sad3ns (8261) puow 17 Inq ‘uasuttseds uo sai0u OU) “peta ANayseq YIe]q 10] pasn,, sONsaIeeY) aansasing HYNI ‘NWO ZIav ZIV winasnyy ‘O-D OUIpeusag uRg OWN ZIav OW ‘SN ‘AN ‘ON on ‘ONaY ZTUV ‘AN on r | on e1reqi2 yj 10 winasn py bL/83-61/11 R£/Z1 1 49/26/01 o/b L/S £6/8/0I OF/S8-81/0I O/LI/L F681 /L2/6 L8/L3/6 8§/9/9 $0/6 L8/60/L G8/L/L aeg 1€ ‘O€-#o# Maqney punwpy S919e%°d “HU a8pog a1ounea RSZI “LSZI# saYOR PTISIOA yas *D Joy pppyud “a bSbbH 1AkIq “AD bOShH/LEOTH 3unIyM “AV 6LIG# J9A0[D Epez[y ggog# seuof snore] GLo# Aaydinyy eAq LI61# ured Adtag Aasjoom “a'f 3/Lg puowarz §1# UTT9390A “MA # 7B 10199[[9) ‘uauads jo Ailjenb s00d 10 o3edeg 916L4 o8edeg bL-a o8edeg o8edeg auoysuoys ‘AaT[EA Yeaq durwmeueg rednseaey tednsearyy rednseaepy gainteg WaYyINoS J9Ary UTSITA ainieg Wwiayinos edeopy ainteg wayinos edeoyw eauoysoys NnsIemMey auoysoyg jeqejnieqn 1 ammyn5 eas ose r : ks .: OMAP] ‘VAOUOS ‘apiaA 070g ‘Tuy “OF Bull “UONPAISIY uerpu] o8edeg ‘jeusaiq Tuy “OD eultg ‘uONRAsasay uerpuy o8edeg ‘uoson] “or eutg ‘uoneAlasay oSedeg ‘yeag NITY WoT paag PTE “OD ou] ‘doystg jo “os rut g “yaar UOSMEY “YOUBY UOSIDATT A “HUY “OD OUTUODIOR ‘s{[et ofeaey eau ‘uodue; rednseaezy muy “OD OUTUOIO,) ‘UOAUET) NSEARL{ outuose;) ‘uodueD rednseaey YEIQ “OED uoWuryse ry “UBT A eperan “OD yelp ‘Auaay uerpuy edeop Bpeaany “OD ARID ‘uoneAlasay UeTpUy edeop FHIED “OD Way ‘uss4 FeO “op: WEY ‘SI ainted “yT}oUCW ‘JHE “OED Way ‘UOp}EM AntyeI07] DaptIsogord vuojfrasvd “d Dept7soqord paopfrasnd “¢ vaojfrasvd “Ww gmiopfiasvd “g viopfraund “g vaprosoqoid “"W vrunsIno] “W varunisino) “We voruptsina) “Ww gwotuntstnoy “qf gvotuntsinoy “W vapisoqod “yw I peursugo ‘DLDgAay PUuv Suunasnid Ut azvI1;saUOpP pajsadBns ay) fo sps0Ias 2]gQISSOg— fh TIAV 143 NABHAN ET AL. May 1981 ‘ade [IA punore PIM “peas yoeq “WD 7] MEP ‘ITY eT[OIOT) ‘peOl WIp JO apts uo uorssardap uy (uUMOId ST alBonsawiop a1ayM) preAasnoy Isasop> WoL WOOT ueTd plim “Fae “wo Zz SMe] ‘paas yuRG “AqieaU UMOL3 Se. -souop Aq asl AT[PINeUaz 40 [R19] aq 01 wey “BAe WO GZ] SMELD UT paas B13 pue ary “YSPM BSOY BIURS UT ‘paas RIG “Hae wo py smep ‘per wp uel prim ‘iaqy Anoyseq se [pas 01 UMOI ‘palesuuy ‘asnoy puryaq uapres ur pawanyno (¢°§1) ,,2d41 p[iM,, papaas yoeiq pue tz sep ‘saqo| uo aydind yim ‘ony e[]O10 “eultd BID Aq saqy AN@YSeq 10} payeaN[Ny “UW OT xX Zz eal ‘AY paas ‘saqoy uo afdind yim ‘airy eTfo107 1O[OD B[JOIOFD “WD FZ SMELD ‘Spaas aiTY MA “WeIp wig] x [ye) wo g° on dn yenuue paeaniny uMOITD “ypied inci sciita ul + pavesayon pue paiaeM siaa) -UNJOA OS|e ‘Uspres preALOOp Ul pares pue paiue]d ‘siaMoyj aed ‘smep> Buoy ‘paas airy “S9ARQ| ABIE] “1aMOT] ajB@q “Anoyseq 10) 1aqy aonpoid 0} paiaieM Sutaq ‘ysenbs etm uspsed ut saaun[o, “waqly Anayseq Jo} redng ul pareany[ny sme] MOT ‘paas aTY AA ‘daqy Anayseq 40y ayepysepD ur pareanny ‘paas Avid-a.1YM WO] paeAN[Nd sBurpprsg “Anayseq Ul pas/) “spjaty Ul papaas-jjas Udy palesajo J “a8e]] 1A ay) reau saoeyd aisem Pue splat Ut PIE M “P22s YET “MepD WOYS “Anayseq Ur pas() “(Buoy payooy)-epnAryey eeyey “SPpal] UT payuR], ‘pees aryM ‘smepd Buoy ‘pyp Ownz “Bursuiq ures pue Burupysty yIIM PaIIBLOsse ‘s|ME IO} ‘SEUTYDEY UT pasy) “splaty UL UaYyM paam paieiajo | “apispeoy ‘ainjewUly sonspaperey) ansaaang ZIAV ZIAV ZIAV ZVaV ZIaV ZIav 9L/1/8 9L/II 9L/01 9L/01 94/01 9L/1/8 9L/96/L 9L/1/01 9L/F1/8 9L/+1/8 9L/$6/L 9L/18/9 ‘9L/61/9 OF/S8-81/01 Ob/S3-81/01 LE/*1/6 aed LIN “YA Woy T]-FRg ueyqen OLSx ueyqeN 69Gx URYqeN cgcx ueyqen pgGX URYGEN pR¢ UEYGEN Geg ueyqeN a9zcx UeYqeN LOpx UeYQeN pue Zumaig copx ueYQqeN pue Zumaig PIng ou WO} §b€ ‘$78 /660E SUNIYM LIE ‘629/660 pue $b§ ‘$278 /660¢ 3untyM SOsha /LEOI Bunty FOSEE/LEOL Sunry 168z8/FS8 Suny # 3 10192][07-) rednsearyy odedeg osedeg odedeg odedeg eultd eultd o3edeg oSedeg oedeg tednseaepyy aypedy rednseaepy rednseaeyy ido anand uoson Ty ur UMOId padg 1] 009'G “7UY “or outuod0’) ‘a8eqTA redng ‘Tay Or) BUTg ‘esOYy BIURS ‘Tay “Or Bulg ‘Rsoy EIUES ‘TU “OF Bulg ‘Rsoy RIUeS “Ty “OD eultg ‘Avy eau ‘UOWIIS URS uoson | Ul UMOIB p2rg “1} OFO'T ‘Ty “oD edoourp ‘apewoy uoson [ut UMOIB Paag "ZU “O07 [eUrTg ‘Bare UONPAIISIY IATY P[I9-a¥ptjoo) “y 00S'S &? Tuy “OD: Bultd “FEL NMED 1} 000°S-008'2 ‘Tuy “O= Bulg ‘wedzuy Y bLEZ Tuy “oD ewrg ‘emodoy ‘Tuy “oO: redeae x “[epsye[) y 009°¢ “Tuy “Oo: OuTUOs0ZD ‘adeI[TA tedng °y 9009'S “zUy “Or) ouTUOI0Z ‘ade]]1A redng "SLOG ‘zuy “oR ofeaen ‘IqreiQ Anpeooy PUM PUM ZaeONSsUIOG [P4197 PUM piim pareanpnd R Neonsawog aeonsaWOG areonsawo0g ae s3wWog a1eonsaui0g aeonsawi0g PEM smieig pasodorg He “‘DIADQLIY Ul MDJI S,jtaap paziytyn Appoanjyjna fO suotj22]102 4NQN— ‘Gh TIAVL. a ae ee Oe iia Vol. 1, No. 1 JOURNAL OF ETHNOBIOLOGY 144 - 2 me —— e200 tog jo waved Spass arg a poxrunuy ‘Te wa Gt-| BUeTd “PPey P y } aaqyeuss * ‘spaas cas 11 wt sued play poreSust- UOISISAIG (v)yojoFy ssorwe wi Z x [eI x []B) Wg" | Ue, “ANayseq Ul pas~ ‘aameWwUT imag ‘ayed Aypensnun siaMoyy ‘play pamord uy 112) WO ¢p-G] “‘Spaas yoeTQ ‘siamoTy WIRaID 0) a[dind-pay *pyay Morey ut JaaunjoA “poss YIEIG “SIIMOT) aed ‘PPy azrew ur yied Apaam ‘yory) ‘1aajuNToA, ‘s19MO]] aed “WINIy WUT *ssOIDe UI ¢"| ‘{[21 WD Og ‘p9aM J99}UNIOA ‘platy Aq apispeoy “J9qty JO} pasaarepy “SME]D 110YS ‘spaas YDE[G “UI OG UT SjURTd ZIE ‘pjaljueeq pue azrew url siaajunjoA juareddy ‘Bny-] “|e. WG uel, ‘WN IBLE] ‘s1aMOT] a[eg “uapsed preA100g “Anayseq Ul pas() ‘21y1]o1d ‘stamoqy ayed ‘saaeay adrey [yer wt f° slUeTg ‘paiean[ns Aypeuonuajur ‘uapsed predsrc0g “1may plo wo snouuadsoideuds ‘s8utppass aanjewwy ‘spue] pjIM Surpunouns ut uaas dor ‘{pleyusod pawuuey Arp ut "spans 9puq ‘SME[D MOYS “Bsa MOTaq ‘sIIqININd Jo play pamold Apues ut saajunyoa [eisasde yuasreddy ‘Anjayseq Ul pasy—) “sIaMOTJUNS ome, ane yun PIey ul ‘spaam oulos * ‘oo) uon -ejndod ut spaas ary yodas sopedy “yoriq spaas ‘Buoy A[awianxa smey> ‘ayed sejforo= ‘yooy J ‘Jaq Anayseq 10] paed ut UMOI “paas airy ‘aderpoy Surppenba -jenba o1 passedins -qyut ‘wieaid ayed ‘e]fo107) Sonsua ey aansadaing a ZIav ZIV ZIav Zrav 42 BL/61/8 8L/61/8 8L/66/8 8L/11/6 8L/9I/L 8L/9I/L SL/LI/L 8L/b/8 8L/91/L BL/91/L ‘8L/21/8 8L/61/8 8L/S1/8 LL/6/8 L£L/G/8 LL/G/8 LL/G/8 ned a 068 URYQeN 688 UeYqeN 668 URYQeEN GOL UeYqeN 9001 ueYqeN gool ueyqen stor ueyqen qgog ueyqeN LOO ueYyqeN LOOT ‘IOI URYqeN os0l UPYGEN Ggol ueYqeN 699 UeYqeN qg9¢ Wor $99 ULYqeN eL19 UeYqeN % My wor Z99 UEYqeN # % 10199110 6 Tee redinseaeyy rednseaeyy tednseaezy osedeg ayoedy aypedy aysedy odedeg aypedy aypedy doy aypedy andaqiy odedeg codedeg osedeg anmyjn> mt ect ae O00'S “fury “OD OUUOOO) ‘ORR A tedng Vy 09'S “ALLY “or) OuTUODI0Z) ‘adeT[T A tedng ‘Y) 009" “zy “Ory OUIUODOT) ‘adETITA redng Y 00G'% ‘ZUy “oD eutg CIEL WNMIYD) adeypiA peusary ‘Vy 008s ‘ZY “OD E[Id ‘sojieD ues ‘Vy 008'S ‘ZY “OD B[ID ‘sopie ueg ‘Vy 008'¢ ‘Ziy “Or OfeaeN “aALIITyY MA ‘Wy 00F'S ‘Tuy “Or Bulg ‘UOsIN] apnry ‘IY 008'S “ALY “ODD EIT ‘Soper ues "Y 009'9 "zy “OD ofeaey ‘andaqrr) Jo sea “tun ¢ ‘Ys 008’ ‘ZuUy “og ofeaen ‘IqieIQ PIO Molag ‘Y OF6'F ‘zuy “O- ofeaen ‘anzaqity 005‘ 1-006 ‘may “or [eurd ‘nypiny:) uoson Ul UMO1S paag “1) OOF] “Zuy “or [eulg ‘Bare 1oIeMyDRTG uoson | Ul UMOIB paag “zy “OD jeurg ‘edoouey eau ulyp-yY¥ Anesoy ‘< Parone ee AVWOUMS SUI? PITM areonsawoeg PIEM PITM PUM PITM AEITSAWOd areonsawiog éPIIM PIIM aleonsauiop dustdrour papess 49e(q a1eNSaWIOG, 2PUM aeonsawog snierg pasodoig ‘panunuods “gp T1AV_T NABHAN ET AL. 145 May 1981 ‘Bny-] “raqy Anayseq se yee MN ‘s192uMO ‘spaas prig ‘ss0.De w | [yer we ¢ swuetg Jaq paam [eispny “yom2H “[fe) wo Og siueTg “WD OZ ME[D ‘Spaas YRIg “platy 1eau aptsproy duoj WD €Z MELD ‘paas aI M4 “SsO.De WZ ‘T[B) WG SUe[d ‘P[ey Ovewo) ul ‘papuai pue paeaniny “Suo] WO g] MEI ‘paas yoRTG “[[e1 WZ oO dn siuetd ‘paean[ns Ayfeuonuajut Apussreddy “Zuo| WO 9¢ ME[D ‘poas sity MA “SsOIDe WI ZZ ‘[]e] WE | SIUe[g ‘pa ~JINBIS ‘Pay /MOpeow ur sued ButaaUN[OA “WD §'6 Apog ‘WD g'9z MeID ‘poss YM “[[e) WW Gg sluel_ ‘platy paresiur ‘paeanno Ayyenued ur sued Buraaiunfo, ‘WD [| Apoqg ‘wo g¢ MelD “paas airy M4 “ssoe WI /*| [ei wp syURTg “OTd Aredai ut pasanesg ‘paiueyd Ajpeuonuaiut ion “‘Djvoun J “sjayseq ainreg Suryew uewom tdozy. sd Inet J id ‘Paty UOJ2aW puke Ukaq UOTSIBAIP IOATY ‘redng woy dn 1y8no1q inj Wo. JaajuNTOA ATqeqoig ‘redeyeM ZIaV ZIav ZIAV ZIaV 6L/1/8 8L/S6/L BL/b3/L BL/b6/L 6L/98/8 6L/96/8 6L/G6/8 6L/F6/8 6L/82/8 8L/S1/8 8L/18/8 8L/06/8 ned 6z01 UeYGeN S201 UeyqeN 81ol weyqen odedeg aypedy-redearx Put [ZOl ueyqeN (gaypedy x) redeaey SIT UeUqeN PITT ueyqen SIT ueyqen 9011 UeYqeN coll ueyqeN p88 UeYqeN ainted "Og SIIMAIYS ainied “0 SIIMATYS auoUu 10 ganteg ainteg ‘og qeqrey idoy doy $06 UBYqeN redeyeM ‘tednseaeyy 006 UeYQeN # % 40199110) rednsearyy amyny y 00¢°% ‘BO TLY “OD PUllg ‘T[2M $,Usane ) 00%'§ BE “ZUy “Or ledeaeX ‘apio, a[pplw VY 0071 (Ro "zy “oD [eulg ‘BOUR]g BSED JO Isom "TW Z Vy OOF T zu “on edoouep ‘|]amoqg IW Moy ‘VY 000'F “BS YRI) “Or UOIZUTIYSE A 'SIIMATYS ‘VY 000" “B9 YRI “or: uo duryse MA ‘SITMATYS 1) 000'F & i lt Ue Or) UOIBUTYSEM “py durary O17 simmarys ‘YY 009'F “e ‘ZUy “OD aaeyory ‘qeqiey ‘VY LLL'b “ty “og ofeaey ‘tdoyus0py 12M07] ‘Y LLL'b “tury “oD outuos07) ‘tdoyusep 19aM07] "ty 008" “ord aaeyoyr ‘sdurds yorag Y 009°¢ “tury “oD ouruos0r) ‘adeyp1A redng AN[eIOT aeonsawog PITIM PUM PUM aeONsaWOg uonean[n:y sapun PIM éPI!™ 10 aiwonsaulo0g [e194 aeonsawog [e4a4 gaeInsawiog [ea4 gemnsawiog [e194 ZaIBINSIUIOG] [P1494 éPIIM smimg pasodosg ‘panunuod ‘qp TV _T. 146 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 Castetter and Bell’s conclusions were based primarily upon several interviews between 1938-40; we have not come across any voucher Nanni collected by them, or notes on the lants themselves. In the interviews, only the white seeded variety is noted as being cultivated, and only the black seeded variety is eaces as growing away from fields. For one interview, notes imply that a Pima farmer responded negatively to the question of whether his people cultivated wild plants, but later acknowledged that devil’s claw is cultivated (Castetter 1939:44). Dobyns (1952:211) concluded cia the Papago had domesticated devil’s claw based on similar (unpublished) observatio Yarnell (1977) has concluded that the distinctive characteristics vi ee cultivated form described by Castetter and Be!l—white vs. black seeds, longer ‘‘pods,”’ plus finer grained and deeper black pod-epidermis—justify its status as a cultigen. He suggests that several centuries of artificial saieiaeaseoty * a eer makarwisee’ for the guraenne of the oS process in devil’s claw cultivation was possibly the isodceatie of the pods and seeds, = that more recently, cultivation has emphasized basketry material production. He does not mention the presence of the domesticate in culture groups other than the Pink Pataes. e feel that with the limited data which Yarnell had available, it would be difficult to refute 2 arguments against his conclusion: 1) How do we know that the white seeded variety is not another “‘wild”’ species of devil’s claw ae into the wray whores is cultivated — the other indigenous not? 2) How sae we know th I not result in the longer, finer claws? Could the be d I vesting of cultivated fruits, which would keep the seed from ripening toa a hiatk color? Thus, is it the treatment of the plants rather than distinct genetic material due to domestication which account for the apparent differences: TABLE 5.—A comparison of relative association with man-made habitats of two races of Proboscidea parviflora. EET, HABITAT WILD RACE DOMESTICATES BLACK SEEDS WHITE SEEDS A. Undisturbed or protected range N=23 N=18 B. Minimally-managed & grazed range or deserts + C. Overgrazed & manipulated range + D. Floodplains 13% © E. Riverbeds & arroyo channels 17.3% 5.5% F. Managed meadows & corrals 4.3% G. Roadsides, paths & cleared areas 30.1% 11% H. Abandoned fields 4.3% 5.5% I. Dumps, houseyards (uncultivated) & plant-processing areas 4.3% 11% J. Cultivated temporal (runoff) fields 13% 5.5% K. Cultivated irrigated fields + 5.5% L. Cultivated (kitchen) gardens 13% 55% sii eon istic Saal es Gee ares cages se ge weediness in related plant taxa (Hart 1976). “+ indicates lack of . ee May 1981 NABHAN ET AL. 147 Additionally, we have discovered that Castetter and Bell’s “clean” correlation of white with agricultural fields, ms black seeds with “‘wild” environments does not hold true in Pima-Papago country t able 5). We have located several fields and gardens where Indians are propagating black seeds, and have also found white seeded fruit on plants growing away from fields, although always within Papago rancherias. We therefore doubt that Yarnell’s inductive reasoning that Proboscidea parviflora must be domesticated has really settled the matter. His contribution is, on the other hand, that he has rought the suggestion of domestication of a native arid land plant to the attention of a wider audience. We would like to answer his challenge, by providing a methodology for evaluating whether devil s claw, or any other plant, has been domesticated. Domestication: Definitions and Testable Principles In using the term cultivated plant, people ones eying = Langs of cultivation (i.e., planting, and tending p e plant itself. By the plant’s status, we mean whether or not its genotype i is different from ie RENOHPS of plants growing in the wild. A propagated plant may have the wild, even though the conditions ina garden environment may influence its phenotype so that it appears different. When the genotype is different due to direct human influences, the plant is often termed a cultivated (or wid domesticated amare In order for us to consider the status of devil: sc our use of the term domesticate. Indeed, t what a domesticated plant is (Table 6). Utilizing differs definitions, one might actually come to conflicting conclusions regarding which of the world’s plants are domesticated. Asa foundation for our study, we will use the explanation offered by Harlan (1975:63-64): “In the case of domesticated plants and animals, we mean that they have been altered genetically from their with man. Since domestication is an +3 sane BROCE, there will be found all ners of plant and animal association with man anda a range forms identical to wild races to fully ae aa races. A fully domesticated plant or animal is completely dependent upon man for survival. Therefore, domestication implies a change in ecological adaptation, and this is usually associated with morphological Saeed.” To wae yoshi explanation, it must be added that the intended human influences such as Selectio joined by “accidental” or indirect pen aaa (Baker 1972 32). The most significant indirect human influence is the m ication of iia Particularly the maintenance of agricultural environments, hase a then underg natural”’ selecti ion. : f. kh t i pte ee enhanc Of oo The da ust be viewed within the context of plant’s natural history and use, nown instances of aclegtion. and other factors. Otherwise, we may be contrasting a weedy race of a species witha ce ina manner in which the weedy race appears to a domesticate. 148 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. | TABLE 6.— Domestication: Alternative definitions and explanations. Quotation Citation “domestication ... complete and regular reproduction of the species Meggitt 1965:23 through more or less controlled and selective breeding in the company of man. .. the crucial feature of domestication is man’s control over the breeding Watson and Watson of his domesticates. He improves his crops by sowing only selected seeds... 1971:5 Domestication implies that the plants or animals have been manipulated to Bender 1975:1.52 such an extent that genetic changes have occurred resulting in new races or species ... Cultivation, with the attendant element of human selectivity, conscious or unconscious, frequently results in genetic changes. Even so, there will be an intermediate stage where plants are sown and h arvested but morphological hence and ‘domesticated’ plants where morphological change has occu .. domesticates show both intended and accidental results from human Baker 1972:32 actions, including selection. The pede of domestication are as follows: Zeuner 1963:63 loose contacts, with free breedin ne ane to human environment, with breeding in captivity. selective breeding organized by man, to obtain certain characteristics, and occasional crossing with wild forms. d) economic considerations of man leading to the planned ‘development’ of breeds with certain desirable properties e) wild ancestors persecuted or piieiased: ee The cultivated plant never originates directly from the wild species, i Schwanitz 1966:63 perfect form, but evolves step by step over a long period of time. The iti it has come along, that is, the earlier it was taken under cultivation or the more intensely bred and selected, the fewer wild characters will be found in it ... Their occurrence [wild-plant characters] in cultivated plants must thus be taken as a sign that a plant has not yet completed its evolution from a wild species to a cultivated plant. The most immediately apparent change under Ens is i Pickersgill and Heiser morphological characters such as size, shape and color, vareeakats of the 1976:55 part of the plant used by man ... Up until now, crop plants have not evolved selection as well as, or sometimes instead of, natural selection. pono we have gleaned from the literature a number of morphological and ecological characteristics which commonly change through the process of domesticating 4 plant (Table. 7). Hypothesizing that these changes would occur in any Proboscidea if domesticated, we can use these indicators to examine the “‘real life situation.” Indivi ual characteristics which may be found in any useful plant, wild or cultivated, or in weeds, will be interpreted in light of these other possibilities We have made 2 major assumptions in applying these indicators to the problem of possible Proboscidea domestication: We have ass umed that if ar s claw has been domesticated, the ucts which have been V¥ 2220-24 most pervasively and intensively ities che bert in mere aie ile aa the seed. Thus we 149 NABHAN ET AL. May 1981 19-09:9261 4astapy pur []1Bs19x21g 09:9261 JastaH pue [PBs1ayI1g ‘ggT:G/61 UELEH 09:9261 42St9F{ pur [[ sxx Id “FE:ZL61 12AeT ‘ZE:9961 TIUEMYDS PEZLEL AE “67:9961 NIUEMYIS bE-bh:9961 TTUEMYDS LEUGL6I UPLEH “bp:9961 DIUeMY IS $F:9961 AIUBMYS ‘ZETGL6L URLIEH ‘bS:ZL61 1eT Sb-8S:9961 AUEMIS “E¢:2261 APT TT ‘L2UG26l UepeH SEI “SLE URLeH LL6T M9UIBA “BETGL6] URE SESL61 JAP “LELGLE1 URE 82 ‘b1:996] ZIUeMYS 1d ‘FE9961 AIUeMY IS LELGL6I URLEH ‘ZE:2261 124PA ‘06:9961 ZIURMYDS UI passndsIp S839 -o1d uoneonsauop U} puaNn [e12U2") x X x x pe xX x x x x x x x x x x x x x x x x x x xX x x x x x ¥ x x x xX x xX x x é JU2UTUOIIAU pareA Is2azeYy YIM pare “hyn ur saanssaid = -Dosse sainssaid = u0N99]98 UFUMY —-13q1j 10] p228 10] 2anD9[98 0) anp PAN2a[ 98 OF aesaqiap Ol anp = =—s paaeonsawop jr aon Popess airym uy Aiqissod aduey = Ajqussod aduey = Ayqissod aBuey = a3 ueya pynoyg = aouasayy1p puaseddy Aarprqeuea snauad ‘UOneINIeW WUOFTUN BOW (swistueysau Buikejap-uoneuruiad jo 40) AQUBULIOp [ENUaIaITP JO sso" sued 2]qrpa jo Airxo) paonpay v a tea ie al aespAyoqie) /utaioid ur aduey wnpoid jo anixa) ut adueysy inpord jo s0709 ut a8ueyy 1gey YIMOIs aeUTULIAap ao; sued azis jeay Ul aseamuy aonpord jueyd pastsap jo luawasie[ua ayeuonsodoidsiq, aunpeag “MD)I $,]10ap OF a2uasafad Ut ‘UOIIDINSaWUOp JuMd Ut Spuas] JOLIUI—'*L TTA. 150 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 hypothesize that devil’s claw was domesticated either for basketry material, for a food product, or for both, and not for other reasons: its value as an ornamental or religious item, the use of the young fruit as a vegetable, etc. Secondly, we have decided to compare the white seeded, supposedly longer clawed devil's claw cultivated by Southwest Indians with the 3 most common annual Proboscidea in the Southwest. In particular, most of our quantitative comparisons are with wild Probosicidea parviflora, as it occurs in Arizona spontaneously, and when brought into cultivation. In doing so, we have ruled out that the annual white seeded devil’s claw 1) belongs in another genus; 2) is more closely related to Proboscidea perennials in either subgenus; 3) is more closely related to other annual Proboscidea in the Southwest, or elsewhere Our emphasis on comparison with Arizona populations of Proboscidea parvi flora i is in part due to logistics, since that material is more readily available to us. However, Table | makes evident that the white seeded devil’s claw is more phenetically similar to wild P. parviflora than to P. fragrans or P. louisianica, as we understand these taxa today. F urther- more, Yarnell’s suggestion that the white seeded devil’s claw is a Proboscidea parviflora cultigen warrants our most critical attention. We will nevertheless note similarities to P; louisianica and P. fragrans whenever possible, and allow as an alternative hypothesis the development of the white seeded devil’s claw from interspecific hybridization oF introgression (Fig. 3). Finally, following Harlan and DeWet (1971:509-517), we will avoid using the terms variety, cultivar, line, strain, type, or kind for the rest of the discussion, due to t their indiscriminate use in the past. Temporarily, we esha refer to the suggested done eeey as the ‘‘white seeded Face: of Proboscidea, gthatitisa ated weedy © y particular species, Also, for the purposes of suevini, we will refer to all black seeded P. parviflora as the P. p even though there may conceivably be domesticated or weedy black seeded races which we are ignorantly lumping into this one category. We will also refer to the spontaneous race as wild or typical black seeded P. parviflora, depending upon the context. Skewed Distributional Range Th ] niche which the whi led should be regarded in light of the distribution of wild Proboscidea in general. Yet it is is somewhat difficult to determine the “natural” distributional range of annual Proboscidea spp- in the Southwest. Whereas there are ‘core geographical areas’ where each species is commonly found (Table 1), the intrinsic dispersibility of their fruit has allowed them to be transported by animals (including man) to many isolated — far away from these cores. hi Jd al Large native her bivor es Vi I n the aiid distance dispersal oF — s claw to disjunct localiti bef. a we process. Nance historians have described the shape of devil’s claw fruit as one all adapted t g in the fetlocks of ungulates. They have hypothesized thal this mechanism was responsible for the dispersal of P. louisianica to South Africa, and toa locality in Great Britain (Bancroft 1932:62-64). The habitats which annual devil’s claw frequent are often corridors which allow further geographical extension of their range by animal, water or wind transport. The habitat preferences e 3 Species of annual Proboscides indicate adaptation to po “7 adsides. Historic human modification of Southwestern floodplain environments, particulary pene: agriculture and road-building may have d from prehisto times. Additionally, such modification maintains niches with senna ‘il wet deliberately transported plants such as P. | nestablish themselves after escaping from cultivation (Robbins 1940: 0 Although the distribution of the P. fragrans, P. louisianica and P. parviflora remain ee NABHAN ET AL. May 1981 DEPIDsOQosq snueS DepI9sOgoig snueb OdiUDISINO| DdIUDISINO| g O10} j1As0d | pragky pepees 8uUM pepiosogoig snub yODIGB | SHUM (GL61) MeEeH sMoT[oOy iat seddn ut puasay ‘areonsawop ayp o1 sdrysuoneyas spood aguas 10] sas DOIUDISINO| 4 40 DOIUDISING| q 40 DsOjpiAsod Y DsO|JIAOd gy $@091,SeWwop aypoddy aaneurify "§ “OM DePIdsogos4 snueh ODI | @LUM DdIUdISINE| q 40 DsO}JiAIOd g HOODIE | SLUM senbiuyoe, }091po4 Buiinbes 40 @/Q1SS0d OU Jejysu0s) eUeD sod 48jSu0.4 1-d9 $@9Dy SNosUDjUOdS 1405 '4 BUIOS ysne| 4D 1 |-d© WIM pessoso eq 9008 Hy ; 9114048 40 ‘}Oyse) ‘sNO;OWOUD hd 4M spugkH 152 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 problematic, th hit ded heless peculiarly skewed in relation to them: 1) it is highly —. to the rancherias of native peoples of the Southwest's true nearby uplands. 2) It appears to extend northwest beyond where annual Proboscidea is commonly found in the wild in northern Arizona, southeastern California, ern Nevada. 3) Its range overlaps to the greatest extent with the range of P. parviflora. Today, the black seeded annual Proboscidea are strongly associated with man-disturbed environments, yet the degree of association is even higher for the white seeded race. A survey of annual Proboscidea specimens collected in one particular area—the aboriginal er of Noxthern — ope in Arizona: ang Sumnre--ru eerie this point (Table 5). A mmnare th location of the white seeded race and oo P. parviflora, regardless of whether or not they were cultivated in those locatio Although the presence of P. reccarae in fields and on pathways around human settlements indicates weediness and dependence on human aspen it ranges beyond these habitats to a greater extent that the yards, the white seeded race has only been collected within ‘disturbance habitats in rancherias. Although it is not possible to prove that all these plants are recent “escapes” from cultivation, subjective information suggests that the plants are feral cultivates. Papago informants have volunteered that white seeded plants growing in their yards “planted themselves” from seed that blew over from nearby devil’s claw processing areas (see Table 8, for processing site explanations). In both cases, large stands of the white seeded race were cultivated hen ae vancheria. If the whi i 1 from P. parviflora, i it is powbe that the status f black seeded n this process. Since black seeded P. parviflora i is considered a tolerated weed in fields Boda the range of devil's claw cultivation today—among the Hopi and Apache in Arizona (Whiting 1939:92 and Anonymous 1976), and among mestizos in eastern Sonora—it is doubtful that in this case 4 weed race evolved as a result of introgression betwen domesticated and spontaneous races (see Harlan 1965:173- 176). Finally, th the whi Jb . =. . : lel : by man, is not an obli Iti lt cet ae to Harlan’s definition, itcan not bea fully domesticated an in the x strict sense, since it can survive to some extent without direct Seed Characteristics it domaesticated for the food value of i its aced, , devil’ s claw : should have undergone poansiel aiUat, /) Oo = 2 3 d) es * a ? 1 ws scctlidite e) change in seed eal pattern; and 1 loss of germination-delaying mechanisms. Several of these s claw was domesticated for the fiber in its fruit, agian Ge and £. compared the number of seed per fruit in the white seeded race (n=69, from 4 ak ae with the number in black seeded P. parviflora (n=50, from 3 populations) there is no statistical difference at the .05 level for the 53.0 + 9.8 seed per fruit of the white seeded race, and the 53.9 + 8.3 seed per fruit of the black seeded race. These sample sizes are relatively small, and the populations analyzed do not allow ‘aniation however, it is apparent that there are no major diferences in the sed number of the 2 kinds of frit Wee a , Se | be Aolit L 1 . ace ae ae . pease with cultivation and harvest. 153 NABHAN ET AL. Spue udisap asym Japun yont sureusad 40 yo drus ‘por punoze dem wards poddrag ay popacu sy ayruy 40 yprd ‘pay ‘wuyyds snotaaid sapun sso) Aisap layseq Out srurtds YOM OF, Joo yor CAE YIOM yoyseq ul ayyuy [oor yo0u (8161 491e) URS Uy ‘sBurdens yuryds epewel 10 asnopy asn on soud isnt 3B ‘yP01 JO UD UT » JO/PUR ayy YIM adepsg aqoy y8noryp wuyds [Ng asn 10) juts ade oy epeures 10 asnopy pepacu sy 1a1eM % [MOG — [40g ut yROS asn arerpauut 10; srutjds Apeos oy SIBOIPIIg wes aumAuy yea WODpses YUL) (yoeus) spaas 1a OL, mnuy papaasap ‘parurtds asnoy 1eaN aumnAuy saaduy “Png uado Aud 8 ypns yn a inuy paruryds wosy spars ysanp 7, aureg Surdduns ary auIM cy ipunq 1 syurds purg asn auniny 40y syutds aro Of, uortsuen Apog /Me]? 01 dn wo ‘naa yim yo swuryds BQY S [99d “19Q1) 2pun [me siurds wio Waytadwiy Ss Bare YOM EpeWeY Suryeos injy Yieay ‘aytuy 40 [My un 2 “ae yim dn MB IS nay Wo4y siaqyy Jo siutjds duns oy shep z 10 1431u i1eM \puasp 93 pues ut Aing ‘sme[o [Mog UT YeOS ‘sMe]D Suryqew-juryds asnoy Ivan ~4Jaa0 ‘papoou sy ayong 10 [MOY JO Jaquinu [jews FOoyU/) jo !quinu posisep yooyuN 40J WN pap Apear % Usjos OF, epewel Japuy) Jaye RB [eA 49A09 JO JUIM | < dooy 19009 10 Burpy asn aye] 10} aasasaid OT, ft dooy out imay uayxo1q Muy palup yooy ‘stuuaprd, asn aaniny 3B ‘stuuiapida paysnzy aules Butfup 1yy duBuUal JO 1INay YSN JO} 1inay ayqesn daay y uaYysrens OF, asnoy Aq 49am B 10} sigUTe}UO pue (o8edeg) Arp ray ‘yse ETM ay JO apispjaty “Trey Aqrens-) daem 10 ‘saysy (eurtg) [Mog Aranod ur yeog =—-19A09 ‘UNS UT TINA Ua—aId aft query ainay daay OL (paunroyap K t JO []BUIS 001) dieys awooaq sdn Sutiayieam 01 10d indy ayqesnuy, — py IM UT JO pray Uy 1SO1] 1811} JIOJIG MPP UaYyM IN Uaedd IsanIe}y Rueryd [pos uaym inay asinboe of rr ‘gjoy punoae aos uapres qsnuq waroud % urefdpooyj uo UI | UT spaam seal? ‘WO g sajoy sp22M 10 urejdpooly Jae] 10 Ae Ypns a[qqIp 40 a0fy =~ Aqueau uonejndod prim pur ur paas jueyd 10 isespeorgy = Burssav01d 104 1NUy jo Apddns aimsut of, S218 uonwoy Aouanbesg aur Sapep2uI0Z) [ELIE sanranoy asoding ‘my? sjtaap fo asn O3vdvg pup vutgy 40f utvy? poL01avYyag—'R TTAVL il ih a a lm SF ee See eee ieee ron oer 154 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 In terms of seed size, the 3 annua 1 species of Proboscidea with which we are conicerned, all fall within the general range of 7-11 mm long x 4-6 mm wide x 2-4 mm thick. Size of a . 1 oe c. Particular seea p y, as well as by maturity of factors. Size variation within a fruit is considerable. We measured seed sizes of all seed in only 2 average-sized fruit of the white seed race, and 2 averaged sized fruit of black seeded P. parviflora, grown in the same irrigated field. Mean sizes an ges at tandard deviation are given in Table 1. These data suggesta slightly greater volume of the white seeds, but without a substantial sample, we will refrain from further speculation. A severalfold difference in seed volume, such as that between domesticated beans and their wild progenitors, is nevertheless not evident with these devil’s claw. In terms of nutritive value, Proboscidea seed are non-toxic with high oil and protein content. Because of interest in the 1950s in developing devil’s claw into a commercial oil seed, numerous chemurgic analyses of Southwestern Proboscidea were undertaken. After compiling protein and oil values in the literature (Earle and Jones 1962: 245; Ghosh and Beal 1979:748), hat th 1of South ] ll bet 35-43% oil, and 20-35% protein. Two acquisitions of the whit led | f the black seeded P. parviflora, grown in the same irrigated field in 1976, have been analyzed by nutritional biochemist Dr. James Berry. The white seeds cultivated by the Pima contained 40.3% oil and 25.5% protein, values remarkably high for Proboscidea. The white seeded race cultivated by the Havasupai, and the black seeded race originally growing wild in their area yielded 39.2% and 38.3% oil, plus 23.9% and 23.2% protein respectively (Barry et al. in press). Thus the white seeds apparently have a slightly higher nutritional content that black seeded P. parviflora, or at least they are at the high end of the range for Proboscidea. It is possible that selective pressures in the cultivated environment, or deliberate human selection for the fruit or seed have resulted in relatively more energy being funneiled into these reproductive parts of the plant. ; We mentioned earlier that the white-gray seed coat of the commonly cultivated race 1s atypical for the genus Proboscidea. In analogy, Yarnell (1977) has pointed out that lighter colored seed distinguishes domesticated Amaranthus from its wild progenitors. In devil's claw, it can either be hypothesized that 1) natives found this character in the wild, and brought it into cultivation; 2) it was expressed after selective pressures associated with harvesting were initiated, or 3) it is a functi ftheg f y of variants, including recessives, which survive in cultivated environments. : It is probable that the lighter color is determined by one or a few major genes, i.€., itisa quantitative character. A crossing program to determine the inheritance of characters such as this is now in progress (Peter Bretting, personal communication). Seed coat morphology study by electron microscope has not yet identified any differences between races. In terms of seed dispersal, Sappenfield (1954:1) has calculated that approximately 10% of wild Proboscidea seed “‘shatter,” or drop as the fruit dry and the claws split and curl. From our simple observations, we estimate that roughly 4-12% of the fruit’s total seed are released as the white seeded fruit begins to dehisce. In spite of these crude estimates, we doubt that there are major differences in the seed dispersal of the various races and species. Certainly, there is not a dramatic difference in fruit dehiscence as there is between wild and domesticated legumes (Harlan 1975:138-139). ; Germination delaying mechanisms in wild Proboscidea include 1) germination inhibitors of the seed and 2) the leathery-textured ovary walls behind which the seed are trapped unless the fruit is physically torn apart. Through differential dormacy wild Proboscidea spp. avoid “putting all their eggs in one basket;” the proverbial basket here being the unpredictable moisture conditions of the Southwest. Anderson (1968:171) has determined that the germination inhibitors in wild Proboscidea include a) seed coat thickness; b) a water soluble chemical inhibitor in the seed coat; and c)a dark requirement, or light sensitivity factor in the embryo. Because of these inhibitors, Ps May 1981 NABHAN ET AL. 155 agronomists have had difficulties getting good field germination with wild Proboscidea brought into cultivation (Quinones, personal communication) Our attempts at utilizing a standard laboratory test to determine possible differences in rate and per cent of germination were somewhat unsatisfactory. At 85° and then at 90°F, we obtained 40% germination in one sample of the white seeded race, but there was no germina- tion of one other sample of white seeds, and 2 samples of black seeded P. parviflora (n=25, at each temperature). Our field plot observations indicat diff rigated conditions. In 1977, one month after an April 21 planting, 65% of the white seed had emerged (n=55, from 9 acquisitions) and 16% of the black seed of 2 P. parviflora had emerged (N#25, wi 4 acquisitions ). uspect that the white-seeded race may have lost at least one of its germination aihiinns possibly due to long term selective pressures associated with planting seed, and utilizing seed from plants in surviving cultivated populations. Further paired tests are needed to determine a) if field emer gent differences are significant for larger | sample sizes and b) if an inhibitor which the bl be isolated. We doubt whether other germination delaying mechanisms, such as the persistance of seed behind the placentae walls, are different for the white seeded race. Floral saolig aieigaeete and paps Flower size, sh d ] 1indirectly through human selection for the economic products of a cae If a plant, through domestication, comes to produce fruit much larger than = of its sbi progenitors, the or size May = increased too in order Or ofte once, so that a flower color may increase in aisle ina population, due to its genic association with a selected character. On the other hand, overall floral design is fairly conservative, and within a species is little affected by short term selective pressures. In addition, floral ecology is certainly affected by cultivation and domestication. For instance, in South America, where wild = domesticated tomatoes originate, they are predominately cross-pollinated by insects; pollination agents, they have evolved into a self pollinating plant (Rick 1976). Such ecological factors may eventually work as selective pressures influencing floral characters. A species variable for Hower i, dependent upon cross pollinatio may swamp bee cultivated in large stands. Particularly bright flowers might have a selective advantage over less intense flowers by attracting a greater percentage of the available bees. Depending on the inheritance of flower color, this may influence the frequency of allelles affecting os opin —_ In terms of flower size, our data indicate tha corolla, calyx and bracts of black seeded P. pate iflora, these characters are sull uihcuay longer in the white seeded race. In fact, the white seeded race overlaps in these characters as much or more with P. fragrans and P. louisianica as with P. parviflora. Several hypotheses can be proposed to explain this situation: 1) In the white seeded race, floral part sizes reflect a closer affinity with P. louisianica or P. fragrans. 2) A larger flower Size has developed i in the white seeded race while being domesticated from P. parviflora; the larger size accommodates the anger fruit. 3) It reflects InLTORTESSION | ee 2 of the species. Flower shape in the whi y che lo Among the largest flowers of ites white seeded race, there i tency slightl ventricose, though not as much as typical P. louisianica and P. Padiens: me is pence that a wild, long clawed (32 cm), black seeded specimen collected on the Gila River Indian Reservation at the Pima village of Sacaton had a similar ventricose flower shape (Peebles, com d an Harrison #75, ARIZ). In addition, its f] Kearney and Peebles (1960:795) Micwesed i its s affinity with P. fragrans even though that Species is nowhere else in Arizona. Again, does this reflect hybridization between different feh 156 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 races or species, or simple introduction? Abberrant flower aap ED including ones with an extra lobe and a wider tube, have been found in low freq n plants within cultivated plots of the white seeded race. Flower color determination in wild annual Proboscidea is not well understood. Perry (1942:43-47) reported that reciprocal crosses between P. fragrans and P. louisianica, and subsequent backcrosses, indicate that purple flower color ibis white flower color. Perry suggested that color inheritance was due to a single gene. However, reciprocal crosses between 4 annual Palins by Hevly (unpubl. notes) do not substantiate that purple flower color is dominate over white, since F! plants were intermediate. F? plants tended to have darker flower colors, but the F? population size was not large enough to suggest genotypic frequencie Most flowers of the white seeded race have similar color patterning and internal orna- mentation as wild P. itil ei however, ay colors ae usually gs intense. Often, cant color is pale cream or white, b ed plants on she Papago pemeanon: However, these darker “flowers were in a population P. parviflora is cultivated. Does the variability in flower color in this white seeded population reflect the introgression of typical P. parviflora in the white seeded race? We should note that white or pale cream flower color is not specific to the white seeded race; it also occurs in P. louisianica, and infrequently in wild P. parviflora. It has been sug- gested that different fruit types—of distinct lengths and shapes—are associated with different flower color types in P. parviflora (Paur 1952:1), but we have noticed no such clear cut relationships. Finally, it is noteworthy that in other floral characters (e.g. corolla orna- mentation, uaa) aia cis and inflorescence position) the white seeded race is most similar to P. parv The pallinsden ia of devil’s claw has received an i recent years, but the picture is far from complete. Hurd and Tinsley: (1963:249- 9.280) reported the apparent cross-pollination of perrenial Proboscidea altheifolia by the corolla-cutting bee Perdita hurdi. However, their repeated examinations of wild P. parviflora flowers failed to show bee visitation for pollen, ora Eelavonsiyp with this bee. Dr. P.H. Timberlake (personal me aware of one example of Perdita hurdi visitation to annual Proboscidea in Mexico. Toc our knowledge, there are yet no reports 0 this bee pollinating wild P. parviflora in the United States. : Thieret (1976:175-176) reports the insect visitors, including pollinators, to P. loutstanica flowers on wild plants in Oklahoma and in his garden in Utah (see Table 1). Preliminary experiments with pollinator exlusion, plus self and cross-pollination, suggest that f. louisianica fruits do not develop if pollinators are excluded of if artificially selfed (Thieret 1976:177). However, other investigators report that hand pollination of P. louisianica yields about 50% fruit set regardless of whether plants are self- or cross-pollinated (Moegenson, personal communication; Phillippii, personal communication). Self-pollination, though still probably not the key pattern an wild populanons. may also ia ton in be effective in black seeded P. percitior: ts In an exy w Mexico, of P. erediors ine other species?) he 500 infl bagged for self eolinetion produced some seed (Anonymous 1953: af). Dr. Floyd Wee has identified for us a few of the fairly f I iltivated plots of the white seeded race (Table 1), but we do not have concrete confirmation of actual pollination by any of these hymenopterids. Most noteworthy is our discovery of Perdita hurd: in the flowers of a large, saat planted houseyard plot of the white seeded race at the Papago village of Santa Exclusion detailed field ob i on both Indian-cultivated pee & the ded d black ded P. pernflora, and in sp y occurring of P. parviflora, are needed tod : Iv sn th lations i eine «at = te AA ee Se hae 6. Ato in the tit selective = May 1981 NABHAN ET AL. 157 white seeded race? 2) Is the frequency sa visitations by various bee species different in cultivated plots as appa edt y occu ring | pop ? 3) If P. hurd? is in fact pollinating h bite ded b Il stands of P. parviflorai in the wild, is this due to greater fetiaality or abundance of reward for the bee, akin to that provided by perennial Proboscidea? Fruit Size and Morphology j Among the features which might be modified, if Proboscidea fruit’s fiber, are: a) a disproportionate increase in the fibrous “claw’”’ part of the fruit; b) changes in texture, color and quality of the fiber; c) a greater yield of fruit per plant d) an altered frequency of unusual fruit shapes surviving. If large seed were selected in the domestication process, changes might include a) a disproportionate increase in the seed- holding ‘“‘body”’ part of the fruit, where the ovaries are; b) reduction in fruit dehiscence (see seed ‘esti discussion); and possibly c and d as above. Additionally, because mean fruit lengths of populations vary within wild Proboscidea species ranges, a bottleneck effect might occur, where the wild populations would be more variable than the domesticated populations. The ‘‘bottleneck”’ in variability would be the original selection of germ plasm undergoing domestication from = a small portion of the ‘‘available’’ genetic variability within compatible races of speci Table 9 indicates that there is nee differences in the claw/body ratios of the fruit of the white seeded race, and typical P. parviflora, in the wild and under cultivation. We defined the “claw” and “body” of the fruit ina woe peerary way, but: were Cometaneeet in how these features were measured. The claw, as I = fruit from which the Indians derive their fiber splints (Fig. 4). Fic. 4. Claw and body measurements of fruit of devil’s claw (drawing by Judy Spencer). 158 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 TABLE 9.—Claw/Body ratios for the white seeded race and the black seeded P. parviflora. Seed Source Locality Grown X claw/X body Sample White seeded race, cultivated: Havasupai Indian Cornville 2.99 36 Apache Indian Cornville 2:59 37 Pima Indian Tucson 2.59 28 Papago Indian Topowa 2.84 35 Black seeded race, cultivated by Indians: Papago Indian Chiawuli Tak 2.32 31 Black seeded race, spontaneously*; or cultivated#: Navajo Indian* Wupatki 2.34 10 Havasupai Indian# Tucson rae 21 Mestizo, for Papago* Nogales 2.29 26 Botanists* Tecoripa 2.21 15 The 2.5+ claw/body ratio of the white seeded race may not necessarily indicate that a dis- proportionate increase in the usable part has occurred via domestication. Hevly has noted that in P. louisianica fruit, the ratio may vary from 1.5-3 (in Correll and Johnston 1969:1449), and it { P. parvifloraand P. fragrans unavailable to us may have fruit which have a ratio greater than 2.5. The white-seeded race could have simply been chosen from such wild material, without selective pressures for wae claws being active within the cultivated environment, It is noteworthy that Gus Rare pc remember wild populations of P. parvifl. pir as 150 km | away | from their present homes. If the white seeded race has had kee val is — not for : seed- holdinie vaactiy. fl e have measured the claw lengths of populations of the wate seeded race, as = a in of typical P. parvif vite when a) harvested from th the wil igated’@ temporal fields (Table 10 pk Table 11. In the analysis of variance in 3 between populations of localities with 5 or more fruit of the cultivated whi black seeded P. parviflora, one or more populations are digtatcs at t the ‘01 level of significance. Utilizing a localities with ee cl more fruit, including those of presumably ‘“‘feral’” white seeds, the distinction betw' populations is still significant. This is due primarily to the extremely high values for the white seeded race under cultivation. The greatest apparent difference in claw lengths is between the white seeded race when under cultivation, and all the other material measured, cultivated or uncultivated. The cultivated white seed oe measure ie 3 cmt + 4.3 cm, whereas all naep gun means fall below 20 black seeded P. parvjione: treatments at the .01 level of confidence. May 1981 NABHAN ET AL. 159 TABLE 10.—Samples of claw length listed by source and locality. Locality and Source Mean (X) Range ( & ) Population (n) A- White seeded race cultivated and/or irrigated 25.3 cm +4.3 cm 249 A-1 Cataract Canyon #(Havasupai) 26.1 3.9 62 A-2 Cataract Canyon (Havasupai) 25.0 0.7 2 A-3 Cataract Canyon *(Havasupai) 34.0 0.0 1 A-4 Camp Verde *(Apache) 26.9 4.1 36 A-5 Moapa, Nevada (Southern Paiute) 24.3 0.0 1 A-6 Kern Co., California (Tubatulabal) 32.7 0.0 1 A-7 Komatke #(Gila River Pima) 23.2 3.3 52 A-8 Casa Blanca #(Gila River Pima) 22.6 1.6 8 A-9 Blackwater (Gila River Pima) 22.6 2.9 17 A-10 Chuichu (Papago) 25.8 1.8 A-11 Santa Rosa (Pa 22.5 3.7 3 A-12 Covered Wells (Papago) 25.8 0.0 1 A-13 Kitt Peak (Papago 20.5 3.0 10 A-14 Ali Chukson (Papago) 27.0 0.0 1 A-15 San Simon (Pa 25.6 3.8 7 A-16 Chiawuli Tak (Papago) 21.0 0.0 1 A-17 Topowa (Papago 28.2 4.7 40 A-18 Ahegam (Papago) E 23.2 0.0 1 A-19 Sells (Papago) 24.1 5.5 3 B- White seeded race cultivated (feral?) 18.5 3.9 6 B-1 Kaka (Papago) 12.4 0.0 1 B-2 Santa Rosa (Papago) 20.0 1.2 ~ B-3 Covered Wells (Papago) 19.4 3.9 3 C- Black seeded P. parviflora, spontaneous 15.7 4.8 127 C-1 Cataract Canyon (Havasupai) © 25.5 0.0 , C2 Wupatki (Navajo) 17.4 2.1 - C-3 Sacaton 32.0 0.0 l C-4 Sacaton 27.5 0.0 1 C-5 Ventana 145 3.3 37 C46 Ventana 9.0 15 7 C7 Wilax 18.5 0.0 : C8 R ont 11.7 4.6 15 C9 Hereford 13.5 1.6 12 C-10 Agua Prieta, Sonora 17.4 0.0 i C-11 Nogales (Mestizo for Papago) 20.2 3.3 - C-12 Tecoripa, Sonora 15.7 3.0 15 D- _ Blackseed, cultivated by Indians 19.6 3.1 - D-1 Chiawuli Tak (Papago) 19.6 5.1 E- Black seed cultivated and/or irrigated 16.5 2.6 or E-1 Cornville® 171 1.6 19 E-2 Sacaton 29.8 1.0 2 E-3 Cataract Coogee # 15.8 2.5 78 E4 Tucson 15.1 1.5 16 E-5 Tucson H 18.3 2.8 17 E46 Southern Arizona # 17.9 2.6 16 a a NS aE a LP re LT 2, . 4 - Srown in 1976 in Cornville, Arizona. #grown in 1976 in Tucson, Arizona. 160 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 TABLE 11. oe evaluation of claw measurements (see Table 10 for identifications of populatons 1. Analys = qh ed PR is exhuded): obse ved F317 = 25.25; greater than tabu le F317 = 5.18. Therefore at least one population is significantly different at the .01 level of confidence. Analysis of variance within and between populations for = localities (Bi is included): observes F4,36 = 30.12; greater than tabular F4 36 = 3.58. .01 level of confidence. 3. Conwast of pooled variance of combined populations via contrast coefficient matrix: (A + B) vs (D + E) - Pooled variance T value = -1.195 Therefore pooled populations not significantly distinct at .01 level of confid Contrast of Indian Serraie (d) black seed vs experimentally cultivated i) gene seeded P. parviflora (see conclusions... D = 19.55 + 1.14 (SE x T) vs E = 516.49 + 0.41 (SE x T) Therefore populations significantly distinct at .05 level of confidence. ed x = = Th f ing black seeded P. parviflora with the black seeds in cultivated treatments is most “revealing. The pooled variance analysis shows no significant difference between the pacelavated oy culisvatedd P. panniilore. One variable interpretation of this analysis is th y affect claw length of P. parviflora. In general, these dat pn eae Ree h a ees . salhy - oso S 5 y bY determined. Th icabl i tothi 1 ® a ma cc } Ww hit te seeded claws. Yet because of our extremely limited See of uncultivated white seeded fruit, we hesitate in considering this a major contradiction of the general trend. Until additional data indicate otherwise, we conclude that the white seeded race is genetically different from P. parviflora in this en ciarncteretie, Sve a ware) is reer feel flow between these taxa. Table I indicates that d quality of the claws and their fiber. These differences have been pointed out to us by native basketmakers, and will be discussed later. It is possible that these tpcome quantitative area Sica been gradually modified ee cultural selects ion Our 4 7 oe ll taxa grown under th nditi i Piweves.’ we have ccanted at least 150 ripening fruitona single plant in a Papago pane: at Sells, and project that its yield could easily surpass 200 fruit over the entire growing season. None of the wild seed which we have brought into cultivation have Pfr this productivity, although several of our white seeded plants yielded at least 80-120 fru There are also little iba on the frequency of fruit variants, or mutants, surviving in wild and cultivated populations of Proboscidea. However, 3- and 4-clawed fruit are a curiosity readily collected by Pima and Papago basketmakers. They have provided us with a multiple clawed fruit with white seeds, and 2 informants have recalled 3-clawed germ plasm nat was supposedly maintained for several generations. We have only come across one 3-clawed black seeded fruit, brought into a Blackwater, Arizona trading post by an Indian. Because of the difference in the relative number of cultivated versus wild fruit we have examined, w¢ cannot yet hypothesize whether the statistical frequency of surviving variants is actually higher among cultivated white seeded fruit Finally, it is notable that a number of Papago basketmakers volunteer that they plant only the seeds of the longest ones, because when the plants come up, ney make more big devil’s claw.” In other words uing. The ang of the Papago and Pima who note this selection al acide White seeds with intrinsical larger fruit. May 1981 NABHAN ET AL. 161 serene eee .. 1 5 l ap ee hy Ey BEBE, | ical data in] Pam wi 1 . + q (Fig. 3), we will attempt to eet a the aia questions: With which established Proboscidea taxa does the white seeded race e show the greatest aieent How does it differ from this taxa? Are th e they the effects of cultivation, or do they indicate true domestication? If so, what oe ‘a domestication process: selective pressures for food or fiber? Although the wentite seeded race has a geographic range which aces a fall completely within the ran parviflora, arid little with r. louisianica or P. fragrans. The area where i itmay extend hevdnd the range of the recognized wild annual Proboscidea is in Nevada, where but one truly wild P. parviflora occurrence has been recorded (Dr. Wesley Niles, personal communication) and parts of eastern California. However, given the ease of dispersibility of devil’s claw, we conclude that geographic range is in itself a poor indicator of affinities within the Probosicidea genus. There is little doubt, however, that in regard to floral morphology, color and ornamenta- tion, the white seeded devil’s claw is most similar to P. parvi ae rather than P. fragrans or P. louisianica. Additionally, the f with P. parviflora but not with P. louisianica or P. fragrans. These features are not always clear on pressed herbaria specimens, so that collections noting white flower color, with relatively large flowers, have often been referred to as P loutstanica on these latter features alone. We are confident, however, that the flowers of the white seeded devil’s claw show much more affinity with P. parviflora than with typical P. ae fare except in terms of flower size, a trait easily influenced by both cultivation and sele Cikee diagnostic vgs pein ‘such as leat shape _ ements hehe bear out an affinity with P. parviflora. Le such number of seed per fruit, oil and protein content also illustrate that the white seeded race and black seeded P. parviflora are within the same general range. he characteristics in which the white seeded race diverges the most from typical P. parviflora are not those which distinguish wild species. from one ange, but those most commonly influenced by domestication. TI economic product (the claw), increase in quality of the product (darker and more pliable), seed ne change, and loss - ate eases Other s ich as yield, leaf size, ] d lla size, and oil content are j in features easily ue re by indirect cultural selection. ost hasta the white seeded race does appear to have been domesticated from wild P. parvi rviflora, since the spontaneous race of P. parviflora does not “take on” these characteristics when simply brought into cultivation. Because the claw has been enlarged to a greater eager — the a none sind " o fruit, we feel that selection for fiber rather than food h 5 tion. Fiber quality has been considerably modified, wh ed number per fruit, dispersibility, and protein have remained relatively the same. These characteristics are usually altered significantly when a plant is domesticated for the food value of its seed. The seed features, e. g.s., loss of delayed germination, white seed color, which have developed in the domesticate could evolve under pressures from cultivation and deliberate human selection for fiber as easily for food. Thus we recognize numerous features which suggest disruptive selection of P. parviflora in cultivated environments and deliberate human selection, resulting in the evolution of a distinct white seeded race. This process is continuing, but to bol knowledge map. pos yet 1 nd > |x developed a fully domesticated, obligatory cultigen. The p p y feral white seeded devil’s claw i in Papago that the d hly associated but not entirel d their intentional EE of seeds. Iti is possible, : om however, thati ie the Kern iis California Neva a, y gewne»re ( 5 162 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 . = 1. fat jE epee | wild annual Proboscidea are commonly found, th was more dependent on cultivation than it is in the Papago rancherias. nally, it is worth emphasizing that the situation is much p more complies than simply having wild black seed and domesticated white seed. Ch ral parts, and more grayish hues in the white seeds suggest that “the domesticated qualities” of Camp Verde Apache devil’s claw are not as those and Havasupai. The black seed which the Papago cultivate have claws 19.6 t $3.1 cm, significantly longer than the black seeded claws which we brought into cultivation (Tables 10 and 11). Does this indicate incipient domestication, or merely that the Papago selected seed from longer claws in the wild to begin with? The frequent association of wild devil’s claw with the gardens of Apache basketmakers on the Fort Apache Indian Reservation (Anonymous 1976), may well illustrate the ‘‘self-domestication” process discussed by Whitaker and Bemis (1975:325-368). Bretting ( nal communication) is undertaking a systematic crossing program of various asaeAeRES of white seeded and black seeded P. parviflora, including some of our collections. Presently, variation within the white sete domesticate’ 8 gene pool, as wala as within P, parviflora in general, is poorly understood. \ variation, eliciting information from Native haskermakers on less obvious characters that they recognize. To clarify the selectiv l’s claw domestication, we urge scientists to actively work in the ssaiiiiek where this process took place - the agricultural fields and gardens of Southwestern rancheria people oilaaiits LITERATURE CITED in Great. Basin Cultural Ecology, ANDERSON, FLORA. 1922. The Development of the 104, flower and embryogeny of Martynia louisiana. Torrey Bot. Club Bull. 49 (5):141-157. ANDERSON, LoRAN C, 1968. Effects of Gibberellic Acid on Geraynation and a Growth hytomorph 18( ] x 166-173. beaker 1953. Improvement of Native New o. Agric. Sta. Annu. Report 64:26. Axownoe us. 1976. Women Preserve Basket Art. pache Scout, May: 7. BAKER, ie G. 1972. Human influence on plant evolution. Econ. Botany 26 (1). 32-46. BANCROFT, HELEN. 1932. A puzzling discovery of a capsule of Martynia louisiana. Torrezia 32 3): 59-64. pte eet 1975. Farming i in Prehistory : i John Midi SLUICE “A vA LIICTOT Baker, London BERRY, JAMES, PrTER BRETTING, GARY NABHAN, AND CHARLES WEBER. In Press. Domesticated Proboscidea parviflora: a potential oilseed crop for arid lands. J. Arid. Environments. LE, JAMES FRANK. 1923. The Pima and His Basket. Arizona Archaeol. Hist. Soc., Tu ucson. BUSKIRK, WILFRED. 1949. Western Apache Sub- sistence Economy. Unpubl. Ph.D. dissert., BYE, ROBERT A., jR. 1972. Ethnobotany of the uthern Paiute Indians in the 1870’s: with a Note on the Early Ethnobotanical Contributions of Dr. Edward Palmer. Pp. 87- bigots D. Fowler, ed.). Desert Res. Inst. Publ. Soc. Sci. 8. CASTETTER, Sth p F. 1939. Notes from Jemez n ethnobotany file for Southwest, Lab., Mus. Anthrop., Univ. Michigan, Ann Arbor. : 1939. Interview with Manuel Lowe, Pima Ca . 1939, Field Notes on Papago Botany and Agri- WILLIS H. BELL. 1942. Pima Papago Indian Agriculture. Univ. New aaa Press, pce apa M.E. Oper. 1936. The Ethnobiology pache Ethnobiol. Stud. Amer. So New Mexico Bull. 297, Biol. Ser. 4 (5): 3-63. UTH M. UNDERHILL. 1935. The Ethnobiology of the Papago Indians. Ethno- biol. E . Southwest 2. Univ. New Mexico Bull. 275, Biol. Ser. 4 (3): 3-84. CorBUSIER, WILLIAM F. 1886. The Apache-Yuma pache-Mohaves. Amer. Antiquarian 8 (6): 325-339. CORRELL, STEWART AND DONOVAN Ne CONRING JOHNSTON. 1970. Manual of - Vascular Plants of Texas. Texas Res. Foun R 1892. The Amer. FREDERICK VERNON. ndians of California. Anthropol. 5:351-359. enner CoviLie. Panamint | May 1981 CurTIN, LENORA SCOTT Muse. 1949. By the the Earth. San Vincente e. 52. Experiment in Con- servation. Erosion Control and Forage Pro- duction on the Papago Indian Reservations in Arizona. In Human Problems in Technologi- cal Change - A Casebook (Edward x Spicer, ed.). Russell Sage Found., New Yor DRUCKER, PHILIP. 1941. mete Element nai v. Californ . AND QUENTIN JONES. 1962. Analysis samples hi 113 plant families. Econ. Botany 16(4):221-250. FELGER, RICHARD S. ae anne Ne 1976. ForpE, C. DaRYLL. 1931. Ethnography of the Yuma Indians. Univ. California Publ. Amer. D SCOTT The Warihio Indians of Sterne ihatechae An Ethno i Survey. Bur. Amer. Ethn. Bull. 186: 61- H, AMITABHA AND JACK L. BEAL. 1979. Seed lipid constitutents of three species of 0 Nat. Products 42 (3): 287-292. GIFFoRD, E. w. 1936. Northeastern and Western “i avapai. Se acai Publ. Amer. Arch. Ethn. Se 24 cl Distril XII ara ee Univ. California Anthropol. Hyg 4: gy . 1965. The possible role of weed races in ae of cultivated plants. Euphytica 14: 173-176. 1975. Crops and Man. Foundations for Modern Crop Science oa Amer. Soc. Agr. Crop Sci. Soc. Madiso ——— AND J.M.J. DW WET. 1971. Toward a rational pap of cultivated plants. Taxon 20 (4): 509-5 spans Rosin. 1976. os index for comparing iness in plants. Taxon bad 5: st 247. ets, RICHARD H. 1969a. ew species o Proboscidea (Martyniaceae) ee Mexico. Brittonia 21: 311-313 Baan ; Nowenclatneal history and typifi- cation of Martynia and Proboscidea (Martyniaceae). Taxon 18 (5): 527-534. 1970. A new species of Proboscidea (Martyniaceae) from Baja California, Mexico. Bana tet 8): 393-3 —. HUDGENS. “MS. Archaeobiology and Paleooloy of Exhausted Cave, Yavapai County, Arizona NABHAN ET AL. 163 HOUuGH, WALTER. 1897. The Hopi in relation to their ore environment. Amer. Anthropol. 10 S. 1905. Notes on the San Carlos Apache. Amer. Anthropol. 7 (3): 480-495. 19006. N he Pi f Ari Amer Anthropol. 8 (1): 39-46. HurRpD, PAUL D., JR.ANDE. GorDoN LINSLEY. 1963. JAEGER, EDMUND C. 1941. Desert Wild Flowers. Stanford Univ. Press. Stanford (Third Ed., 968). . 1963. Archaeoethnobotany of Cordova Cave, Mexico, Econ. Botany 17 (4): 350-359. KEARNEY; THOMAS H. AND R.H. PEEBLES. | Arizona Flora, 2nd edition with supplement by ohn Thomas Howell, Elizabeth McClintock et al., Univ. California Press, Berke KELLy, ISABEL. 1964. Southern Paiute Ethno- phy. Univ. Utah Anthropol. Papers 69 KIssELL, MARY Lots. 1916. Basketry of the Papago d Pima. Amer. Mus. Nat. Hist. Anthropol. cota 17 — 1 fogs) 1959. Cochiti: A New Mexico Puch, Past eee Present. Univ. Texas Press, KAPLAN, L. "OTIS TuFTON. oes ib se ge cg Basketry. Smithsoni Report for 1902: 173-548 & sensi (Rereleased in 1976 by Peregrine Smith, Inc., Santa Barbara). ae America cKEE, BARBARA, EDWIN McKEE AND JOYCE HEROLD. 1975. Havasupai Baskets and Their Makers: 1930-1940. Northland Press, Flagstaff. McKENNAN, R. 1935. Basketry. Pp. 1-293, in MecciTT, M.J. 1965. The Association Between Australian Aborigines and Dingoes. Jn Man, Culture and Animals oo. Leeds and Andrew P. Vayda, eds.). Amer. Assoc, Advan. Sei; Se DG MERRIAM, C. HAR 1903. Some aged Known Basket Materials. eSdienice 17 (438): 923. Plants put In 20 agate in 1970 by Recta Sachi Picetias, N P. AND GORDON L. Fritz. 1977. Devil’s s claw (Proboscidea): cash crop in the industry. (abstract). J. No S. Dept. Report of the Commissioner of Agric. for the 164 JOURNAL OF ETHNOBIOLOGY year 1870. U.S. Government Printing Office, Washington, sg 404-428. serene f . 1875. Martenia proboscidea. Amer. Natur. me 42. PAUR, maton 1952. Four native New Mexico gric. College Agric. Mech. Arts State College. PENNINGTON, CAMPBELL. 1980. The Pima Bajo of Cen utral Sonora, Mexico: The Material Culture. k PERRY, BRUCE ALLEN. 1942. Genetic and Cytological Studies on the Euphorbiaceae, Martyniaceae and Malvaceae. Virginia Univ bstracts Dissert. 1942: 43-47. PICKERSGILL, BARBARA AND C.B. HEISER, JR. 1976. Cytogenetics and Evolutionary Change aa near Philosoph. Trans. Rev. London 275:55-67. RICK, ee 1976. Tomato. Pp. 268-272, in Evolution of Crop Plants (N.W. Simmonds, ed.), Longman, New York. Rossins, W.W. 1940. Alien plants growing withou t cultivation in California. Univ. Californ apa sta Sta. Bull. 637: 1-128. ROBERTS, HELE agen seemmriend of the San Carlos Apac abe Am us. at. Hist. Anthropol. Papers eS (2) ie ROBBINS, WILFRED W., JOHN PEABODY HARRINGTON AND eal FREIRE-MARRECO 1916. preggo of the Tewa Indians. Bir. Amer. Ethn. B RUSSELL, eek nn ‘The Pima Indians. . Ethn. Annu. Report 26 aa in 1975 by Univ. Arizona Press, Tucson). oo WILLIAM P. 1954. jaiteey 29 letter Ru Branch of the National Farm ptlionts Council, Inc., New York regarding devil’s claw research at State College, New Mexico 1-2. WANITZ, FRA 1 Cultivated > £ = INTRODUCTION The Kayapo Indians are one of the largest remaining tribes in Brazil’s Amazonian Basin. Their well-earned reputation for belligerance and violence (cf. Wagley 1977:31) kept them insulated from encroaching western society until 1937. In that year the first missionaries established permanent contact with the Gorotire Kayapo. The Gorotire represented only one of several schismatic groups, all of which had once been united in a powerful and populous ancestral village, Pyka-to-ti} (Posey 1979b). Once the Gorotire js been “‘pacified’”’ with Western trade items and medicines, other Kaya apo grou ased their warfare and established contact with Brazilian Indian Foundation (FUNAI) officials. The last group to be pacified was the Mekrangoti Kayapo, who have now had less than 15 years of sporatic contact with the outside world (Verswijver 1978). Most of the data analyzed in this paper were collected in Gorotire, the largest of the northern Kayap6o villages. Gorotire was the base camp for this 14-month project because of its accessibility and the presence of some bilingual (Kayapo and Portuguese) Indians. Gorotire was originally established as an “‘attraction” village that was well-stocked with medicines and trade items to ‘‘attract” unpacified ve th groups. As a result, the | Gorotire ae i isa emis Sar group. Nearly 2 20 yap6 group), 1% are non-K llyc id Kayapo), and 10% have immigrated ie Gorotire from opher Kayapé groups ¥ wiiue the ee years. This lends to Gorotire a “‘syncretic” th over whose version of a story or ceremony is the see one. Thus it should not Ho aieuned that Gorotire i is a village that agrees even upon ats own lore and mythology. isovan eager: % ayapo ] » however, g idl y d ned, pe, if predictable manner. Thi ith the principles underlying she Keane a ea classification system. A iced ac ye RS ] L Pd ma thic mattern: Ecological Profile The Kayapo have traditionally been considered “marginal” peoples poorly adapted to their environment (Steward and Faron 1959). They have been pictured as exiles from savannas and inadequately adjusted to the region of Central Brazil (Levi-Strauss 1958). Bamberger (1967) refuted this misconception by pointing out that sociological factors, not ecological limitations, were responsible for the size of Kayapo villages. The Kayapo are abundantly adapted to the diversity of the campo-mato ecosystems in which they are found and dietary essentials are obtained with minimal effort and time (Posey 1979a). There is evidence that aboriginally the Gorotire population was 8-10 times larger than today (Posey 166 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 1979b). The great amount of time spent in the p i fintri d time-consuming artifact production, plus frequent felal ituals and ials, hardly seem to reflect a group pushed to . the brink of marginal survival. The village of Gorotire is located on the broad, flat campo next to the Rio Fresco (7°48'S, 51°7’W). To the east are vast expanses of “campo cerrado” and “‘cerradao”’; in other directions, deciduous forest called ‘“‘mato de segunda classe.’’ Along the Rio Fresco is found “gallery forest’ (See Cole 1960, and Hueck 1966, fora discussion of these ecological types). Classification of soils, climat d cage, h point of obscuring any variations in the B Kayapo villages have traditionally aa located near both campo and n mato. This allows exploitation of various ecosystem types and maximizes the potential for utilization of natural products and game. This diversity has given the Kayapo a greatly varied diet that requires minimal effort. Elevation at Gorotire is approximately 1000 m. There is a marked dry season (May to 7o ° , h is still lacl I Gorotire is one of 7 northern Kayapo villages iocaten § in the — indigena Kayapo (see map). The total Kayapo population is now r 1,900,000 ha. METHODS Research was at first limited to work with the 6 men and 3 women who spoke Portuguese. Although an attempt to learn and utilize Kayap6 was made from the onset of the project, it was 7 months before eliciting could be carried out in the indigenous language. The type of data gathered reflects these stages of the A >see One of the first tasks begin — to t insect collection. Frequent field trips _ taken for the sole p diffi the Indians loosely grouped togeth Four to 5 Indians accompanied me researcher mm cores ting forays. The researcher bi ae the process by capturing a large grasshopper. dozens and dozens of other grasshoppers. The jebearcher attempted to — the selective process by capturing a dragonfly. The K of captured dragonflies. The researcher ‘Contiaed to try to widen the parameters ral “acceptable” things by pointing out butterflies, then beetles, and finally cicadas. “Are these relatives?”’ the researcher we onsrig pointing to the insects ube! collected and those still uncaptured i in an effort to d a notion of existed. ‘“‘Yes,”’ responded the Kayapo assistants. ‘“Then capture all of the relatives of these (saunas to insects already collected) you can!”’ The results was hundreds and hundreds of the same insects, depending upon the frequency of certain insects at the time. = was impossible to explain to the assistants why 300 of the same thing the range of “relatives of insects” (consistently anita “maja’’) expanded in 1 what was assumed to be a reflection native ideas of relatedne So £ After 3 months of this cue llecting, i dthe1 l ion of the category was completed. The category included ‘all insects, scorpions, spiders, = centipedes, aitieahe. crayfish, and pseudoscorpions. The category maja has one-to-one correspondence with the scientific category of Phylum Arthro} As the collection progressed, it became apparent that most iene were grouped into ibs flange ss categories. If there were no consistent sub-groupings (i.e., NO named oF unnamed differentiations), the specimens in that group were boxed and sent to the Museu Goeldi for classification and sraite in the Museu collections. fae any evidence of subdivisions did exist, however, t 1 in the village for further study. In the village, informants were asked to a) name each specimen, and b) group thos May 1981 POSEY 167 specimens sas were sas same cabenhen), o simliar fomusge In this manner, it was determined th t that correspond in a one-to-one fashion with the aie Class Arthropoda (Table 1). Further sub-groupings were few, except for the covert category corresponding to the scientific Class Insecta. Eighteen sub-classes (“forms”) were found in this category (Table ach s ee was numbered and each SPREE 21 was recorded in a master notebook. This igecbiouk cs contained essential field data on the 5 ptanrtiog plus a sketch or field identification notation if possible. If appropriate, entries g the cultural use of the TABLE 1.—Arthropod groups. CLASS/ORDER COMMON NAME KAPAPO NAME CORRELATION hnoidea 0 Pf ee Pay sctet oR EE: Rieti a eS Ee (a) Scorpionida ion e 1:1 (b) Pseudoscorpionida pseudoscorpion makkryre ? 1:1 (c) Phalangida harvesters hehpati 1:1 (d) Aran spiders heh ane 1:1 (e) Acarina mites/ticks ten 1:] stac crawfish maj 1:1 Diploda milipede morokreruti 1:1 Chilopoda centipede ekek 1:1 Insecta insects (covert) 1:1 TABLE 2.—Levels of correspondence for insects. B.O.L. CORRESPONDENCE CATEGORIES* COMMON NAME LEVELS CORRELATION # Focal Forms: ( 1) mara beetle Order (Coleoptera) 1:1 ( 2) ipoi true bug Order (Hemiptera) 1:1 ( 3) kapo roac (Family: Blattidae) # ( 4) sii tee grasshopper, cricket Order (Orthoptera) 1:1 (5) w butterfly, moth (Various neg a ( 6) nepal dragonfly Order (Odonata) 1:1 ( 7) kokot leafhopper, cicada Order hensaiietiie) 1:1 ; aa ae fly Order (Diptera) 1:1 (10) rorot termite Order (Isoptera) 1:1 1) (Family: Formicidae) # (12) amuh social wasp (Family: Various) # (13) mehn bee (Family: Apidae) # Collective Forms (14) ngoire minute insects (Various) Aberrant Forms: (15) karere earwig Order (Dermaptera) 1:1 Transititional Forms: (16) kapoti giant roach, mantid a Smeg # (17) kungont solitary bee & wasp = (Var # (18) mehnkamamuh honey wasp hac ‘alee # ‘asses fenruensoruteiemaihaseinssisioummmsiisiessoe sss cen, *B.O.L. (Basic oe Level Corer relation *Correlations differentiation). at the scientific level of Order (# indicates an over-differentiation; - is under- 168 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 insect or any peculiar ci d hich th lected. (Often Indians would bring a saeclines to be examined because they thought it interesting, unusual or particularly significant). Groupings of insects were tabulated initially for 6 men and 3 women; the maximal number of insects utilized in these sorting experiments was 635. Informants conducted the ouping activities on 3 different hansen each time with actual insect specimens. The bog ors scetath num miber of each specimen grouped was recorded ss each category. “Informant error’ was treated as erobleretie since patterns in ‘error’ were soon evident and vetniiialls predictable. Based on these data, 4 types of “forms” were identified (Table 2): 1) Focal forms, those consistently labeled and grouped in the same way and considered “typical” of the category. These forms are best illustrated as “fuzzy sets” (cf. Gardner 1976; Kempton 1978) with certain members being more focal and others being more peripheral. 2) Transitional fcrms, those consistently ‘‘mislabeled’’ between 2 categories. These forms are viewed as being “‘like’’ 2 groups that are contiguous categories in a morphological uence. 3) Aberrant forms, those consistently labeled in one category, but given a special name because of unusual morphological characteristic 4) Collective forms, those oe gigs ‘the or eccaeesan and grouped together, although i y not “really” be the same. In the one collective form discussed in the paper, small flies (ngdire), members of the category were considered too small to have significant morphological features and were illustrated with small dots. Uulizing tabulated resy d inf ing resp it was possible to hak 998 ‘eae Th the criteria of “basic iin level’’ “categories (cf. Dougherty 1978; Rosch et al. on Informant drawings and statements showed that the underlying patterns of these subordinate groupings were based on recognition of gross morphological features. DISCUSSION Patterns in Folk Entomological Classification For the Kayapé all things are divided into 4 categories: 1) things that move and grow, 2) things that grow but do not move, 3) things that neither move nor grow, and 4) man, 4 creature that is akin to all sos gs yet unique and more powerful than most animals because of his social organizatio It is the covert (unnamed) a of ‘‘animal” with which this paper is particularly concerned. All animals are sub-divided into 2 named groups: those with “flesh” (called by the name “‘mry’’), and those with “‘shells’’ and no flesh (called ‘‘maja’’). This latter ans animals with shells and no flesh, coincides with the scientific Phylum Arthropoda. F urthe folk subdivisions correlate with the 5 scientific classes of Arthropoda (Table 1). Although the folk grouping that corresponds with “insects” is covert, there is a 1:1 relationship with the sctentific Class Insecta. There are 4 morphological “sequences within this grouping (Fig. 1). The term “morphol ogical sequence’ refers toacontinuum of continuum, or there may be interruptions in the continuum. To bridge this gap, named transitional forms may occur to produce intermediate categories (Table 2). Sequence 1: Let us look at Sequence 1(Fig. 1) as an example. There is acontinuum of gross morphological form from the OVATE “polar form” to the OBLONG “polar form. ” Within this sequence can be found 2 distinct complexes: _ Complex A. This includes that per of the overall Sequence from beetles ies er win kapo) protective wing covers; their general pia ranges from ovate to oblong. “Considerable May 1981 POSEY SEQUENCES: (1) (Covert) 169 Complex A Complex B mara ipoi kapo 7 kapoti ; krytkanet wewe katiénet > karere 1 | | | (aberrant) | (transitional) | (2) (Covert) (3) (Covert) kokot (kryre) kokot ngoire pure kopre “ oe @ e @ a (collective) merc nl Ey One es ‘ees (4) “ny” (social insects) Complex A Complex B a ungont, rorot mrum amuh { mehnkamamuh j mehn \ f | i | | | i | { i [ a ~~» | (transitional) i ATR aeRO ot Be ‘ Fic. 1.—Insect sequences and complexes (based on drawings by Ira Kayapo). 170 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 Fic. 2.—A drawing by Ifa Kayapo of the wasp nest (amuh Tiriikwa). : ‘ lvcl ified as } Qf, thatis, } ambiguity mara and ipoi, but never is there overlap between mara and kapo. Likewise many ipoi are classified as mara, but also as kapo. No kapo, therefore, are co-classified with mara. The earwig karere is an aberrant form. It is consistently classified as a type of kapo, Las is singled out because of its morpohologi cal distinctiveness (mainly because it . rudimentary wings and “pincers” on its abdomen) and given a special monomial labe The overall sequence is interrupted with the transition from kapo to krytkanet, i.€., eit cockroaches to grasshoppers, although the morphological form continues tow war elongation. This break is clearly due to the presence of saa wings that become sufficiently conspicuous to define the perimeters of the animal's sha There is a transitional group, kapoti or g hes, that brid p. The we. wings and elongated bodies of this group canse them to be map os i kapo a krytkanet. This transitional form has a distinctive name and coincides with the scienuHc Family Blattidae. Complex B. The Sequence (Sequence 1) continues the second Complex (Complex B). In Complex B we have 3 overlapping genera: grasshoppers (krytkanet), butterflies (od dragonflies (kanenet). The polar form is the dragonfly, whose form is its extremely elongated abdomen and 4 wings. : Sequence 2: This sequence consists of a single complex called kokot. The capo within the complex is one of smallness to largeness — the leafhoppers being considered “children” of the larger cicadas. There is something of a form sequence from a si y rounded leafhoppers to the ovate cicadas, but this is insignificant to most infor gt mer Sequence 3: This sequence consists of a single complex of flies. It includes 2 object . ee nen May 1981 POSEY 171 categories: tiny flies (kopre), and mosquitoes (pure), biting flies and pium. There is, as is expected, overlapping between contiguous categories and minor morphological form gradation from ngoire (tiny flies, which are drawn as small dots) and more slender mosquitoes. Sequence 4: This sequence is composed of 3 distinct object level x A: termites (rorot), ants (mrum), and wasps lames): Complex B is composed of the single category honey bees (mehn). The break in and bees. This is attributable to the anomalous nature of bees, for they are the only shelled animal maja with major economic benefit. There are intermediate forms to bridge this functional gap. These intermediate forms are bees that make no honey and are solitary kungont, and social wasps that do produce wax and honey (mehnkamamuh). This is the oe named se ani aire ee “y.”’ This name refers to the social nature of sip insects; the f (1 1 pupae) that th Indians sa 1about like child he insects’ “villages” (or urukwa). The “fly” or er insects are seen to be in a special relationship to man because of their communal: nature. All “ty” colonies (urukwa) are thought to have a chief (@-benadjyra) and be organized into family units just like the Kayap6. They are known to have warriors and the sounds of their movements are likened to Kayapé movements and singing. ae is Kayapo are aware that some “ny” really Hive alone — sere is, there are solitary forms. t they ina ‘‘village”’ but for some reason now live alone. Solitary bees and wasps are like certain Kayapo who go off alone maybe for years on spirit quests, or or like ne ea = are solitary by nature. These insects are associated with th of shamans. In short, dicit dtiaata iil nature in relation to other social Hymenoptera and Isoptera make them important tools in the manipulation of natural powers by shamens. These aberrant forms are labeled with primary lexemes, although they are consistently classified as a sub- -group of the category amuh, social wasps. Except for Sequence 4 amma ants, hee and wasps), sping eye are few for Insects; subspecies are even fewer. Affixes d texture, size (or agi general feature are frequently attached to the primary (1°) lexemic label of ee generic category. An informant may choose any of a number to describe a specimen. Thus, (mara-tyk-ti) means big, black bee and the label my aby to any one of many beetles that are big and black. But the same b ‘ara-kra-ti), big-headed beetle, if it were black and also had a big head. Occasionally a descriptive (or secondary lexeme) label may be reserved for a particular, limited set of insects. Within the beetle category is such an example, (maratire) or dung beetles (Scarabidae). Each insect group (basic object level category) hasa “father” (bam). The “father” is usually the largest member of the group. The “father” of the (maratire) is the i impressive Rhinocerous Beetle (Strataegus, Scarabaeidae). It is called the (kra-kam-djware) and is also considered the “chief” (6-benadjware) of all insects (really all maja). There are, however, only a few examples of “as specific naming in Kayapé insect classification — except, as I have said, within the Sequence (4) of “fy,” the social in: There are 32 sub-groupings of (mrum) ants; 48 eee of wasps (amuh); and57 sub sroupings of bees (mehn). These specific and su primey secondary (2°) lexemes. But why does this specialized “classification occur seein the Sequence “iy?” The importance of bees is obvious: they are sources of honey and wax. But of what Significance are wasps and ants? Already we know these animals are like man because they live in societies like the Kayapo: they have villages, chiefs, and rs do termites, yet there are only 4 sub-divisions of termites (rorot). This is certainly not due to a paucity of termite types in the Kayapo area To understand this situation, we must understand one of the most significant of Kayapo myths: the story of the ancient fight with the giant beetle, the kra-kam-djware. 172 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 In ancient times the Kayapo lived in the sky with other animals. The Kayapo were then like other animals and Indians could understand animal languages. But in these ancient days, the Kayapo were weak and did not live in villages or have societies. Indians were not more powerful than other animals and certain animals, especially the beetles (mara) under the leadership of their “chief,” the kra-kam-djware, waged i In th ient day in the sky, the Kayapo] d ganize tk Ives into grou] 1 live in villages like the “ny” (wasps and ants). Then in a great battle in those ancient times, the valiant and fearless warriors of the Kayapo defeated the kra-kam-djware. That defeat established man as a creature more powerful than other animals because of 2 things: 1) the power came from the social organization, and 2) the great strength and valiance of the Indian warriors that had also come from the wasps. The Kayapo had learned the wasps’ secrets by carefully observing the behavior of wasps and had learned of their “power” that could be gotten through their potent stings. The venom of the wasps had been the secret; the aggressive, fearless attacks of the wasps had been the model for Indian warriors. Today, on regular occasions the Kayap 5 th juisti f these secrets and their victory over the kra-kam-djware. They are tantly searching for tk f the most powerful and aggressive wasp (the amuh-dja-ken: Polistes testacolor). When a nest is found that is sufficiently large (usually 1.5 m long, 0.5 m in diameter), scaffolding is erected (by night when the wasps are inactive) to prepare for a re-enactment of the ancient event. In the numbing cold of a gray pre-dawn haze, the entire village goes solemnly to the site. The warriors dance at the foot of the scaffolding and sing of the secret strength they received from the wasps to defeat the giant beetle. The women wail ceremonially in high-pitched, emotional gasps as the warriors, two-by-two, ascend the platform to strike with their bare hands the massive hive. Over and over again they strike the hive to receive the stings of the wasps until they are semi-conscious from the venomous pain. This ceremony is one of the most important to the Kayapo: it is a re-affirmation of their humanity, a statement of their place in the universe, and a communion with the past. Time and space collapses to provide the unity of being — the continuity of life, history, identity and knowledge The wasp’s nest itself is a symbolic statement of this unity. Its three-dimensional shape illustrates the relationships between the polar forms of the classification morphology — the ovate and elongate forms (Fig. 2). A cross-sectional view — or view from above or below — shows the circular form; a lateral view shows the elongate form. The nest is a graphic study of the relationship between these sha Even more importantly, the general structure of the hive itself serves as a model of the universe. The hive is divided into parallel “plates” that seem to float just like the layers of the universe. The Kayapo say that today they live on one of the middle plates. But in ancient days, they believe they lived on another plate above the sky. Some Kayapo still live on an upper plate the tribal elders say, and their campfires are the stars in the sky (Fig. 3). And below? From lower plates comes the “worthless men’’ (non-Kayapo, kuben-kakrit). Many kuben-kakrit still live below, though most have already ascended to “this earth layer” through a termite mound. Oo non-Kayapo “worthless people.” (A fourth subgrouping labels the termite that lives in the mound through which came the kuben-kakrit). And what of ants? They lil h the ground. The Kayapo beli } h =: iL dhunton wasn aes fol, ctings. r 0 the power f ts is more useful on man’s hunting ally — the dog. Ants are nee i ; maLlea h os See nd to in Many concoctions to di to the g - — May 198] POSEY 173 (Side View) sky layers earth layers Oo Ltn, Copece? RO SCED are te CPE ESRB E SS DESO BEDE ZED ae FIG. 3.—Cross-section of a wasp nest (drawing by Ira Kayapo). (Bottom View) make him aggressive. Some ants are seen as excellent hunters, so often man and dog are adorned for the hunt with the sacred red urucu paint mixed with ant parts. To be good hunters, therefore, the Kayapo must know 4 hey must know ps to be b fearless warriors. CONCLUSION In conclusion, I believe ethnomethodology can lead the ethnographer into fields of investigation along natural (emic) paths. Folk taxonomies are in and of themselves cultural statements, but it appears that these taxonomies may reflect deeper cultural patterns. This analysis indicates that insects are encoded at a “basic object level” with the predominating characteristic being gross morphology (shape) that grades from the ovate form to the elongate form. These 2 “polar forms,” and the relationships b een th f become an underlying principle for Kayapo folk entomological classification as well as a spatial and structural theme in the belief system. It is therefore suggested that the correlations bety ) basic shay i forms, b) belief s) ' dc)cl : Principles may be more closely integrated than previously expected. It appears that belief Systems can play an important role in classification patterns and that such patterns can, in turn, offer an emic guide to cultural realities of perception. ACKNOWLEDGMENTS Funding for this project was made by the Wenner-Gren Foundation for Anthropological Research. Brazilian sponsors of the project were the Conselho Nacional de Pesquisas (CNP q), the Instituto Nacional de Pesquisas de AmazOnia (INPA), the Museu Paraense ‘Emilio Goeldi,’ and the Fundacao Nacional do Indio (FUNAI). I am most grateful to the Institutions for their support. 174 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 LITERATURE CITED BAMBERGER, JOAN. 1967. Environmental and ah Classification: A Study of the apo. Ph.D. dissert. histinony vad Univ CoLe, Monica. | Cerr rao. Catinga and Pantaanal: @incthiiinc and origin of the savanna vegetation of Brazil. Geograph. J. 126 (2): 186-179. DOUGHERTY, J.W.D. 1978. Salience and relativity in classification. Amer. Ethnol. -80. GARDNER, PETER. 1976. Birds, canis and a equiem for the greene as informant. Amer. Ethnol. 3:446-4 HUuECK, K. 1966. a ‘Dicircinda iia-Kbwarics do Sul: Ecologia, Composicao e mportancia Economica. Brasilia: Editora Universidade da Brasili KE a. . MPTON, WILLETT. 1978. Category grading and taxonomic relations: a mug is a sort of a cup. Amer. Ethnol. 5 1958. Anthropologia L. 1979a. Kayaos Controla Inseto com Uso pear do Ambiente. Revista da Atualidade Indigena III (14): vice _____. 1979b. Pyka-t6-ti: Kayapé Mostra Aldeia de Gb a8 Revista da Atualidade tedieeend Il (15): 50-57. puree ioe a. MERVIS, W. ea D. JOHNSON . Boy ES-BRAEM. 1976 objects in aa categories. Cognitive eck 8:382-439 STEWARD, J.H., AND L.C. FA 59. Native Peoples of South America. McGraw-Hill, New York. STOUT, MICKEY AND RUTH THOMSON. ee Kay yapo eaphas Interntl. mer Linguistics 37 (4): 250-256. 1974. Fonemica Txukuhamei (Kayapo). Serie Linguistica 3:153-176. VERSWIJVER, GUSTAAF. 1978. Enquete Ethno- graphique Chez lesKayapo-Mekrangoti: con- tribution a Vetude de la dynamique des groupes locaux (scissions et regroupements). Unpubl. Thesis, Ecole des Hautes Etudes en Sciences Sociales, P. WAGLEY, CHARLES. 197, Frielciwee of Tears: The Tapirape Indians of Central Brazil. Oxford Univ. Press, New York. NOores I'Th > ee E paper for ees words is the official | a phe ment version developed in 1974 hth onjunction wit ummer Institute ie Lie {SI.). For further information on language, see Stout and Thompson yore 1974), 2A collection of nearly 6,000 insect specimens was deposited with the Museu Paraense ‘Emilio Goeldi’ (Belém-Para), under the supervision of Dr. William L. Overal, head of the inven oology se ction. I am indebted to Dr. Overal fo “es Sexe assistance in identification of sot see tion J. Ethnobiol. 1 (1): 175-181 May 1981 USE OF OPAL PHYTOLITHS IN PALEOENVIRONMENTAL RECON STRUCTION RHODA OWEN LEwISs University of Wyoming, Department of Anthropology, Wyoming Recreation Commission Laramie, Wyoming 82071 STRACT.—Soil and climatic conditions in the Wyoming-Nebraska-Colorado area of the High Plains duci pollen preservation. Opal phytoliths, which are present in many achaeological sites in sufficient quantities to provide information concerning both natural and introduced vegetation, are proving to be a viable al i pollen in pal environmental reconstruction. The results of phytolith studies can amplify the archaeological data in regard to environmental, cultural, and geological processes. Examination of soil samples from archaeological sites on the High Plains is on-going research by the Department of Anthropology and the Wyoming Recreation Commission, University of Wyoming. Thi deals with y d h bei ducted and the results of these studies. iy INTRODUCTION One of the goals in recent archaeological excavations has been to obtain data useful for reconstructing paleoenvironments. Palynology has been the best and most widely used method for this purpose. However, soil and climati litions in the Wyoming-Nebraska Colorado area of the High Plains are often not conducive to pollen preservation. To find a viable alternative to pollen studies in paleoenvironmental reconstruction, the Department of Anthropology, University of Wyoming, has been conducting opal phytolith research. Opal phytoliths are present in many archaeological sites in sufficient quantities to provide information on vegetation, both nat 1 introduced. The results of phytolith studies can amplify the archaeological data in regard to environmental, cultural, and geological processes. Phytolith studies can provide data about types of grasses growing on the site area and on nearby grazing areas, changes in vegetational types, and changes in moisture levels. It may be able to determine use of buffalo chips for fuel, primary butchering fg imals, Sleeping areas with grass pads, and types of grasses ground on metates. Humid grass phytoliths may indicate a previous water source, and phytoliths in fill ial provide information about vegetation from areas draining into the site. Rovner (1971:343-344) states: For any fossil system to be useful to th material must with id position, pokes taxonomic significance, and provide sufficient quantities to reflect the nature of the entire assemblage from which it is derived. haeologist at least three criteria must be met. The t=) is FE cs } Although phytoliths may not meet all 3 criteria in all situations, they can provide valuable insight and their contribution to the archaeological record must be considered. _ Opal phytoliths form as plants take up soluble silica. The soluble silica forms around and in plant cells producing distinctive shapes (Yeck and Gray 1972:639). However, the shape of the silicon bodies may not be the same as its cell, making phytoliths difficult to classify. Pollen is produced in a single repetitive form by each plant, phytoliths are produced in a wide range of sizes and shapes within any specific plant and parts of plants. When the plants decay, the resistant silicon forms are deposited in the soil. Twiss et al. (1969:111-112) produced a morphological classification of grass phytoliths of 4 classes: Chloridoid (short grass), Panicoid (tall grass), F id (humid grass), and Elongate which appear in all grasses (Fig. 1). Amounts of opal phytoliths available in samples differ from SMe to site depending on the variables present. A percentage count of phytolith types 176 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 UBO BY Vio oy ee ea RN ho 4.5) Ff PANICOID oa oe, FESTUCOID Se a Ol a, ELONGATE Fic. 1.—Morphological classification of grass phytoliths. appearing in each slide ascertains with some degree of reliability the g predominentin the site area. Opal phytolith studies can provide information concerning not only the climate of an area during the time a site was occupied, but can also show changes that have occurred in an area over a period of time. Changes in vegetation are shown by changes in class type of phytoliths. Phytolith size is also indicative of changes i i (Yeck and Gray 1972:639). Phytoliths are not indestructible; they can be fused by fires, eroded by soil action, and are susceptable to destruction by the soil chemistry. DISCUSSION In the past 4 years, samples from a variety of sites in this High Plains region have been examined for phytoliths. The sites range in age from Folsom to Late Prehistoric, spatially cover most of Wyoming and j f h Nebras! 1 h n Colorado, and represent various functions as campsites, kill sites, and rockshelters. Some sites have produced excellent phytolith evidence, others have been totally void. This paper presents a broad overview of phytolith studies that have been conducted (Fig. 2). Th Mountains near Shell, Wyoming, is a prehistoric campsite. The site has been dat 10,700+670 B.P. (RL-374) (Frison 1978:23) and is culturally affiliated with the Folsom complex. Phytolith evidence from Hanson was very poor. Samples from a soil column 1? son I yielded nothing. However, a sample taken in the occupation level from the —_ fill below a Bison humerus in Hanson II produced representative samples of tall grass 4" humid grass phytoliths. ; Another site which I d ly poo phy toliths the Agate Basin site located we eastern Wyoming in the Cheyenne River drainage close to the southern Black Hills 0 South Dakota. Agate Basin dates 10,430+570 B.P. (RL-557) (Frison 1978:23) and is the type i i a May 1981 LEWIS ~” HANSON a es oA at n -4 2 wy - = LADDIE > a <= CREEK $ i sity ca é 2 a ye 7% ay SOUTHSIDER ‘ THE WATERS es AGATE BASIN a AN . DFAS, Am ~. SBR i NON on a - Ma a oS HUDSON-MENG z “pn ore af = aC te s oat Pts rai nae 77, - = Se ~. “an BS a . ail pal =” ~~ = ¥ y ’ 5 ey yy ~ a JONES-MILLER Fic. 2.—Location of sites discussed in this study. site for the Agate Basin point. Itisa bison kill and p ing site. Samp il profil and within the bone bed produced rod-shaped phytoliths but none that could be classified. One €xception is a sample from Unit 4 in a soil profile. This sample is within the Agate Basin level in the site and contained short grass prairie phytoliths. Hudson-Meng, a bison kill and processing site, is] dafewk Agate Basin in northwestern Nebraska near Crawford on the northern slope of the Pine Ridge Escarpment in the Hat Creek drainage (Agenbroad 1978). The site dates 9820+ 160 B.P. (SMU-224) (Frison 1978:23) and is associated with the Alberta cultural complex. Soil samples were taken from Hudson-Meng during the 1975 field Samples f trench at the west edge of the site (Fig. 3), in the bone bed, and from a hearth within the bone bed produced the following results. A high percentage of Festucoid class phytoliths were in all the productive trench samples indicating the constant presence of humid grass, possibly 4 microenvironment caused by a spring or stream. Trench samples 1, 3, and 4 imply a Period of increased moisture when tall grass prairies would be present. Trench samples 6 (the Alberta bone bed level) and 7 show reversal of this trend with short grass phytoliths being more common. Samples 2 and 5 ially void of phytoliths which may it an extremely dry period. These interpretations are supported by the pollen analysis at the site (Agenbroad 1978:117). Extrapolated figures from the bone bed, hearth area, and sample 6 are compared in Fig. 4. Seated € 6 was taken outside the butchering he high percentage of Festucoid phytoliths Indicates a nearby water source. The phytolith counts of Festucoid and Panicoid classes in - bone bed samp! with sample 6. The contents of the bison viscera deposited dur ing butchering account for the significant increase in Chloridoid class phytoliths. The evidence in the bone bed sample implies that the bison had been grazing on short grass Prairie prior to the kill. Increased phytolith counts from the hearth area, high percentages of Chloridoid class phytoliths, and lack of charcoal at the site strongly suggest the use of buffalo chips for fuel. +h = I id 4 Phat vnict 1aicate é 178 pe esses seeses ne 7 e, { he OO mr a a a 3 Pesepeeeaeeeeee eee Peeeeeveeeeeeenee ta le OG Sm cee eo om 4 - JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 Brown-gray silt, fine sand, and root mat. Pale, brown, silty sand with medium prismatic structure. Dark brown-gray, silty, fine sand with coarse, prismatic structure and active rootlets. Light, brown-gray, silty, fine, sand: very soft; darker than the unit below it. Pale, brown-gray, silty, fine sand; very soft. Aes erosionsal contact. e, brown, calcitic, fine sand with calcite nodules. Weak cleavate with truncated rootmolds. = Very pale, gray-brown (white); very calcitic; very fine, silty sand with numerous rootmolds; dry. Very pale, brown, fine sand. ‘Transitional. Very pale, brown, fine sand. Weak, organic bone; 2-3 cm thick. ees rta bone bed. brown to black carbc wganic, fine sand. cater ee with outlines of insect aa ys and casi Very pale, brown, calcitic, & clayey, silty sand with ce root Casts. ‘ ape 50% 22 10 60 14 204 28 fragments < 50% 10 5 51 12 38 5 split at s 25 12 65 15 125 17 charred or pale charred 1 + 1 + 62. 9 Total 217 101% 430 100% 718 99% tes d iorati However, if i b bundant, the effect of temperature appears to be minimal. Over the course of a Aiea a Dcomating winter by gtheg bacteria. The amount of loss during stable mid year ith al fom ‘ wetting from summer rains and drying might cause rapid disintegration. If the soil is slightly acidic, this will slow deterioration by impeding the growth of bacteria and mold. The second examination of deterioration brought about by exposure to the elements was drawn from the Walpi archaeobotanical record. The time differences in species counts and ] their conditions in 3 asia at the pueblo were measured. These rooms are, 1) nee a pikih whose fill p 1920- 1965, I elements i. 10 y years, 2) Room 83, ast whose fill db 1930-1950, and has been exposed 25-35 years, and 3) Room 187, also a storage room whose fill was deposited much earlier, between 1880-1920. The fill in room 187 has been exposed for approximately 65 years. The first test of this data base involved lant counts by genera in each of the 3 rooms. Hypothetically, one would expect that the 1 room exposed for the longest duration would contain the least plant remains. Such was not the case. This study produced results (Table 3) which indicated an inverse oe where more plant remains were in the room exposed 2 65 years. This does not seem plausible, and an alternative needs to be examined. For instance, actual counts may me misleading as ieatribers oye reflect the cultural bias of erential discard. Thus, n Room 187 than in the other 2 room A more viable test of thes data was to ane the oo of the 2 eg in each of the 3 rooms. A pl ’s condition i f the cultural experience than i Lis NuMerical ease We are bidder bicte' in any of the non-cultural effects such as rodent and insect gnawing and other : a oe of plant remains 1 in a Table 3 details the results of this examination of t _ each of the 3 rooms. These data indicate only a few relationships one would expect of remains which had ering lenigtlis of capes ure ie the elements. For example, 49 assemblage of plant to degr ade with length of exposure. Seeds and other plant parts which were slightly degraded, fragmented less than 50%, split at suture, OF charred or partially charred, occurred with greater frequencies as the length of exposure increased. This implies a steady and rather constant rate of degradation. In some examples this was not always the case. Unaltered remains and fragments greater than 50% were less May 1981 GASSER AND ADAMS 187 common in the rooms i pea sii 10 and65y han they in tl posed for 25 35 years. This might b li the eff. f if , butit does indicate that there is not a direct selanionship: where organic remains throughout a pueblo disintegrate at an even rate. Rodent and Insect Activity.—Another significant fact that appeared as a result of this study was that rodent and insect gnawing apparently decreased a i Sadteaa of exposure increased. R 48% of allt only 10 years, on 16% in the room exposed for 25-35 years, and 9% in the room exposed f or 65 years. These data inc — ——- i api and consume large amounts of plants food supply. Rodents bias the record extensively. The data on Table 3 ee that as much as 50% of some discarded organic trash is partially consumed by rodent Some seeds are probably consumed in their entirety by rodents and nce others seem to be gnawed or cracked only enough to gain access to the soft portion of the seed encased in its Shell. Figure 2 illustrates watermelon seeds from Walpi and reveals partial gnawing. A laboratory produced example of rodent gnawing on watermelon seeds was obtained at the beginning of the Walpi botanical analysis and enabled identification of some rodent gnawing on viclquansisis'acsaie seeds. For this sainan —— gnawing was determined by the presence of incisor f the seed or seed coat, such as that illustrated with the watermelon, bottle-gourd, and juniper seeds in Fig. 2. When some of this seed damage might be attributed to insects as well, the combined category of rodent-insect ehawing, was used. Fao only auprs ted vy hen QR\yi bh 1 d ill di Figure 3 which shows corn betaacks on cobs which have b Gee ee fd id beetles. The Ss cobs with dermestid infested kernels shown here ltered when Ad from Room 12] 2 years ago. Dermestid larva accompanied the remains back to the Museum, and did this amount of damage in the intervening period. The larva of this beetle enter the kernel at its anterior end and consume the soft starchy interior, often leaving only the ne behind. The thin pericarp would probably disintegrate leaving no remains of the ernels Rodents and insects seem to prefer different foods, ra taking - toll i in —— degrading and pipiens = in sites. camine be ol rodent and insect dam age led i heat ‘(Table 4). Fables indicates those species gba hale and lieoct activity, the species total seed count in the sample, and the number and percent of which were either rodent or insect gnawed. In some cases it was impossible to identify the gnawer, hence the category rodent-insect Snawed. Table 4 indicates that 42% of all of the corn at Walpi, excepting shucked cobs, husks, and other non-kernel parts, eres sees by bose activity. oe or rodents damaged over half of the bottle-gourd see seeds of the Cucurbitaceae. Rodents damaged over ia of the more than 24,000 watermelon seeds investigated by this analysis. Rodents (or insects) also damaged 75% of the juniper seeds. Rodent gnawing was also evident on many stones of fruits such as cherries, plums, z and mes sino pits, which were abundantat Walpi, seem the rodent predato: An wits of 11 rodent nests and one rodent cache (Table 5) is helpful in determining Preferred rodent foods. Table 5 indicates that 75% of these rodent contexts contained watermelon seeds and half or more contained sunflower, unidentifiable squash, and melon - Corn was found in only 2 of the 12 contexts, beans in only one. In general, ihe “ie contained few seeds, and the small seed counts in tl to find preferred foods. A rodent cache (e.8., Lockard and Lockard 1971:221) in Room 112a Ipiw ferred by rodents. The cache was inasmal subfloor pe hepsi: It contained 1468 seeds of 15 genera, many of which were roden Snawed, and a large amount of rodent feces. Almost half of the seeds in this cache wer 188 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 TABLE 4.—Tablulation of rodent and insect gnawed plant parts at Walpi. TOTAL #INSECT % INSECT # RODENT % RODENT # RODENT- % RODENT- SPECIES COUNT GNAWED GNAW GNAWED GNAWED INSECT INSECT GNAWED GNAWED Bottle-Gourd seeds 462 240 52 Watermelon seeds 24,746 13,738 5G seeds 1,155 25 9 Maxima squash, 138 70 banana type seeds 96 Maxima soap 23 6 26 S. American type seeds M 3 q ae os 9 Pepo squash seeds 55 3 5 Mixta squash 926 59 6 seeds Unknown type A 9 1 1] squash seed Not identifiable 22 3 ll squash seeds Beans 509 3 + Corn —_ with 46 41 89 kerne Cob set 10 3 with kernels kernels 1,692 642 38 Peach pits 2,960 1 _ Plum pits 23 9 Apart pa 43 mT 26 pits 9 5 Almond hull 4 1 Juniper seeds 700 593 75 Pinyon testa 1,954 93 le ndENY Sou Wedel SN ieee watermelon, almost a third were corn kernels. The only other signifi pecies were melon and cushaw squas seeds. The data in Tables 4 and 5 are indicators of foods which rodents preter when foraging on Hopi trash deposits. The last experiment undertaken involved a more direct test of rodent seed preference. Live mice were fed some seeds which were commonly found at Walpi to determine which were preferred over others. Four domesticated mice were purchased from the Ba store: =e 2 domestic noun mice (Mt us Lpomglgsigads a species wie occurs at Walpi, The results of this srnaulane ent are portrayed i in Table 6. It is clear bom this evidence that given a choice 0 5 species which included blue flour corn, honeydew melon seeds, Mammoth Russian sunflower achenes, Halloween field pumpkin seeds, or white, common, kidney, or tepary beans, the “average’”’ mouse preferred ee cont. In se 2 experiments with the domesticated mice, they consumed over 40% of th d practically ignoring other species, and ate all of the 25 8 of com within 5 days. Once the’ corti started to diminish significantly, the mice They chewed the seeds from margin, never splitting them along the suture to extract the owe interior as might be expected. The white beans and pumpkin seeds suffered no significant weight loss. After 3 days a few of the beans and pumpkin seeds were slighty chewed by the mice, resulting in nothing beyond slight damage to the seed. Th owed that after 3 days they, too, avoided the beans and aclecied the melon, corn, and Scaflower seeds in nearly equal amounts GASSER AND ADAMS 189 May 1981 ey hanes Tre NO I specs Ou I Gs oO ‘Sal0adS G1IM Jamoyung uopPaw uOauIaIeM, ‘VaOV.LICMNOND rOZST 9IZSH E22SH C8ISHT G68ISH FGISH F9IST sIIsd 682SH 6984 LOZS4 LOISA Sa1adS SLSIN 1d] 10 Sisau [] PUY aYyIvI JUapOL aUO UI SJUNOD paas fo UMOPYyvaLg—'G TTAV_L 190 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 1 TABLE 6.—Seed weight loss due to consumption by mice at 48, 72, and 120 hour intervals. at 48 hours at 120 hours EXPERIMENT #1 seed weight % loss seed weight % loss 2 domestic mice with free water 25 g flour corn 1] 56 0 100 25 g melon 23 8 18 28 25 g sunflower 24 4 20 20 25 g pepo squash 25 0 25 0 25 g white beans 25 0 25 0 EXPERIMENT #2 2 domestic mice with free water 25 g flour corn 15 40 0 100 25 g melon 22 12 16 36 25 g sunflower 24 4 24 4 25 g pepo squash 25 0 24 4 25 g white beans 25 0 25 0 EXPERIMENT #3 at 72 hours seed weight % loss 2 house mice without free water 25 g flour corn 23 6 25 g melon 29 12 25 g sunflower 23 6 25 g kidney bean 25 0 25 g tepary bean 25 0 Other — have ‘experimented cares feening native rodent species (Perognathus Spp and pursued here (Mares and Williams 1977; Lockard and Lockard 1971; ‘Rosentweig and Sterner 1970). These studies f p ll types of f seeds presented, but might specialize i in small or medium-sized varieties if resources were especially abundant (Mares and Williams 1977:1188; Rosenzweig and Sterner 1970:223). Lockard and Lockar (1971:219) found, in addition, that if specialization occurred, it generally favored seeds high in carbohydrates, oils, proteins, _ A apECES | with a thin seed coat. Hu man trash deposits provide abundant nearby resourc n rodent foraging a behavior. It is worth noting, too, grey even in areas of abundant seed resources, ee is not enough seed selection involved to prevent a number of rodent species from nthe same environment (Rosenzweig and Sterner 1970: 222-223). Hence, an abundant habitat such as‘a trash midden could support a number of species of rodents. These ecological studies recorded some information \ that t appears to. contradict — reached here. For instance, Lockarda kangaroo rat, preferred — beans and macet other available species to sunflower - 60 ON PREDICTING HUMAN DIETS, H. Ronald Pulliam ..........2+0ee008> ... 61- 68 RESOURCE UTILIZATION AND FOOD TABOOS OF SONORAN DESERT UNS Ales AE ee a Sn ee ons se eee 69- 83 DIETARY MINERAL ECOLOGY OF THE HOPI, Harriet V. Kuhnlein .....--+++> 84- 94 THE PERCEPTUAL BASES OF ETHNOBIOLOGICAL CLASSIFICATION: EVIDENCE FROM AGUARUNA JIVARO ORNITHOLOGY, Brent Berlin, James Shiles Banter. ‘and Joka PO’ Ned) os 5 i a ss Pn He a eee 95-108 QUELITES — ETHNOECOLOGY OF EDIBLE GREENS — PAST, PRESENT, AND FUTURE Rib A be, ec a es a . 109-123 ON THE RELATIVE CONTRIBUTION OF MEN AND WOMEN TO SUBSISTENCE AMONG HUNTER-GATHERERS OF THE COLUMBIA PLATEAU: A COMPARISON WITH ETHNOGRAPHIC ATLAS SUMMARIES, Eugene S. Hunn ......---++++* 124-134 DEVIL’s CLAW DOMESTICATION: EVIDENCE FROM SOUTHWESTERN INDIAN FIELDS, Gary Nabhan, Alfred Whiting, Henry Dobyns, Richard Hevly and Robert Euler 4c... eis i 135-164 WASPS, WARRIORS AND FEARLESS MEN: ETHNOENTOMOLOGY OF THE KAYAPO INDIANS OF CENTRAL BRAZIL, Darrell A. Posey... -2.0-+2+00000" 165-174 UsE OF OPAL PH IN P INSTRUCTION, Rhoda Owen Lewis ..........-+----> FO EG a ee. 175-181 7 A rT OF PLANT REMAINS IN ARCHAEOLOGICAL sires. Tae Watt ARCHAEOLOGICAL PROJECT Robert E. Gasser and eR ee ee ee ee eee 182-192 = Journal of VOLUME 1, NUMBER 2 DECEMBER 1981 JOURNAL ORGANIZATION CO-DIRECTORS: Steven D. Emslie and Steven A. Weber, Center for Western Studies, Inc., P.O. Box 1145, Flagstaff, Arizona 86002. EDITOR: Steven D. Emslie NEWS AND COMMENTS EDITOR: Eugene Hunn, Department of Anthro- pology, DH-05, University of Washington, Seattle, Washington 98195. EDITORIAL BOARD BRENT BERLIN, Department of Anthropology, Columbia University, New York; ethnotaxonomies, linguistics. ROBERT A. BYE, JR., Department of Environmental, Population and Organismic Biology, University of Colorado, Boulder; ethnobotany, cul- tural ecology. RICHARD S. FELGER, Senior Research Scientist, Arizona-Sonora Desert Museum, Tucson; arid land ethnobotany, desert ecology. RICHARD I. FORD, Director, Museum of Anthropology, University of Michigan, Ann Arbor; archaeobotany, cultural ecology. B. MILES GILBERT, Adjunct Research Associate, Division of Vertebrate Paleontology, University of Kansas, Lawrence; zooarchaeology. TERENCE E. HAYS, Department of Anthropology and Geography, Rhode Island College, Providence; ethnobotany, ethnotaxonomies. RICHARD H. HEVLY, Department of Biological Sciences, Northern Arizona University, Flagstaff; archaeobotany, palynology. EUGENE HUNN, Department of Anthropology, University of Washington, Seattle; ethnotaxonomies, zooarchaeology, cultural ecology. HARRIET V. KUHNLEIN, Division of Human Nutrition, University of British Columbia, Vancouver; ethnonutrition. GARY P. NABHAN, Meals for Millions Foundation, Tucson; cultural ecology » plant domestication. DARRELL A. POSEY, Center for Latin American Studies, University of Pittsburgh: ethnoentomology, tropical cultural ecology. AMADEO M. a Curator of Birds and Mammals, San Diego Museum of Natural History; , cult ural ecology. i of Ethnobiology i is published semi-annually. Menecpty: oe ee and information for the “News and Comments” as on the inside to the oe td back cover of this issue. © Center for Western Studies, Inc. i a Journal of Ethnobiology Proceedings of the Fourth Ethnobiology Conference 13-14 March 1981 University of Missouri Columbia, Missouri VOLUME 1, NUMBER 2 DECEMBER 1981 Missouri BOTANICAL Maanen Eiaeza j. Ethnobiol. 1 (2): 193-199 December 1981 EARLY ACCEPTANCE OF WATERMELON BY INDIANS OF THE UNITED STATES LEONARD W. BLAKE Department of Anthropology, Washington University St. Louis, MO 63130 ABSTRACT.—Modern authorities agree that watermelon (Citrullus lanatus) is an Old World plant that probably originated in Africa. Watermelons were grown by the Spanish in the Southeastern United States before 1576 and their presence was noted 50 leagues inland and in the Southwest before the end of the sixteenth century. Over the next 100 years there is increasing historical evidence of their use by Indians over a wide area. Some early accounts indicate that more than one kind of watermelon was grown. It is suggested that there may have been multiple introductions and that the plant was readily accepted by the Indians because it could be successfully cultivated by the same methods used for the native squash (Cucurbita spp.) Pioneer ethnobotanist Melvin Gilmore, writing in 1914 (1977:68-77), was so im- pressed by accounts of his Omaha Indian informants that they had known watermelon “from time immemorial” that he suggested it might have been present in America before the coming of Europeans. This possibility was reinforced for him by reading early historic accounts of the use of watermelons by Indians over a wide area in the United States and by documented instances of the plant’s ability to volunteer. Lyman Carrier (1923:67-78), an agronomist with the United States Department of iculture, was greatly impressed by the early historical evidence of watermelon in America, particularly by LaHontan’s account of the late seventeenth century that this tropical plant had been adapted for cultivation as far north as Canada by the Hurons. Although he mentioned the possibility that seeds of watermelon may have been distri- buted from a Spanish introduction in the West Indies, he concluded that “‘At present the evidence favors the American Indian as the discoverer and domesticator of the edible watermelon’’. Carrier was also of the opinion that the descriptions in the early herbals of “angu- ria” or “citrull”, which are often claimed to be watermelon, really are those of the colocynth or bitter melon. He further claimed that the description of “Virginia Water- melons” in the revised edition of Gerard’s Herbal, published in 1636, but written in 1621, “is unquestionably the true watermelon” (Carrier 1923:65-68). Carrier’s argument was to help support his contention that the watermelon originated in America. It is men tioned here, without passing judgment, because it does give an indication of the approx mate time when the watermelon became generally known and grown in Northern Europe- Gilmore’s and Carrier’s historical research on the use of watermelons by the Indians is extensive and useful, but we have been unable to find publications by other botanists that agree with their conclusions. No direct relatives of the watermelon (C. lanatus) = known to be native to the Americas and, to our knowledge, there is no archaeological evidence of its presence there before Europeans reached the New World. Modern botanical writers generally agree that the watermelon (C. lanatus) originated in Africa and that its use is of considerable antiquity. It has not yet been possible, however, to pin down a specific time and place of domestication (Harlan et al. 1976:145 David 1976:254; Purseglove 1976:294). Whitaker and Davis (1962:2) suggest that there appears to have been a strong secondary center of diversification in India. It may pene coincidence that sandia, the Spanish name for watermelon taken from the Arabic, refers to an origin in India. (Dozy and Englemann 1915:339). 194 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 At the risk of belaboring the point that watermelon is of Old World origin, it may be pointed out that seeds identified as C. vulgaris (now lanatus) were recovered in archaeo- logical excavations from the Inyanga ruins in Africa which date from the eighth century or ig (Shaw 1976:114). addition, the Moor D’Ibn-Al-Awam of Seville, Spain, in his Book of Agriculture ee in 1158, describes six kinds of melons, one of which appears to fit two varieties of watermelon: 6. The melon in the shape of a jar, because it resembles this sort of vessel; the melon of Palestine, which is the melon of Constantinople, the melon of India or the peo includes two er ib the one has a black seed and the rind of this one is very dark green passing to black; ther one has a pure red seed and the green color of it’s rind passes to yellow slams deg tr. 1864, Vol. I:17, 1866, Vol. II, Part 1:216). (Clément-Mullet’ s French translation from the Arabic will be found in Appendix I.) Historical references in this paper are ones in which the English word “watermelon” or the Spanish sandia was used directly or in reliable translation. They also include cases where watermelon was clearly indicated by descriptive terms or statements such as “it is nothing but water”. References in which there appears to be some doubt whether water- melon or some other cucurbit was intended have been omitted. In 1576, a farmer named Juan Serrana testified in a Law court that the soil of Santa Elena Island was good for growing maize, pumpkins and watermelons (sandias) (Connor 1925:159). In 1597, de Salas, a Spanish soldier, traveled 50 leagues inland to the Indian town of Cute and there found Indians growing watermelons (Serrano y Sans 1912:144). This was in the province of Ocute located on the lower reaches of the Ocmulgee River, a tributary of the Altamaha, which had been visited by de Soto (Swanton 1946:138). This indicates that the aboriginal cultivation of watermelon had already begun to move inland. the west, in Mexico near the border of the present state of Texas, Espejo in 1582, reported watermelons among the Indians of the Conchos nation (Bolton 1963: 170). Sixteen years later in 1598, Ofiate declared that watermelons were generally grown by Indians of the Pueblo area in the Southwest (Ibid. 1963:217). There is archaeological confirmation of the watermelon in the Southwest, sometime between 1626 and 1675, by Volney Jones’s identification of two kinds of watermelon seeds recovered from the turkey pens excavated by Toulouse at Abo Mission, southeast of Albuquerque (Jones 1949:30). It has been pointed out above that the watermelon began to be generally known in northern Europe in the first quarter of the seventeenth century. The earliest reference that we have been able to find for watermelons in the Northeast is the 1629 statement of Master Graves (1968:124) that in New ——- ‘‘we abound with . . . sundries sorts of fruits as musk-millions, water-millions . A 1634 report by Father ‘hinted White on his way to Maryland shows that ships from Europe enroute to the colonies sometimes stopped in the West Indies where water- melons had been previously introduced (Hall 1959:38). In the same year, it was reported in a Relation of Maryland that settlers had made trial of “‘Musk-mellons, Water-mellons, Cow-cumbers . . .” (Hall 1959:82). In a 1649 account of New Netherland, the writers Temarked that watermelons could be grown in the fields, but in the Netherlands, they require particular attention in gardens (Brodhead 1856:277). By the period of 1665- 1670 watermelons had reached the Great Lakes, for Perrot (1911:113) noted their use among the Heron at that time and, shortly after in 1673, they were reported among the Illinois by Marquette (1966:129). From the seventeenth century and later there are historical references to the culti- vation of watermelons by Indians from the lower Mississippi Valley north into Canada and from the east coast west to California. Kino, the explorer priest, reported in October December 1981 BLAKE 195 1700 that he saw watermelons growing ‘‘at the foot of a hill, from the top of which, California is plainly visible”, while he was in the country of the Yuma Indians (Bolton 1919, Vol. 1:249). The watermelon appears to have been in use among the Natchez some time before the early part of the eighteenth century, when the French had regular contact with these Indians, for they called June “‘the watermelon moon” (LePage duPratz 1975: 338). (See also Appendix II for additional early historic references to watermelon in the United States.) have been unable to find any reliable mention of watermelon in the North and Northeast before the accounts given above of about 1629-1649, hence transmission to the Huron and the Illinois appears to have taken a relatively short time. The spread to the north was probably hastened by the method in use among the Huron described by Sagarde. Seeds of squash were sprouted by placing them in a box filled with rotted wood, which was moistened and suspended over the smoke of a fire (Kinietz 1965:19). A similar method was described by G. L. Wilson’s Hidatsa informant as late as 1914 (Wilson 1977:68). Charlevoix (1763:237) observed that this method was also used for watermelons. He said, ‘“‘Sun-Flowers, Water-Melons, and Pomkins are set by themselves; and before they sow the Seed, they make it shoot in Smoke, in light and black Earth”. Accounts of the seventeenth and early eighteenth century sometimes indicate that more than one kind of watermelon was grown. It will be recalled that Jones ( 1949:30) recovered two kinds of watermelon seeds at Abo Mission. Marquette (1966:129) said of the Illinois in 1673, “They also sow beans and melons, which are excellent, especially those that have red seeds’. Some watermelon seeds are “reddish black’’ and some are “white, yellow, brown, green, black”. (Whitaker and Davis 1962:38). Marquette’s state- ment appears to indicate that more than one kind was grown by the Illinois. John Banis- ter (1678-1692) spoke of “watermelons red, yellow and white meated” (Ewan and Ewan 1970:350). LePage duPratz (1975) described several kinds of watermelons grown in Louisiana that varied in shape from round to long and in size from 10 to 30 pounds. The seeds of ‘‘Some are black and others red” (1975:230, 231). There is, of course, no certainty that more than one variety reached all Indians or, if they did, that they con- tinued to be grown. Watermelon seeds could have reached the Indians by diffusion from Spanish settle- ments in Florida, Mexico or the Southwest, or possibly from traders or missionaries in th orth or East. The priest Kino did distribute seeds, including those of the water- melon, while in the country of the Pima and the Papago (Castetter and Bell 1942:74), but we know of no documented instances of this practice by early missionaries or traders in the North, Incidently, Kino found that the Indians already had watermelons, eve? those near the present day Cocklebur, Arizona, “although never in that village had there entered another white face or Spaniard” (Bolton 1936:398). Among collections of plant remains recovered from archaeological sites sent to the Missouri Botanical Garden for identification were five that contained seeds of wateT melon (C. lanatus) (Cutler and Blake 1976:13, 14, 45, 46). These are listed with me surements of seeds, cultural affiliation and dates in Table 1. A brief additional statement on each may be in order, however. Zimmerman, located on the Illinois River opposite Starved Rock, was possibly the village visited by Marquette in 1673 (Brown 1975:2). King Hill is located on the Missouri River bluffs in a residential area of St. J oseph, Missouri (Ruppert 1974:2-11). tz is on the south bank of the Missouri River north of Sedalia, Missouri. It was the home of the historic Missouri Indians until about 1714 (Berry and Chapman 1942; Bray 1978). Rhoads is on a tributary of the Sangamon River north of Springfield, Illinois. It w4s occupied by Kickapoo Indians in close touch with the British (Klippel 1973). Coal Pit, a village of the Little Osage near the western border of Missouri was prob- ably the one visited by Pike’s Lieutenant Wilkinson (Chapman 1974:120). While there on A Ay TT ee ee ee 196 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 TABLE 1. — Watermelon seeds from five Midwestern historic and p histori haeological si (Carbonized, unless otherwise noted.) Site name Zimmerman King Hill Utz Rhoads Coal Pit Site number 11Ls3 23Bnl1 23Sa2 11L08 23Ve4 Approx. period A.D. Ca.A.D. Ca.A.D. A.D. A.D. of occupation 1673-1691 1700 1600-1714 1760-1820 1790-1820 Culture or Historic Late Oneota Late Oneota Historic Historic tribe Kaskaskia Kansa (?) Hist. Missouri Kickapoo Little Osage No. of seeds 52 2 2 7 5 Sizes of 7.8 by 4.9 8.8 by 5.1 10.0 by 5.6 9.4 by 5.0 11.0 by 5.6 seeds (mm) 9.0 by 5.6 9.4 by 5.3 9.5 by 4.7 9.4 by 5.2 5 by 5.0 9.9 by 5.0 12.3 by 8.2 (*) 9.9 by 5.2 (*) Not carbonized. 18 August 1806, Wilkinson was given “watermelons about the size of a twenty-four pound shot, which though small, were highly flavored” (Jackson 1966, Vol. II:5). Since a 24 pound shot is only about 14 cm in diameter, these were quite small watermelons. ariety of watermelon has survived up to recent times among some . Gilmore (1977:68) described the old watermelon of the Omaha as “‘small, round and green, having a thin rind we red flesh, with small, black, shining seeds’’. Bohrer (1960:200) described a similar v. grown by the Zuni, which was 17.1 cm in diameter. aia bes and Bell (1942: 119) al es spoke of a small, spherical melon with pink flesh, 20.3 cm in diameter, which was grown by the Pima prea for rapid acceptance of the watermelon by Indians over a wide area, often with different cultures, appears to lie in the fact that methods of me — are similar to those of the native squashes (Cucurbita spp.) with which they w Also, the long, hot continental summers over much of the United States eps its th. tributing mrp were neigh sweet and refreshing taste, which was probably e onotonous, and the excellent keeping a alites of some varieties. Whiting (1939:92) SUBAREA an old Hopi type that kept until mid- February. Storage for winter use was noted by Robbins et al. (1916:112) among the Tewa and by Bohrer (1960:200) among the Zuni. Peter Kalm described methods used by the settlers in Pennsylvania to keep watermelons “during a great part of the winter” (Benson 1966:5 16). Since watermelons appear to have been adopted by the Indians for use fairly early, and since they could be grown locally and did not have to be imported as did the usual trade goods, there would seem to be a good chance that watermelon seeds may turn up On sites where no trade goods are recovered. The increasing use of flotation techniques enhances this possibility. In any event, they should be of particular interest to archaeo- logists for they are as good an indication of the effect of European presence as glass eads and articles of brass or iron. Also, with a greater recovery of seeds, it may be Possible to distinguish archaeologically a number of places where more than one variety was grown. 3 December 1981 BLAKE 197 APPENDIX I Clément-Mullet’s French translation from the Arabic 6. le melon en forme de jarre, parce Qu’il reasemble a cette sorte de vase: le melon de Palestine, qui est le melon de Constantinople, le melon de I’Inde ou du Scinde, comprenant deux variétés; l'une a la grainé noire et (l’écorce) d’un vert trés-foncé passant au noir; l’autre a la graine d’un rouge pur, et la couleur verte de son €corce passe au jaune (Clement-Mullet 1866, Tome II, Pt. 1:216) APPENDIX II Additional early historic references to watermelon in the United States Date Reference 1663 John Josselyn noted watermelons in New England in the account of his second voyage (1865:60,101). 1679 John Nanieier wrote a letter in this year telling about watermelons in Virginia, (Ewan and Ewan 1970:4 1687 Joutel was wate watermelon to eat by the Arkansas and later by the Kaskaskia Indians (1962:143,146,156,163). 1690 De Leon saw watermelons near the present day town of Crocket in Houston County, Texas (Bolton 1963:415). 1691 Casanas noted watermelon among the Tejas and Asinai Indians of Texas (Swanton 1942: 128,243). 1691 Espinosa mentioned that watermelons were grown in the province of Texas (Swanton 1942:243), 1697 Mange said that watermelons were grown at San Agustin de Oiar, which is near present day Tucson, Arizona (Burris 1971:215 1697 Cadillac said that the harvest of the suvhiein Indians included watermelons (1962:12). 1699 Penicaut was given watermelon to eat at the village of the Pascagoulas (McWilliams 1953: 18). 1705 Liette said that the Illinois harvested ‘a great many fine watermelons .. . many of them as big as a water bucket” (1962:126). 1705 a spoke of several varieties of watermelon in Virginia (1947: 14 : 1748- Peter Kalm mentioned watermelons a number of times in noting their presence in Canada 1749 alias North America and in the Illinois Country (Benson 1966: 59,508,509, 515,516,617). LITERATURE CITED BENSON, A.B. 1966. Peter Kalm’s travels in BRODHEAD, J.R. 1856. Documents relative North America. Dover Publ., Inc., New York to the colonial history of the State of New BEVERLEY, R. 1947. The history and present state of Virginia. (L.B. Wright, ed.) Univ. of North Carolina Press, Chapel Hill. BERRY, B. and C. CHAPMAN. 1942. An One- ota site in Missouri. Amer. Antiquity 7:290- 305 BOHRER, V.L. 1960. Zuni agriculture. El Pal- 3 BOLTON, H.E. 1919. Kino’s historical memoir of Pimeria Alta, 1683-1711. Arthur Clark Co., Cleveland. ———. 1936. Rim of Christendom. The Macinillan Co., New York. > 596 Spanish exploration in the Southwest, 1542-1706. Barnes and Noble, New York. BRAY, R.T. 1978. European trade goods from the Utz site and the search for Fort Orleans, Missouri Archaeol. 39:1-175. York. (E.B. O'Callaghan, ed.) Weed, Parsons and Co., Albany, Vol. I. : BROWN, M.K. 1975. The Zimmerman site. Illinois State Mus. Reports of Invest. 32:1- 124. BURRIS, E.J. 1971. Kino and Mange explorers of Sonora and Arizona. Jesuit Hist. Inst., St. Louis Univ., St. Louis CADILLAC, L. 1962. “The memoir of Lamothe Cadillac. Pp. 3-83, in The Western Country m the 17th Century (M.M. Quaife, ed.). Citadel Press, New York. CARRIER, L. 1923. The beginnings of agtt en : America. McGraw Hill Book Co- New Belcan E.F. and W.W. BELL. 1942. Pima and Papago Indian Agriculture. Univ. of New Mexico Press, Albuquerque. CHAPMAN, C.H. 1974. Osage Indians III, the Origin of the Osage Indian Tribe. Garland Publ. 198 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 LITERATURE CITED (Continued) Co., New York. CHARLEVOIX. 1763. Letters to the Dutchess Lesdiguierres: giving the account of a voyage to Canada. Printed for R. Goadby, London. CLEMENT-MULLET, J.J. 1864, 1866. Le livre de agriculture par D’lbn-Al-Awam. Libraire 1925. Colonial records of Vol. I Letters and reports of governors and secular persons, 1570-1577. Deland. CUTLER, H.C. and L.W. BLAKE. 1976. Plants from archaeological sites east of the Rockies. Amer. Arch. Reports No. 1, Microfiche, Amer. Arch. Div., Univ. of Missouri, Columbia. DAVID, N. 1976. History of crops and peoples in North Cameroon to A.D 223- 267, in Origins of African plant domestication (J.R. Harlan, J.M.J. DeWet and A.B.L. Stemler, eds.). Aldine Publ. Co., Chicago. DOZY, R.P.A. and W.H. ENGELMANN. 1915. Glossaire on mots Espagnols et Portugais . John Bantites and his natural history of Virginia, 1678-1692. Univ. of Illinois Press, Urbana. GILMORE, M.R. 1977. Uses of plants by the Indians of the Missouri River region. Univ. of Nebraska Press, Lincoln (hepetnted from the 33rd Annu. Report Bur. Amer. Ethn., Washing- ton, 1919), GRAVES, M. 1968. A letter sent from New England. P. 124, in New England’s plantation, written in the year 1629. Pp. 117-124, in Collections of the Massachusetts Historical Society for the year 1792, Series, Vol. I. 1959. iaeatioey of early Marland, Barnes and Noble, New York. HARLAN, J.R., J.MJ. DeWet, and A.B.L. 1976. Plant seers and indigenous African agricultu Pp. 3-19, in Origins of African Plant pionralienie (J.R. Harlan, J.M.J. DeWet, and ‘ie Stemler, eds.), Aldine Publ. Co., Chic JACKSON, D. 1966. The pase of Zebulon Montgomery Pike, Vol. II. Univ. of Oklahoma “—" Notes on some organic Appendix 2, y J.H. Toulouse). School of Amer. Res. Mo onogr. No. 13, Santa Fe. JOSSELYN, J. 1865. An account of two voy- ages to New England made during the years 1638, 1663. William Veazie, Boston JOUTEL, H. 1962. Joutel’s journal of LaSalle’s last voyage. Corinth Books, New York. (Copy of edition published in London in 1714. KINIETZ, W.E. . The Indians of the western Great Lakes, 1615-1760. Univ. of Michigan Press, Ann Arbor. KLIPPEL, W.E. 1973. Recent native heritage of central Illinois. The Living Museum 35:206- 208 LePAGE duPRATZ, M. 1975. The history of Louisiana. (J.G. Tregle, Jr., ed.). Louisiana State Univ. Press, Baton Route. LIET EE, P: 62. The memoir of Pierre Lictte on the Illinois country. Pp. 87-171, in The Western Country in the 17th Century. (M.M. Quaife, ed.). Citadel Press, New York. McWILLIAMS, R.G. 1953. Fleur de lys and calumet, being the Penicaut narrative of adven- ture in Louisiana. Louisiana State Univ. Press, Baton Rouge. MARQUETTE, J. 1966. Voyages of Marquette in the Jesuit relation, 59. Univ. Microfilms, . 25-272, in The Indian Tribes of the —- Missin Ne and the Region of The Great Lakes Blair, ed.). Arthur H. pone snd ig Coveted. PURSGLOVE, J.W. The origins and mi- grations of crops in tropical Africa. Pp. 291- 309, in Origins of African plant domestication (J.R. Harlan, J.M.J. DeWet, and A.B.L. Stemiler, eds.). Aldine Publishing Co., Chicago. ROBBINS, W.W., J.P. HARRINGTON, and B. FREIRE-MARRECO. 1916. Ethnobotany of the aie rage Bur. Amer. Ethn. Bull. 55, Washin Picimoins ng ME. 1974. Analysis of the verte- brate faunal remains from the King Hill site, 23BN1. Unpubl. M.A. thesis, Univ. of Nebraska, Lincoln. SERRANO Y SANZ, M. 1912. Documentos a re la Florida y Luisiana siglos XVI al drid. SHAW, ie "1976. Early crops in Africa: a review of the evidence. Pp. 107-153, in Origins of African Plant Domestication (J.R. fin, J.M.J. DeWet, and A.B Stemler, eds.). Aldine Publ. Co., Chicago. SWANTON, J.R. 1942. Source material on the history and ethnology of the Caddo Indians. Bur. Amer. Ethn. Bull. 132, Washington 946. The Indians of the South- eastern United States. Bur. Amer. Ethn. Bull. 137, Washington December 1981 BLAKE 199 LITERATURE CITED (Continued) WHITAKER, T.W. and G.N. DAVIS. oni WILSON, G.L. 1977. Agriculture of the Hidatsa Cucurbits botany, cultivation and utilizat Indians. Reprints in Anthropology, Vol. 5, World Crops Books, Interscience Publ. J. and L. Reprint Co., Lincoln, Nebraska. New York. (Reprint of Univ. of Minnesota Stud. in Social WHITING, A.F. 1939. Ethnobotany of the Sci. No. 9, 1917.) Hopi. Mus. N. Arizona Bull. 15, Flagstaff. J. Ethnobiol. 1 (2): 200-207 December 1981 A “LOST” VIKING CEREAL GRAIN LISA CARLSON GRIFFIN State Historical —— of Missouri olumbia, MO 65211 - RALPH M. ROWLETT Department of Anthropology, University of Missouri-Columbia Columbia, MO 65211 ABSTRACT.—Lyme grass (Elymus arenarius), a consfal wild grass of the artic and sub-artic of the northern hemisphere, occurs in carbonized form in Viking archaeological sites, especially in Iceland and Greenland. There is ae an increase in elymus pollen contempo- rary with the Viking homesteads at L’Anse-aux-Meadows, Newfoundland. In Iceland, lyme grass grain was the main source of bread flour until the eighteenth century, when imports from Europe of wheat flour replaced it. Folk tradition in Iceland held that lyme grass bread was both more tasty and nutritious than wheat bread. Comparison of some basic nutritional values of lyme grass grain with some other standard foodstuffs, including ama- ranth, shows that lyme grass has considerable nutritional value for human beings. The most remarkable aspect of the restoration of lyme grass as a foodstuff in the future would be that it would form a cereal crop which can be grown in the artic regions where no other suitable agricultural crop is forthcoming. If the world food shortage becomes more acute in the next few decades, the cultivation of lyme grass would open up millions of acres of food produc- tion INTRODUCTION Several peoples of the world have been recorded as having used lyme grass (Elymus arenarius L.) as a flour. While this grass (Fig. 1) normally grows in circumpolar contexts, primarily near marine environments, Newberry (1857) cites its collection by Pacific coast “Digger” Indians as far south as northern California. It has also been collected by peoples of the northern Soviet Union (Komarov 1963). The greatest users of lyme grass, how- ever, were the Vikings, especially those Norsemen who came to Iceland, Greenland, and Newfoundland. Different folk names are “‘strand oats’ and “strand wheat” in English, Strandhvede (Strand wheat), Sandhavre (Sand Oats), Vild Hvede (wild wheat), Melur and Sand Melgras (Sand Meal-grass) in Norwegian and Icelandic. These folk-names (Fer- nald 1910) reveal a notion of lyme grass bearing an edible grain The evidence for lyme grass from mainland North Aeaciies comes from two sources. The first source is the reference to “‘self-sown wheat” in the Vinland discovery sagas (Magnusson and Palsson 1965: 52). The sig source of evidence is from the excava- tions at L’Anse-aux-Meadows, Newfoundland, where there is a distinct jump in easily Tecognizable Elymus pollen in several zones at hi site, although the genus was in New- fo ig thousands of years before the Norse settlements (Henningmoen 1977). n the pollen at L’Anse-aux-Meadows pond Elymus pollen is present at the deep 160- = cm levels with a C14 date of 3,890 +4110 B.P., but it rapidly drops out of the record at two loci. It picks up again later, at the 60-80 cm levels, roughly at the position of the Norse occupation, judging from the date cited, and another date at 2,000 B.P. obtained from the column. A third sample has: Elymus only at the 60-70 cm zone where the radiocarbon date falls between 1,130 to 1,450 B.P. at a 95% confidence level. December 1981 GRIFFIN AND ROWLETT 201 FIG. 1—Glume-covered and bare grains of lyme grass (Elymus arenarius). Even clearer indications come from the Palsa Bog on terraces immediately behind the Norse houses. Elymus is lacking here at the deep 170 cm level with a date of 5,320 = 60 B.P., but is plentiful at the 40 cm level and up; a sample from the 35 cm level has a date of 460 +80 B.P. It appears again at the 65-80 cm level, which should be the zone repre- senting the Norse settlemént according to Henningmoen (1977:314-315). Thirty cm om of House F, Elymus occurs only at the 50 cm level where there is a corresponding 14 date of 1,280-1,680 B.P. in the levels below. 2 From the archaeological features themselves, at House A, North wall, Elymus first appears at the 40 cm level, where the wall is about 4 cm above the old land surface. lt occurs sporadically inside the wall turves, reaching a maximum abundance at the 18 — level on top of the surviving wall, reducing to a minimum presence at the 13 cm level in the soil overlying the top of the turf wall of the house. This suggests the pollen is conce™ trated on the ledge of the standing wall. Also occuring outside the houses is the Euro- pean weed Rumex, known only in modem or very recent pollen deposits (Henningsmoen 1977:328-333). Location of pollen in the southern wall of this house is less clear, ok though the 3 cm deep occurrence of Elymus in the diagram is coded “top of the wall despite the shallow depth. Similarly, the pollen diagram for House F, northern wall, 18 difficult to equate with a published section, but the sole occurrence of Elymus here the 20-30 cm level of the wall has an unambiguous accompany C14 date of 1,050 to 850 B.P., making it clear that the Elymus, like the Rumex, associates with the Viking homestead on Newfoundland. € most intensive use of lyme grass seems to have been in Iceland and Greenland. Historic records imply that at about A.D. 1000 lyme grass was collected from thick w stands to supplement the crops of wheat and barley which sometimes failed this far north. Apparently, as the climate of the northern hemisphere deteriorated by the = Middle Ages and eventually led to abandment of the Greenland settlements and increas¢ 202 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 hardship in Iceland, lyme grass replaced the more normal cereals of Western Civilization in cultivation. Archaeological excavations of Norse settlements in Greenland were done mostly before World War II, so the lack of a fine chronology prevents any glimpses of changing incidences of lyme grass use. As late as 1783 the English botanist Sir William Johnson Hooker (quoted in Fernald 1910) spoke of “Fields used to produce the Sea-lyme grass’’ and noted that devastation by volcanic eruption would have dire consequences for the food supply since very little “corn” (undoubtedly normal wheat grain) was im- ported into the cities and those isolated regions. The somewhat vague evidence from the sagas concerning lyme grain in Greenland finds support in the distribution of wild stands of Elymus in Greenland. Lyme grass in the narrowest sense (Elymus arenarius L. [2n=42] variety villosus) grows primarily on the ruins of old Norse and ancient Eskimo settlements. Other lyme grass (Elymus aren- arius mollis [2n=28]), the Neoarctic variety, is found from Greenland to northwest Asia (Pedersen 1972:76-77; Fredskild 1973) and occurs with less density but more wide- spread coastal distribution. It is still uncertain as to whether or not lyme grass was ever domesticated in the Viking homeland. Certainly it is notably lacking in the bizarre gruels which were con- cocted from cultivated grains, some questionable domestics such as Chenopodium, and weeds. These gruels were consumed as last meals by the time-of-Christ bog bodies such as Graubelle, Tollund and Borremose men (Helbaek 1958; Brandt 1950), who clearly antedate the Vikings by nearly a millennium. Olsson (1974:30-32) notes that lyme grass has spread anthropogenously from the coasts all over southern Sweden, exhibiting both a normal littoral as well as xerophytic varieties. He considers the key ecosystem elements in the Elymus communities to be human agency and nitrophilia. Decline in the consumption of lyme grass began with the rise in commercialism in the sixteenth century. After 1800 with the blossoming international trading system and the greatly enhanced food production in Europe and North America, lyme grass grain was slowly replaced by imports of wheat, rye and barley. This process was encouraged by § regulations from the kingdoms of Norway and Denmark (Né¢rlund 1936), which seeeuhate exercised sovereignty over Iceland. Lyme grass flour was not abandoned completely, however, until early this century (Sigurbjornsson 1960:52). At first consideration, it seems only normal for the Icelanders to prefer wheat flour and the like over lyme grass meal for bread and other bakery products. Despite the factors mentioned above, the replacement of lyme grass by imported flour remains something a mystery. According to some Icelanders, the melur flour from lyme ass was more tasty than wheat flour and more nutritious as well. Hooker, the botanist (in Fernald iin. 27-28), shared this opinion concerning the relative taste and the Ice- landers had not yet perfected the technique of “‘drying and preparing the grain.” Is this folkloric insistence on the superior gustatory and nutritional qualities of lyme grass the mere bling of some conservative diehards? § Hooker’s assertions were strongly put and folklore often contains a grain of truth hidden therein. Rather than dismiss these traditions, we have tested the lyme grass grain for some aspects of its nutritional value. Our motives were not merely to vindicate Vikings, who seldom needed help from anyone, but also included the evaluation of lyme grass as a possible agriculture crop for the arctic, a climatic zone which hitherto has been only seldom utilized for food production. METHODS funding at our disposal, we tested the content of lyme grass seeds for iron, seven ees pas and 17 amino acids (Table 1) at the Experiment Station Chemical Labora- tories of the College of Agriculture, University of Missouri-Columbia. A comparison was December 1981 GRIFFIN AND ROWLETT 203 made between dry, uncooked lyme grass grain and several other prominent foodstuffs including the top four crops (wheat, rice, maize, and potatoes) which produce more ton- nage for the world’s food then do the next most prominent 26 crops combined (Harlan 1976). The amino acid assays were made on hulled, twice-ground samples and were saponified and analyzed by liquid chromatography (Benson and Patterson 1971). Heptacecanoic acid (C17:0) comprised the internal standard used to quantify fatty acids. No indigenous heptodecanoic acid had been found in a trial sample. A computer interfaced with the GLC set was used for identification and quantification of the fatty acids. The iron con- tent was assayed by a nitric perchloric acid wet-ash digestion follows by an atomic absorption determination. TABLE 1.- Amino Acids, Iron and Fatty Content of Lyme Grass (Elymus arenarius). Sample Sample Sample Sample Sample Sample 1 2 3 1 2 3 AMINO ACID Phenylalanine 98 85 1.17 (protein per 100 g) Histidine 44 47 51 Aspartic Acid 94 1.34 1.06 Lysine 51 75 59 Serine 87 81 88 TOTAL Glutamic Acid 5.84 4.78 6.62 PROTEIN 18.23 17.38 20.76 2.58 Z aoe z71 FATTY ACID Glycine 87 88 94 brag pee 166 Al ns ikl . wi te Palmitic 218 1.78 2.04 Cystine 18 12 16 eee 08 ubiakade 56 39 86 Palmitoleic .08 - : ' Stearic 13 .09 14 Ammonia _ — Ol 2.83 9.14 2.74 Arginine 82 100 1001 a ; 90 ; Linoleic 10.10 5.20 10 Threonine 61 .64 65 : , 4.14 is 78 84 94 Linolenic 4.85 6.30 : Methionine 24 23 26 asia rs 0.12 15.51 20.04 Isoleucine 55 61 71 i ; ; Leucine 121 1.16 143 IRON (mg per 100 g) 5.50 5.40 5.80 ae See DISCUSSION value but lyme grass total protein content challenges even salmon. Among the essen amino acids amaranth may have twice as much leucine and lysine as Elymus, although Amaranthus edulis with a high lysine content which provided protein for Aztecs (Ortiz 1978), has a leucine content somewhat inferior to that of Elymus (Downton 1973) even though the absolute content is only slightly higher. Amaranth’s leaves, of course, provide a human and animal foodstuff which though less storable, is even more nutritious. The leaves and glumes of lyme grass would not suitable for human consumption, but make an excellent hay for livestock. According to data compiled by Sigurbjornsson (1960), sheep, beef cattle, and horses thrive on the 204 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 TABLE 2.—Comparison of Lyme Grass With Other Foodstuffs. : : sze 4 «4 § a g E S 7 2 §28222 2 e8ig £2 2 2 2 = 3 FE gani 3s 5 £RE 2 a a : 5 z Py = AMINO ACID (protein per 100 g) Aspartic Acid 1.20 1.38 .60 Serine 85 72 46 Glutamic ci 5.95 2.92 2.338 Proline 2.26 1.15 98 Glycine .90 96 34 A .84 ce 2 Cystine AS 25 30 21 Ammonia — Arginine -94 1.39 93 $1 Threonine 64 99 71 34 ne .84 1.39 .90 48 Methionine .25 .30 al 27 Isoleucine -64 1.35 53 .37 Leucine 1.27 241 1.01 1.35 Ph 1.00 1.28 67 60 Histidine 48 36.66 46 31 ysine 62 1.26 68 20 TOTAL PROTEIN 18.80 14.5 14.6 14.4 93 12.9 8.5 21 14.2 222 20:2 28.6 FATTY ACIDS (mg 100 g) 1366 5 KX %X 48: & Be + B TA 15 125: 164 IRON (mg per 100 g) ‘6 52 8 =< 35 % 13 29S © 45 22) 4, Be (Most non-Elymus values are taken from Watt et al. 1975 Tables 1, 3. Soviet hybrids and wheat are from Ivanovskaya 1960:89; high-protein maize-corn amino acids are from Misra et al. 1972:1426; aranth amino acid means are from Olivera and Carvalho 1975:258.) cut body of the plant, the food value of which far exceeds that of the straw of wheat and other familiar Eurasiatic cereals. Lyme grass does not endure close grazing by live- stock and would not be suitable for a pasture grass, even though it makes excellent fodder as the by-product of grain production. : ; Lyme grass seems to compare with several other crops as poorest in terms of yield (Table 3). One must bear in mind, however, that high yielding crops such as rice, maize, have some potential as to productivity as well as nutritional value. As Sigurbjornsson (1960: 51) points out, the low yield is relatively unimportant, since the crop can be December 1981 GRIFFIN AND ROWLETT 205 TABLE 3.—Yield per hectare of Lyme grass and other crops (world averages). Crop Yield, kg per hectare Crop Yield, kg per hectare Crop Yield, kg per hectare Rice 2,300 Wheat 1,560 Amaranth 683-983 Maize-corn 2,400 Oats 1,660 Lyme grass 600-800 Barley 1,910 Soybean 1,740 (Source Amaranth, Organic Gardening and Farming Research Center and Downton 1973, Tsitsin and Petrova 1952 quoted in Sigurbjornsson 1960:50; others, U.S. Department of Agriculture, modern world averages.) grown on land otherwise useless for agriculture. Jack Harlan (personal communications) assures us, moreover, that this yield is identical to that of domesticated cereals, such as rice, maize, and wheat, under conditions of subsistence agriculture with no soil inputs. In any case, lyme grass has potential as a cultivar in the high artic, north of the ar sho agri regions would add immensely to the world land acreage available for food production at a time when some feel (Brink et al. 1977) that ne is hardly any more new lan for farming. This would include vast areas of the Soviet Union, a region notorious for inadequate grain supplies, as well as millions of acres in North America. In Alaska alone there are an estimated 16 million acres useable for agriculture (Wooding et al. 1974). Figure 2 shows the region of prime potential for lyme grass farming in North America, since stands of wild lyme grass grow there now. Our conclusion then is that the Icelandic folklore and tradition is correct, that it was a highly nutritious cereal and therefore it must have made a considerable contribution to Norse and Viking diets of the North Atlantic region. Also, the potential of lyme grass to ameliorate the shortage of world food supplies emerged from this study, perhaps the most significant conclusion we can draw. Its ability to thrive in marginal climates and its high nutritional values should bring lyme grass to the attention of both botanists and agricul- tural economists. Vol. 1, No. 2 JOURNAL OF ETHNOBIOLOGY 206 (*‘pueypunojmay, pur purjusei5 10J suoNvolyipour WIM dnvy J91Jy) “suOTIPUOD payfoNUOD JapuN payNnuapt ssvrd aurAT JO SUOHPIO[ MOYs SJOp ‘sseIB JUIAT 1OJ IIqQUIINS SJEIIqGeY so}eoIpUT wae padding ‘voLIOUTY YON UI (SsMuA)qZ) sseIB dUIAT JO UOTNGUISIp JUasoid Jo dew—z “OTA Vis yOaVYEVT \ Be {t; %, § punjueei9 ** Ava f NOSQ@NH ea oi VAISVIV fe, D e] December 1981 GRIFFIN AND ROWLETT 207 ACKNOWLEDGEMENTS This research was supported by the hing = Missouri Research Council Grant 1851 (1977). usly submitted by the following laboratories: vn Universitets botaniske Museum, Gr¢gnlands Botaniske Undersogelse, K@benhavn, Newfoundland Forest Research Centre, St. Joh Universitet i Oslo, Botanisk Hage, Oslo, Norway ne nma: es Universite de Montreal, Institut tie Herbier Marie-Victorin, Montreal, Quebec, Canada ada n’s, Newfoundland, Can We have been greatly aided by Sippues comments from Alan Leavitt, Charles W. Gehrke, Larry L. Wall, Klaus O. Gerhardt, Richard A. Sm mand Cornellier, Alexander W. Roberson, pon th, Paul R Wilon, Reed C. exroad, Clair Kucera, Eythor Einarsson, Nor- Rollins, Elmar Marker, C. John Burk, =a Fredskild, Hanna Stazewska, Deborah Pearsall, Per Stormer, Jack R. Harlan, J.M.J. de Wet, wa David Griffin LITERATURE CITED BENSON, J.V., JR. Ju, PATEE sey 1971. New teiaess in Amino Acid, Pep and Protein Analysis. Ann Arbor Science oe > £950. Planterester fraet moselig fra aeldre Jernalder (Borremose). Arboger for ewe Oldkyndighed and His- torie: 342-34 BRINK, R.A., = W.D ene on : ne HILL. 1977. Soil sane Growing World Demand for Food. Brees a 625-630. DOWNTON, W.J.S. 1973. Amaranthus edulis: A og Lysine Grain Amaranth. World Crops 25: gous ite M.L. 1910. Notes on the Plants of Wineland the Good. J. New England Bot. Club 12:7-38. FREDSKILD, BENT. 1973. Studies on Vegetational History of Greenland. Meddelser om Grgnland 198:No. 4 1-247. HARLAN, J.R. 1967. A Wild Wheat Harvest in Turkey. Archaeology 20:197-201. 6. The Plants and Animals that Nourish Man. LEN Comparative Evolution of PRIC 73 Cereals. Evolution 27:311-325. HELBAEK, 1958. Grauballemandens sidste Maltid. Kuml: 83-116. HENNINGSMOEN, KARI EA. - :3977, ae America (Anne S. — ed.). Universitets- forlaget, Olso-Bergen-Tromso. IVANOVSKAYA, E.V. “1960. The Method of Raising Embryos on an Artificial Nutrient Medium and its Application to Wide Hybridi- Pp. 83-187 in Wide Hybridization of National Science Foundation, Washington, D.C. KOMAROV, L.V., ed. 1963. Flora of the U.S.S.R., vol. II. Israel Program for Scientific Translation, Ltd., Jerusalem. MAGNUSSON, MAGNUS and HERMANN PAL- SSON, translators. 1965. The Vinland Sagas Peguin Books, Baltimore. gore PREM S., RAMAMURTHI JAMBUNA- HAN, EDWIN T. MERTZ, DAVID V. GLO- i HELIO BARBOSA and KENNETH S. McWHIRTER. 1972. Endosperm Protein Increa' a 1857. Pacific Railroad Report oh Yam acon, POL. 1936. Viking Settlers in —— Cambridge University Press, Lond OLIVERA, J. SANTOS and M. FIDALGO DE CARVALHO. 1975. Nutritional Value of — Edible Leaves Used in Mozambique. Eco Studies on South Swedish Sand ere Acta oe graphica Sueica, vol. 60:1-17 ORTIZ TELLAN eS BERNARD © R. 1978. Aztec Homarer ie An_ Ecological Necessity? Science, 200:611-617 PEDERSEN, ANFRED. 1972. _Adventtious ~ 196 ; the pasty Elymus. Ph.D. Cornell Univers! University Microfilms, Ing. Ann Arbor, Mich- i TSITSIN, N.V., Ed. 1960. Wide Hybridization in Plants. Akademiya Nauk SSSR Translated from Russian sayy ational Science Founda- tion, ee MILLER. 1975. Handbook o tional Contents of Foods. Dover Larne 8 New York. (Reprint of U.S. Handbook No. 8, 1963. WOODING, Oe 1974, Small Grains on Agricul- in Remote Areas of Alaska. Agro borealis et & THE ANU AND THE MACA TIMOTHY JOHNS Division of Biological Sciences and Museum of Anthropology University of Michigan, Ann Arbor, MI 48109 ABSTRACT.-—Tropaeolum tuberosum, anu, and Lepidium meyenii, maca, are cultivated in the Andes mountains for their edible underground parts. Cultural and medicinal associa- tions between the plants are supported by their similarity in secondary chemistry, and by the pharmacological properties of the isothiocyanates released upon hydrolysis of the glucosinolates present. J. tuberosum has been reported to contain p-methoxybenzy] glu- cosinolate; L. meyenii is reported here to contain benzyl and p-methoxybenzy] glucosino- lates. The likelihood that human selection for specific flavor and medicinal properties has altered the secondary chemistry of, at least, the amu raises questions concerned with both human taste perception and plant domestication. INTRODUCTION The relationship between the use of plants for food and medicine and the chemical constituents of these plants involves a combination of biological and cultural factors. € choice of particular plants in many cases reflects their obvious and well studied nutritional and pharmacological properties (Arnason et al. 1981). Where plants are used in ways which have no apparent western scientific basis, their use is thought to be prin- cipally of symbolic and cultural significance (Ford 1981). Needless to say effective medicinal agents have considerable cultural significance as well; plants may possess Properties which are not yet defined by western scientific methods. The question of the initial discovery of empirically used plants is intriguing (Ford 1981). The means by which humans (and animals) perceive beneficial or harmful con- stituents in relation to physiological homeostasis is interesting but not well understood. Plants of solely ritual and mythological importance must be considered in the broader context of human spiritual and social values. However the association of the western empirical and the cultural values of plants with their chemical constituents are both likely, Physiological perception primarily through taste and smell provides the basis for such association. Lévi-Strauss (1966) discusses briefly the systematization of such sensory data. Although his chemical treatment is rudimentary, he provides insight into the processes by which primitive man might form structures that are ultimately uncovered by science. The differentiation of plants on the basis of organoleptically detectable physical pro- Perties and the translation of perceptional differences into culturally important cate- the o be a combination of both empirically definable and cultural concepts. Both plants con- tain similar constituents which are readily detectable by taste and which have a physical basis of action in many cases. December 1981 JOHNS 209 The anu and the maca,in the Tropaeolaceae and Brassicaceae respectively, are two species of plants from Andean South America cultivated for their edible underground parts and for their medicinal uses. The anu is presently known from southern Venezuela to northern Argentina although use of the tuber as food is relatively localized in compari- son to other tuber crops such as potato (Montaldo 1977). The aziu grows best between 2500 and 3700 meters above sea level. The varied medicinal uses of the plant have been summarized recently and their efficacy in many cases has been substantiated (Johns et al. 1981). The maca is more restricted in distribution. Although it may have been more widespread at the time of the Spanish conquest, it is presently cultivated for its edible root in the Departments of Pasco and Junin, Central Peru, between 3500 and 4000 meters above sea level (Leon 1964). Historically the plants probably grew sympatrically over a much wider geographical range than at present. These two plants correspond strikingly in terms of the historical and modern folk beliefs associated with their putative effects on human reproductive potential. These beliefs correspond in turn with the similarities in phytochemistry. Although the two amilies, Tropaeolaceae and Brassicaceae, are usually classified quite separately by sys- tematists, they are both typified by having glucosinolates, the mustard oil glucosides, as their major secondary metabolites. The numerous reports on the supposed effects of anu in enhancing female fertility and as an anti-aphrodisiac and anti-reproductive agent in males have been summarized (Johns et al. 1981). References to the maca are more scarce; Leon (1964) provides the most accessible and recent overview of its biology and ethnobotany. It is reported by the chroniquilists in the time of the Spanish conquest that the Indians recommended feeding maca to domestic animals to combat low reproductive rates at high altitudes, and that the Spanish noticed the positive effects. Leon (1964) reports that maca is now eaten by Indian and white women who want to have children. It is sold in the market for this purpose. More recent visitors to the area around Lake Junin (Michael F. Brown, Jefferey Parsons, Kent V. Flannery, personal communications) report that belief in the fertility effects are widespread. However the fact that the belief applies particularly to male fertility seems to contradict the beliefs listed previously. Maca may be eaten fresh at the time of harvest, but is more commonly dried for long term preservation. It is prepared similarly for both food and medicine. Dried roots are cooked in milk and/or water and are served either in the cooking liquid with perhaps 4 little sugar, or in a cocktail with aguardiente. Anu is usually boiled before use and retains much of its characteristic flavor. It is occasionally preserved in a drying process similar to the production of chuno from potatoes; in Bolivia this product is known as taiacha (Fernandez 1973). The effects of preparation on medicinal properties or chem! cal constituents are unknown, but in the case of maca and boiled a7iu they appear inslg- nificant. The glucosinolates characteristic of both of these plants undergo enzyme hydrolysis upon damage of the tissue and release the volatile and distinct tasting isothiocyanates, or mustard oils (Fig. 1). These are the compounds responsible for taste in cruciferous vegetables (Macleod 1976); they are, as well, biologically active (Benn 1977). A variety of naturally occurring isothiocyanates (and parent glucosinolates) are known. These os be distinguished chemically on the basis of side-chains. Although they all have the sharp taste of mustard, they are also distinguishable by taste to some extent. Tropaeolum tuberosum has been differentiated into a wild and a cultivated sub- species. This classification is supported chemotaxonomically (Johns and Towers 198 1). The obligate cultigen, subsp. tuberosum contains only p-methoxybenzy! isothiocy anate (Fig. 2). The wild subsp. silvestre is characterized by benzyl, 2-propyl, and 2-butyl ixothiocyanates (Fig. 2) (Kjaer et al. 1978; Johns and Towers 1981). i The literature contains no reports of phytochemical studies of Lepidium meyentt. Lepidium species, as members of the family Brassicaceae, are known to contain glucost nolates. Species from other parts of the world have been studied and found to contain a 210 JOURNAL OF ETHNOBIOLOGY Voli, No. 2 N-OSO3 H20 N-OSO3 oo S n-c” . SR : ‘“s “Glu 7 ser a4 NCS + $04 myrosinase m isothiocyanate glucosinolate Cg H120¢6 FIG. 1—Formation of isothiocyanates by enzymatic breakdown. R-N=C=S 2-propyl R=(CH3)2-CH- 2-butyl R=C2H5~CH- CH3 benzyl! R= CH2- = 1 R= p-methoxybenzy naco{ ore: FIG. 2—Isothiocyanates of T. ropaeolum tuberosum and Lepidium meyenii. variety of glucosinolates including benzyl glucosinolate reported above from 7. tubcro- sum. Variation comprising alkyl and alkenyl derivatives occur within the genus although aromatic glucosinolates, with or without hydroxy and methoxy substitutions in m- and p- positions, prevail (Kjaer and Wagniere 1971). 3,4,5-Trimethoxybenzy! glucosinolate occurs in L. sordida A. Gray (Kjaer and Wagniere 1971) and L. hyssopifolitum Desv. (Kjaer et al. 1971). The only species studied from South America, L. bonariense L. from Argentina, is reported to contain p-hydroxy and p-methoxybenzy] isothiocyanates. MATERIALS AND METHODS Roots of L. meyenii collected in Wayri, Department of Junin, Peru on July 15, 1973 by Michael F. Brown and subsequently preserved in p-dichlorobenzene and deposited at room temperature in the Museum of Anthropology, University of Michigan were exam- ined in 1980. Isothiocyanates liberated enzymatically from ground root material (7 g) were studied using the methods described previously for T. tuberosum (Johns and Towers 198 1). Extracts were examined for isothiocyanates, thiocyanate ions and cyclic oxazo- lidinethiones (Ettlinger and Thompson 1962). RESULTS Lepidium meyenii gave a negative test for thiocyanates and cyclic oxazolidine- thiones. Therefore the plant does not contain p-hydroxybenzy] isothiocyanate. Paper chromatography (PC) of thiourea derivatives using a solvent system of benzene-ethanol- water (5:1:2) (Ettlinger et al. 1966) showed only one spot corresponding to benzy] isothiocyanate. Reverse phase High Performance Liquid Chromatography (HPLC) showed one large Peak Corresponding to benzyl or p-methoxybenzyl isothiocyanates, and one smaller unidentified peak. By normal phase HPLC this sample was resolved into four peaks. The largest of these corresponded to benzyl isothiocyanate and a smaller one to p-meth- oxybenzyl isothiocyanate. The area of the ‘benzyl’ peak in reverse phase was 63%, while December 1981 JOHNS 211 the area of the combined benzyl and p-methoxybenzyl] peaks in normal phase was 65%. This rough measure supports the supposition that the two peaks were resolved from the major peak in the reverse phase system. The identity of the two other peaks remains unknown, PC and HPLC data combined indicate that L. meyenii contains benzyl isothiocyanate as its principal isothiocyanate and p-methoxybenzy] isothiocyanate in relatively smaller amounts. Because only one thiourea spot was seen by PC the unidentified spots on HPLC are likely not isothiocyanates. Until more samples are examined and other methods of analysis can be used to confirm the identity of the compounds present, these results must be viewed as preliminary. DISCUSSION The parallels between the anu and the maca as agents affecting fertility appear more than coincidental. Reproductive rates are indeed lower and a concern at high altitudes (Sobrevilla et al. 1968; Buck et al. 1968) and folk beliefs associated with fertility are to be expected. However, the association of two glucosinolate-containing ‘root’ crops with this concern is highly suggestive. Chemical analysis shows that both plants are character- ized by aromatic glucosinolates. ey appear to both produce p-methoxybenzyl isothio- cyanate while the maca also produces benzyl isothiocyanate. At least in the conception of Andean peoples there is a relationship between aromatic isothiocyanates and human reproductive processes. The overlap of constituents between the two plants suggests that association may be as specific as between p-methoxybenzy] isothiocyanate and human reproduction. The use of T. tuberosum subsp. tuberosum, and aviu, to negatively affect male reproductive processes was supported by pharmacological studies with rats (Johns et al. 1981). The mechanism for this activity, while apparently indirect, supports the empirical use of the plant by Andean peoples. The proposed mechanism is likely to account for similar effects for the maca, as well as for any isothiocyanate containing plants, whether they be aromatic or not. Therefore, although the use of isothiocyanates in general has a western scientific basis, the specific emphasis on aromatic isothiocyanates seems culturally determined. It may be strictly coincidental that both ‘root’ crops contained these compounds, and that associations were easily drawn by people familiar with both plants. However, studies on the botanical origin of the cultigen, 7. tuberosum subsp. tuberosum, indicates that the situation is not so straight forward (Johns and Towers 1981). Although a hybrid origin of the cultigen from the wild taxon is likely, this process has resulted in the = placement of the three constituents of subsp. silvestre with p-methoxybenzy! isothio- cyanate. Although this is conceivable without human intervention, in light of the selec- tion that has gone on in producing the afiu, a plant cultural artifact (Ford 1980), it seems likely that human selection for the particular chemistry of the cultigen has played a role. Such selection with regards to cultural concerns of flavor and medicinal use underlines the association of the avu and the maca. The origin of Lepidium meyenii is not known, and whether it has as well been chemically selected by humans is an intriguing question. CONCLUSION Considerable work remains to explore fully the exciting implications of this associa- tion between the maca and the a7iu. Ethnobotany has traditionally linked the interests of botanists and anthropologists. Problems of this sort require the collaborative efforts of phytochemists and anthropologists as well as of physiological psychologists. a investigations are necessary to understand the biological aspects of human perception ferentiated and the process by which perceived chemicals are interpreted in relation to human chemical taxonomies. 212 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 ACKNOWLEDGEMENTS I would like to thank Prof. G.H.N. Towers, University of British Columbia, for support and advice with the experimental portion of this paper, and Prof. R.I. Ford, University of Michigan, for his helpful interpretative insights. I am grateful to the Natural Sciences and Engineering Research Council Canada for support in the form of a Post- graduate Scholarship. LITERATURE CITED ARNASON, T., R.J. HEBDA, and T. JOHNS. JOHNS, T., and G.H.N. TOWERS. 1981. Iso- 1981. Use of plants for food and medicine by thiocyanates and thioureas of Tropaeo native peoples of Eastern Canada. Canadian t sum from Andean South America. J. Botany: In pre Phytochemistry 20: In pre BENN, M. 1977. ‘Ginssnciiies Pure Appl. KJAER, A., J.O. MADSEN, and Y. MAEDA. Chem. 49:197-210. BERLIN, B., D.E. BREEDLOVE, and P.H. 1974. Principles of Tzeltal plant Classification. Academic Press, New York. UCK, A.A., T.T. SASAKI, and R.I. ANDER- SON. 1968. Health and disease in four Peru- vian villages. The Johns Hopkins Press, Balti- re. ETTLINGER, M.G. and C.P. THOMPSON. 1962. Studies of mustard oil glucosides (II). Final Report Contract DA-19-129-QM1689, U.S. Dept. Commerce AD 290 747. ETTLINGER, M.G., A. KJAER, C.P. THOMP- SON, and M. WAGNIERES. 1966. Veratryl isothiocyanates, a new mustard oil from Helio- Philia longifolia DC (Cruciferae). Acta Chem. Scand. 20:1778-1782. FERNANDEZ, J. 1973. Sobre la dispersion meridional de Tropaeolum tuberosum R.&P Boletin de la Sociedad Argentina de Botanica. . 1980. ‘Artifacts’ that grew: their roots in Mexico, Early Man 2:19-23. I. 1981. Ethnobotany in North America: An historical a Ti me Nega per- spective. Canadian JcB , T., W.D. KITTS, F. NEWSOME, and G.H.N. TOWERS. 1981 Anti-reproductive and other medicinal effects of Tropaeolum tuberosum. J. Ethnopharmac.: In press. 1978. Seed volatiles within the family Tro- paeolaceae. Phytochemistry 17:1285-1287. KJAER, A., and A. SCHUSTER. 1968. Gluco- sinolates in Lepidium bonariense L. Phyto- chemistry 7:1663-1666. KJAER, A., A. SCHUSTER, and R.J. PARK. 1971. Glucosinolates in Lepidium species from Queensland. Phytochemistry 10:455- 457. KJAER, A., and M. WAGNIERE. 1971. 3,4,5- Trimethoxybenzylglucosinolate: a constituent of Lepidium sordidum. Phytochemistry 10: 2195-2198. LEON, J. 1964. The “Maca” (Lepidium meye- nii), a little known food plant of Peru. Econ. Botany 18:122-127. LEVI-STRAUSS, C. 1966. The ‘Savage Mind. Weidenfeld and Nicolson, London MACL J. 1976. Volatile —— com- A eds.). Academic Press, New York. 1 Cultivo de raices y agree tropicals. Instituto Interamericano de oe s Agricolas de la OEA, San Jose, Costa Rica. soaeevint A L.A., I. ROMERO, F. KRUGER, ieee MBURY. ge Low estrogen excretion d pregn: at ia altitude. Amer. J. Obst. ae 120: alien J. Ethnobiol. 1 (2): 213-220 December 1981 NUTRITIONAL CONTENT OF SELECTED ABORIGINAL FOODS IN NORTHEASTERN COLORADO: BUFFALO (BISON BISON) AND WILD ONIONS (ALLIUM SPP.) ELIZABETH ANN MORRIS Department of Anthropology, Colorado State University, Ft. Collins, CO 80523 W. MAX WITKIND U.S. Corps of Engineers, Little Rock, AR 72200 RALPH L. DIX Department of Botany, Colorado State University, Ft. Collins, CO 80523 JUDITH JACOBSON Department of History, Colorado State University, Ft. Collins, CO 80523 ABSTRACT.—An examination of the nutritional content of wild onions and of bison meat is an outgrowth of rari ie research conducted by Colorado State University. Archa- ecological aie’ in north-central Colorado indicate more than 10,000 years of occupation er hu unting and gathering ate American Indians. The sites are located at nee g from over 12,000 ft. to less mck 4,000 ft. in the Rocky Mountains and in th of the Great Plains. Typically they consist of complexes of stone tools : association with fireplaces, and sometimes ceramics, stone rings and bone debris. The pre- historic occupants are interpreted to be nomads harvesting wild foods according to their seasonal availability. Ethnographic ares ea include the Ute and Arapahoe tribes. Efforts to examine the nutritional content of wild plant and animal foods have yielded interesting results. Wild onion (Allium), and buffalo meat (Bison) data are presented with comments a where available. Conditions and a effects of preservation are discussed. These foods are evaluated in terms of omni minimal daily requirements. Data collection strategy and directions for further research are indicated. INTRODUCTION For e than ten years faculty and students at Colorado State University, Fort Collins, see been engaged in archaeological research in north-central Colo rado under the direction of Elizabeth Ann Morris. The geographic area involved is the ecotone between — front range of the Rocky Mountains and the high Great Plains of the eastern portion of the state. During the early years of our research our concern was locating and prin v6 ai prehistoric sites in this region. We are developing a chronology glo P utilizin oa ocarbon dates, stratigraphic depositional occurrence of taxonomically tinct act types, and the rare occurrence of historic trade goods. We also are accumu lating poets pattern data, and learning to recognize functionally different site Se Additionally, we hypothesize aspects of the prehistoric life style based on “hard” evi- dence of the material culture and “soft” inferences from observations on ethnographic eoples. To this end, we have surveyed h igs of sites and excavated several that promised to be informative with regard to the viously mentioned research goals. Our d ata consist largely of locational data for ees sites, artifacts of stone and rarely other materials, and diverse midden debris including bones of the animals that were eaten. The reconstructed life style of the people hypothesizes at least 11,000 years of semi-nomadic wandering by eee or extended families and probably bands. These groups moved about an area familiar to them, camping by streams and springs, and har- vesting seasonally available plant om animal foods. In the warm summer months they —= Sl SES TS Ce -—s SS 6 hh) | Se ee re ee Oe ee ee ee ee nn ee ee a ee 214 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 would move to the forests and tundra of the mountains to fish and hunt and collect the berries and greens of high altitudes. In the fall they would return to lower elevations and some of them would participate in one or more buffalo hunts. Whether the animals were driven over a cliff or into a man-made or natural trap, large numbers would be eaten on the spot or, importantly, preserved for the cold, lean winter months in the form of jerky or pemmican. The winter months would be spent, by a least many groups, in the dry foothills-plains border in the rain shadow of the Rocky Mountains where snow fell less often than on the western slope or the high plains, and where the warm dry “‘Chin- nok” winds would ameliorate the cold temperatures. The coming of warmer tempera- tures and the longer days of spring caused greening of the countryside, to the relief of the man and animals alike, and the beginning of a new year’s food supply. In recent years our interest in the details of their subsistence economy has intensi- fied. The animal bones in the trash deposits at archaeological sites are a factual indication of which species were eaten. Bones have not been preserved in all of the sites that we have excavated but where they have been, they have been predominately buffalo (Bison) (Kainer 1976; Metcalf 1974; Morris and Kainer 1975; Morris et al. 1979; Ohr et al. 1979; Witkind 1971). There is one notable exception, the Owl Canyon rock shelter where the people ate rabbits almost exclusively (Burgess 1981). Some rabbit bones occur in most sites, as well as other animals such as deer, antelope, and canid. However, deer is second most numerous after buffalo in the bone producing sites. Additionally, the bones in most sites do not represent many diverse animals, but are of these few species suggesting a cultural preference. Apparently, there was often no desire or need to eat other available meat such as porcupine, badger, prairie dog, bear, lion, bobcat or birds. The Packrat Rockshelter (5LR170) is an interesting exception with a very diverse fauna represented (Emslie 1981; Morris et al. 1981). Interestingly, evidence for the cannibalisation of Homo sapiens is also absent (Garn and Block 1970; Wurf 1976). Research on tet nutritional content of bison meat has been presented elsewhere (Witkind and Morris In Pre ep.) and we will present only a brief summary here. Witkind (1971) describes he patti buffalo jump and presents the original form of the nutri- tional content data. Bones in the site represented both adult and immature animals. After a considerable search, a freshly deceased bison was located that had not been fed a supplementary diet. Meat samples from five different muscles of a range- -fed 500 lb. bull calf were analyzed for their caloric, protein, and moisture content and for the amount of eight nutritionally important trace elements. These were calcium, phos- Phorus, sodium, potassium, magnesium, zinc, copper and cobalt. The samples tested contained 67% to 73% moisture. Drying the meat to make jerky would not only keep it from spoiling but it would reduce its weight by two-thirds to three-quarters. This would be an important consideration for people who carried their supplies or depended on dog travois to move all of the belongings that they did not cache. Only moisture is eliminated in drying; the fat, protein, minerals and trace ele- ments remain. In the five muscles tests, 4-9% of the fresh weight was fat. The fatness of an animal would vary a good deal with season of the year and the richness of its Pasture. From 25-29% of the fresh bison meat samples was protein. Clearly, buffalo meat is mostly water and protein. The total caloric content of the samples varied from 83 cal/g in shank meat to 2.14 cal/g in loin meat. The great range reflects the variation in the amount of fat (with its high caloric content) in the individual samples. DISCUSSION Almost all of the eight trace elements (calcium, phosphorus, sodium, potassium, Magnesium, zinc, copper and cobalt) were present in greater quantities in buffalo hump meat than in the other muscles tested. It is interesting to note that hump meat was so Prized by Indians and early white settlers alike on the plains. It not only tasted better, December 1981 MORRIS ET AL. 215 but it was better for them. Additionally, buffalo meat contained much more per unit volume of two trace elements that were tested in common in the onion sample that is discussed below. They are phosphorus and sodium. Copper, zinc, calcium, magnesium and cobalt are apparently present in onions in greater quantities per unit volume than in buffalo meat (Table 2). This may be a sampling error, in which dirt adhering to the plants was included in the analysis. In summary, if we consider the nutritional content of jerky made from buffalo hump meat, less than one-tenth of a pound of jerky would supply the minimum daily require- ments of phosphorus, sodium, potassium, magnesium, zinc and copper. From five- to seven-tenths of a pound would supply enough calcium and 1.2 Ibs. would supply enough calories. In view of the fact that bison bones are so often prevalent in local archaeological sites, it is interesting to compare the amount of meat available from a bison compared to the amount available from other animals. A selected sample indicates that the total weight of buffalo bulls ranges between 1,600-1,800 lbs., and buffalo cows range between 800-900 Ibs. A whole bull elk weighed by the Wyoming Game and Fish measured 529 Ibs., approximately half the weight of a buffalo cow and a third of the weight of a buffalo Deer and antelope are even smaller. Wild onions (Allium spp.) were selected as an example of a vegetable food available in the mountain-prairie ecotone. Samples were arbitrarily picked because they grew commonly in the site area during the summer of 1980, and because they were easily recognized by the non-botanists on the crew. Ralph Dix, Professor of Botany, designed the sampling procedure and trained Judy Jacobson, historic preservation major, to do the field observations and laboratory analysis. We have no concrete evidence that any of the specific cultures whose remains we recovered ate onions but it seems possible if not likely. In central Wyoming, Frison (1975:396) reports the outside husks of onions covering a rock shelter floor that had a Middle Archaic occupation. Our site, Pack Rat rock shelter, also contains Middle Archaic remains, but we have no perishable items preserved (Morris et al. 1981). Harrington (1967:345) describes onions being used to flavor soups, stews and meats. Eight species of Allium are mentioned by Rogers (1980). The bulbs and sometimes leaves of all can be eaten fresh or cooked, and they store well. The juice boiled down to a syrup is good for coughs and colds. Allium spp. were col- lected in spring and early summer by the Gosiute Indians of Utah, but they were not saved for winter use (Chamberlin 1911). Weiner (1972:74) also mentions that the Dakotas and Winnebagos applied crushed onions to insect bites to ease the pain and wages for a scalp massage, to strengthen the heart and restore sexual potency ! lf dote. Students and staff alike became extremely sensitive to the differences between Death Camus and wild onions as the Indians must have been in their time. The study area was divided into four parts, because preliminary inspection indicated that the differing vegetational cover might effect the frequency of occurrence of onions in each. These areas were grassland (short grass prairie), shrubland, gully, and rocky slope. All four vegetational areas contained some plants in common. Sandy soils varied only slightly from area to area. Differing density of plants, moisture availability and slope varied between the areas and are thought to be major factors affecting the varying dominant vegetation and the onion occurrence. Thirty 1 Squares were sampled in each of two areas, the shrubland and the grassland. The location of each was determined by throwing a 1 m stick and measuring off a 1x1 m square where it landed. Twenty-five squares on the slope were sampled and 15 squares on the gully floor. The frequency of Allium plants per square ranged from 216 JOURNAL OF ETHNOBIOLOGY Vol. 1,No. 2 a minimum of .26 bulbs in the grassland, to 1.76 plants per square in the shrubland. Total number, partly affected by the differing sample size, ranged from eight in the gully to 214 onions in the shrubland. In summary, Allium bulbs were numerous within a few hundred meters of the site, and tended to concentrate in shrubland and slope areas where they were not crowded out by the grasses. Other nonsystematic counts of onion fre- quency were also made and all observations confirm the high frequency of onions in the area in June 1980. Onions were removed from five quadrants in each vegetational zone. The bulbs were measured and weighed, with slight differences per zone observed. The average weight of each whole plant was 1.00 g. The average weight per bulb was .39 g. Shrubland and slope areas not only produced many more Allium plants, but they were slightly larger bulbs than those in the grassland and gully areas. In an effort to determine the nutritional content of Allium a sample of whole onion plants was sent to the Raltech Scientific Service in Madison, Wisconsin. The results were presented in terms of g/100 g or, they can be thought of as percentages of any volume or amount. Moisture 67.9 g/100 g (or %) Protein 2 gor% 0.4gor% Ash 2.6 gor% Crude fiber 6.1 gor% Carbohydrates 20.8 g or % There were 95.6 cal/100 g of onions, or remembering that the average weight per fresh onion plant was 1 g—each plant would have .956 cal, as well as minute amounts of fat, protein and ash. Any size of fresh onion sample would be just over two-thirds water. The ash was analyzed for its content of the following important nutritional elements: Calcium, Phosphorus, Potassium, Magnesium, Sodium, Aluminum, Barium, Iron, Stron- tium, Boron, Copper, Zinc, Manganese and Chromium. The results were: Mg per onion (Average wt. = 1 g) Mg/100 g onions Calcium 4.3770 437.70 Phosphorus 0.3096 30.96 Potassium 2.7200 272.00 Magnesium 0.4423 44,23 Sodium < 0.1500 <15.00 Aluminum 0.1259 12.59 Bari 0.0246 2.46 Iron 0.0850 8.50 Strontium 0.0030 0.30 0.0035 0.35 Copper 0.0872 8.72 Zinc 0.0523 5.23 Manganese 0.0114 1.14 Chromium 0.0014 0.14 In the interests of learning what wild onions would contribute to the Minimum Daily Nutritional Requirement for humans, Table 1 was compiled. The MDR figures are for both sexes and for individuals at least one year old. It must be remembered that the MDR es were computed in the United States. Other nations, including even Canada, have computed different figures for some of these and other nutritional components. F ermore, it has been widely suggested that peoples of different ethnic backgrounds December 1981 MORRIS ET AL. 217 Table 1.—Minimum Daily Requirement of calories and selected elements and minerals required by humans of at least one year of age. Average computed weight of one onion equals one gram. MDR figures are for the U.S. Element Minimum Daily Amount in Amount in Requirement lg 100 g Calcium 800-1200 mg 4.3770 mg 437.70 mg Phosphorus 800-1200 mg 3096 mg 30.96 mg Potassium 550-5625 mg 2.7200 mg 272.00 mg Magnesium 150450 mg 4423 mg 44.23 mg Sodium 325-3300 mg <.1500 mg 15.00 mg Aluminum No figures 1259 mg 12.59 mg Barium No figures 0246 mg 2.46 mg Iron 10-18 mg 0850 mg 8.50 mg Strontium No figures 0030 mg 30 mg Boron No figures 0035 mg 35 mg Copper 1.0-2.5 mg 0872 mg 8.72 mg Manganese 1.3 mg 0114 mg 1.14 mg Zinc 10-25 mg 0525 mg 5.23 mg Chromium 02-.20 mg 0014 mg 14 mg Calories 2400 560 95.6 —————— may have different minimum daily nutritional requirements. The daily total energy requirement would be greater in any case. However, keeping these possibilities in mind, the figures are presented in Table 1 as a point of reference. It may seem that even if 100 g of fresh wild onions consisting of an average of 100 onion plants, only the MDR of copper and for some people chromium and manganese would be met. However, substantial amounts of the MDR for calcium, zinc, potassium, and iron would be consumed and useful amounts of magnesium as well. Useful amounts of the minimum daily caloric requirement would be taken. Onions contain so little phosphorus and sodium that they provide negligible contributions of these nutrients. le who needed a low sodium diet. In summary, collec: elements but very small portions of phosphorus, sodium, protein, fat, and calories. An abori deficient the resource would quickly be exhausted in any given area. , Analysis of vitamin C content in wild onions is not included in our analysis at this time. However, the U.S. Bureau of Agriculture Handbook 8 lists the ascorbic acid con ent of commercially grown raw onions and whole raw onion plants as ranging from “en 138 mg/100 g (Watt and Merrill 1975). Meats generally contain no Vitamin C and this no doubt would have given added appeal for onions to an aboriginal tribal diet in the spring. If we compare the nutritional value of wild onions to that of buffalo meat, inter esting and not entirely unexpected results emerge (Table 2). JOURNAL OF ETHNOBIOLOGY Vol. 1 No. 2 TABLE 2.— Comparison of the nutritive value per 100 g sample of fresh wild onion and buffalo 7 Ne hump meat. With the exception of the calories, figures may be read as g/100 g, o Ingredient Wild Onions Buffalo Hump Meat Calories 95.60 138.00 Moisture 67.90 g 67.00 g otein 2.20¢g 25.00 g at 40g 5.00 g cium 437.70 mg 2.60 mg Phosphorus 30.96 mg 399.00 mg Potassium 272.00 mg 33.50 mg Magnesium 44.23 mg 17.00 mg Sodium 15.00 mg 76.50 mg Aluminum 12.59 mg -—<—- Barium 2.46 mg ee So Iron 8.50 mg = Strontium .30 mg <= Boron .35 mg Copper 8.72 mg 60 mg M 1.14 mg ee in 5.23 mg 2.50 mg Chromium 14mg eee ee Cobalt 1.20 mg Buffalo hump meat compared to an equivalent weight of wild onions has about the same amount of moisture, much more protein and fat and half again as many calories. Considering comparable measurements of the minerals and trace elements, bison meat contains more phosphorus and sodium than does wild onion, and less calcium, potassium, magnesium, copper and zinc. A diet featuring buffalo meat flavored with onions would supply a high proportion of basic nutritional requirements. able 3 is a selected list of wild foods with their nutritive values as published by the U.S. Bureau of Agriculture (Adams 1975; Watt and Merrill 1975). The items were selec- ted by us for being in the natural undomesticated, unfertilized state but we cannot be Positive of this. Some of the results are interesting indeed and offer useful suggestions for direction for future research. We are interested in pursuing these aboriginal nutritional studies. Some floral data is available already such as the nutritive value of prickly pear cactus, pinyon nuts, yucca plants and numerous wild fruits (Watt and Merrill 1975). Certain other wild animals besides bison have been measured, at least once. Examples are muskrat, beaver, rabbit and caribou. Obvious directions for future research are to assess the quantities available and nutritive value of local plants that appear frequently in the sag oo if not the archaeological record. Ponderosa Pine (Pinus ponderosa) tree products, ring glory root (Ipomea spp.), Indian rice grass (Oryzopsis hymenoides) seed, ‘aaa Sumac (Rhus trilobata) and wild plums (Prunus americana) are easily available. Using buffalo meat and wild onions as initial studies we expect an interesting, and rewarding class of information to emerge in the future. 219 MORRIS ET AL. December 1981 ~ a! = 606 or 070 OF O'S 9aT OL oot Mei‘ ATUO JVIUI URI] “UOSTUS A, ie = = me = 0°0 g" 861 68 G8. oor med ‘U99I8 ‘API L, ae x ie = - 0°0 ae & | GIS S61 ¢$°99 Oot MBI ‘pedy]eeis 10 MOquTes “NOI, = = = 996 we 0°0 Vs 6 61 IOT ag 3 oOoT mei ‘Yoolg ‘no1L, = -. os si 0% rT ToT 06 G6L OOT MBI “TRUS fie = - - 0°0 Pr 3°63 Gas SPs OOT payseor ‘payoos ‘uoos0ey = = sai a = 0°0 O°S OTS SéI OSL 00T mei ‘ATUO YsaTy PIM “IIqqey = = = = =~ L9 = ag SIZ 9LT og9 O0T $121 q18 ‘mes ‘Trend 991 G ¢ 83 06 6°OT T > oP 0°88 0OT med ‘sreadAT41I1d = = oT TZI g Bg oLt L’s Ost 1 3 8S ‘ZO | - paT[ays ‘uoutg ‘synusut is ia ac sie a 0°0 l'? oLS GT $°L9 0oT paiseor ‘payood “yerysnyy TOZT cb 9 OFF 69 S63 Ls 9°8 6ST T'68 PoP “qI T - Mea ‘suroorysnyy oe ca oP T8I IP 80S as 9°8 OFS #8 bob “QI T > Sysny qnoyiM ‘me ‘(satiaqasoo03 -adeo 10 eyod) satiiayopunoiry = is “ = _ 0°0 og ae SéT 8°02 Oot Ajuo ysalg ‘Yond = sie oT 10é LL raat ¢ 9FT ol S38 Oot qayemysaay ‘YstARID 7 “= “i 3 = 1'c6 0 = 8'9ZS—s« OBB 919d 9°9 $oPp “q] [ - PeTfays-sanuyssoq = i = a a 0°0 1°39 Gost Gert 3'°9G PoP “QI T - (pa1seo1) payoos ‘12avag SIPS _ €°S 896 69 So = =#'T SIT ol O'T6 PSP “I IT - y38ua] ‘urT Jo saoatd ojut 7nd ‘mes ‘sJOOYSs OOquIeg PI8T = fit bOS Iter G66 $36 6ST 91 6°98 PGP “q[ [ - SaAvay Mes ‘yjUBIeUTYy su 3ul 3ul But Bu 3u 3 3 SaTIoTer) % sures) (™) (®N) (24) (d) (8D) ajeip Ay Wy ula}01g. ABIQUa «1978 M 131m pue -oqie’) poog sj1un ‘sainsvau a3ewTxoidde ‘poo spoo] jo 11ed a[qIpa 10} sanjeA (SL6T 1MsaW Pur 130M SOLE] Suopy) Suydus ayy us papjosy -UOI JOU SDM S1Y4] ING YuawmyIiua dazyiysaf 40 [10S UsapOU JNOYUM spoof aaiou quasaidas 07 paunsasd St DIDG spoof pajoajas fo sanjpa aat4gnnN—¢§ ATAVL JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 ACKNOWLEDGEMENTS Portions of this manuscript were presented at the 3rd and 4th Ethnobiology Conferences in io es a, and Columbia, Missouri. The authors would like to express their appreciation to Richar Blakeslee, Department of Anthropolo aio Bo Nutrition, Colorado State University, for useful suggestions in early drafts. ogy and to Coerene Jansen, Department of Food LITERATURE CITED ADAMS, CATHERINE F., Value of American Foods in 1975. Nutritive iia wien vet ubl. M.A, Thesis (Anthrop.), Colorado State Univ., Ft. Collins. CHAMBERLIN, RALPH V. 1911. The Ethno- botany of the Goshiute Indians of Utah. Amer. Anthropol. Assoc. Mem. No. 11, Lan- caster. EMSLIE, STEVEN D. 1981. Preliminary Analy- sis of Faunal Remains from the Pack Rat Rock Shelter (5LR170). Report on file the Archaeology Laboratory, Dept. po - ‘Cale tado State Uniy., Ft. Collins FRISON, GEORGE Cc. 1975. Prehistoric Hun- 1976. sige Inve vestigations at the Spring Gulch Site (5LR- 252). . M.A. iS amg ee ), Couo- tado State Univ., Ft. Colli METCALF, MICHAEL D. ie ipso Excavations at Dipper Gap, Lo ounty, Colorado. Unpubl. M.A. Thesis (Anthrop.), Colorado State Univ., Ft. Collins. MORRIS, ELIZABETH A. and RONALD E. KAINER. 1975, The ange Site (5LG122), A Disturbed Bison Kill o South Platte River, a Pa Southwestern Lore 41(1):1 MORRIS, E. A., K. KVAMME, N. T. OHR, M. METCALF, H. DAVIDSON, R. KAINER and R. BURGESS, 1979. The Archaeology of the ceo Project: A Water Control Project r Co North Central Colorado. . MAYO, L N. 1981. Archaeolog- k Rat Rock Shelter ai 1979. The Lykins Valley Site (5LR263): A Stratified Locality on Boxelder Creek, Larimer Rag Interagency Archaeol. ROGERS, DILWYN J. 1980. Edible, Medicinal, Useful, and Poisonous Plants of the Northern Great Plains, South Dakota Region. Biology Dept., Augustana College, Sioux Falls, South Dakota. WATT, BERNICE K. and ANNABEL L. MER- RILL. 1975. Composition of Foods; Raw, ape gra = la Agric. Handbook No. 8, U.S. B WEINER, ee & 1972. Earth Medicine— r Interpretation of the Roberts e, Larimer County, Colorado. Unpubl. M.A is (Anthrop.), Colorado State Univ., t. Collins. Redes , W. MAX and E.A. MORRIS. In Preparation nsideration of the Nutri- tional Benefits Derived from Bison Meat. WURF, KARL. 1976. book for People. Owlswick Press, Philadelphia. J. Ethnobiol. 1 (2): 221-230 December 1981 FACTORS INFLUENCING BOTANICAL RESOURCE PERCEPTION AMONG THE HUASTEC: SUGGESTIONS FOR FUTURE ETHNOBOTANICAL INQUIRY JANIS B. ALCORN Department of Botany, University of Texas Austin, TX 78712 ABSTRACT.-— Patterns abstracted from interview sessions, discussions, and observed be- havior during 13 months of fieldwork among the Huastec in northeastern Mexico suggest some of the factors shaping Huastec botanical resource perception. The recognition of a plant as a particular sort of resource ._ depends on the interaction of a number of factors. Among the many factors discussed here are the biological, physical, and chemical properties of the available plants, human biological and cultural needs, Huastec perception of their social environment, the subsistence system, household demography, economic strategies, politico-economic standing, historical trends, and illness and curing beliefs. Factors influencing the resource status of a subset of plants used by the Huastec for fire- wood, construction, and medicinal purposes are discussed. An understanding of plant use context is necessary to interpret Huastec plant use data, to study Huastec plant manage- ment systems, to investigate the impact of human activities on plants and plant communities and to evaluate the adaptive functions of ethnobiological knowledge. It is suggested that inquiry into the reasons for species’ inclusion in useful plant lists can bring to ethnobotany a much needed focus for organizing systematic, multidisciplinary research yielding inte- grable data. INTRODUCTION This paper focuses on the botanical resource perception of individual human actors who sustain themselves in a moist tropical northeastern Mexican environment emicly understood on Huastec terms. The questions to be addressed are: What makes a given plant a particular kind of resource? What kinds of needs do resources fill? What factors influence the perception and choice of resources to fulfill these needs? In order to answer the questions I have raised, my discussion will develop the meaning of “resource” and “need” in the context of action-related decisions made by individual Huastec actors. Before the discussion is begun, however, I will briefly delimit the term “resource percep- tion”. As it is used here, “resource perception” refers to the process of assigning 2 parti- cular resource role, or “‘use” to a plant by evaluating that plant’s possible utility and the consequences of using it. _ Bye ucidating the factors influencing resource perception, I hope to accomplish two gs. First I hope to provide a new perspective for interpreting the data in useful plant lists, and secondly, I hope to contribute to the development of appropriate methods for evaluating the tive functions of ethnobotanical knowledge. Useful plant lists collected from indigenous people are often touted as the empirically valuable results of illenia of native experimentation designed to fine-tune the human to his environment. It is generally felt that, unless a “supersititious” basis for a plant’s use is clear, the plants are listed as specific kinds of resources because they have the physical properties which answer standard human needs and would very likely serve these purposes well in any context where they are available. In most cases, no further criteria beyond the folklore vs. functional distinction are applied for understanding the value or meaning of a plant’s use. The Huastec data suggest that the value of the information in these useful plant lists, so long associated with the term ethnobotany, has been both under and over esti- mated. — ln TT 222 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 In addition to containing ‘‘uses”’, ethnobotanical lists usually include native plant names. During the past few decades cognitive and linguistic anthropologists have focused on the plant names in these lists to the de facto exclusion of their accompanying uses. Investigators (Berlin et al. 1973; D’Andrade 1976; Hunn 1977; Randall 1976), however, have debated the value, structure and functioning of classification systems generated using ethnoscientific methods. Randall (1976) has stressed that the motive for classifica- tion must be examined in the context of real-life situations if classification behavior and the underlying mental organization of the information relevant to the classification are to be understood; i.e., plant names and plant uses need to be tied back together and studied within real contexts where plant names are signs for information about more than just the plant’s morphology. Hays (1981) has called for testing of assumptions that folk biolog- ical classification systems are adaptive. If ‘‘uses” are to be interpreted as “behavioral responses” as Hays (1974; 1981) suggests, then we must learn to recognize and under- stand these responses by developing a deeper understanding of the real-life contexts in which the responses occur. The following consideration of Huastec resource perception Suggests the complexity pursuant to developing appropriate methods for evaluating the adaptive functions of ethnobiological knowledge and folk biological classification sys- tems, RESEARCH SETTING AND METHODS The Huastec are Mayan language speakers who in pre-Conquest times controlled a large ecologically diverse area including most of San Luis Potosi and large areas of Hid- algo, Tamaulipas, Queretaro, and Veracruz states (Laughlin 1969). Today approximately 60,000 Huastec speakers live southwest of Tampico in the Sierra Madre Oriental foothills of southeastern San Luis Potosi and northern Veracruz, primarily a climax vegetation zone of tropical rainforest, Bosque Tropical Peuaiitin (Rzedowski 1966). The vege- tation today, however, consists almost entirely of successional communities that reflect human management (henceforth referred to as anthropogenic vegetation zones). Huastec resources labeling decisions occur within the context of the natural and Families generally live in dispersed household units within corporate communities and receive their subsistence from the products of their land, cash cropping and wage labor. Anthropogenic vegetation zones usually maintained by a Huastec family include cash Producing sugarcane, managed forest plots, and cornfield-fallow cycling fields managed as milpa by slash and burn methods. The details of Huastec botanical resource manage- ment within these anthropogenic zones have been discussed elsewhere (Alcorn 1981). Land use patterns are affected by individual and community concerns, local interpre- tation of government regulations about using available land, and locally administered government loan policies. Cash cropping centers around sugarcane which is processed into raw sugar by individual family operations and sold at low prices into a mestizo con- trolled market. Henequen and coffee are also important cash crops in some ar Research was designed to identify Huastec botanical resources, to eae the methods and impact of Huastec resource management, and to construct and integrate Cognized and operational models of the Huastec natural and social environment in order to generate hypotheses about the impact of Huastec world view on ethnobotanical Processes influen ncing elements of the Huastec vegetational environment. Methods used uring 13 months of fieldwork included structured and informal discussions, the admin- istration of interview schedules, the drawing of land use maps in consultation with Huastec land holders, and participant observation. Interviews were conducted in Spanish and Huastec. Extensive data about plant uses, plant names, and plant management were Provided by 50 informants from 20 communities. Eighty-four individuals were formally December 1981 ALCORN 293 interviewed but many more people provided valuable insights. One Huastec collaborated with me throughout the entire project and participated in most interview sessions. To date, 2000 plant specimens have been collected in scientifically identifiable con- dition (i.e., flowering or fruiting specimens). I have identified the majority by using pertinent floras, monographs, and the University of Texas herbarium. Specimens that were difficult to identify were sent to appropriate specialists for determination. Different botanical taxa collected and identified from the Huastec habitat during this project now total 910. Vouchers will be deposited in the following herbaria: CHAPA, INIF, MEXU, TEX, and the Instituto Nacional de Investigaciones sobre Recursos Bioticos in Jalapa, Veracruz. RESULTS The Huastec depend on the plant world for many raw materials. Men, women and children quickly discriminate between plant resources and nonresources every hour of the day. Specific uses have been recorded for 65% of the botanical taxa I collected. Plant utility is reflected in three of the four Huastec botanical life-form cover terms. The word for tree, te’1, also means a branch, a pole, or a piece of wood, be it a loom, a housepost, or a quickly fashioned hook for retrieving a desired fruit. Ts’a:h, the cover term for vine, also refers to any lashing material, be it vine or rope. Ts’oho:l, the cover term for herba- ceous plant, is also used to mean medicine derived from any plant source (tree, vine, bark, root, etc.). Today to:m only refers to grass. Grass is rarely used for thatch in the Huas- teca today but in Tzeltal, a related Maya language, tom refers to the grass bundles pre- pared for thatching a house. A term for grass thatch bundles was not elicited in Huastec. When presented with a fresh vegetative shoot, Huastec informants attempted to recognize and identify the specimen by evaluating characteristics which would make it useful. Leaves were usually crushed and smelled (chemical evaluation). Questions were asked about the fruit, habit and habitat. Knowledge of all these characteristics has potential value for resource assessment as well as for identification. Unless they were large and showy, edible, or otherwise useful, flowers were rarely discussed. Once identi- fication was made, the informant usually volunteered uses for the plant and add qualifying statements about preparation, value, and problems involved in its use. The resource value of a particular plant for a particular use often hinged upon the context of the use and the user. Although “resource” means something which is used to satisfy social, biological, and physical needs, the Huastec acknowledge that trade-offs are inhe- rent in the use of any resource. Plants are clearly resources for the Huastec. But what kinds of resources are neces sary and what makes “plant X” a specific kind of resource? When a Huastec informant was asked, for example, “Is ‘plant X’ a resource for ‘use Y’, a wide range of responses were given, “Yes” and “no” answers were rare. Characteristic answers included: “My ancestors needed it, I don’t.” “Only other people know.” “I can’t use it, people are too invidious.” “I’ve heard that it can be used, but I’ve never tried it so I don’t know.” i. “I don’t know. Maybe that’s why mestizo merchants buy it from us in the market. ee well yes. It could be used, but I use ‘plant Z’ because I have it here by the ouse.”” The range of answers to this question gives us some clues about the Huastec resource labeling process. The utility of a resource is assessed by the individual from 4 shared Huastec bias and a personal idiosyncratic bias. Respondents’ replies indicate 4 personal consideration of actions implicit in the choice, the appropriateness of the action, and the constraints limiting the action. Nonetheless, a given respondent’s answer is not neces” sarily optimally adaptive. His decision obtains from his calculation of the interplay effects of factors which also shape that context. 224 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 Patterns abstracted from conversation, interview sessions, and observed behavior suggest some of the factors shaping the context of Huastec resource perception. Based on the investigative focus necessary to study them in further detail, I have chosen to lump the factors into four general catagories: biological and physical; cultural; economic; and agi and social. sic to the Huastec resource evaluation process are the empirically measurable ee or chemical qualities which qualify or disqualify a plant from the use in ques- tion. Huastec, however, not only assess the physical and chemical attributes of “plant X”’ but also consider its spatial and temporal position in their vegetational environment. The plant’s ecological requirements, its membership in a particular community, its life ev its reproductive biology, its speed of growth as well as other aspects of the plant’s ology are considered, as far as they are known to the individual, within the context of ihe existing time and space investment patterns which characterize his land management System. Because his existing land management system is designed to meet many other needs apart from the one under question, the consideration of the changes necessary for the integration of “plant X” into that system is critical to his decision. Ecological changes potentially caused by the integration of “plant X”’, the scarcity of ‘‘plant X’’, and the scarcity of resources necessary to maintain “plant X” are also considered and com- pared to similar criteria for available substitutes. Other empirically measurable variables considered include the biological needs of the individual, his household, and his domestic animals, There are also, however, less easily measurable factors that are important in the resource perception process. Culturally generated needs create the requirement for cer- tain resources (e.g., specific ritual items). The perceived context for plant use is shaped by Huastec world view. Native inquiries into a plant’s properties and the interpretation of the results are shaped by native epistemological biases. Cultural sanctions against the use of “plant X” in a given way may remove it from the resource category of “‘use Y”’. In some cases a use or a plant may be identified with a particular cultural identity and thus appropriate or inappropriate for the person being questioned. For example, some plants are identified with mestizo as opposed to Indian identity. Other plants/uses are hollow lengths of carrizo [Arundo donax L., Pennisetum bambusiforme (Fourn.) Hemsl. (paka:b) | is associated with curers. Other considerations are economic. Allocation of time and space necessary for the maintenance, acquisition and/or use of the possible resource is considered within the context of the individual’s present life strategy and against the value assigned free time. The known “opportunity costs” of opting for a “new” resource are evaluated against the probable gains. The risks and uncertainties that surround the maintenance of “plant X” were it to be acknowledged as a resource are considered within the context of the land management system presently in operation. Risks are also evaluated in a social context. Some resources are more easily appropriated by other people, and the i indi ivid- ual must assess the ease of appropriation as well as the losses and benefits which might attend such a transaction. Limitations on the individual’s land management system may leave particular needs unfilled and transactional resources become necessary to fulfill these needs. The cash generating potential of a resource, the labor investment necessary, and the stability of the market for ‘plant X’’ may also be considered in this context. A al category of considerations are those peculiar to each individual. The in- dividuals’s household demography affects the household’s needs and resources, and, of course, changes over time. The individual’s personal knowledge of “plant X”’ and alter- Natives, as well as his knowledge of particular resource categories, clearly affects his answer to our question. His knowledge depends upon the form of existing information ge It also depends on the individual’s own investigations of the plant. Age, personality, €xperience, and status affect the individual’s knowledge as these factors change with December 1981 ALCORN 295 time. The individual’s particular management skills and his personality also affect the decision to perceive ‘plant X” to be a resource. For example, an individual’s definition of plant resources is affected by his personal response to the traditional wealth leveling pressures of the community, his personal desire for power, and the paths he chooses to achieve it. The individual may rely on others for certain services/products and have no need to know about resources necessary to them. The factors considered by the in- dividual in resource labeling decisions reflect the fact that the Huastec derive their plant resources from two sources: directly from the natural environment to which they have access, or indirectly through transactions with other people. Resources maintenance activities include the management and cultivation of social relationships as well as the management and cultivation of vegetation. Methods open to the individual for attaining goods indirectly through other people influence his resource perception and vegetation management. The factors which I have listed here clearly do not exist in isolation. My isolation and classification of them can only be heuristic because these and other interrelated factors interact in shaping resource perception. The individual’s answer will vary over time as any of the factors mentioned above, and thus the interrelationship between fac- tors, is altered. In addition, past interactions have created historical trends of resource definition and “‘use’’ that shape the possible present-day choices. Huastec plant resource ‘needs” and ‘‘uses” are restricted to those that are part of particular present-day Huastec to operate in today’s Huastec habitat and within Huastec social organization. At the same time, however, ‘“‘needs” and “‘uses’” bear marks of the historical strategies out of which they developed. Conflicts and constraints limit the use of available resources. Needs and the choices of certain resources may conflict with the choice of others. Such conflict imposes constraints on the use of particular plants. Other constraints are generated by peculiar intracommunal resource access regulation. For example, while each family has a specific, others. Finally, the politico-economic environment of the individual as an Indian within the Mexican sector of the world economic system limits and shapes his decisions. His choices can only be made within the context created by the decisions of those of higher authority. The ethnobotanical list of Huastec plants and their uses generated by my research reflects the interaction of factors influencing Huastec resource perception. A brief treatment of a few selected species in each of three “use” categories important for sur- vival (firewood, construction materials, and medicine?) will illustrate this interaction. The species commonly recognized as firewood include: Acacia angustissima (Mill.) Kuntze (shi:shit), Acacia cornigera (L.) Willd. (thobem), Adelia barbinervis Schlecht. & Cham. (ata’), Calliandra houstoniana (Mill.) Standl. (wi:t’ot’), Callicarpa acuminata H.B.K. (et te’), Conostegia xalapensis (Bonpl.) D. Don. (chikab te’), Croton reflexifolius H.B.K. (olih), Cupania dentata DC. (ts’aw’), Guazuma ulmifolia lam. (akich), Leucaen@ pulverulenta (Schlecht.) Benth. (thuk’), Lippia myriocephala Schlecht. & Cham. (anam te’), Nectandra loeseneri Mez. (oh te’), and Sapindus saponaria L. (walul). Although one might expect the firewood designation to be awarded to heavy wood which burns slowly to produce long lasting, hot coals, some of these species produce firewood of a very poor quality. What these species do share in common is their membership in the fast growing successional community of fallow cornfields. Depending on its age and structure, the vegetation which covers the fallow can be a resource to be burned, grazed, or kept fos future use. If the decision is made to slash and burn the vegetation, then the option is 226 JOURNAL OF ETHNOBIOLOGY Vol.1, No. 2 where to burn it and for the production of what. It can either be burned in the field to provide ash for fertilizing the corn to be planted or it can be collected and burned in the hearth to provide heat for cooking. My preliminary data suggest that the amount of land devoted to the cornfield-fallow cycled fields depends more on the amount of firewood needed than on the amount of corn needed.” Management practices often emphasize firewood production over corn production. Deliberate light burning of fields and removal of wood before buring increase the firewood yields but cut the yield of corn (Fig. 1). Pollarding of firewood-producing trees and opting not to carry out the traditional one- time weeding of the milpa speed up firewood production. Thus, factors influencing agricultural patterns, the availability of alternate firewood resources, species’ response to e agriculture regime, and species’ representation in the successional fallow vegetation are ee aii in the labeling of these species as firewood resources. cies oa used for the main houseposts in house construction include: Cordia Pw (R. & P.) Oken (wish te’), Diphysa robinoides Benth. (chichath), Harpa- lyce arborescens A, Gray (k’an te’), Nectandra loeseneri Mez. (oh te’), and Piscidia piscipula L. Sarg. (tsi:hoi). The physical requirements for houseposts strictly limits the species that could potentially fulfill this resource role. The species listed share strong wood, a straight bole, and decay resistance. They also share the ability to grow well on agriculturally poor ridgetop soils or as isolated individuals spared in the sugarcane or milpa-fallow cycle fields. The species used for roofing materials are: Jmperata brasiliensis Trin. (ata: to:m), Licaria capitata (Schlecht. & Cham.) Kosterm. (sholim te’), Sabal mexicana Martius (oto: mal, apats’) and Saccharum officinarum L. (pakab). The superior material is the thatch of Licaria capitata leaves. A Licaria leaf roof is said to last for 30 or more years, keep the house cool, and be impervious to rain. Today, however, few people recognize FIG, 1~ —Planting milpa after firewood harvest. Stacked between the new milpa and the stand of flowing co m in the background is the firewood collected before burning the slash. Light aie left more firewood to be gathered later. Also visible are spared palms for thatch production and develop- ing housepost trees that were pollarded for firewood during preparations for planting corn. December 1981 ALCORN 220 or use Licaria capitata as a roofing resource ostensibly because increasingly intensive land use has caused Licaria capitata to become a scarce forest tree in man es. thatch from Sabal mexicana has largely displaced Licaria and Imperata brasiliense thatch. Sabal mexicana is a multipurpose species that has been purposefully introduced into many local areas in the past 50 years. Palms neatly fit into existing patterns of land use because, once established, they can be spared during milpa clearing or integrated into sugarcane fields. ple who are really pressed for land and can’t afford to devote space in their fields to palms or those who can’t afford to get palm leaves from others choose to use the inferior thatch of sugarcane leaves, instead of using the sugarcane leaves as a resource to mulch and fertilize their cane fields. Thus, the list of thatch resources is not a list of four functionally equivalent species but rather a list that reflects a trade-off be- tween land use and utilitarian considerations. The list of Huastec medicinals can not simply be viewed as a list of drug plants used to treat biomedical diseases. Medicines, as a Huastec category, have the innate power to transform, or to be empowered by a curer or witch to transform a person’s state of being. They may be used to poison or to sicken someone as well as to cure a person of an illness, or to prevent illness from striking. An illness, moreover, is not just a biomedical malfunc- tion but also an event within a social field, and both of these aspects of illness are treated by curative medicines. Medicines may do their work by direct application, by ingestion, by merely being swept over a person, or by burning at midnight in the pathway to a person’s home. Huastec illness etiologies include such agencies as embedded foreign objects sent by witches, interference by dead spirits, the action of stars, and soul loss by frigh While knowledge of the same construction and firewood resources is shared by most adult Huastec, the knowledge of many medicinal plants is not. Over half of the 900 plants I collected had medicinal uses but agreement on the use and the means of appli- cation varied widely. There are many reasons for this. A person may learn about medi- cines from treatments applied by relatives, neighbors, and curers. Visible signs of use (e.g., seeing scars on a tree’s trunk) might lead the individual to ask someone what kind of resouce the tree is. At the same time, however, a person may also have a plant’s use revealed to him or her in a dream. Some people also experiment with plants to determine their effects on the body, and the information derived is then processed according to Huastec beliefs. Such individually derived information is not freely shared. Another factor influencing medicinal plant resource recognition is the increasing availability of manufactured medicines which substitute for some herbal remedies. People may know that a given plant has a given medicinal use but not perceive it to be resource for them- selves because they prefer some other plant or because they prefer to get injections OF take pills instead. In order to illustrate briefly the complexity of evaluating Huastec medicinal plants, I will discuss one commonly used species. Cissampelos pareira L. (k’on k’ach), a pantrop- ical vine found in a variety of habitats, is used to treat ichich in the Huasteca. Ichich 1s similar to the concept of “evil eye”, but rather than referring to eyes, it refers to hearts. The heart of an older, more serious person, for example, naturally saps energy from the heart of a younger, more lighthearted person, causing ichich. Chidren are especially vulnerable to this malady, but adults and even pigs also suffer from it. Ichich is often e initial diagnosis made when someone feels ill, especially from gastro-intestinal prob- lems. The patient who is suspected to be suffering from ichich may be swept with the leaves of k’on k’ach or a number of other plants. In the case of k’on k’ach, the leaves are then crushed in a small amount of water by rubbing between the hands. If the liquid gels, as it invariably does, it is seen as a positive diagnosis of ichich detected and removed by the sage, omniscient plant. Of the 34 people questioned about ko” k’ach, only three stated that this diagnostic procedure could be followed by the ingestion ofa root decoction made from the same plant. K’on k’ach is a relative of the South American plant Chondrodendron tomentosum Ruiz & Pav. from which tubocurarine and other 228 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 alkaloids that paralyze voluntary muscles are extracted for biomedical use. The roots of k’on k’ach contain some of the same alkaloids and have been exported as substitutes for Chondrodendron tomentosum (Morton 1977). In other parts of the world, native people have taken advantage of the muscle relaxant properties of k’on k’ach for a number of purposes including the expulsion of intestinal worms (Uphof 1968). But, although the root is pharmacologically active, the Huastec primarily perceive the plant to be a medi- cinal resource because of its predictable, diagnostically valuable (as culturally defined) gelling properties. Huastec who have chosen to identify themselves as mestizos assert their new identity by denying the existence of ichich and the resource value of ichich treatment plants like k’on k’ach. ange in the resource status of a plant can alter its management and thereby the ethnobotanical process to which it is subjected. Huastec informants assign resource value to plants not only while acting to use particular plants but also while making plant management decisions. Reasiicaeds percent of native non-com plants available to the Huastec are currently “managed for” by some individuals, and resource status is an important determinant - how a particular plant is managed (Alcorn 1981). Resource designation, then, potentially affects at least two types of “behavioral responses’’: the usage of a plant and the management of a plant. DISCUSSION The Huastec data fulfill the expectation that indigenous people know a great deal about their environment. But the Huastec data also demonstrate that the specific “uses” to which plants are put at any given time derive from a complex plant resource evaluation process operating from a well developed knowledge base that includes contextually related/dependent data. Resource evaluation is not an objective consideration of a plant’s material qualities abstract from context, and resource perception is influenced not only by the individual’s context but also by his understanding of that context. The “needs” whose fulfillment is being sought include biological needs of the individ- ual and household, as well as aA needs, both of which vary over time. The use of one particular resource, be it plant or rials may in turn create the need for other specific resources. Consideration op ‘costs’? of recognizing a particular resource include the assessment of the energy expenditure to fulfill the ‘ oe and the possibility of los- ing this investment to others with or without compensation. The interpretation of what “fulfills” a need also depends on biological and cultural factors. Because of culturally determined preferences, a particular food, a particular style of house, or a particular kind of medical treatment ma y “fulfill” needs better than other available items of equal or better functional value . Useful plants lists isolated from their context may be of limited value to economic botanists seeking empirically valuable data about the useful characteristics of particular species. On the other hand, these lists provide an indispensable vocabulary for studying the grammar of human ecological relations. For example, once resources are known studies of the management of these plant resources can contribute to the evaluation of of ian cognition as evidenced in ethnobiological classification systems must come to Stips bites the fact that the “uses” or “behavioral responses” to plants are not so simple y have been understood by many investigators. Assigning “functional equivalence” (Hays 1 1974; 1981) to plants is problematic when “use” and appropriateness of use may vary depending on specific contexts. December 1981 ALCORN 229 CONCLUSION Ethnobiology is a rich, relatively unexploited domain that could yield important information on human ecology. But if the adaptive value of ethnobotanical knowledge is to be tested in any meaningful way, plant “use” must be analyzed as a text that derives part of its meaning from the cultural, natural and social context in which it occurs and serves its function. The complex of factors influencing resource perception described here forces us to recognize that meaningful investigation into the adaptive value of botan- ical resource use requires not only the collaborative efforts of botanical, ecological, bio- medical, pharmacological, economic, and nutritional approaches, but also anthropolo- gical study of potentially adaptive functions of resources used in the social and politico- economic aspects of the human’s ecosystem. Despite recent redefinitions of ethnobotany to include linguistic, epistemological, and evolutionary approaches to plant-human interrelationships, ethnobotany quixotically remains an ill-defined discourse without a unifying theme (Ford 1978). More workers are bringing the techniques of their particular disciplines to bear on different aspects of plant-human interrelationships, but their fragmentary contributions are not being syn- thesized in a way that makes their results useful or meaningful to workers in other dis- ciplines. A renewed focus on the useful plant lists that traditionally defined ethnobotany may provide the important and necessary starting point for the systematic, multidisci- plinary inquiry that is the unrealized potential of ethnobotany. Understanding the dynamic process leading to the inclusion of plants in useful palnt lists provides the blue- print for work to flesh out the bare bones of these lists so that their potential contribu- tion to human ecology can be fulfilled. Knowledge of the factors important to individ- of their diverse disciplines. Such a multidisciplinary approach would mutually enrich the participating disciplines and add new depth for archaeological and linguistic interpreta- tion of plant-related data. On the applied level, the integration of such ethnobotanical inquiry with current efforts in peasant agricultural decision-making research (e.g., Barlett 1980) could make a ignificant contribution to the development of locally adapted sustained yield agro- ecosystems that provide appropriate resources to meet the needs generated by the physi- cal, biological, social and politico-economic realities of the local ecosystem. ACKNOWLEDGEMENTS Special thanks are given to the people of Te:nek Tsaba:! for their participation in Huastec ethno- botanical research. The collaboration of Candido Hernandez Vidales and Alphonsa Rodriguez Orta is gratefully acknowledged. I also wish to thank Brian Stross and Terence Hays for commenting 07 the original paper presented at the 4th Ethnobiology Conference in Columbia, Missouri, and Barbara Edmonson for commenting on a later draft. For their continuing moral and intellectual support throughout this project, I would like to express my appreciation to Marshall Alcorn, Robert Bye, Richard I. Ford, Alcinda Lewis, Marshall C. Johnston, Brian Stross, and Molly Whalen. Support for field research August 1978-1979 was provided by a Social Science Research Council International Doctoral Research Fellowship, National Science Foundation Dissertation Improvement Grant (DEB 78-05968), an E.D. Farmer International Fellowship, and grants from the University of Texas Insti- tute of Latin American Studies and the Office of Graduate Studies. An International Summer Fel- lowship from the International Student and Faculty Exchange Office at the University of Texas supported fieldwork during June 1980. The Huastec ethnobotany project is being done in associa- tion with Instituto Nacional de Investigaciones sobre Recursos Bioticos (INIREB) and the Flora of Mexico project. The conclusions of this paper, however, remain my own responsibility. ——— = = 230 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 LITERATURE CITED ALCORN, J.B. 1981. Huastec non-crop resource management: implications for prehistoric rain col.: In press. tions to Rural Development. Academic Press, New York. B., D.E. BREEDLOVE, and P.H. 1973, General principles of classifi- cation and nomenclature in folk biology. Amer. Anthropol. 75: 214-242. BYE, R.A. 1979. Incipient domestication of mustards in Northwest Mexico. Kiva 44 237-2 D’ANDRADE, RG. 1976. = bilabial trill SCREAMERS, Anhimidae Anhima cornuta ‘Horned Screamer’ Ag: amuntai [iju.iik. i@u. 4au. diu.iik. sac aaie amt. Hu: amuntai [a.mun.tai. Hau. a.mun.tai. #.au [a.mun.tai. ii.au.] falsetto, lilting, overloud HAWKS, Accipitridae Ictinia plumbea ‘Plumbeous Kite’ Ag: (no data) tA oe ra ee Hu: isip [i-si. isi. i.si. | falsetto Spizaetus tyrannus ‘Black Hawk-eagle’ reenter ee Perce MnO SRR: Ag: ukukti [u.ku.ku.ku-ku.kui. u.ku.ku.ku.ku.kui] Hu: ukukiéi [u.ku.kui. u.ku.kui. ku.ku.ku.u.ku.kui] falsetto 242 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 a FALCONS AND CARACARAS, Falconidae Daptrius ater ‘Black Caracara’ ; : : ] i Ag: Sana’na [Sa. Sa. $a. |] slight falsetto : | le " Hu: Sanasnha [¥a. $a. Sa. | _ Falco rufigularis ‘Bat Falcon’ Ag: tiu tiu (not recorded) Pe Hu: ¢iartik [tsi.tsi.tsi.tsi.tsi.tsi] falsetto CHACHALACAS, GUANS, and CURASSOWS, Cracidae a : Ortalis guttata ‘Speckled Chachalaca’ Reese Soe we. 4 Ag: wakag [sa.sa.sa.sa.sa.sa.sa.sa.sa. a rn é us.ta.taka. ua.ta¥a.ka. ua.ta.ra.ka ‘a Hu: wakag [ua.ta.tak.a. ua.ta.rak.a. ua.ta.rak.a. kau] a Penelope jaquacu ‘Spix’s Guan’ ‘ | yu Tu I ae Ag: aung [a-yui. auui. a.uui. auR,uaR. uaR.uaR.] : a | oY falsetto, first phrases lilting ‘ I = — I | | Hu: aun¢g = [au.uui. au.uui. au.wui. auR.auR.auR | ey Cy ay falsetto, first phrases Aburria pipile ‘Blue-throated Piping x 2D Ag: kuyu[kun. Su. kun. tui ki.i iu.ui. kiiu.ui. kiiu. ui. "hie tsu. i. Pat Hu: kuyu a ku.ku.ku.kun.tsai] Aburria aburri ‘Wattled Guan’ Ag: uwacau [u.uai.ua. u.uai.ua. | gentle rising to 2nd syllable, a falling, vowels heavily sustained SR at ve Hu: awaca [t¥a.¥4i.ua. tSa.rai. ua. tSa.rai. ua] Crax globulosa ‘Wattled Curassow’ A oe [= tt. ee byt ot ] > piiwi is.pi. pis.pi. p4s.pi uu.uu.uu. g: piiwi [pis.pi. pis.pi. pis.pi pag.yu.yu.y Hu: pji (bird name known, but no knowledge of its vocalization) December 1981 BERLIN AND O’NEILL HOATZIN, Opsthocomidae Opisthomcomus hoazin Ag: saasa (not recorded) ore eee 8, Pee Hu: saasa _[Sa.sa.sa.sa.sa.sa.sa.sa RAIL, Rallidae Aramides cajanea ‘Gray-necked Wood-Rail’ Ag: kuacau (not recorded) Hu: kuncar (kun. tsi kun.tYbe ki kG kG ket. [eon ie bun Re RMR] lilting tempo Anurolimnas castaneiceps ‘Chestnut-headed Crake’ Ag: pilhuak (not recorded) ON 1 Vv ete |v Hu: piturcigki [pi.tu.ru. pi.tu.ru. pi.tu.ru. 5 5 nee pi.tu.ru] lilting tempo, falsetto SUNBITTERN, Eurypygidae Eurypyga helias Ae Sra Te Ob gas ee Oe Ag: tiinkin [tin.tin.tin. tin.tin.tin. tin.tin.tin.] regular tempo Hu: tinkia (not recorded) PLOVERS, Charadriidae Hoploxypterus cayanus ‘Pied Lapwing’ oe Poa ee We he Se Ag: tiutiu [ t#U.tiu.tiu.tiu.tiu. tiu.tiu.tiu.tiu.tiu. | Hu: tuintui (no knowledge of call) SANDPIPERS, Scolopacidae Actitis macularia ‘Spotted Sandpiper’ ao Pies ! ae, ee Ag: piampia ~ tinkin [pi.am.pia.pia.pia. | pee eae Pi.am.pia.pia.pia.| rapid tempo ie Pep a Hu: piampia [sui.sui.sui.sui.sui.sui.sui. | 243 et ee ooo a 244 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 DOVES AND PIGEONS, Columbidae Geotrygon saphirina ‘Sapphire Quail-dove’ , , ‘is j Ag: pukui [pu.kui. pu.kui. pu.kui.] Hu: pupui (hummed imitation) PARROTS, Psittacidae Ara manilata ‘Red bellied Macaw’ ‘ Ps l 2 7 Ag: kayak [kia.ku.ka.ia:ku. ki.ia.ku.ki-ia.ku.] ° im) ° ey ° Cy ° lilting falsetto a ~ y: ~ a ~ on Hu: kayak ([ka.iak. ka.iak. ka.iak.| oO aa yD falsetto, irregular Aratinga leucophthalmus ‘White-eyed Parakeet’ ‘i ae Ag: tipi [Ki KLKLKLKLKLK:] falsetto , a Se ee Hu: kiki [ki.ki.ki.ki.ki.ki.ki.] rapidly, falsetto Aratinga weddellit ‘Dusky-headed Parakeet’ Ag: (not in Cenepa area?) —__ viv. ¢ ‘i ve? vivivi vt Aes Hu: sisi ~ pirtunyis ~\ santanta ([Si-si.si. si.si.si. wa VN ‘Sisi.si.] falsetto Pyrrhura picta ‘Painted Parakeet’ 2 Lu% ee. ! ! Ag: mangiit [tsit.tsit. tsit.tsit. tsit. tsit.] irregular 2 ae ae 1st | Hu: ¢ip [tsit.tsit. tsit.tsit. tsit.tsit. tsit.] irregular falsetto Forpus xanthopterygius ‘Blue-winged Parrotlet’ ‘s fe Se ! . ! : Ag: siwim [Si.ui. Siui. Sicuil irregular tempo ono omo om Hu: nuinui [sic] (nui.nui.nui. nui.nui.nui. nui.nui.nui] cf. Ag Touit huetii falsetto Brotogeris cyanoptera ‘Cobalt-winged Parakeet’ Ag: kihus [kidd kidhkik. im. dim, tim) Hu: ciimp [tt Fem, yen. Mi.vim ] first syllable in each phrase clipped (Note: the last segment of the vocalization of this species in Aguaruna is cognate with the name of the bird in Huambisa.) December 1981 BERLIN AND O’NEILL Touit huetii ‘Scarlet-shouldered Parrotlet’ - em kee” Ga ae . . . » * Vs . Ag: nuinui [ Stk Sik. Sik.sik. Sai. ait ani. hi a ) Vv [sa.kas.kas.kas-kas. i in tw 8 In lw I~ v4, os ee We te Oe Hu: sai Sai.sdi.sai. sai.sai.sai.] (Note: Huambisa name is identical to part of the vocalization in Aguaruna.) Pionites melanocephalus ‘Black-headed Parrot’ Oe v, vi viv! v, vi Ag: cirikas [tsi.rik. tsi.rik. tsi.rik] irregular falsetto Hu: Cirikas (vocal information not recorded) Pionopsitta barrabandi ‘Orange-cheeked Parrot’ Ag: (bird not known) “a A ! a at ee Hu: mui [muimui. mui.mui. Mui.mui. mui.mui. Pionus menstruss ‘Blue-headed Parrot’ re ' \ 1 i i Ag: tuis [tuis.tuis.tuis.tuis.tuis.tuis. | ‘ ! [kU ia.tsik. kii.ia .t3ik.] wy ly Hu: tuis [tuis. tui.tuis.tuiS.tuis.tuis. | vowels clipped Amazona ochrocephala ‘Y ellow-headed Parrot’ Ag: kawau [tsa.tsa.tsa.tsa.ii cka.Ya.ta.ra.ra] ’ —. v v ay Hu: awarmas [au.fa.au.ra.au.ra.au.ra| falsetto Amazona festiva ‘Festive Parrot’ Ag: Cawit (not recorded) 2 vf wen eee 1 Hu: cawit [tsa.ui.ta.tsa.ui.ta.tsaui.ta. ka.ran. nN a oO [ka.ran .ka.ran .ka.ran.] falsetto Amazona amazonica ‘Orange-winged Parrot’ ’, Pee Pete a a ; Ag: pahai [pai.pai.pai. pai. pai.pai. pai] irregular tempo 7 — —————— ——— 246 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 Hu: parai [sa.ta. Sa.ta. Sa.ta. Sa.ta.] falsetto, lilting CUCKOOS, Cuculidae Crotophaga major ‘Greater Ani’ 1 vee Ag: kuakua [qxR.qxR.qxR.qxR.sa.ra] Hu: kuakua [sak. au. -kuR ] Crotophaga ani ‘Smooth-billed Ani’ Ag: bait —_ [mai.mai.mai.mai.] falsetto Hu: mawai [ma.uai.ma.uai.ma.uai] falsetto OWLS, Strigidae Glaucidium brasilianum ‘Ferruginous Pygmy-Owl’ P ie: ae S| See eg Ag: takint [t@n.taq.tanq. ta.kin.ta. ta.kin.ta] repeated Hu: (not recorded) POTOOS, Nyctibiidae Nyctibius ‘Great Potoo’ —s. | I | I oe a a tes ee , : ! Hu: kau [qxauR. ja.tsu.ru. ja.tsuru. ia.tsu.ru. qxauR] falsetto ooo | > im) ooo Nyctibius aethereus ‘Long-tailed Potoo’ e ptab—tah Soy -— ; Ag: autaém [auu. auu. auu.] note distinct vocoids, falsetto Hu: (not recorded) Nyctibius griseus ‘C on Potoo’ Pan ~ od 7 Ag: athu 4. ai8,ua.dua.gua. 2. ais.ud.sua.sua falsetto ia) 4, fal en aapiaiats fal ae aii Hu: auhu§ [au. uu.uu.uudu. au. uu.uu-uu Ot, | falsetto AAA A AAAA December 1981 BERLIN AND O’NEILL 247 NIGHTJARS, Caprimulgidae Hydrosalis climacocerca ‘Ladder-tailed Nightjar’ ; Ae ee : Ag: papahu aa = ae a irregular Hu: papar _— (not recorded) SWIFTS, Apodidae Reinarda squamata ‘Fork-tailed Palm-Swift’ Ag: (not in, or very rare in, Cenepa area) ve, VF x ! l | | ve Hu: acunmaya surpip [™bis. ™bis. Mbis. ™bis. su.ru.ru.] HUMMINGBIRDS, Trochilidae Eutoxeres aquila, E. condamini ‘Sicklebilled Hummingbirds’ Pe gag Se s l “ee Ag: jimpictau [4im. pt. wis.w4s.wis.was. iim. pi.tsau.wis.wis. eee ony ee ee [4im. pi wis. wis.wis. } melodic Hu: uhuh jimpi (no known call) TROGONS, Trogonidae Trogon vtridis ‘White-tailed Trogon’ Ag: tawai [tau.tau.tau.tau.tau.tau.] staccato falsetto oe) ; 1 ; Hu: tawai [ta.ta.ta.ta.t ta.ta.ta.] rapid tempo falsetto Trogon collaris ‘Red-bellied Trogon’ (perhaps other red-bellied species as well) eaDeMeG naps" Ag: cakua_ [tsa.kua.ktia.kta.kta. tsa.kua.kua.kua.kua.] falsetto, some whisper Hu: (no information) KINGFISHERS, Alcedinidae Ceryle torquata ‘Ringed Kingfisher’ Ag: mun Cahi (not recorded) Hu: taras [ta tas. tasuy. ta.rAs.] Chloroceryle amazona ‘Amazon Kingfisher’ Ag: Cahi (not recorded) Hu: Gdhi [Bad SLT G] ro | oO (2) io) rapid regular tempo, falsetto ee 248 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 MOTMOTS, Momotidae Ba yphthengus ruficapillus (martii) ‘Rufous Motmot’ Ag: yukuhu [uu tu. uh Sa, wifu. uu.ru. uur¥] cd oy A Hu: yukuru fuu.ru. u.ru. ° A A PUFFBIRDS, Bucconidae Bucco capensis ‘Collared Puffbird’ ? Ag: bukitau —[bu.kia.tau. bu.kia.thu. bu.kia.thu] lilting Hu: maukatarar [mo.ka.ta.rar. mo.ka.ta.rar. mo.ka.ta.rar] Monasa morphoeus ‘White-fronted Nunbird’ , ae oe ae oe | ! Ag: tawikuru ~ tihu [iiu.iiu.iiu.iiu.iu.iiu.ta.ui.kruk. ° ° ° ° ° ° i) Cy es Ge Oy Ee ! ! ! ! ! [tayik.ta.uik.ta.uik.ta.wik. ta.ul. kuru. ] a [ta.ui.ku.ru.] I | ! Vw Hu: tawikuru [ta.uik.ta.uik.ta.ui.ku.rur. a oO 9 J Vv~ ta-ui.ku.rur] BARBETS, Capitonidae Capito aurovirens ‘Scarlet-crowned Barbet’ Ag: (not present in Cenepa area; inhabits flooded forest lacking there) Hu: apu pur [ pur.pur.puf.pur-put.] falsetto Eubucco richardsonii ‘Lemon-throated Barbet’ Ag: puuy [bupububu] Hu: wincu pir [br.br.br.br.br.br.] falsetto Eubucco bourcierii ‘Red-headed Barbet’ Ag: tiwa [tiu. ti. tia. tia] ys Oy , l Hu: tiwa = [ti.ua. tiua. tiua.] falsetto December 1981 BERLIN AND O’NEILL TOUCANS, Ramphastidae Aulacorhynchus derbyanus ‘Chestnut-tipped Toucanet’ ? we walle ik Ag: ikauk ~ waak (tak tak Wak Wak] continuous breathy | | | | Hu: ikiak [tak Wak Wak Wak] continuous Pteroglossus castanotis ‘Chestnut-eared Aracari’ Ag: piristan pininc ([pris., pris. ,pris. pris. prjs] 2 ve ie ab a Hu: piristin [piris. pisis. pi.ris. pi-ris] Pteroglossus flavirostris ‘Ivory-billed Aracari’ YQ iB aaron ! I I fixi. = Ag; saitam pininc [ka.ka.ka.ka.ka.ka.ka.ka.ka.ka.ka.ya. ion) eT [ka.ka.ka.ya. ka.ka.ka.ka.ka.ka] Hu: kakarpag (karkAtkarkat. kaY.kdt. kar.kat. kar.kat.] vowels slightly whispered Selenidera reinwardtiit ‘Golden-collared Toucanet’ Ag: kahungam (kau.kau.kau.kau.] rasping Hu: karun¢gam (kau.kau.kdu.k4u] rasping Ramphastos culminatus ‘Yellow-ridged Toucan’ , 1 | ! 1 Ag: kithua [kKiau. kiau. kiau. kiau. kiéau.] falsetto, breathy ¥ ly ly ly Hu: k#rua_(Kirua. kirua. ki.rua] R. ambiguus ‘Black-mandibled Toucan’ oN dh Vins vl Vv jb Ag: Saatak [sa.ta.rak.ta.rak.ta-rak. 3a.ta.rak.tarak. ON . [tarak. Satarak. Sa.taxhk.taxhk.] — cee falsetto ~ a Hu: Sartik [Sar.tik.tik.tik. ar.tik.tik.tik. ! [Sar.tik.tik.tik.] falsetto Ramphastos cuvierit ‘Cuvier’s Toucan’ Ag: yarika gukay ka wu apu gukan ka “Ml fiatikakakhka. iukhkd. oO JOURNAL-OF ETHNOBIOLOGY Vol. 1, No. 2 250 . vi . vi és eS Pe ees djari.ka.jia.ri.ka. diu.ka.ka.ka. ia.ri-ka] P rey tee ee Sk ee Hu: apu ¢ukan ka [ia-kuéy -kuig .kuty. ia.kuéq -kuta .] continuous WOODPECKERS, Picidae Celeus elegans ‘Chestnut Woodpecker’ ia , e a a ol ~ Ag: yawa sawaki [ia.ua.ua.ua.ua.ua.] breathy falsetto Ci CY EN. Cy 64. OR se , ae (Bae : A . . Hu: apu sawakia [sa.ua.kia> Sa.ua.kia] cy rapid fall on final syllable Melanerpes cruentatus ‘Yellow-tufted Woodpecker’ Ag: tihasa (no vocalization, whistled only) o* oO oN Hu: tiraksa [tictak.sisa. ti¥ak.siSh. tivak.si.sa.] falsetto WOODCREEPERS, Dendrocolaptidae Campylorhamphus trochilirostris ‘Red-billed Scythebill’ ne ae a RY A Ee ae Ag: bikuamkuas [bi.kuam.kuam.kuam. bi.kuam.kuam.kuam. A ! J J Vie iNa iy [bi.kuam.kuam.kuam. tsi.tsi.tsi. [bi.kuam.kudm.kuam. ] eee Oe eee eee vy - ! | ! ! Hu: saawia [tSia.tsiau.tsik.tsik.] ANTBIRDS, Formicariidae Thamnophilus schistaceus ‘Black-capped Antshrike’ Vow v4 A ice ve yer oe yen Ag: cihikiu ~ cikiu [tsi.kiau. tsi-kiau. tsi.kiau. Ivbulovivivl vee [tSi.tsi.tsi. tSi.tsi.tsi. tsi-kiau] falsetto throughout 4 io) Hu: Cicikia [tsitsikia. tSik. tsik. o oO ° ° tsik.tsik. t8ikia] . vowels strongly clipped Thamnomanes ardesiacus ‘Dusky-throated Antshrike’ ‘ rt ‘ Ag: kuncacam [kun.tsa.tsat. kun.t$a.t3at. December 1981 BERLIN AND O’NEILL 951 [sui.Sui.suiiuiiuiiuiuiui] falling tone can Hu: (no information) Myrmotherula (striped species, ie brachyura or obscura) , Pot tite Ag: cuncuikit [ tsun. tsui.ki.ki.ki.ki. a ae ts tsun. tsui.ki.ki.ki.ki] Hu: Cuncuikit (not recorded) Mymotherula axillaris ‘White-flanked oe v. bod Simp ven ies Ag: ciatas [tsia.tas. tsia.tas. tsia.tas] Hu: kiatsa [Kia.tsa. kia.tsa.tsa. kia.tsa.tsa.] Myrmotherula schisticolor ‘Slaty Antwren’ a ieee at Ag: cucup _[uis.uis.uis.uis.uis. tsu.tsu.pi. Cy OG} Gy GY tf) ° o”o [ts pd 7: tid F su.tsu.pi. tsu. pi] rapid, short syllables , | oym | - Hu: Cucup = [tsik.tsu. tsik.tSu] vowels whispered Myrmoborus myotherinus ‘Black-faced Antbird’ falsetto Hu: pisak [ pi.sak. pi.sak. pi.sak] falsetto lew! 1 my, [pi-si-ri. pi.si-ri] danger call Hylophalax naevia ‘Spot-backed Antbird’ OMe 4 eae: ly ly ly Ag: -pIS. uim.pis.uim.pis.uim.pi g: wiimpis [uim pis. uim.pis.uim pis-uim.pis] staccato ie oe ee Hu: pisipis [ pis.pis.pis.pis.pisu] ° ° ° ° °o-o Chamaeza sp. ‘Antthrush species’ Ag: (no data) Hu: tuas [tua.tuatuatudtuaththth. thk.thk.tbk.] falsetto, dropping pitch on last three syllables —mrennatiaie itil. itis: nasa cai ttt a NO RC LA CCN. I LE EON LL, LE ALLS — EE A — ae ameniiiiiie svelte 252 JOURNAL OF ETHNOBIOLOGY Formi carius analis ‘Black-faced Antthrush’ Ag: takinc [tin-ki. oa. tin.ki.aa. tsu.kip. tsu.kip] A O, Hu: tukimp = [takimp. takimp. t&kimp. Skim] Myrmothera campanisona ‘Thrush-like Antpitta’ Pan Ag: puampua [puam puam.puam.puam:puam] Hu: pudmpua _ [pu.am.pu.am.pam.pam.pam.] rapid tempo, pitch gradually falling COTINGAS, MOURNERS AND PIHAS, Cotingidae Rhytipterna simplex(?) ‘Grayish Mourner’ Ag: ukuntu¢ {u.ku.ku.ku.tu.tsia.tsia.tsia. Speman a ee u.ku.ku.ku.tu.t3ia.tsia.tsia] /_ ee ee he ee Hu: ukurpip [ku.ru.ku.pi.pi.pip. 4.kan.tsam.kan.tsam] repeated Lipaugus cinerascens ‘Screaming Piha’ a) oN ae: ; , I. ow Ag: papainc [ pai.pain.t8a. tu.pia.tu.uia. pai.pain.tsa. ] [tu.uia.tu.uia] Hu: papaine [ pa.pain.tsa. pa.pain.tsa.] falsetto [tu.uéa.tu.uia. tu.uia.] oO f i=) Querula purpurata ‘Purple-throated Fruit-crow’ ; 4, i lg —_— : Ag: pauwai [ pau.uai. pau.uai. pau.uai. [ii.ia. ii.ia. pau.uai] irregular tempo oOo ry on ry ae ee in Hu: paucinki (no data) Phoenicircus nigricollis ‘Black-necked Red Cotinga’ a Ag: piga [tsia. tsia. tsia. tsia.] regular tempo “OE ORES, ME 3 ged Hu: gadné [tsants. tsir-tsir-tsir-tsir.tsants bee tv ‘ [tsants. tsir.tsir.tsir.tsir.tsant3] rapid tempo Vol. 1, No. 2 December 1981 BERLIN AND O’NEILL 253 Rupicola peruviana ‘Andean Cock-of-the-Rock’ ae Ag: tnatinwn bY asYasY aiYa.] su nka Hu: iatinu [iadadada] a falsetto, overloud, sinka yell-like MANAKINS, Pipridae aoe eee eee Ag: kaawia [ki.uia?.ka.uia? 55.5. | ka.uia?.ka.uia?.p.p.p.] Hu: kaawia (not recorded) Tyranneutes stolzmanni ‘Dwarf Tyrant-Manakin’ | { | Pipra pipra ‘White-crowned Manakin’ ‘ ) Pa ee | 1 1 | Ag: cuup [ tsuu.pi?. tsuu.pi?. tsuu.pi?] slight whisper Hu: cuup (not recorded) FLYCATCHERS, Tyrannidae Contopus virens ‘Eastern Wood Pewee’ Ag: (no data) , 7 ‘ . Hu: tiwi — [ti-ul. ti.ui. ticui,] falsetto SWALLOWS, Hirundinidae Stelgidopteryx ruficollis ‘Rough-winged Swallow’ , Le er | Ag: Cinim [Sich eS. eB.) Ve v! vi vl v! v! vi vl Hu: cinim _[Si.si.si.si.si.si.si. Atticora fasciata ‘White-banded Swallow’ ae viv vvivly v.v.vl Ag: suimpip [Sisis. Sisisis. 3isisis.] oi ae / slight whisper ee ae a ON OR See Hu: namakaya surpip [su.rip.su.rip.su.rip.su.rip | slight falsetto, lilting 254 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 JAYS, Corvidae Cyanocorax violaceus ‘Violaceous Jay’ Ag: kihuanéam [kiau. kia oe kau. ] some falsetto Hu: kithuantam (kia. ee kia. a ] WRENS, Troglodytidae Troglodytes aedon ‘House Wren’ Ag: Cutuin —[tsubtsaitsar tSultuit¥ui. kudn . [kuan .kuan .kuan .] rapid tempo Hu: cuicuin (not recorded) Microcerculus Sui acer far seg ene ce wae e mo, Ag: tinkis (ttheLeltie. ‘shetehelskes repeated Hu: tin kis [theleltitied] pitch falls and tempo slows toward end of call AMERICAN ORIOLES AND BLACKBIRDS, Icteridae Scaphidura oryzivora ‘Giant Cowbird’ Ag: ¢angingt (not recorded) Hu: gingangs _[ tsin.tsan.tsi.tsi.tsi.tsi.tsi. tsin.tsan.tsi] falsetto Clypicterus oseryi ‘Casqued Oropendola’ Ag: wauk [uaaa. uak. | Ss A | iii. ii Gbi.ibi. falsetto ° Qo fe} 1°) ° a eS Net eM | Hu: wauk [uau.kik. uau acd. uau. kik. t$a.ya.kak. fe ee eee falsetto Psarocolius decumanus ‘Crested Oropendola’ | I. Ag: suak tuwi [ tSuak.t3udk.t¥uak. tSua. tua.ta.uai. 5 fo) fo) oO [tua.ta.uai. sax. sax. sax.] December 1981 BERLIN AND O’NEILL 255 Pa { | \ \ ly, Hu: gan kt [tsan. tsan. tsa. tsar. tsirik. edik, ta.rik.] fortis onset (P. decumanus is referred to by non-cognate forms here, though each term is onomatopoeic for their respective languages.) Cacicus cela ‘Yellow-rumped Cacique’ Vly Ww wily oly Ag: tiis [ ti. ti5.tis. tis. tis.tis. tis.] irregular - oy ~~ ~ _“~ Hu: éuwikit [sa.sa. sa.sa. sa.sa. sa.sa.] irregular Icterus icterus ‘Troupial’ Ag: (no data) , | | Hu: huitam [uui.uui. uuiuui, uui.uui.] repeated cyt ry 3% a TANAGERS, Thraupidae Euphonia rufiventris ‘Rufous-bellied ecard Ag: tama uugap [us .pa. Min kta ita Sen Fut Sal bisa, on is wade. cca, Saenne kn: kia. kia] Hu: tuma usap [sic] (no knowledge of call) Tangara chilensis ‘Paradise Tanager’ eR f Ti. We Sol Ag: simancuk [sé.tsik. Si.t8ik. si.tSik.] staccato 4 cy Hu: sica (not recorded) Wetmorethraupis sterrhopteron ‘Orange-throated Tanager’ Ag: incituc [in.tsi.tu.tsia. in.tsi.tu.tsia.] melodic (sustained ingression of air, then call repeated) Hu: sancipu (not recorded) Thraupis eptscopus bax os: ang Ag: suwic iaiel eee irregular fy > gee pee Hu: suwic [su.ui. su.ui. Su.ui] final syllable falsetto a ea 256 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 Ramphocelus nigrogularis ‘Masked Crimson Tanager’ ve i I | 1 I I | Ag: canké [tsan. tsa) .tsaQ. tsaQ. tsan.tsan] irregular, staccato Hu: ednké [sit8an. sitar. t84/) .ké.sk.t3an. tr) .ki] vowels strongly clipped Piranga rubra, P. olivaceus ‘Summer and Scarlet Tanagers’ Ag: (no data) N ae ee ae ee | me Hu: picurkik —_[pi.tsur.kik. pi.tsur.ki.ki, pi.tsur. apg, oe ee eS [pi.tsur.ki.ki.ki] Tachyphonus surinamus ‘Fulvous-crested Tanager’ Ag: wampan kit [tsit.tsi.tsi.tsi.tsi.] irregular tempo vf vl Deen Hu: cangim [ tsan. tsi.tst.tsi.] ~ [tsAn.tsin.tsin.tsi. tsi.tsé ] 0 6 86 Cissopis laveriana ‘Magpie Tanager’ : ! 1 ! e ° Ag: pis [pist. pist. pist.] rapid, irregular Hu: pisi [pis. sh pis. si.] last syllable falsetto Unidentified onomatopoeic bird species, Aguaruna kunnki. (probably an antbird) i 7 [kun.ki. iGididi, kug.ki, Sasa.sa $4.34. ii. oo0o°0 7 Via Ve V ns Vow, ii ku). ki. sa.sa.sa.sa.sa.] oy wincuncu (probably an antbird) _ aca [na.ya.tu.witat.mi. uin.tsun.tsu. ] repeated, melodic quality tan tan (probably an antbird) | ' 1 1 1 1 [ta). tan. tan. tan. tan. tan.] tuatua (probably an antbird) ae ! at 1 1 1 i 1 [tuat. tuat. tuat. tuat. tu. tu. tu.] December 1981 BERLIN AND O’NEILL / 4¢4 wisui (a cotinga?) ' ! ! ! ! ! {uis. uis. uls. uls. uis. uis.] pidndik (a flycatcher?) a eC aatatad vl pep y~ vl [piun.tsik. piun.tsik. piun.tsik.] [t3ik] syllable strongly staccato kihim (a flycatcher) [KiKLKLKLKLKI. © kikiki.kiki-ki.] slightly falsetto, slightly nasalized kistonkai (yellow breasted flycatcher) a eee : ee (tip.tip.tip. tip.tip.tip. kis.ton.kai. tip.tip.tip. i] kis.ton.kai] short, irregular tempo 4 wiswiswis (yellow breasted flycatcher) Ee SE ee A ee [uis.uis.uis.uis. uis.uis.uis.uis. uis.uis.uis.uis. ] maakua (a hawk, probably Herpetotheres cachinnans ‘Laughing Falcon’ [ma.kua, ma.kua. ma.kua. ua. ua. ua.] soft, whispered quality kauta (a hawk, probably Micrastur sp.) au. au. au.] breathy quality Siiki_ (a parrot) rapid tempo 4 Ped wawik~ apu Siik Notharchus macrorhynchus ‘White-necked Puffbird’ 1 I ! ' ' [ua.uik. ua.uik. ua.uik. ua.uik.ua.uik.] breathy, increasing in tempo toward end of call wiu (a puffbird) 2 ‘ , ’ —- call begins with a whistle— [uiu. iu. enon Pe ye ey a 258 JOURNAL OF ETHNOBIOLOGY éais (a tanager?) [tSais. t¥ais. t¥ais. t¥ais. tSais.] irregular tempo Citakaim (a tanager?) we . . . . . . . [ tsi.tsi. tSi.tSi. tSi.tsi. tsi.] irregular tempo ° ° fe} ° ° ° ° dahun (a trogon?) (da.ua. diud. ddut. ddut. da.ui.] staccato A a a A a 4 . yamakiu (a trogon?) ' ye coe aera [yamak. kiaa. Kia, kiaa. ya.mak. kiau.kiau.kiau.] kap (a thrush) é fo teed a." ! [ku.pi ku.pi. ui.ki.pu.uus.ai.si. ku.pi.] is] a wium (a flycatcher) ee" Ge ee ee ee. ee . (uiu.uiu.uiu.uiu.uit. u.uiu.] rapid tempo 74 piisa (a flycatcher) [ piu’. piu’. pius.pius.] irregular tempo Unidentified onomatopoeic bird species, Huambisa pardiparfi (a flycatcher?) [pa.fai.pa.Fai.pa Yai. pa.tai.pa.rai.pa Yai. [pa.tai.pa.rai.pa.rai.] falsetto i¢itukaip (a flycatcher?) ces ! ! } [#tsir.tu.kaip.tsir.tkaip.tsir.tkaip] Vol. 1, No. 2 December 1981 BERLIN AND O’NEILL 259 CONCLUSION The foregoing data demonstrate that onomatopoeia in Aguaruna and Huambisa ornithological vocabulary is pervasive. More than a third of the terms of both languages for bird species are onomatopoetic in origin. These figures are comparable to those obtained for several other languages spoken by peoples of comparable socio-technological level of development. The Tzeltal Maya show 49% onomatopoetic bird names (Hunn 1977:84, Berlin n.d.). In Kaluli, a language of Highland Papua New Guinea, Feld reports that 49 of the 125 bird names (39%) are onomatopoetic (Feld 1979:149). These species as a group are recognized in Kaluli as those that “‘say their names” (ibid.:156). In Selepet, another Papua New Guinea language, McElhanon (1977) describes 131 of 355 bird names to be onomatopoetic (37%). Speck reports that 23 of the 63 bird names he recorded in Canadian Delaware (some 37%) were derived from the perceived sounds of the birds’ calls (Speck 1946). While onomatopoeia is most obvious in ornithological vocabulary, it is also highly productive in the naming of frogs, toads, some mammals, and some insects in both Aguaruna and Huambisa. Hunn has suggested for Tzeltal ethnobiological vocabulary that “The distribution of [onomatopoetic] terms closely parallels the distributions of highly developed auditory signaling behavior among animal forms” (1977:83-84). We believe that onomatopoetic naming plays a productive mnemonic role in naming certain animals. Our evidence is anecdotal but suggestive. During the process of our inves- tigations (Berlin et al. 1981; in preparation), large numbers of Aguaruna and Huambisa subjects participated in experiments where they were asked to name specimens of animals sequentially arranged along long work tables. Many subjects, when confronted with a specimen whose name they had temporarily forgotten, be it a bird or a frog, would (un- consciously?) begin to vocalize the animal’s call, then, with a flash of recognition, proud- ly pronounce the appropriate name. It appeared to us that the process of phonological vocalization aided these subjects in recall. Jespersen, noting the tendency of children to use stereotypic vocalizations for certain animals as their first names for the creature itself appears to be talking about a similar process. ‘‘These words [such as quack quack] are an imperfect representation of the birds’ natural cry, but from their likeness to it they are easier for the child to seize than an entirely arbitrary name such as duck” (1921:150). In languages spoken by peoples of small-scale, technologically simple, non-literate societies, one might expect to find that the natural sound signaling habits of many creatures are replicated in the actual names people assign to these animals, thus forming a direct link between the linguistic designation of the organism and an important aspect of their behavior. We speculate that such non-arbitrary names are easier to remember, and probably less difficult to learn. Naming animals after their calls would appear to be an efficient way of reducing the cognitive effort required of peoples of non-literate itions who must learn, retain, and actively employ rather sizable ethnobiological vocabularies. Conversely, we further argue that the functional load carried by ethno- biological onomatopoeia will lessen and ultimately be lost altogether as societies move to higher and higher levels of socio-technological complexity and individuals become less and less aware of their biological environment. Ethnobiological onomatopoeia might serve, then, as a useful index of cultural evolution, the process being highly elaborated in the languages of non-literate peoples and gradually diminishing with the growth an development of complex literate traditions. It is our hope that future ethnobiological research will be conducted to test the validity of such speculations. ACKNOWLEDGEMENTS The research reported here was made possible through the generous financial support of the National Science Foundation in the form of grant number BNS76-17485, “Field Research in Ethno- biological Anthropology”, Brent Berlin and James L. Patton, Co-Principal Investigators. Collegues 260 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 who have a us special support in the development of this paper include Leanne sensi John Ohala, the staff of the Phonology Labora ratory, University of California, aa Louan the Laboratory of Ornithology, Cornell University; Eugene Hunn, David Fre ecil ee, ch, James anny and Brown. We dedicate our paper to the memory of Alberto Espejo, pentane ethnobiologist. NOTES lWe gi not addressed the problem of infor- mant nen in this paper. We do not know arto extent the onomatopoetic bird vocali- zations ew from the four and isa informants who participated in me study are shared for the population as a w This could, of course, be tested in several ways, that they identify the bird represented. We have reason to believe, however, that the vocalizations represent rather entrenched features in Huambisa an aruna ethno-ornithology by virtue of the striking ities in the overall patterning of calls in both languages. Furthermore, a compari- son of the Aguaruna and Huambisa with those entered in a dictionary of Shuar of Ecuador (Bolla n.d.) show striking parallels, —" ee probably characteristic of proto-Jivaroan. LITERATURE CITED BEASLEY, DAVID and KENNETH L. PIKE. 1957. Notes on Huambisa phonemics. Lingua Posaniensis 1:1-7, BERLIN, BRENT. n.d. Notes on Tzeltal bird calls. Unpubl. MS. BERLIN, BRENT, JAMES SHILTS BOSTER and JOHN P. O’NEILL. In preparation. A compara- tive Jivaro ornithology: The bird = system of the Aguaruna and Huam' of Amazonas, P . 1981, The ign ee Ree = ethnobiological classificatio: > aruna Jivaro ae hy : sano Prey 5-108. BOLLA, LUIS, n.d., DICCIONARIO SHUAR- CASTELLANO. unpubl. MS. Centro de Docu- mentacion, ietinrion y Publicaciones. Juc- la, Ecuador, 1980. Una Paz Incierta. Direccién de Organizaciones Rural. Lima, Peru. FAST, GERHARD an d MILDRED L. sone 1974, Introduccién al Idioma stituto de Linguistico de Verano, ae FELD, STEVEN. 1979. Sound and Sentiment: Birds, Weeping, Poetics, and Song Expression, Unpubl. Ph.D. dmarastin (An- throp.). Indiana Univ., Bloomington. peor esawiee 1960. The general classificatio and South American ages. ee on “194, in Men and Cultures: ifth International niv. Pennsylv: siti, GRIMES, BARBARA (ed.). 1978. Ethnologue ae edition). Wycliffe Bible Trans., Inc., tington Beach. a eeuaee JOSE MARIA. 1964. Nomen- clatura Jibaro- cies de aves Aguaruna en el Alto Maranon. Biota 5:210-222. HUNN, EUGENE. 1977, Tzeltal Folk Zoology: The Classification of Discontinuities in Nature. JESPERSON, OTTO. Language: Its Nature, Development, and Origin. MacMillan Co., New York LARSON, MILDRED L. 1957. Comparacion de los vocabularios A Tradi- cion 19-20:147-168. Cuzco, Peru. . 1963. Emic classes which manifest the obligatory tagmemes in major independent pes of Aguaruna. Pp. 1-36, in Studies in Peruvian Indian Languages (Benjamin Ellson, ed. Summer Inst. Linguistics, Inc. Huntington Beach. 1966. Vocabulario Aguaruna de Amazonas. Serie Lingiiistica Peruana, 3. Insti- tuto Lingiiistico de Verano, Yarinacocha, Peru. . 1978. The Functions of Reported Speech in Discourse. Linguistics No. 59. Summer Inst. Linguistics, Inc., Dallas. December 1981 LOUKOTKA, CESTMIR. 1968. Classification of South American Indian Languages. Univ. California Center for Latin Amer. Studies, Los Angeles, McELHANNON, K 1977. The siege one of birds by the ke New Guinea. Ocean 48(1):64-74. MEYER, DE SCHAUENSEE, R. 1966. The Species of Birds of South America with their distribution. Livingston Publ. Co., Narberth, Pennsylvani Mc — pecans A. 1955. The — us languages of Latin America. Amer. Anth- pee) 57:501-570. PAYNE, DAVID. 1974. Nasality in Aguaruna. Unpubl. M.A. Thesis (Linguistics). Univ. Texas, Arlington PIKE, KENNETH L. and MILDRED L. LAR- SON. 1964. Hyperphonemes and non-system- BERLIN AND O’NEILL 261 atic features of Aguaruna a ics. Pp. 55- 67, in Studies in Languages and Linguistics (A.H. Marckwardt, ed.), aw. pean Press, Ann Arbor. eres OLIVE A. and MARY RUTH WISE. Grupos tai omaticos del Perd (2nd enon Universidad Nacional Mayor de San Marcos y Instituto sinaiiition de Verano, ma ima. SPECK, FRANK G. 1946. Bird nomenclature and song interpretation of the Canadian Dela- ware: An essay in ethno-ornithology. Wash- ington Acad. Sci. 36:249-258 TURNER, GLEN D. 1962. — phonemics visnacients summary. Unpubl. MS. URIARTE, LUIS. 1977. Poblaciones nativas de la Amazonfa Peruana. Amazonia Peruana 1:9- 58. 262 JOURNAL OF ETHNOBIOLOGY Vol. 1, No. 2 NEWS AND COMMENTS This section of the Journal of Ethnobiology is intended as a “‘bulletin board’”’ for upcoming conferences, brief reports of research in progress, recent publications of inter- est, and requests for exchange of information or opinion. You all are encouraged to make use of this space. Bear in mind a probable delay of several months or more between receipt of material by the ‘“‘News and Comments” editor and its appearance in print. Contributions for the next issue, Vol. 2 No. 1, should be received by the ‘News and Comments” editor no later than 1 March 1982. This issue will be released in May 1982. Contributions for Vol. 2 No. 2 should be received no later than 1 September 1982. This issue will be released in December 1982. Refer to the inside back cover of this issue for the mailing address for the ‘“News and Comments” editor. Comments are solicited also, critical of or supplementary to previous Journal of Ethnobiology articles or further- ing debate on other issues relevant to the readership. In the instance of a critical r joinder to a Journal of Ethnobiology article, the original author(s) will be aie to respond before publication. The relevance of materials submitted will be evaluated by this editor with the assistance of knowledgeable editorial board members. NEWS Harold C. Conklin’s Folk Classification: A Topically Arranged Bibliography of Contemporary and Background References Through 1971 has been reprinted (1980) with corrections and the addi- tion of an author index. It is available from the Department of Anthropology, Yale University, New Haven, CT 06520 OGICAI The fourth International Archaeozoological Conference is to be held at the Institute of Archae- ology, University of London, April 18-23, 1982. The theme of the conference is “The Contribution of Faunal Analysis to the Study of Man”; it is expected that the proceedings will be published in three volumes dealing with hunters and their prey, shell middens and the exploitation of the sea and sea shore, and the use of domesticated animals. Abstracted from the Newsletter of the International Council of lacy nace sity i Donald K. Grayson, Depart me Anthropology, ms of aaa Seattle, WA 981 FIFTH ETHNOBIOLOGY CONFERENCE The Fifth Annual Ethnobiology Conference will be held in Balboa Park, San Diego, — Co-sponsored by the San pe Natural History Museum and the Museum of Man. A call for per be issued in January 1982. Registration will be Wednesday evening 21 April. Paper sessions will be Thursday and F riday, 22- 23 April. An ethni nic Sena t featuring native Japanese plant and animal Notice received from Amadeo M. Rea, grin of Birds and Mammals, Museum Natural History, P.O. Box 1390, .. Diego, CA 92112. SOCIETY OF ETHNOBIOLOGY are currently filing papers to establish the Society of Ethnobiology, In n which will oversee publication of the Journal of Ethnobiology and organization ‘of sm annual December 1981 NEWS AND COMMENTS 263 COMMENTS When the Journal of Ethnobiology was first oo three of the present editorial board mem- bers, B. Berlin, T. Hays, and E. Hunn, were editing an occasional newsletter dubbed the Folk Classifi- cation Bulletin. Publishing the Folk Classification Bulletin poe to be a struggle dependent entirely begged and borrowed time and money. The Folk Classification Bulletin content had been largely ethnobiological—reasonable given the predilections of its editors—and it seemed logical for the Folk Classification Bulletin to join forces with the Journal of Ethnobiology. Thus the Folk Classification Bulletin ceased publication after four issues, Volume 1, number 1 (Fall 1977), and number 2 (Spring 1978), Volume 2, number 1 (Fall 1978), and Volume 3, number 1 (Fall 1979), the last under the editorial guidance of James Boster. e last two issues of the Folk gary hy rsa highlighted a debate on the relevance of Set “Fuzzy Set Theory” (Lofti A. Zadeh, “Fuzzy me tacts ah Control 8:338-353, 1965) for constructing formal models of folk biological classifi n. Hunn touched off the debate with a dis- cussion entitled “Fuzzy Sets and Folk Biology” (rc “oft -3, 1978) in which he evaluated the utility of an extension of ponte notion of “fuzzy subset” proposed by Willett Kempton to account for ertain anomalous results he encountered in CS, to apply the original notion of fuzzy subset (Zadeh 1965:340) to the classification of cups and mugs (“Category Grading and Taxonomic Rela- tions: A Mug is a Sort of a Cup,” American Panic 5 en 1978). Hunn concluded that neither Zadeh’s nor ee. s nn of “fuzzy set inclusion” was appr priate for ei hilien e structure of folk biological taxonomies. In the paca issue (FCB 3[1]: is 13) weighted features was faulty, and that what is needed is less formal manipulation and more empirical ch. In same issue, Boster defended Hunn’s analysis (FCB 3[1]:13-15) by oman a more explicit interpretation of the graphic heuristic Hunn had devised. Subsequently David Reas of the University Ken Classification.” By way of preview, I will briefly summarize Reason’s reasons — here, with the understanding that the full text of this debate with additional commentary will b e published i ne as subsequent issue of the Journal of Ethnobiology. ason and Hunn agree that the fact that the notion of a set (e.g., a basic folk biological taxon) is logically primitive contributes to the inadequacy of set theoretic models of folk biological classi- fication. re basic folk biological taxon, as for example “raccoon,” is taken as given. Our various attempts to formally analyze the structure of taxonomies address only the logic whereby these sets might be related to one another to construct a taxonomic hierarchy of such sets. Hunn has argue that ethnobiologists must account not only for the taxonomic structure, but also for the existence of the basic taxa in the first place (“Toward a Perceptual Model of Folk Biological Classification,” American Ethnologist 3:508-524, 1976). Why “raccoon”? Why not “ringed-tailed quadruped” (a concept both more or less inclusive than “raccoon”)? According to = neither fuzzy set theory nor classical set theory provides an adequate framework for understanding how basic sets are con- structed. Rather, Hunn has suggested, one should take the perception of sacra and difference as primitive (1976:515-520). Then formally construct sets from those primitive relatio Reason disagrees, arguing that these basic sets of which folk biological canines are built are derivative instead a social and historical conditions governing their appearance. Thus to under- stand the true meaning ne a folk system of concepts is a task “intrinsically antithetical to the exercize of formal representation.” are interested in contributing to this debate, copies of the relevant pieces may be obtained from the “‘News and Comments” editor. A contribution of $1 will be welcome to cover duplication and mailing expenses. Sear, Me ee OT Pee ae Sa Re BS ORR IRS TS Ne ee ee ee AE hy NOTICE TO AUTHORS The Journal of Ethnobiology accepts papers on original research in ethnotaxonomy and folk classification, ethnobotany, ethnozoology, cultural ecology, plant domestica- tion, zooarchaeology, archaeobotany, palynology, iat daira and ethnomedicine. Authors should follow the format for article organizati ibli hies f: articles in this issue. All papers should be typed double-spaced hes pica or dite type on 8% x 11 inch paper with at least one inch margins on all sides. The ratio of tables and figures to text pages should not exceed 1: 2-3. Tables should not duplicate material in either the text or graphs. All illustrations are considered figures and should be submitted reduced to a size which can be published within a journal page without further reduction. Photo- graphs should be glossy prints of good contrast and sharpness with metric scales included when appropriate. All illustrations should have the author(s) name(s) written on the back with the figure number and a designation for the top of the figure. Legends for figures should be typed on a separate page at the end of the manuscript. Place all tables and figures in consecutive order at the end of the manuscript. Do not place footnotes at the bottom of text pages; list these in order on a separate sheet at the end of the manuscript. Metric units should be used in all measurements. Type author(s) name(s) at the top left corner of each manuscript page; designate by handwritten notes in the left margin of manuscript pages where tables and graphs should appear. Submit 2 copies of the manu- script plus the original copy and original figures to: JOURNAL OF ETHNOBIOLOGY P.O. Box 1145 Flagstaff, Arizona 86002 NEWS AND COMMENTS Individuals with information for the “News and Comments” section of the journal should submit all appropriate material to Eugene Hunn, Department of Anthropology, DH-05, University of Washington, Seattle, Washington 98195. Please note that the former Folk Classification Bulletin has been incorporated into this section. SUBSCRIPTIONS Subscriptions to the Journal of Ethnobiology should be addressed to P.O. Box 1145, Flagstaff, Arizona 86002. Subscription rates are $22.00, institutional; $15.00 regular membership, for U.S., Canada, and Mexico; foreign subscribers add $6.00. Write checks Payable to Journal of Ethnobiology. Defective copies or copies lost in shipment will be replaced if written request is received within one year of issue. CONTENTS EARLY ACCEPTANCE OF WATERMELON BY INDIANS OF Fre Ur ca Ale, Deonird W. Blake... 2. ee ew eee eS 193-199 A “LOST” VIKING CEREAL GRAIN, Lisa Carlson Griffin ee ee ee eee ee ee 200-207 THE ANU AND THE MACA, Timothy Jones ........-........ 0s es 208-212 NUTRITIONAL CONTENT OF SELECTED ABORIGINAL FOODS IN NORTHEASTERN COLORADO: BUFFALO (BISON BISON) AND WILD ONIONS (ALLIUM SPP). Elizabeth Ann Morris, W. Max Witkind, Ralph L. Dix ete FC i ee a a en ae ee ee 213-220 FACTORS INFLUENCING BOTANICAL RESOURCE PERCEPTION AMONG THE HUASTEC: SUGGESTIONS FOR FUTURE ETHNOBOTANICAL INQUIRY, RES oo Sage ate es ne ee ere re ee 221-230 ELEMENTS OF THE PUREPECHA MYCOLOGICAL CLASSIFICATION, Cristina Mapes, Gaston Guzman and Suet ee 231-237 THE PERVASIVENESS OF ONOMATOPOEIA IN AGUARUNA AND HUAMBISA BIRD NAMES, Drent Werien ae fone PO Nee ass kG es ge oi ee we ee 238-261 NEWS AND COMMENTS ...... ees ss rs wee 262-263 | Journal of Ethnobiology VOLUME 2, NUMBER 1 MAY 1982 MiSSouR! BOTAN; AL JOURNAL ORGANIZATION EDITORS: Steven A. Weber and Steven D. Emslie, Center for Western Studies, Inc., P.O. Box 1145, Flagstaff, Arizona 86002. NEWS AND COMMENTS EDITOR: Eugene Hunn, Department of Anthro- pology, DH-05, University of Washington, Seattle, Washington 98195. EDITORIAL BOARD BRENT BERLIN, Department of Anthropology, Columbia University, New York; ethnotaxonomies, linguistics. ROBERT A. BYE, JR., Department of Environmental, Population and Organismic Biology, University of Colorado, Boulder; ethnobotany, ethno- ecology. RICHARD S. FELGER, Senior Research Scientist, Arizona-Sonora Desert Museum, Tucson; arid land ethnobotany, desert ecology. RICHARD I. FORD, Director, Museum of Anthropology, University of Michigan, Ann Arbor; archaeobotany, cultural ecology. B. MILES GILBERT, Adjunct Research Associate, Division of Vertebrate Paleontology, University of Kansas, Lawrence; zooarchaeology. TERENCE E. HAYS, Department of Anthropology and Geography, Rhode Island College, Providence; ethnobotany, ethnotaxonomies. Ss mrctas: Dore — of Biological Sciences, Northern Arizona Uawenin 4 Flagstaff; 3a _ EUGENE HUNN, Department of Anthropology, University of Washington, Seattle; ethnotaxonomies, zooarchaeology, cultural ecology. HARRIET V. KUHNLEIN, Division of Human Nutrition, University of British Columbia, Vancouver; ethnonutrition. GARY P. NABHAN, Meals for ee Foundation, Tucson; cultural ecology » plant domestication. DARRELL A. POSEY, Center for Latin American Studies, University of : Pees. ethnoen 4 il cultural ecology. AMADEO M. M. REA, See of F Birds and Mammals, San Diego Museum of Natural History; ethnotaxonomies, zooarchaeology, cultural ecology. 2 ae of Ethnobiology is published semi- — anually. Aanuscripts for neem: and information for the “News and Comments” ane nhs sent to the appr. yn the inside aa. oe a Se ee | _ Journal of Ethnobiology VOLUME 2, NUMBER 1 MAY 1982 J. Ethnobiol. 2(1):1-15 May 1982 ANIMAL DOMESTICATION AND OSCILLATING CLIMATES BRIAN HESSE Department of Anthropology, University of Alabama in Birmingham Birmingham, AL 35294 ABSTRACT.—Two case studies, one set - western Iran, the other in northern Chile, are described in an effort to reintroduce climate as a factor in the prehistoric adoption of goat and llama herding. This paper focuses on he likelihood of short-term variability in precipi- tation in these arid regions dramatically affecting game availability in a non-density depen- in a context of re-orientation of social values from those associated with hunters to those associated with pastoralists. INTRODUCTION This paper presents an argument that irregular short oscillations in climate can be isolated as key variables in two aon separated episodes in the record of animal domes- tication—one in the Zagros Mountains of western Iran, and the other on the western slopes of the Andes in a aes While climate has often been suggested as an agent in the origins of pastoralism and agriculture, the causative links binding environment and human behavior have been insufficiently specified (Harris 1977:184). Either the link is too diffused, that is, able to predict the onset of herding/farming within a broad regional forward, emphasizing the technological and economic aspects of change at the expense of the social and ideological contexts in which it is imbedded. The existence of these flaws has deflected interest from climate as a rimary causative element and in the direc- tion of population pressure models (Cohen 1977). Alternatively, the differences between domesticating and non-domesticating societies have been downplayed (Higgs and Jarman 1972), recasting the Neolithic revolution more as a step than a threshold in human history. discussion is aimed at oe climate as an element in the origins of animal domestication. Climate is viewed as ort of boundary mechanism, a frame that des- cribes the limits within which social a reload options are played. The key to the linkage between climate and domestication is oscillation—the occurrence of erratic short-term environmental variability. I try to show how changes in the distribution and availability of game, that can be hypothesized given our knowledge of the climatic history of each region, discouraged the social and technical institutions associated with hunting and encouraged those associated with pastoralism. DOMESTICATION There are two basic approaches to defining the term domestication as a cultural Process. The first views domestication as a temporally elongated sequence of techno- logical innovations in the methods of extracting resources from animals (Bray 1976: 90). Each innovation has effects on the selective factors conditioning animal reproduc- tion with the result that domestic stock gradually diverge morphologically and behav- lorally from their wild ancestors. Hunting and herding, in this view, are distinct sets of ecological relations. Domestication, or the process of moving from one to the other, is modelled by schemes that view culture primarily as an adaptive system designed to wrest resources from the environment. Commonly these schemes include resource Stress as a motivating variable, since domestication is seen as a response involving extra work to solve an economic problem. Z HESSE ‘Vol. 2, No. 1 The cgntinuum of development in doemstication is usually broken into segment , and ‘full domestication’ among many others (Higgs and Jarman , Zeuner 3). Many of the segments are associated with expectations, concerning alterations in animal morphology, or the equipment required to manage animals in a particular way, that can be recognized in the archaeological record. Further, each segment or form of man/animal relationship is arranged along another continuum of productive potential with each historical innovation—taming, secondary products, selective breeding—linked to larger yields of animal products. However, a case can be made that some of the initial steps in the development of pastoralism did not necessarily produce increased yields of animal products. This per- haps surprising statement is based on the behavior of many modern pastoralists. The management goal of these herdsmen is not to maximize production of such commodities as meat, milk, and wool. It is rather to maximize the size of the herd. Any production directed activity such as slaughter, milking, and castration potentially reduces to one _ degree or another the herd growth rate. Said another way, pastoralists try to avoid taking products from their herds preferring instead to have the maximum possible unharvested resource on hand. Texts on animal management emphasize the point that the technological features associated with husbandry, most of which serve to minimize the impact on herds of mortality factors other than man, disturb the system of checks and balances within an environment by reducing its diversity and so its resilience. Herdsmen eliminate com- peting predators such as the large canids and felids and tend to the health and security of their animals. The result is that the pastoral herd can have a population structure that has even more mature animals than a wild herd. In that case a greater proportion of forage is going toward animal maintenance than growth in the domestic flocks compared to the wild herd. Productive potential is actually reduced. The desire of pastoralists to increase their herd sizes also places regional systems at risk through the destructive effect on pasturage of over grazing. Despite this well recognized risk herders tend to continue to expand their flocks in competition with each other, particularly in societies with frag- mented, ‘family’ based, forms of pastoral production (Bates 1973:143). This tendency is at least part of the reason large areas of the Near East have been deforested through over-grazing. It is therefore more accurate to say that pastoralism is potentially more productive. Whether that potential is realized is a separate issue. The question is largely one of whether pastoralists can be induced to cull their herds in such a way as to maxi- mize the number of young growing animals e answer depends on the pastoralists’ perception of risk. Another primary benefit of pastoralism is the freedom to schedule the extraction of animal products with a cer- tainty not available to hunters. Meat, for example, is available year-round if the herds- man is willing to slaughter one of his animals. However, that willingness is tempered by the realization that, in contrast to the exploitation of plants, harvesting animals does not lead to resource renewal (Ingold 1980). Therefore, in the absence of an institution capable of transforming animal products into a form of wealth or credit, later redeem- their herds to provide insurance for themselves and their dependents. Allen (1977), following Jacobs (1969), suggests that the institution which can serve this function best is the village, a nexus where a wide range of resources may be exchanged. Under these circumstances the productive potential of animal management can be released. The economic benefit of herding is thus two-fold. The immediate gain is stability, insurance against times of stress. In those cases where a reliable institution of exchange vailable, management can lead to greater productive levels. The process of domesti- cation in economic terms, however, is initially one of risk redirection. The hunter’s gamble on the availability of game is replaced with the herdsman’s worries over disease, predators and stock loss. j ; , May 1982 HESSE 3 The second approach to defining domestication emphasizes the social component of the association between human and animal populations. For example Ducos (1978: 54) f “domestication can be said to exist when living animals are integrated as objects into fins socio-economic organization of the human group, in the sense that, when living, those animals are objects for ownership, inheritance, exchange, trade, em as are the other objects (or persons) with which human oups have something to do.”” Domestica- tion in these terms is a social process, a transformation of the rules that structure not just animal populations, but human groups as well. For animals the transformation is primar- ily one of taming (Hediger 1968:108).. When more than one animal is included, it also involves the restructuring of the age and sex distributions of their groups with resultant alterations in social behavior. For human groups the social transformation moves along two paths. The first path establishes specific bonds between individual domesticators and individual domesticated animals—a process that encompasses petkeeping, the use of animals for labor, and milk production. ae nai the transformation establishes bonds between populations of domestic animals and populations of domesticators with a blurring of the ties that bind any two indian together. This second path is associ- ated with what Ingold (1980) has called carnivorous pastoralism, or, on a more elaborate scale, ranching. Neither axis necessarily presupposes the other, and the order of appear- ance is probably due to the migratory habits of the animal involved—the more migratory, the more likely the first will precede the second, the less migratory, the second may pre- cede the first. Each of the paths is associated with different human emotional relationships to animals. For instance, Bennett (1964) examined the attitudes of farmers and ranchers on the Great Plains who were raising cattle for meat. He observed that “when animals are herded in fairly large numbers . . . utilitarian attitudes toward the animals tend to become dominant” (p. 37) and “gieihaden toward cattle contain almost no element of sympathy or oe and very little tendency toward the establishment of relation- ships between single men and single animals” (p. 42). This attitude contrasts strongly with the nS nica or see Soni relationship that describes the bonds of these ran- chers with their horses and dogs, or milkers with their dairy animals. difference in attitudes along the two paths also exists in the realm of social struc- for cementing social ties. The cattle of African pastoral systems, which are used for milk and blood production, bear the load of establishing complex intersocietal links, where goats, which are raised primarily for meat, do not (Ingold 1980:186). The social and ideological factors associated with domestication can also be con- trasted with those of hunting groups in a broad way. Ingold (1974, 1980) and Paine (1971) have isolated several important contrasts. First, hunting groups tend to have undivided access to the animals they hunt. The hunting territory and the animals it Contains, in other words, are viewed as shared resources though buffer zones between hunting groups may be present (Hickerson 1965). Second, hunting groups tend to have complex redistributional rules that extend outside the family and insure the maxi- mum dispersal of food. These rules for game need not extend to gathered resources. Third, for hunters, expertise in the chase is the important source of prestige. Prestige is not generated by possession of the sah but by the right to distribute it (Paine 1971: 158). The unit of production is corpor Herders, on the other hand, ee access to the resources. The production unit and the circle of redistribution is subsocietal. Prestige is generated through the number of animals held. Successful herders are disinclined to gather dependents around them & o” Ee 5 cs a) be J “A ° Lae) a f 5 Q. 5 _— (and insurance against hard times), the flock. Pastoralist groups are, therefore, more Tigid as the individual entrepreneurs balance stock levels with staff to maximize their Positions, Explaining the origins of domestication on the social level may therefore be 4 HESSE Vol. 2, No. 1 characterized as offering a plausible solution to the problem of why the concept of value is shifted from a focus on a hunter’s skill rather than his property, to a focus on a herds- man’s property rather than his skill in accumulating it Modeling the Process of Animal Domestication The process of domestication thus operates in two spheres—productive systems and social institutions. A successful model must explain the transformation in each that is brought on by the shift from hunting to pastoralism. Here I will consider one possible pathway Climate change can affect the structure of productive systems by altering in a non- density dependent way the availability of various plant and animal resources. As such, climate has been included among the considerations involved in resources stress explana- tions for the origins of plant and animal domestication (Binford 1968). Resource stress acts to cause the evolution by creating an imbalance between residents and food sup- ply. Climate change is an efficient factor in this regard because it is non density depen- dent and does not allow productive systems to adjust their demand requirements in res- ponse to the stress. Population pressure is less efficient since predator-prey (or hunter- hunted) relationships are density dependent. That is, the rate of the hunter’s kill is proportional to the availability of game. Most predator-prey interactions tend toward either equilibrium points of population abundance or stable limit cycles where the opulation levels of predator and prey traverse a loop of positions. In either case the density dependent nature of the relationship tends to buffer any population pressure unless an intrinsic rate of population increase for the predator can be hypothesized. Resolution of the stress can take three forms—abandonment of the region, shifting speculate on what conditions would work against selection of either of the first two. The second option, turning to new resources, would be foreclosed in an environment where all the resource options were being restricted by the climate change. This would attractive when the stress factor was of short or intermediate duration: that is, when experience had taught the residents that bad times tend to be brief, and that game levels will rebound in a relatively brief time. This would be true especially when the game species involved have high natural herd growth rates and when tia mortality is not primarily conditioned by density-dependent factors like predat To summarize, under conditions of short range resource stress in a simplified icant the option of husbandry or pastoralism appears to be an attractive way to buffer food supplies The other side of the problem, how resource stress can act on social institutions o bring about the change in values outlined previously, has been discussed by Ingol peer 527) with respect to Lappish reindeer specialists: Traditionally, Lappish reindeer hunting and fishing groups were organized on a territorial (stida) basis. The territory, together with its resources, was viewed as the joint property of traditional hunting economy, cag “should be regarded as capital in an emic and extra De context” (Paine 1971:169). Heuristically we may suppose that increasing scar city of wild deer would assign to | the deer a strictly economic value over and above that of the aor: itself. Placed in the context of resource stress, once climatic variation had reduced game levels to point below which even the greatest hunter could not count on success, value and pres- tige would come to be attached to the possession of game itself. This shift is also a root May 1982 HESSE 5 of the different perception that hunters and herders have of their resources. Hunting groups share access to the animals they rely on for food, herding groups divide access to them. During periods of stress those hunters who also possessed tame animals would be in a strong position. Reliance on the offspring of these culturally controlled animals during periods of resource stress would establish the pattern of numerous independent pastoral units within the society, since the axis of ee binds individual animals to individual herders rather than populations to population return to climate, insofar as people would peice that game levels were likely to continue to oscillate due to unpredictable environmental changes, they would be encouraged to establish insurance herds against this certainty. Groups caught in this pattern of resource oscillation would experience conflicts both between alternative views of property and prestige, and between modes of production based, on the one hand, on shared access to the resources, and on the other, divided access. These tensions would be likely to persist until the disruptive effect on uncontrolled pastoral herd growth effec- tively eliminated wild game as aresource alternative. Because of the tension, transitional occupations are likely to be rare in the archaeological record, and pastoral management systems will seem to emerge suddenly from hunting societies. The Mahidasht Region (Fig. 1) The Mahidasht or the valley regions near Kermanshah, Iran (Levine 1976:487) has been the scene of investigations into the origins of domestication and settled village life since the 1950’s. It is part of the Zagros mountain system, a northwest-southwest trend- ing band of rugged terrain in Iraq, Iran and Turkey that is crosscut by a complex drainage System. The joint effect of these physiographic factors has been to produce a topo characterized by independent valley systems connected by narrow and precipitous defiles eoengpe 1965). The valley floor of the Mahidasht is located at an altitude of approxi- y 1350 m. The region overlaps two important vegetational zones—the oak forest aa a almond/pistacio savannah. The area was occupied during the Pleistocene with samples reported from such sites as Warwasi (Turnbull 1975), Bisitun Cave (Coon 1951), and Ghar-i-Khar (Young and Smith 1966). Until recently the only known early Holocene sites were Sarab, Asiab (Braidwood et al. 1961) and Tepe Ganj Dareh (Smith 1978). However, a recent survey by Smith and Mortenson d.) anges that condition is an 25 km 4 ee aes FIG, 1—Location of late paleolithic and early neolithic sites in the Mahidasht region of west-central Iran, 6 HESSE Vol. 2, No. 1 artifact of concentrating the search for sites primarily on the alluvial plain. In their sur- vey, three new sites of early Neolithic character were discovered, all located in narrow “nearly inaccessible valleys” within two or three hours walk of Tepe Ganj Dareh. On the basis of this discovery, they suggest that the initial steps toward sedentary life, pastora- lism and agriculture, took place in the small, ecologically diverse valleys surrounding the Kermanshah Valley. e earliest evidence of animal husbandry comes from the site of Tepe Ganj Dareh, excavated over five seasons by Philip E. L. Smith of the University of Montreal. On the asis of radiocarbon determinations, the site was occupied in the eighth and perhaps ninth millenium B.C. The site contains five superposed occupations. The earliest, labeled Level E, is characterized by a series of basins excavated into culturally sterile soil, no evidence of permanent architecture, and some indication that the site was occupied seasonally, probably in the spring (Hesse 1979). The upper four levels, D - A, contain mud-brick architecture, ceramic storage facilities, and stone mortars and a few fragments of pottery. Evidence for animal husbandry is of two sorts (Hesse 1978). What seem to be goat footprints in some of the bricks from Level D argue that tame animals must have roamed the village. The harvest profiles, estimates of the age and sex of the slaughtered animals, and, therefore, evidence of the system of production, calculated for Levels E and D - A contrast. On the basis of mandibular tooth wear (Payne 1973), the proportion of young sheep and goats (12-24 months) slaughtered during the architectural phases of the occupation increased compared to the earlier material (see Fig. 2). Considering the two species (Ovis orientalis and Capra aegagrus) separately, on the basis of the relative frequency of fused long bone epiphyses, it appears that only the harvest pattern for the goats changed. Using the fact that goat sexual dimorphism is reflected in the dimensions of many of the bones of the skeleton, it was possible to create individual harvest profiles for both bucks and does (Hesse 1982a, following the suggestions of Higham 1968). The harvest profile for Level E compares favorably with the age and sex proportions found in a wild goat nursery herd—an almost total absence of mature males in a sample dominated by male and female kids and mature females. Tepe Ganj Dareh Level E is roughly con- temporaneous with Asiab (BOkényi 1978) where mature males dominate the samples. en compared, the samples produce a picture of seasonal hunters exploiting sub-popu- lations of goats, taking those age and sex categories that would be expected to be nearby each of the sites based on the seasonal topographic preferences of the species. Tepe Ganj Dareh Level D, on the other hand, has a harvest profile for goats that agrees with the slaughtering patterns of pastoralists—a reduction in infant mortality and elimination of does which fail to kid (Bates 1973:147). Lu T — + : 4 .s ' 2612 24 36 48 72 96 months FIG. 2—Goat/Sheep harvest profiles from the basal(—) and architectural (————-) —— at Tepe Ganj Dare’ curves represent estimated percentage survivorship at the ages indicat d and are based on tooth wear (Payne 1973). [SS a Oe Oe ee ee ee May 1982 HESSE 7 To draw climate into the discussion about this episode of animal domestication, two First, the evidence suggests that the diversity of the herbivore fauna was reduced from Pleistocene conditions. The faunal samples from Warwasi and Bisitun indicate that Pleis- tocene subsistence was based on a mix of equids, red deer, gazelle, sheep and goats. In the Holocene, however, equids are absent from Tepe Ganj Dareh, as well as the three new sites discovered by Smith and Mortensen, and present in only very small quantities at Asiab (BGk6nyi 1978:6). It is not unlikely that the growth of an oak/pistacio forest on the western slopes of the Zagrow may have been partially responsible for this. Ona- gers, gazelles, sheep and goats are not well adapted to wooded conditions. Some forms of sheep and goats, in fact, show a marked reluctance to enter forested regions (Geist 1971). In fact, the development of this vegetational pattern may have been an important deter- minant in creating the distribution of sheep phenotypes in modern Iran (Valdez et al. 1978). However, sheep and goats are vertical rather than horizontal migrators (F ormozov 1969). As such they can adapt to seasonable variations in climate by utilizing steep grad- ients in relatively small areas, rather than moving across extensive tracts of open country. Equids are not so adapted. They tend to migrate significant distances in regions where rainfall is seasonal (Klingel 1974:130). Male equids maintain large territories which they dominate through their visual presence (Ibid.:127). Of all the equid species the least known is the Asiatic wild ass or onager likely to be represented in the Mahidasht samples. However, Clabby (1976:34) notes, “all races of the Asiatic Wild ass prefer the unrestrict- €d views and wide expanses of the open country where their alertness and speed give them maximum protection.” Equids, of course, did not disappear from southwest Asia. What I suggest is that by interfering with migratory routes, the vegetational changes that accompanied the early Holocene caused a rearrangement of faunal communities. The new faunal distributions were zoned, with the foothills on either side of the Zagros mountains supporting the open country herbivores (equids and probably the gazelles), while the mountainous regions were occupied by sheep and goats, species that could successfully exploit islands of rugged terrain in a growing forest. This trend would have forced the occupants of the Mahidasht to specialize more and more on caprines as modern climatic conditions emer- d. Our knowledge of post-Pleistocene climate suggests factors that could induce varia- tions in the remaining game biomass. Biomass is linked to variations in annual precipi- tation. If the regression equation published by Coe et al. (1976:348) based on twelve African ecosystems is taken as an estimate of the trend of change (not an estimate of actual biomass) expectable in the dry conditions of the Zagros, the modern range ° annual rainfall for the Mahidasht (378-490 mm) translates into a 25-30% variation in biomass, Van Zeist (1969) indicates that early Holocene rainfall was greater than Pleistocene levels, but was concentrated more seasonally, while mean temperature was lower in the Period than today. A study of the alluvial deposits in the Mahidasht suggests that the Precipitation that did occur in the early Holocene would have been stormier than today (Vita-Finzi 1969). Bobek’s (1963) study of Zagros snowlines reports that they oscil- lated several times during the Holocene. From these observations I tentatively conclude that Holocene conditions in the Mahidasht were characterized by irregular and violent Weather. The cooler temperatures implies that much of the precipitation fell as snow. The serious effects that snowfall can have on sheep or goat populations has been discussed by Formozov (1969). Both snow depth and nast—a hard icy layer formed by Conditions of repeated melting and refreezing—prevent animals from getting at forage. “Ne most disastrous effects occur when the heavy snowfalls come late in winter. it Compounds the problems of already exhausted animals and is particularly devastating a the new born. The picture is clear in the following passage, which describes condi- Hons in a region to the north of the Mahidasht. 8 HESSE Vol. 2,No. 1 In the previous hard winters, many of these animals perished from lack of food, and those remaining e were greatly exhausted by hunger, and they became the prey of wolves. Radde (1862-1863) showed that the snowy winter of 1831 destroyed the remaining Trans- because of various conditions in which man is completely innocent, even species of large animals can become extinct at least locally.” (Formozov 1969:63). The fact that snowfall can seriously affect sheep populations was emphasized by Murie (1944:65-67, 87-88). Murphy and Whitten (1976) have been able to plot the relationship between snowfall and population levels for the Mt. McKinley Dall Sheep. They conclude that every major population decline is associated with an episode of heavy snow. The ability of these animals to revive from near extermination lies in their high natural herd growth rates, and their relative invulnerability to predators. In fact, the sheep populations studied by Murphy and Whitten experienced several cycles of popula- tion growth and decline in the fifty years their data covers. Local oscillations in density are unlikely to be overcome by migrations since sheep and goats are vertical migrants, exploiting botanical successions in precipitous habitats. They show a considerable at- tachment to home ranges and population exchange between habitats is slow. While these reports have dealt specifically with sheep, it is reasonable to assume that similar patterns would characterize the Near Eastern wild goat, since it seems to be a behavioral analog of New World sheep. Also, the descriptions for Alpine Ibex (Nievergelt 1966) and Bezoars (Schaller 1977) suggest broad similarity between the species. While the evidence is circumstantial, I believe that it is a reasonable conclusion that the Holocene occupants exploiting the goat habitats fringing the Mahidasht experienced recurrent oscillations in game availability. These oscillations probably took place within the lifetimes of single individuals. This led to two related changes—a need to adapt the mode of production to include the provision of an insurance herd, and a shift in the way resources, prestige, property and the social unit of production were defined. What could have been the origin of the insurance herd? One explanation (Bokonyi 1973) is that hunters would have simply captured herds, tamed them, then herded and husbanded them. The difficulty with this sort of model is that it is hard to see how the captured herd could be divided among the households of the community as individual properties to form the basis of competing rather than cooperating productive units. In Ingold’s (1980) discussion of the origins of pastoralism, the argument is presented that the insurance herd is not bred directly from hunted animals, but are the offspring of tame animals already incorporated into the households. As the property of households, when the insurance herd was used as food during periods of resource stress, it woul fall under different redistributive rules than hunted animals in which all members of the band had a stake. In the case of reindeer pastoralism, which is the subject of Ingold’s treatment, the source of the tame household animals was the need for transport stock to move the camp along the nomadic round. ether a parallel existence of tame household goats in preneolithic Zagros or munities can be demonstrated is problematic though some morphologically domestic goats are present at Asiab (B6k6nyi 1973). However, the bulk of the slaughtered ani- mals do not seem to have been herded (Hesse 1982a). Three speculations might be offered. First, the footprints found in the Level D bricks suggest that some adult Caprini were highly socialized in the Tepe Ganj Dareh community during the development of pastoralism. Two tentative functional explanations for their presence can be offered— as ritual animals as suggested by the use of horn cores (Ovis orientalis) as decorative devices, or as producers of dung for fuel, the nature of their droppings being such tha it can only be effectively collected when the animals are penned together. Either, ad- mittedly unsupported, suggestion would place an insurance herd in the Tepe Ganj Dareh community under different conditions than hunted animals. May 1982 HESSE 9 Rapid increase in the insurance herd, and goats can increase at a rate of 33% per year (Dahl and Hjort 1976:231), would lead to ecological destabilization, reorientation of social priorities, and the establishment of the pastoral way of life. $ - q Y FIG. 3—Location of third millenium B.C. sites in the Salar de Atacama region of northern Chile. 10 HESSE Vol 2, No. 1 The Salar de Atacama Region The second case study concerns the arid mountainous regions bordering the Salar de Atacama in northern Chile. The study region, which is centered around the town of San Pedro de Atacama contains three broadly defined habitats. The first of these is the Salar, a basin with a floor located at an elevation of approximately 2500 m. Above 4000 m is the Puna de Atacama, a rugged area dotted with lakes that are the habitat of several varieties of flamingo. According the Niifiez and Dillehay (1979:39-40), this region should not be considered similar to puna regions to the north in Peru. They emphasize the harsher nature of the more southerly climate, the more dissected nature of the terrain, which is dotted with volcanoes, though both areas share general climatic instability (Winterhalder and Thomas 1978). Between the puna and the salar is a sloping region cut by a series of small canyons through which water flows year round. Between these canyons the slopes are extremely desertic at the lower elevations, though they do contain important resources including raw material for stone tools. The earliest occupation of the region that is known from excavations is represented by the sites of Tuina, located to the north and west of the Salar, and San Lorenzo, locat- ed to the east, both of which have radiocarbon determinations indicating habitation in the eighth and ninth millenium B.C. Following the occupation of these sites there is a hiatus, possibly brought on by volcanic activity (L. Niifiez, personal communication), until approximately 3000 B.C. The herbivore resources of the Atacama region are more restricted than what is found in the more watered regions to the north. None of the fourteen sites in the region which I have studied have any deer remains. Only one fragmentary metopodial is known from the Chiichiad complex located just to the north (M. Druss, personal communica- tion). Meat supplies were procured through the exploitation of camelids with supple- cluding the flamingo, the Andean goose and the tinamou. The relative frequency distri- bution of the remains of these species in the various quebrada and salar sites indicate clearly that the hunting techniques of their occupants were finely tuned to the biological subtleties of the region. tological issue), apparently diverge somewhat in habitat preference, diet, and social behavior (Koford 1957; Raedeke 1977; Rick 1980). The vicuna tends to higher altitudes, as a more rigidly defined territorial system, and prefers grasses, whereas the guanaco 1s a browser with a looser social structure that occupies somewhat lower elevations. In terms of the Atacama region, vicuna would be common in the puna, guanaco in the quebradas. Because of these behavioral divergences, it is reasonable to assume that hunters would employ somewhat different strategies in killing them. The unfortunate result of this line of reasoning is that unsegregated archaeological samples will tend to mix the evidence of different human behaviors. Wing (1972, 1977), however, has shown that the size difference between the two species can be used to separate archaeological (Fig. 4). The sites of interest are Tulan-52, one of a cluster of Archaic sites located just = 3000 m along a quebrada leading to the south and east of of the salar, and as ‘ located at about 3200 m above a canyon that leads to the north and east (Ninez 1981). May 1982 HESSE 11 % 100- 9077 i 3° Ss 27 a Age stage FIG. 4—Camelid harvest profiles from Tulan-52 (—-) and Puripica-1 (————). The percentage sur- vivorship is plotted against a series of relative age stages following the suggestions of Wing (1972). Radiocarbon dates for each site suggest occupation during the mid to late third mille- nium B.C. The assemblages from both sides include simple stone shelters and mortars. In the case of Tulan-52, it is estimated that there were about 30 semi-subterranean dwel- lings; in the case of Puripica-1, 40-50. The lithic assemblage from Tulan-52, however, has a higher proportion of projectile points. The excavations at Puripica-1 produced a series of petroglyphs depicting camelids. Both sites contained a mixture of large and small camelids (presumably guanacos and vicunas). The vicunas, however, were represented almost exclusively by toe bones. I would suggest that this represents butchering of these small camelids at some distance from the sites, probably reflective of the facts that vicufia territory would have been located at some distance above the sites and only the skins were being imported. A similar interpretation has been drawn by Rick (1980:273) and Simons (1980) for samples that probably contain a mixture of large and small came- lids. The majority of these animals were mature at death, at least based on the percentage of fused phalanges, a harvest profile in keeping with wild herd demographics (Hesse 1982b). The harvest profiles for the guanacos, on the other hand, contrast between the two sites, with a much greater proportion of the Puripica guanacos being taken young. More specifically, the shift to this emphasis on young large camelids occurs between levels III and II in the occupation. If young kill is accepted as evidence for domestication regions to the north where pastoralism is much earlier documented. Nothing in the archaeological record for the site, however, argues for extensive extra-regional contact. How is climate involved in this? Two factors stand out. Snowfall on the puna has been identified as a significant cause of camelid mortality. Koford (1957:164) alludes to the effects of snow and hail which is particularly frequent during the birth season on the puna, Alternatively (Jane Wheeler, personal communication), oscillating periods of €xtreme drought which are recorded in the puna of southern Peru have been identified a8 primary causes of camelid mortality. Browman (1974:191) notes the effects of both Conditions on puna herds of domestic camelids: 12 HESSE Vol. 2,No. 1 Loss by freezes and snows and through drought and consequent pasture failure also runs high in some years. Several of Diez de San Miguel’s 16th century informants reported losses of animals killed by freezes and heavy snows; the Lurinsaya of Acora claim the loss of the entire communal herd of 1,000 in a freeze in 1565. Even though the wet season is the summer season, at the high elevations (4000 m and above) the rain turns to snow, which prevents the flocks from finding forage and causes high losses among the weaker newborn animals. Further, in the classic ethnography of Andean pastoralism, Flores-Ochoa (1979:95) em- phasizes the danger of droughts which are characterized as a main hazard for the llama and alpaca herders he describes. Bowman (1924:30) reports the history of 19th century precipitation for the Ata- cama region. Heavy downpours of snowfalls occurred in 1819, 1823, 1952, 1859, 1878, 1885, 1903, and 1911 (from Druss 1978:116). Such precipitation would have created short-lived explosions in the desert vegetation at elevations below the puna, but would ve had a devastating effect on the high altitude camelids. If Nijiez and Dillehay (1979) are right in their assessment of the inhospitable nature of the Puna de Atacama for year- round habitation, then we can expect that the hunters moving up and down the quebra- das to the east of the salar would experience irregular periods of lack of game during their visits to the highlands. The best evidence that the occupants of the Atacama region were experiencing oscillations in moisture availability has been published by Druss (1977, 78). His study of the Chidchiti complex produced correlations linking settlement type, location and duration with climatic regimes inferred from artifactual and ecofactual remains (1977:Table 4). He concluded that third and second millenium B.C. settlement patterns were controlled by oscillations in moisture. in, while the case is circumstantial in that direct evidence of resource stress Can- not be achanciaeeaty documented, I believe that it is a reasonable conclusion that the ancient inhabitants of the Atacama region experienced serious fluctuations in game avail- ability brought on by expectable but not predictable climatic variation. In the case of locally abundant mineral resources and trade in vicuna skins. CONCLUSION The two case studies just presented have a number of features in commo episodes of pastoral development took place in arid or semi-arid ecosystems ae ae large herbivore population was dominated by a minimum number of species, one basically a browser, the other a grazer. The sites which contain the evidence for ee are located in complex topographies adjacent to larger open spaces. Each environment w probably afflicted by irregular climatic conditions capable of catastrophically oe game availability. Domestication has both a technological aspect, the methods of taming and breeding, and a social/ideological one, the intersection of animals with a web of values and inter- personal organization. The explanation of its onset must account in some manner for both of these aspects. Oscillating game availability can be linked to cultural systems both in terms of an effect on productive technologies and on social institutions. The potential result of these effects is carnivorous pastoralism, a predatory adaptation capable of in- suring its success by destroying the potential for other systems. May 1982 HESSE 13 ACKNOWLEDGEMENTS The research reported here was supported by the Smithsonian Institution in the form of a Pre- doctoral Fellowship and grants from the Paleoclimatology and Latin American funds, as well as by a Faculty Research Grant from the i of Alabama in Birmingham. I w real for encouraging me to and Lautaro ae Atencio of the Universidad del Norte in Antofagasta for ful and considered comments contributed much to whatever merit the paper has. The deficiencies, of course, remain the property of the author. LITERATURE CITED ALLEN, T.F.H. 1977. Neolithic urban pri- m0 the case against the invention of agri- culture. J. Theoretical Biol. 66:169-180. BATES, pr 19 Nomads and Farmers: A Study of the Yoruk of Southeastern Tur- _ Univ. Michigan, Anthropol. Papers, Arbor 52. Ann rt aoe W. 1964. Attitudes towards animals a ature in a Great Plains com- munity. sie ae 9(23):3747. BINFORD, LEWIS R. 1968. Post-pleistocene adaptations. Pp. 313-41 in New pple in Archaeology (S.R. Binford and L.R. Bin ford, eds.), Aldine Publ. Co., Chicago. BOBEK, H. 1963. Nature and implications of 1975. Some Act ed.), Akademiai Kiadé, Budape 1978. The Animal Remains From Four Sites i in the Kermanshah Valley. Brit. Archaeol. a Supplem. Ser. No. 34, Brit. - Repts ndon. 1924. Desert Trails of the Atacama. ii Geogr. Soc. Spec. Publ. No. 5, New Y BRAIDWOOD, come G., BRUCE HOWE, and CHARLES A. REED. 1961. The Iranian Prehistoric project. Science 133:2008-2010. BRAY, WARWICK. 1976. From predation to .E. Pega eds.), Westview Press, Boulder, BROWMAN, DAVID L. 1974. Pastoral nomad- ism in the Andes. Current Anthropol. 15(2): 188-96, CLABBY »J- 1976. The Natural History of the Horse, Taplinger Publ. Co., New York. COE, MJ., D.H. CUMMING, and J. PHILLIP- SON. 1976. Biomass and production of large African herbivores in relation to rain- imary production. Ocecologia (Berl.) 22:341-54. st as MARK. 1977. The Food a pe in . Yale Univ. Press, New 951. Cave sacs ibs: in ak: 1949, Museum Monographs, The University Museum, Philadelphia DAHL, GUDRUN, andl ANDERS HJORT. Univ. of Stockholm DRUSS, MARK. 1977. Computer analysis of Chitchid Pak ops settlement patterns. El Dorado ae he orp:s i Environment, Subsistence my, pie Settlement Patterns of the PhD. dissert. jts recognition in faunal assemblages. Pea- body Mus. B 153-5 FORES-OCHOA, JORGE A. 1979. Pastora- lists of the Andes. Trans. by Ralph Bolton. i Boreal a haya Publ. No. 1, Univ. Alberta, Edm ered VALERIUS. "1971. eer Sheep. Univ. Chicago Press, Chic HARRIS, Saati pi ‘ viel Aieratv path- ways toward agriculture. 79-243, in pene on Agriculture (Charles . Reed, ed.), Mouton Publ., The Hague. HEDIGER, H. 1968. The Psychology and Behavior of Animals in Zoos and Circuses. Dover Publ. Inc., New York. 1978. Evidence for Hus- eolithic Site of G Unpubl. Ph.D dissert. a ), Columbia Univ. 9, Rodent remains and seden- 14 HESSE Vol. 2, No, 1 LITERATURE CITED (continued) tism in the neolithic: evidence from Tepe Gahj Dareh, western Iran. J. Mamm. 60: 856-857. 1982a,. Slaughter patterns and siesenieation: the a of pastora- lism in western Iran, Man, in pre 1982b. haben te for melid exploitation in the Chilean Andes. Stsietieisiate Mitteilungen, in press. HICKERSON, HAROLD. 1965. The Virginia Deer and intertribal buffer zones in the upper a banat Pp. 43-65, in Man, Culture and Animals (Anthony Leeds and papas P, Vays, eds.), Publ. 78, AAAS, Washington HIGGS, ERIC A ane M.R. JARMAN. 1972. The origins of animal and plant husbandry. 3-13, in Papers in Economic Prehistory (Eric i hat ed.), Cambridge Univ. Press, Cambri HIGHAM ~ w. F, 1968. Trends in European INGOLD, TIM. 1974. On reindeer and men. Man (n.s.) 9(4):523-38. ———. 1980. Hunters, Pastoralists and eae Cambridge Univ. Press bridge. sae — 1969. The Economy of Citi andom House, New York. seer, ro 1974. A comparison of the social igen of the ike sits 124-32, t Relati Management (V. Geist and F. Walters, cs) IUCN (n.s.) No, 24, Morges, Switzerlan earobs, tee 1957. The vicuna and the puna. Ecol. Monogr. 27(2):153-219. LEVINE, LOUIS D. 1976. Archaeological pita eg in the Mahidasht. Paleorient 2(2) -90. eee gaitaced 1944. The Wolves of Mount McKinley U.S. Govt. Printing Office, Washington. MURPHY, E.C., and K.R. WHITTEN. 1976. a0 rig oc in McKinley Park (Capra ibex L.) in seinem Lebensraum. Mammalia Depicta, Verlag Paul Parey, Ha eae rg. NUNEZ, LAUTARO. 1981. Paleoindian and archaic cultu periods in the arid and semiarid regions of north Chile. Unpubl. d TOM D. DILLEHAY. 1979. emai’ Giratoria, Armonia Social y Des- arrollo en los Andes Meridionales: Patrones de Trafico e Interaccion Economic. Direc- al de ce Universidad del Norte, Antofagast ANDER, T. as. The Zagros Streams. Syracuse Geographical Ser. No. 1, Syracuse Univ. Press, Syracuse. : PAINE, R. 1971. Animals as capital: compari- sons among the northern nomadic herders d hunt pve as Quart. 44(3): PAYNE, SEBASTIAN. 1973. Kill-off patterns in sheep and goats: the mandibles from svan Kale. Anat. Stud. 23:281-304. RAEDEKE, oa j. 1977. The repro- ductive ecology o ic6e) of Tierra aa Fuego, Chile. Paper apes at the 57th Annual Meeting of the r. Soc. of Mammalogists. RICK, 70nN W. 1980. Prehistoric Hunters of the High Andes. Academic Press, New York, SCHALLER, GEORGE B. 1977. Mountain Monarchs, Univ. Chicago Press, Chicago. SIMONS, DWIGHT D. 1980. Man and guan- Archae SMITH, PEEL on Ganj Dareh Tene, Iran. Amer. J. Arch. 82(4): ait -540. . and PEDER MORTENSON. 1.4. Three new ‘early neolithic’ sites in western ran. Curr. Anthropol., in press ls ag rig F. 1975. The mam- malian fauna of Warw ock shelter, west- central me Se oraaiedl 33(8):141- 155. VALDEZ, R., F. NADLER, and T. D. og sare: Evolution of wild sheep . Evolution 32:56-72 VAN mee. W. 1969. Reflections on prehis- Dimbleby, eds.), Duckworth, 5S VITA-FINZI, 9, Late Quatern uvial chronology of Iran ecu luvi Rundschau 38(3): 951-973. WING, ELIZABETH S. 1972. agen of mal K. Terada, eds. Jt Univ. Tokyo Press, Tokyo. . 1977. Animal domestication in the psithe’s Pp. sae in Origins of Agricul- May 1982 HESSE LITERATURE CITED (continued) YOUNG, T. CUYLER, JR., and PHILIP E.L. SMITH. 1966. Research in the prehistory ture (Charles A. Reed, ed.), Mouton Publ., e. of central-western Iran. Science 153:386- WINTERHALDER, B.P. and R.B. THOMAS. 1978. Geoecology of Southern Highland Peru. Institute of Arctic and Alpine Re- search, Univ. Colorado Occ. Paper 27. ZEUNER, F.E. 1963. A History of Domes- ticated Animals. Hutchinson, London, ON men ieieesrnessneites J. Ethnobiol. 2(1): 17-38 May 1982 TRADITIONAL USE OF DEVIL’S-CLUB (OPLOPANAX HORRIDUS; ARALIACEAE) BY NATIVE PEOPLES IN WESTERN NORTH AMERICA NCY J. TURNE Research Associate, pee Columbia Provincial Museum Victoria, B.C., Canada V8V 1X4 ABSTRACT.—Devil’s-club (Oplopanax — Araliaceae) is a deciduous, spiny shrub which was and still is an important medicinal plant for many Indian peoples in western North America. Its traditional uses involve both physical and spiritual realms of medicine. The inner b were used eat rheumatism and arthritis, stomach and digestive ailments, tuberculosis, colds, skin disorders, diabetes, and many other ailments. Extracts powers. Special ae powers were attributed to it, presumably because of its prickli- ness. Its wood was used for — lures and the charcoal as a pigment in a protective face paint for ee dancers. Devil’s-club was named in almost every Native language use within its geographic range. "Theis are some 13 to 15 known separate etymons for it in more than 25 different languages. In most languages, the derivation of the name is presently unknown. More pharmacological research on this plant is needed, ““... Behold! there was a devil’s-club tree larger than any other tree in the whole world. He [the son of Devil’s-club] took his stone ax and felled the great devil’s-club tree; and after it was down, he took all the sap and bark; and ... he carried it down to his town... Then he started to wash his body with the bark of the devil’s-club and its sap, and he ate some to purify himself, He did so for forty days...” (from a Tsimshian myth—Boas 1916:175). INTRODUCTION Devil’s-club (Oplopanax horridus [J.E. Smith] Miq.; Araliaceae) is a well-known shrub of western North American forests. The stems and foliage are densely armed with stiff spines that “ . . . break off at once on entering the skin or clothing and make life a burden to the prospector, explorer, or mountain-limber . . . ” (Gorman 1896:73). Nevertheless, despite its sharp, menacing spines—or perhaps in part because of them—it was respected as a protective agent and important medicinal plant by many indigenous Peoples in western North America. Few medicinal plants were more widely and con- sistently used within their geographic range. Devil’s-club wood was also used in tradi- tional fishing technology along the Northwest Coast and the charcoal was used as a decor- ative and protective pigment. This paper summarizes the many uses of this plant in Native cultures, and stresses the medicinal properties implied by its sana ae usage. It is potentially valuable to modern medicine because preliminary research (Justice 1966) indicates that at least some of the traditional remedies involving devil ‘echub may have a sound biochemical basis. DISCUSSION Botanical description of Devil’s-Club Oplopanax horridus is sometimes cited in botanical and cna yee literature under the Synonyms: Fatsia horrida Benth. & Hook., Panax horridum J.E. Smith, and Echino- panax horridum Decne. & Planch. It is in the ginseng family, only one other member of which is indigenous to western North America, namely Aralia nudicaulis L., wild sarsa- Parilla. Both plants are related to the true ginseng, Panax quinquefolius L. oe Aralia 18 TURNER Vol. 2, No. 1 quinquefolia Gray), well known in folk medicine and native to eastern North America. The Oriental ginseng, P. genseng C.A. Mey, is even more prized in Chinese folk medicine (Li Shih-Chen 1973; National Academy of Sciences 1975:102) Devil’s-club is a deciduous shrub of 1-3 m (or more), with long, thick, meeonaegl or decumbent stems and large, palmately lobed leaves with blades up to wide, irregularly serrate margins, and long petioles (Fi ig. 1). They superficially resemble greenish-white flowers bloom from May to July, depending on elevation and latitude. They are subsessile in compact umbels borne in elongate racemes or panicles up to 25 cm long. The fruits are bright red, fleshy berries, somewhat compressed, ellipsoid, and often spiny (Hitchcock et al. 1961 (Pt. 3):506). FIG. 1—Devil’s-club (Oplopanax horridus). Approximately 1/6 natural size. This shrub often grows in dense, nearly impenetrable thickets in moist, rich soil in coniferous woods. It is especially common near streams and occurs from near sea level the coast and on the west side of the Cascade range to southern Oregon, and eastwards to the Rocky Mountains including parts of Idaho, Montana, and Alberta. It also occurs in a small enclave in northern Michigan and the Thunder Bay district of Ontario. Another species of Oplopanax, O. japonicus (Nakai) Nakai (sometimes considered a subspecies of O. horridus), occurs in Japan. Hultén (1968:696) provides a distribution map for O. horridus. The somatic chromosome number for the species is 2n=48 (Tay lor and MacBryde 1977:53), Deovil’s-club in Folk Medicine In Native cultures of northwestern North America, health and the maintenance of well-being seem to assume two general aspects: 1) physical, i.e., the use of various May 1982 TURNER 19 edicinal preparations, usual- ly herbs, whic ini practices of shamans, or “In- dian doctors,” who deal more with the evil spirits associated with illness than with its physical manifestations (Tur- ner et al. 1 7150; Turner and Efrat In Press; Turner et - In Press). Devil’s-club played, and still plays, an important role in both these types of medicine, al- though in some instances the two are so closely inter- mingled that separating them completely would be unrealis- tic. Nevertheless, for the purposes of this paper, it is a useful dichotomy and exam- ples of the use of devil’s-club in these two aspects of medi- cine are given in Tables 1 and 2 respectively. ae FIG.2— Devil’s-club stem, showing thin, sharp spines. Approximately 1/5 natural size. “Physical” Attributes of Devil’s-Club Medicine The chemical properties of devil’s-club, as they might relate to the uses in Table 1, have not been thoroughly investigated. Japanese researchers have isolated a sesquiter- pene, a sesquiterpene alcohol, and a sesquiterpene ketone from the closely related O. japonicus (“‘haribuki”). These are, respectively, echinopanacen (Cj5 H24), echinopana- col (Cis H25 OH), and oplopanone (C15 H96 Og) (Takeda et al. 1966:219). A deriva- tive of oplopanone is used commercially in Japan as an antipyretic and antitussive drug for coughs and colds. Undoubtedly these compounds are also present in O. horridus, but further details of the chemical composition of this plant are apparently not known. Extracts of Oplopanax, like those of its relative, ginseng, have marked hypoglycemic Properties (Lewis and Elvin-Lewis 1977:218; Justice 1966;37). The hypoglycemic attri- utes undoubtedly contribute to the use of devil’s-club to treat diabetes: “Our attention was brought to this material through the examination by one of us of a Surgical patient who on hospitalization, developed marked symptoms of diabetes. This Person, it was learned, had kept in apparent good health for several years by oral doses of an infusion of this root bark, and is in fact still leading a normal life with the aid of this infustion” (Brocklesby and Large 1938:32). 20 TURNER TABLE 1.—Medicinal Uses of Devil’s-Club (Physical Aspects) * Native Group Tanaina (Kenai) Tanaina (Upper Inlet) Tanaina (Upper Inlet) Eyak Tlingit Tlingit Tlingit ingit Kaigani Haida Tlingit or Kaigani Haida Tlingit or Kaigani Haida Tlingit or Kaigani Haida Tlingit or Kaigani Haida Tlingit, Haida or Tsimshian Tlingit, Haida or Tsimshian Haida Details of Use dec of st drunk for fever dec of inner bk drunk for tuberculosis, stomach trouble, coughs, colds, and fever inner bk of rt baked, powdered, used as poultice on swollen glands, boils, sores, other infections | dec (?) drunk as emetic, purgative warm dec of seal oil and inner bk drunk as emetic, cathartic inner bk chewed, tied onto wounds to relieve pain, prevent blood poisoning ashes used for sores inf of bk, rt drunk for general strength, colds, chest pains, arthritis, black eyes, gall stones, ulcers, constipation, tuberculosis inf of inner bk drunk for cancer inner bk chewed, spit on wounds as emergency analgesic inner bk laid on skin over fracture to reduce pain, swelling inner bk or rt dried, pulverized with pitch, applied to skin abrasions dried inner bk laid into tooth cavity for pain relief inner bk pulverized, mixed with oil, eaten for pain relief dec of inner bk in sea water solution drunk for 9 days for rheumatism, arthritis; laxative VoL?2 Nov! Reference Kari 1977:62 Kari 1977:62 Kari 1977:62 Smith 1973:330 Smith 1973:330 Smith 1973:330 Krause 1956:284 Justice 1966:36 Justice 1966:36 Justice 1966:36 Justice 1966:36 Justice 1966:36 Justice 1966:36 Justice 1966 :38 Turner 1970:66 *The following abbreviations are used: dec - decoction; inf - infusion; st - stem(s); tt - root(s); and bk - bark May 1982 TABLE 1.—Continued Native Group Haida Haida Haida Haida Tsimshian Gitksan Gitksan Gitksan Gitksan Southern Carrier Northern Carrier Central Carrier Central Carrier Carrier Carrier Bella Coola TURNER Details of Use dec of inner bk drunk for ‘‘tuberculosis of the bone” bk chewed, juice swallowed for bad cold or general sickness berries rubbed on hair and scalp of children against lice, dandruff st used to beat rheumatic limbs as counter-irritant dec drunk for unspecified illness dec of st taken as purgative in treating gonorrhoea dec of st taken to knit broken bones dec of st, with Viburnum, taken as diuretic, purgative for “strangury,” rupture, or any sickness bk mashed with fern rt, Abies bk, Pinus or Picea gum, Lysichitum rt, and applied warm to boils, ulcers, for rheumatism, lung haemorrhage dec of bk drunk as purgative before and after childbirth inner bk swallowed for stomach and intes- tinal cramps, esp. after taking a purgative; itself a purgative inner bk swallowed for general sickness bk scraped, plastered over sore area bk used by women after childbirth; mashed, swallowed immediately after to help expell afterbirth bk mashed, swallowed as purgative inner bk, esp. of rt chewed as emetic; taken with water Reference Turner 1970:66-68 Turner 1970:66-68 Turner 1970:66-68 Turner 1970:66-68 Smith 1973:330 Smith 1928:62 Smith 1928:62 Smith 1928:62 Smith 1928:62 Smith 1928:62 Smith 1928:62 Carrier Linguistic Committee 1973:82 Carrier Linguistic Committee 1973:82 Morice 1893:132 Morice 1893:132 Smith 1928:62 22 TABLE 1.—(Continued) Native Group Bella Coola Bella Coola Bella Coola Bella Coola Bella Coola Bella Coola Heiltsuq (Bella Bella) Heiltsuq (Bella Bella) Heiltsuq (Bella Bella) Southern Kwakiutl Southern Kwakiutl Southern Kwakiutl Southern Kwakiutl Ohiat Nootka Nitinaht Nitinaht Mainland Comox Mainland Comox TURNER Details of Use inf of st in sea water drunk as emetic dec of st used in steambath for stomach trouble, rheumatism dec of bk of rt and st drunk as purgative, and for rheumatism dec of st with Ribes rt drunk as general tonic dec of inner bk or rt with Sorbus bk, Ribes st used for steambathing, e.g. for lameness inf or dec of rt or st drunk or used in steambath for many illnesses inf of rt drunk for diabetes inf of inner bk drunk as laxative, used for bathing inner bk chewed, then salt water drunk, as laxative 4 pieces of rt held in mouth, juice swallowed, for stomach pains, constipation inf of bk drunk for tuberculosis, other ailments ashes mixed with oil, rubbed on swellings dec of despined bk, with Lomatium seeds, or with sea water, urine, used in steambath for body pains dec used in bath for arthritis, rheumatism inf of despined st drunk for arthritis inf of bk taken for rheumatism; with Alnus, Abies bk for tuberculosis bk, rt inf in bath as skin tonic inf of bk drunk to stop internal haemor- rhaging, sometimes taken with Ledum tea Vol. 2, No. 1 Reference Smith 1928:62 Smith 1928 :62 Smith 1928:62 Bouchard 1975- 77:B.,5 Bouchard 1975- 77:B., 5a Bouchard 1975-77: B, 5a MacDermott 1949:181 B. Rigsby, pers. comm. 1981 B. Rigsby, pers. comm. 1981 Turner and Bell 1973:278 Turner and Bell 1973:278 Boas 1966:382 Turner and Bell 1973:278 Rollins 1972:25b Turner et al. In Press Rollins 1972:25b Bouchard 1973 | Bouchard 1973:7 May 1982 TABLE 1.—(Continued) Native Group Sechelt Sechelt Sechelt Sechelt Sechelt Sechelt Sechelt Squamish Squamish Squamish Cowichan Halkomelem Cowichan Halkomelem Upriver Halkomelem Lummi Skagit Cowlitz Quileute Lillooet Thompson TURNER Details of Use inf of inner bk used in steambath for lameness, arthritis, rheumatism weak dec of inner bk, rt drunk for diabetes dec of bk, rt used as wash for skin disease inf of inner bk drunk as “tonic” charcoal, with oil, poultice for burns dec of bk, rt taken for rheumatism, other ailments dec of bk, rt applied externally for skin disease dec of inner bk, with Abies, taken for diabetes inner bk used in steambath for rheumatism inner bk chewed to clear throat used in sweatbath to drive away sickness, for colds, conditioning dec of bk drunk for measles, esp. in children st taken for arthritis inner bk laid on women’s breasts to stop excessive lactation dec of bk, with Chimaphila, Rhamnus, drunk for tuberculosis, and to start post-partum menstrual flow dec of bk drunk for colds; used to wash rheumatic limbs plant as unspecified medicine dec of despiged st drunk for arthritis inf of st drunk for indigestion, stomach troubles; dec as tonic, blood purifier Reference Bouchard 1977:9; Rollins 1972:25a Bouchard 1978:8 Rollins 1972:25a Bouchard 1977:8 Bouchard 1977:8 Turner and Timmers Turner and Timmers 72:8 Bouchard and Turner 1976:71-72 Rollins 1972:25a Bouchard and Turner 19 76:71-72 Rollins 1972:3 Rollins 1972:10 Galloway 1979:7 Gunther 1973:41 Gunther 1973:41 Gunther 1973:41 Reagan 1934:65 Turner 1972:13 Turner et al. In Press TABLE 1.—(Continued) Native Group Thompson Thompson Okanagan-Colville Shuswap Kootenay Crow, Cheyenne Sahaptin TURNER Vol. 2, No.1 Details of Use Reference charcoal mixed with grease, as salve for Steedman 1930:459 swellings, sores inf of despined st drunk for flu, weight loss, Annie York, pers. and other ailments ‘comm. 1981 inf of rt or despined st drunk for Turner et al. 1980: tuberculosis, dry coug 73 dec drunk by some women for several Teit 1909:584 days after childbirth dec taken as medicine for any illness Hart et al. 1981:54 ?rt spaneeis with tobacco for headache Johnston 1970:316 (id ent. uncertain in orig. source - Slankistido, 1905:12 dec of wood, inner bk drunk for D. French, pers. “tuberculosis” (splitting blood) comm, 1981 TABLE 2.—Uses of Devil’s-Club in “Spiritual” Medicine* Native Group Eyak Tlingit Tlingit ? Tlingit Tlingit Haida Haida Haida Haida ‘rt eaten by novice shamans for purification Details of Use Reference important in magic Smith 1973:330 Krause 1956:195 st used to whip suspected witches Krause 1956:203 dried bk mixed with red ochre as love charm Justice 1966:36 rt chewed by shamans to augment hypnotic Gorman 1896:73 powers st hung over doorways to protect against Turner 1970:67 witchcraft st eaten, with Moneses to gain supernatural Swanton 1905 :212 powers st, bk eaten to bring luck in gambling Newcombe, unpbl., ca. 1901 st, bk chewed for ritual purification of Newcombe, unpubl., gamblers, hunters, sick people ca 19 *For abbreviations used, see Table 1 footnote May 1982 TURNER 25 TABLE 2.—(Continued) Native Group Details of Use Reference Tsimshian inf of inner bk for removing odours M. Seguin, pers. (see also Table 4) comm. 1981 Tsimshian dec drunk, used in bath to gain supernatural Barbeau 1961:73 Power Tsimshian inner bk chewed, rubbed on body to bring Boas 1916:172 luck in hunting Bella Coola st charm against supernatural power Turner 1973:201 Bella Coola st hung in house, used as fumigant, to ward Bouchard 1975- off “‘strong sickness” 77th5 Southern Kwakiutl st attributed magical powers Turner and Bell 1973:278 Southern Kwakiutl st hung with Veratrum rt around child’s neck Turner and Bell to ward off sickness 1973:274 Central Nootka ashes mixed with water drunk for strength Fenn et al. 1979:35 Nitinaht charcoal used in protective face paint for Turner et al. In Press ceremonial dancers Upriver charred st, mixed with grease, used as Galloway 1979 Halkomelem protective face paint Lummi charcoal used, often with red ochre, as J. Thomas, pers. ceremonial face paint; associated with death comm. 1981 Okanagan- medicine (see Table 1) must be made in Turner et al. 1980:73 Colville secret; would lose effectiveness if even another person’s shadow passed over it Crow, Cheyenne, ? used by medicine-men in their incanta- Johnston 1970:316 Blackfoot tions (ident. undertain - see Table 1) A similar observation was made by another doctor, G.E. Darby, at Bella Bella, who reported that the local Indians and at least one non-Indian were using an infusion of the roots for diabetes (MacDermot 1949). Margaret Siwallace, a Kimsquit woman living at Bella Coola, also knew of a local Caucasian woman who took devil’s-club (probably as an infusion) for diabetes (Turner 1973:201). Another study (Graham and Noble 1955), reported by Justice (1966:37), found that the dried roots and stalks of devil’s-club contained a drug that substantially inhibited the effects of a pregnant mare’s serum upon the growth of a rat’s ovaries; this property may well relate to the use of devil’s-club as a post-partum treatment for women (Table 1). MacDermot (1949:18 1) noted that the plant has “. . . apparently a hygroscopic and detumescent effect on swellings,” but did not elaborate. His comment may be based on 26 TURNER Vol. 2, No. 1 personal observation, or may simply be his conclusions from knowledge of how it was used. The fresh plant and extracts made from it have a characteristic sweetish odour. The late George Young, a Haida man from Skidegate who had taken the “devil’s-club treatment’? for arthritis, apparently with remarkable success, re- called that shortly after one had drunk the decoction of devil’s-club, he could smell it from his joints (Turner 1970). Annie York, a Thompson woman keeps a supply of de-spined, dried devil’s-club sticks on hand to use when required (Fig. 3). She makes an infu- sion by steeping four short (2-3 cm) pieces in about a liter of boiling water. (i.e., half a cupful) before meals, to relieve weight loss, flu, and other ail- ments. She warns that it can cause too much weight gain. She had heard that the roots could be taken for diabetes (A. ork, personal communication FIG. 3—Dried, de-spined devil’s-club sticks, kept 1981). for use as medicine for flu, excess weight loss, and Aside from Young’s and York’s, other ailments, by Annie York, a Thompson there have been many testimonies as woman from Spuzzum, B.C. Approximately 1/4 : ; : natural size. to the efficacy of devil’s-club as a medi- cine. Justice (1966:38) notes several. One was a Chief of one of the Alaskan villages who took it for a red, painfully swollen finger that was unreleived by the prescribed treatment of aspirin, raising the hand, and heat. He took one glass of devil’s-club extract, which relieved the symptoms completely in eight hours. Another was a case of four teenagers who used the dried inner bark laid directly into a tooth cavity and experienced prompt pain relief. Adult males reported that they had applied the stalk strips to axe wounds received in the bush, sufficiently relieving the pain to enable them to continue on until they came to medical attention. Yet another case is described by Justice (1966:38) where a male patient with metastatic adenocarcinoma [secondary malignant tumour] was discharged from the hospital with a few month’s prognosis and a terminal supply of morphine. Three years later, he had regained his health and strength after extensive treatment with devil’s-club extract. owever, Justice (1966:38) also notes instances where devil’s-club had no notice- able effect as a medicine. A patient with advanced rheumatoid arthritis and ankylosis of most of her small joints reported no benefit from the extract. A 54-year-old woman with Hodgkins disease was taking the extract regularly and was also receiving more Con” ventional medical treatment at the hospital but no mention was made of the status of her condition, with the implication that the treatments were having no apparent effect. Although devil’s-club extract is not known to be toxic to humans or animals, peop! who drink it regularly report that upon beginning the treatment one may have diarrhea and feel very weak, and that greater weakness is experienced if alcoholic beverages are taken concurrently (Justice 1966:37). Furthermore, hares given the devil’s-club extract in the tests by Brocklesby and Large (1938) had more fatty degeneration of the liver than control animals. No increased tolerance was observed after repeated tests. Effects on the liver of humans are presently unknown. May 1982 TURNER 27 “Spiritual” Aspects of Devil’s-club Medicine various protective and purifying properties attributed to devil’s-club, as shown in Table 2, seem closely related to its laxative and emetic properties, and to its most striking physical feature—its prickliness External and internal cleansing was of paramount importance in Native cultures in the quest for guardian spirit power (to bring success in hunting, gambling and other activities), in the acquisition of shamanistic powers, and for living in general. For exam- ple, ritual scrubbing and bathing, fasting, and the drinking of cathartic tonics often ac- companied the adulthood training of young men and women reaching puberty (Turner et al. In Press). ence, it is difficult to distinguish between the use of the cathartic qualities of devil’s-club as a physical treatment for sickness and their use as a psychological aid to obtain supernatural powers. As Justice (1966:37) points out, the hypoglycemic proper- ties of the plant may well have promoted the abilities of shamans (and initiates) to enter a trance-like state, conducive to having visions of supernatural spirits. He further notes that, “The legend [see later discussion] of the shaman’s increased strength after one week of only the extract [of devil’s-club] for food may be related to increasing tolerance to the hypoglycemic effects.” The prickly, or “sharp” quality of devil’s-club seems closely associated with the plant’s ability to provide immunity against “‘witchcraft”, evil spirits, or people with malicious intent and to bring luck and “power” to the user of the plant (Table 2). A similar protective role is assumed by other species of thorny or prickly plants in western Native cultures. These include: wild roses (Rosa spp.), Rocky Mountain juniper (Juni- perus scopulorum Sarg.), Oregon-grape (Berberis aquifolium Pursh), black hawthorn (Crateagus douglasii Lindl.), swamp gooseberry (Ribes lacustre Poir.), thistles (Cirstum spp.), and trailing wild blackberry (Rubus ursinus Cham. & Schlecht.) [Rollins 1972; Turner and Bell 1971; Turner 1973; Turner et al. 1980, In Press). With all of these plants, including devil’s-club, it is not the prickles or spines per se that give protection; rather it is some innate quality that is manifested in an infusion or decoction of the plant (Turner et al. 1981 :131), or even in smoke from burning it (Turner 1973:206) he close relationship of the protective powers of devil’s-club with its prickliness is alluded to by John Thomas, a Nitinaht speaker from the west coast of Vancouver Island: “The reason they use this kind of wood [as charcoal face paint for ceremonial dancers — see Table 2] is because it’s sharp. When you see somebody with that kind of paint, you couldn’t look them in the eye, their power is so strong...” (Turner et al. In Press). John Thomas (personal communication, 1981) also explained that within his own thin (Nitinaht), and among neighbouring Coast Salish groups, devil’s-club is considered red.” Along with red ochre paint, it is considered to be a link between the ordinary, - si world, and the supernatural, or the spirit world. He pointed out that in a recent reburial ceremony at Lummi, Washington (Coast Salish territory), which was film- €d and shown on local television, devil’s-club charcoal and red ochre were used both as face paint and sprinkled over the graves. Similarly, a Cowichan (Halkomelem Coast Salish) man described a sweat- bath for Purification of a “sick” person who had been made ill by a malicious Indian doctor from another area. First, a canoe was filled with “sharp things,” including thistle, devil’s-club, black hawthorn, and other thorny plants. Water was poured onto these plants, a bulrush mat laid down to protect the patient, and the sick person laid down in the water on the mat. Then hot rocks were placed in the water-filled canoe until the heat was unbearable. The thorns in the plants were thought to prickle and drive sickness out of the bather (Rollins 1972:25a ‘ 28 TURNER Vol, 2, No, 1 The protective or supernatural powers attributed to devil’s-club are also reflected in Northwest Coast mythology and oral tradition, particularly among the Haida, Tsimshian, and Tlingit. A good example of this is in a story told by the late Willie Matthews, a Haida speaker and Hereditary Chief of Masset on the Queen Charlotte Islands: One of his ances- tors had been fasting out in the forest for several days. Eventually, he came across a giant devil’s-club plant with a trunk about 0.5 m (1% ft.) in diameter and leaves almost 2 m (5 ft.) across. He ate the inner bark from it, and immediately lost consciousness. Upon awakening, he saw a supernatural being, similar to a “‘fairy,”” who was thenceforth his ardian spirit. Ever since then, Willie eee family had many names alluding to “fairies”. A Tsimshian myth, quoted in part at the beginning of this paper, provides a similar episode (Boas 1916:172), as does a Tlingit myth recounted by Swanton (1909: 136) There are many other mythological references to hunters and others seeking purifi- cation or supernatural help by eating devil’s-club or drinking or bathing in an infusion of the plant (Barbeau 1953:414, 1961:73; Boas 1912:166-7; Krause 1956:188; Swanton 1905:212, 1909:308). In a Tlingit myth, devil’s-club and red [ochre ?] paint were found at the entrance to a supernatural house (Swanton 1909:95), and later, a woman being pursued threw a devil’s-club stick behind her and it immediately grew into a dense thicket of devil’s-club (Swanton 1909:95). In another, similar account, a devil’s-club comb was dropped to become a thicket, thus obstructing pursuers (Swanton 1909:383). Summary of Medicinal Uses There is a remarkable consistency in the various medicinal uses of devil’s-club, even among cultural groups that are totally distinct linguistically and geographically. In Table 3, medicinal uses are summarized by cultural groups, with the most widespread applica- tions shown first. It can be seen that the use of devil’s-club in treating arthritis and (or) rheumatism is, or was, almost universal along the Northwest coast. It is likely that its more widespread than indicated, since several coastal groups, including Coast Tsimshian, Haisla, and Heiltsuq (Bella Bella), have been little studied ethnobotani- cally. The use of devil’s-club as a dermatological aid is also widespread, as is its use in treating ailments of the respiratory and digestive systems. TABLE 3.—Summary of Medicinal Uses of Devil’s-Club* Type of Ailment Native Groups using Devil’s-Club as Treatment Arthritis and (or) Tlingit and (or) Kaigani Haida; Haida; Gitksan; Bella Coola; rheumatism Southern Kwakiutl; Nootka; Nitinaht; Squamish; Sechelt; Halkomelem (Upper Stalo); Cowlitz; Lillooet alia and (or) Tsetsaut; Eyak; Tlingit; Haida; Tsimshian; Bella Coola; Heilt- Purification suq; Southern Kwakiutl; Nootka; Nitinaht; Halkomelem; Lummi (and probably other Salish groups - cf. Table 4) General tonic or Tlingit; Haida; Gitksan; Bella Coola; Southern Kwakiutl; unspecified illness Sechelt; Halkomelem (Cowichan); Thompson; Central Car- rier; Kootenay Dermatological aid Tanaina; Tlingit and (or) Kaigani Haida; Gitksan; Comox (wounds, burns, (Mainland) ; Sechelt; Thompson; Central Carrier; (? Sahaptin) infections, etc.) *References for individual groups given in Tables 1 and 2 May 1982 TABLE 3.—(Continued) Type of Ailment Stomach and Digestive tract Tuberculosis TURNER 29 Native Groups using Devil’s-Club as Treatment Tanaina; Tlingit and (or) Kaigani Haida; Bella Coola; Heiltsuq; Southern Kwakiutl; Thompson; Northern Carrier Tanaina; Tlingit and (or) Kaigani Haida; Haida; Kwakiutl; Nitinaht; Skagit; Okanagan-Colville Cold or cough Tanaina; Tlingit and (or) Kaigani Haida; Haida; Squamish; Halkomelem (Cowichan); Cowlitz; Okanagan-Colville Purgative or emetic Eyak; Tlingit and (or) Kaigani Haida; Haida; Gitksan; Bella Coola; Carrier Childbirth (Post-partum) Carrier (Central and Southern); Skagit; Lummi; Shuswap Diabetes Heiltsuq (Bella Bella); Sechelt; Squamish; (Bella Coola—knew from use by non-Indians); Thompson Internal haemorrhaging Gitksan; Mainland Comox Broken bones Tlingit and (or) Kaigani Haida; Gitksan Analgesic Tlingit and (or) Kaigani Haida Measles Halkomelem (Cowichan) Gonorrhoea Gitksan Fever Tanaina Dandruff, lice Haida Headache ? Crow, Cheyenne (probably mistaken identification) Other Uses of Devil’s-Club in Native Cultures In Table 4, various non-medicinal uses of devil’s-club are summarized. The wood, which is soft and lightweight, was often used to make various kinds of fishing lures. Devil’s-club lures are said to have the property of spinning through the water as if they were alive, and were apparently very effective. The Nitinaht people used it for at least two types of lures, one of which—the cod-fish lure—was actually named after devil ’s-club (Turner et al., In Press). It consisted of a streamlined piece of cedar with a flat strip of devil’s-club lashed around it endwise, forming two rounded wings which gave a pro- Peller-like motion to the lure (Fig. 4). The lure was thrust down into the water from 4 Canoe with the aid of a long pole. It was then dislodged and allowed to spin to the sur- face. Cod-fish, hungry or curious, would follow it up, and were then speared by the waiting fisherman. The second type of lure consisted of a small, fish-shaped piece of devil’s-club wood to which a hook was fixed and a line attached. This was drawn through the water and functioned in the same way as a modern fish-shaped lure. It was especially 800d for catching “‘sea-bass” (Turner et al. In Press). TURNER TABLE 4.—Other (Non-medicinal) Uses of Devil’s-Club Native Group Haida Tsimshian Heiltsuq and/ or Tsimshian Nootka (Hesquiat) Nootka (Hesquiat) Nootka (Manhousat) Nitinaht Nitinaht Makah Clallam Lummi Straits Salish Upper Cowlitz (Taitnapam dialect, Yakima) Part of Plant wood inner bark leaves, es singed off wood wood charcoal wood charcoal charcoal bark Details of Use used to make black-cod lures infusion used for removing odours, e.g., for washing fishing nets that were not catching any fish, and were suspected of having been carelessly or maliciously urinated or defecated on. used in bunches, like steel wool, in water, to remove human scent from hunters used to make fish lures and octopus spears shavings boiled in water with various kinds of berries to make stain for basket materials and other objects used for fish lures for greenlings and rockfish used for cod and ‘‘sea-bass”’ ceremonial face paint (see Table 2) used for fishing lures, e.g., for ‘‘bass’’ used for fish lures for bass and other fish ceremonial face paint (see Vol. 2, No. 1 Reference Turner 1970:67 M. Seguin, pers. comm. 1981 B: Rigsby, pers. comm. 1981 Turner & Efrat n Press Turner & Efrat In Press Ellis & Turner 1976:7 Turner et al. In Press Turner et al. In Press vo Gill, pers. comm, 1981 Fleischer 1980:197; Gunther 1973:41 Gunther 1973:41; Table 2) John Thomas, pers: co ceremonial face paint (black) or bluish coloured tattoo pigment dried, pulverized for baby talc or perfume Turner & Bell 1971:78 Gunther 1973:41 May 1982 TURNER 31 TABLE 4.—(Continued) Native Group Part of Details of Use Reference Plant Green River bark dried, pulverized as deodorant Gunther 1973:41 Squamish charcoal __ used recently as black face paint Bouchard & (mixed with bear grease) Turner 1976:72 Halkomelem charcoal ceremonial face paint John Thomas, pers. (Upper Stalo) comm. 1981 The uses as face paint, perfume, baby talc, deodorant, and even in preparing a stain for basket materials and other objects (Table 4) may actually relate to the “protective” powers attributed to the plant (Table 2), although the protective aspect was not alluded to in the references cited in the former table. John Thomas (personal communication 1981) confirmed that the Lummi and other Coast Salish peoples who used the face paint in ceremonial dances were, and are, well aware of the underlying, protective purpose of its use (see also previous discussio n). Devil’s-club, in at least two Northwest coast cultures, was associated with bears. The Tlingit apparently based their original use of the plant as medicine on the observa- tion of two — attempting to soothe battle wounds by einai devil ’s-club roots e and believed that skein ate the berries and used the branches for bedding. Addition- ally, at least one Sahaptin person believed that devil’s-club is eaten by bears (D. French personal Hig tes 1981 Among the Haida, “Devil’ 's-Club” was both a place name (a village on the Queen Charlotte Islands) and the name of a Chief (C.F. Newcombe unpubl., ca. 1903). The importance of devil’s-club is reflected by the fact that throughout its range it had a name in almost every Native language spoken. The distribution of nomenclatural recognition and use of devil’s-club in Native cultures in western North America is shown in Figure 5. The various Native names are listed in Table 5. From a preliminary inspection, one can distinguish some 13 to 15 separate etymons (i.e., names with a single, unique source). Several of the Salishan languages (e.g., Lillooet, Thompson, Comox, Sechelt, and Squam ish; Halkomelem and Straits Salish; Green River, Skagit, and Swinomish; and sidkstiiey Stosawai and Okanagan), and the three Tsimshian languages (Coast Tsimshian, Nisgha, and Gitksan), have names of the same etymon (i.e., of common origin), but no such relationships can be s in the names from languages of different families. ont Niti- naht and Nootka, cee related languages of the Wakashan Family that share many words of common origin, have distinct, unrelated names for devil’s-club. of the names for devil’s-club have a “plant” suffix incorporated le -£., -mapt (Nootka); -— (Nitinaht); -ay (Comox and Sechelt); -e+p (Halkomelem); -az (Lillooet); and -wu?k “wood, bush” (Kootenay)]. But for the majority of names the stem of the word has no ieee meaning; its derivation has been forgotten or obscured with time. Exceptions are the Bella Coola, Tanaina, Nitinaht, Sahaptin, and Shuswap names, whose meanings are given in Table 5. This situation seems to indicate a long-standing association of the plant within the various languages and cultures, particularly those such as Haida, Southern Kwakiutl, and Nootka, where the name is unrelated to that in any other langu- age, $2 TURNER Vol. 2, No. 1 FIG. 4—Cod-fish lure used by Nitinaht and Nootka peoples. The elongated piece of wood is western red cedar (Thuja plicata), and the lashing of Sitka spruce root | Sate criti The two propeller- like appendages, which extend around the cedar wood, are almost certainly of devil’s-club wood, although the material is not identified in the cataloguing tale (Lure collected ca. 1911 at Dodge's Cove, Vancouver Island, by C. F. Newcombe; photo'by R. Bethell, Catalogue No. 2224, Ethnology Division, British Columbia Provincial Museum.) Approximately 2/3 natural size. TABLE 5.—Names for Devil’s-Club in Western North American Indian Languages* Language Family Language Name Reference Athapaskan Tanaina _heshkeghka’a (“‘thorn big big”’), Kari 1977:62 or heshkegh (‘‘thorn ala! ’) (also refers to wild rose in Outer and Upper uae ‘denied Athapaskan Tlingit s!Axt! Swanton 1909:383 achta Krause 1956:254 Athapaskan Carrier hwu heal Morice 1892-3:132 Haida Haida c’itonjaw (xIl) Turner & Levine (Skidegate 1972a:9 and and 1972b:11 Masset) Tsimshian Coast wooms Boas 1912:260; Tsimshian M. Seguin, pers. comm. 1981; Dunn 1978:110 *Orthography of original reference source has been retained | | May 1982 TABLE 5.—(Continued) Language Family Tsimshian Tsimshian Wakashan Wakashan Wakashan Wakashan Wakashan Salish Salish (Coast) Salish ( Coast) Salish (Coast) Salish (I nterior) Salish (Interior) Salish (Coast) Language Nisgha Gitksan Heiltsuq, ella Southern Kwakiutl Nootka (Hesquiat, Manhousat) Nitinaht Makah Bella Coola Comox (Mainland) Sechelt Squamish Lillooet Thompson Halkome- em (Cowichan) TURNER Name wurums, or warums hu?ums, or wuPums wiqas ixwmi ha* pa* #mapt Gayx”qWapt (‘‘cod-fish lure plant” ?a* Patbap tsk’alhkw (cf. stsk’ Douglas-fir bark slivers), or sk’alhk ch’i7t’ay ch’é7at’ay (or ch’a7at’ay) ch’Atiyay k’Atlaz k’étye?P qa ?patp Reference B. Rigsby, pers. comm. 1981 Hindle and Rigsby 1973:18; B. Rigsby, pers. comm, 1981 B. Rigsby, pers. comm. 1981 Turner and Bell 1973:278 Turner and Efrat In Press; Ellis and Turner 1976:7 Turner et al. In Press S. Gill, pers. comm. 1981 Turner 1973:201 R. Bouchard Unpubl. field notes 1973-76 Turner and Timmers 197238; Bouchard 1977-78 Bouchard and Turner 976:71-72 Turner 1972:13 Turner et al. In Prep. Turner and Bell 1971:78 34 TABLE 5—(Continued) Language Family Salish (Coast) Salish (Coast) Salish (Coast) Salish (Coast) Salish (Coast) Salish (Coast) Salish (Coast) Salish (Coast) Salish (Coast) Chimakuan Salish (Interior) Salish (Interior) Kootenay Sahaptin Sahaptin Language Halkomelem (Upriver) Straits Lummi Clallam Green River Skagit Swinomish Upper Cowlitz (Taitnapam dialect of Yakima) Snuqualmi Quileute Shuswap Okanagan- Colville Kootenay Sahaptin (Warm Springs) Sah (Columbia River TURNER Name qwo:pelhp qwarpatp qwu’n’numpt puqté xaxadi’a’ts xadi‘ats xadi’ats sqaipqa’ipas tcitca’tc lu”’j che-chah-pulth xwuxwalekw (refers to the clean smell of branch; cf. xw7uxw “any smell”) xaxagaylhp (sometimes also refers to Ribes, lacustre), or xwuxwugwaylhp naliygaxawurk x’naSwaakut (‘currant [prob. R. lacustre ] -like’) shkapk4pnuwash Vol. 2, No. 1 Reference Galloway 1979:7 Turner and Bell 1971:78 Gunther 1973:41 Fleischer 1980:207 Gunther 1973:41 Gunther 1973:41 Gunther 1973:41 Gunther 1973:41 Gunther 1973:41 Reagan 1934:55 Palmer 1975:58 Turner et al. 1980:73 Hart et al. 1981 253 D. French, pers. comm. 198 Hunn 1979:12 May 1982 TURNER Y Unconfirmed WY Ws 800 Ec} » Z RS: =. LL” ee cd “ete, — F a e . i oe. i ° | e S is) | S. i ; i ee Se D Peewee x ee i Bais a \ { \; FIG. 5—The distribution of nomenclatural recognition and use of devil’s-club in Native cultures in western North America. 56 TURNER Vol. 2, No. 1 CONCLUSION The many medicinal uses of devil’s-club may have originated from the protective qualities attributed to it because of its spines, and also from its cathartic properties. Its unusual odour may also have invited medicinal experimentation. It is also possible that its use by animals, such as the fighting bears observed by the Tlingit (Justice 1966: 36) prompted its use by people. Other herbal medicines, such as fern fronds used by the Hesquiat on the west coast of Vancouver Island, were also said to have originated from use by animals (Turner and Efrat In Press). Although the medicinal use of devil’s-club is widespread in western North America, the peoples of the northern Northwest Coast, particularly the Haida and Tlingit, seem to have had the greatest number of medicinal uses for it, and among these groups and the Tsimshian, it played a prominent role in mythology. This could be an indication that at least some of its medicinal uses diffused from north to south and from the coast to the interior. But the fact that many other languages have specific, unique names for the plant indicates that knowledge of it must have been long-standing in many areas. The use of the charcoal as face paint, both protective and decorative, for ceremonial dancers seems centered in the territories of the central Coast Salish and Nitinaht peoples, as does the use In those language groups having a common eytmon for devil’s-club, it is difficult to determine whether the names evolved independently during the course of natural dif- ferentiation of the languages, or whether the name was borrowed after the languages had already diffused. The latter would be obvious in languages from different families, but for this species, related names are found only in related languages within the same family. It is likely that the origins of both its names and uses have been permanently obscured and are thus untraceable. The use of devil’s-club among Native peoples has continued, and in some areas may actually be on the increase as interest in cultural heritage among younger generations is revived (Justice 1966:38). Sometimes the elders who formerly used it are unable to go out and collect it anymore, but do use it whenever they can get it from others Devil’s-club charcoal is still being used as a protective face paint for —" dancers (John Thomas, personal communication 1981). Sometimes vaseline is substi- tuted for the animal fat traditionally used as a base for it (Gunther 1973:41). In the early 1970’s, some Haida people still kept a stick of devil’s-club under their mattresses or across the top of their doorways to protect the household against evil influences. The effectiveness of devil’s-club as a medicine for arthritis, skin ailments, malignant u considering the widespread and continuing usage of devil’s-club among Native and even on-Native peoples, that its chemical composition and pharmacological properties have not been more thoroughly studied to date. There is always a danger in placing too much trust or faith in a medicine such as devil’s-club. Justice (1966:38) cites an example at Yakutat in Alaska, where a female cancer control program was rejected, presumably because of the belief in the efficacy of devil’s-club as a cancer treatment. Nevertheless, assuming that attitude and positive feeling about a medical treatment are important to its success, consideration should be given to the incorporation of well-tried traditional remedies, such as devil’s-club, with modern scientific treatments in medical programs involving North American Native Peoples. ACKNOWLEDGEMENTS The following people are PERS! acknowledged for their help in providing information ‘A this snatthg or for editorial suggestions and advice: William C. Turner, Robert D. Tumer, Dr. Harn Vv. in, Randy Bouchard, cia I.D. Kennedy, Dr. Robert T. Ogilvie, Dr. Andrea sone May 1982 Dr. Bruce Rigsby, Dr. Dale Kinkade, Dr. Laurence C. a TURNER 37 GA. digest Dr. Ni Seguin, te ei Dr. Brent Galloway, Dr. M. ry Thompson, Dr. Barry arlson, Dr. David H. French, John Thomas, Annie York — sl te and the late Chict W: bo Matthews. The photograph in Figure 4 is reproduced with the permission of the Ethnology Division, British Columbia Provincial Museum. The other photographs and the map are by Robert D. Turn LITERATURE CITED BARBEAU, M. 1953, Haida Myths, Natl. Mus. Canada Bull. 127 King’s Printer, Ottawa, ——. 1961. Tsimsyan Myths. Natl Canada Bull, 174 King’s Printer, Ottawa os Native Economic Plants of Montana. Montana Agric. Experi- BOAS, F. 1912. Tsimshian Texts (New Series). ubl. Amer. Ethnol. Soc. Vol. II. E.J. Brill, Leyden. : 916. Tsimshian Mythology. Bur. Amer, Ethn. 31st Annu. Report, 1909-10. Smithsonian ae Washington, D.C 966. . Codere, sik Kw akiutl hnapiiilies. Uni. ‘Ghicago Pre oo BOUCHARD, R. 1973-76. Mai Ethnobotany, fect Field Notes, ‘British Columbia I Cc. Unpul. Field Notes. British Columbia Indian Language Proj., Victori ae ST eee, ee Tanouie. Un- publ. Field Notes. British Columbia Indian anguage ie yond a, B.C, —-. JJ. TURNER. 1976. Squamish ihccesone: ae - ms., British Columbia Indian Language Proj., Victoria, B.C. BROCKLESBY, H.N., and R.G LARGE. 1938. A Hypoglycaemic Substance From the Roots of the Devil’s-Club. . Canadian Medical ay » July, 1938:32 (as quoted in Justice, CARRIER ee COMMITTEE. 1973. Hanuyeh Ghun ’Utnii. Plants of Carrier ral Carrier Language. Ft. St. DUNN, J.A. 1978. A Practical Dictionary of the Coast Tsimshian Language. Natl. Mus. of stabi Canadian Ethn. Ser LLIS, D. and N,J. TURNER. 1976. Nootka Plant ia Manhousat (Hot Springs Cove). Unpubl. ms., Victoria FLEISCHER, M. 0. The Ethnobotnay of the Clallam Indians of Western Washington. Northwestern Anthropol. ne Notes 14(2): 192-210. FENN, L.A., M.A. NORRIS, and N.J. TUR- NER. 1979. Uses of Plants by Native Peoples of the Pacific Rim National Park Area. Un- publ. ms., Western Region, Parks Canada, C Alta. GALLOWAY, B. 1979. Upriver Halq’emey- lem Ethnobotany, Unpubl. ms., Coquale- . 1896. Economic Botany of arden lets prere’y Pittonia, III, Pt. 14: 64-85. GRAHAM, R.C.B., and R.L. NOBLE. 1955. mparison of In Vivo Activity of Variou tice, 1966; no further reference information given). GUNTHER, E. 1973. The Ethnobotany of estern ets aes Univ. Washington Nd TURNER, and L.R. MOR- GAN. oem Ethnobotany of the Kootenai Indians of Western North America. Unpubl. Report to the Kootenay Indian Area Coun- cil, Cranbrook, B.C. HINDLE, L., and B. RIGSBY. 1973. A OWNBEY, and J.W. boring iieatendiek Stanford Univ. as Stanford. UNN, G.H. 1979. Sahaptin Plant Terms (preliminary version). Xeroxed ms., pre- pared for the Yakima Indian Nation, Janu- iv. Washing RI, P.R. 1977. Dena’ena K’et’una. i i ay Plantlore. sei Literacy Labor- KRAUSE, A. 1956. pe by E. Gunther). Tlingit Indians. Univ. Washington Press, Seattle. TURNER Vol. 2, No. 1 LITERATURE CITED (continued) LEWIS, W.H., and M.P.F. ELVIN-LEWIS. 1977. Medical Botany. Plants Affecting Man’s Health. John Wiley & Sons, New York. LI SHIH-CHEN. 1973. (transl. by F.P. Smith and G.A. Stuart). Chinese Medicinal Herbs. MACDERMOT, J.H. 1949. Food and medi- Columbia. Canadian Med. Assoc. J. 61(2): 177-183 RICE, REV. FATHER A.G. 1893. Notes oe Industrial, and Sociological the Western Dénés. Transactions of the porary apg Session 1892-93, Pchioe ACADEMY OF SCIENCES. 1975. as Pharmacology in the People’s ca alge of China. Natl. Acad. Sci., Wash- NEWCOMBE, C.F. 1901-1903. Haida Ethno- logy, Unpubl. Field Notes. Provincial Ar- i ictoria. 75. Shuswap Indian ethno- botany. Syesis 8:29-81. REAGAN, A.B. 1934. Plants used by the Hoh and Quileute Indians. Trans. Kansas Acad. Sci. 37:55-70 ROLLINS, D. 1972. Materia Medica of the Northwest Coast Indians of British Colum- Brit ish Columbia ria. SMITH, G.W. 1973. Arctic Pharmacognosia. Precis I. 1928. Materia Medica of the Bella ok and Neighbouring Tribes of nage Columbia. Natl. Mus. Canada Bull. 6, King’s Printer, Ottawa, Ont. cont oe E.V. (ed.) 1930. The Ethno- botany of the Thompson Indians of British Columbia. bnaine on Field Notes by James A. Teit, . Ethn. 30th Annu. Report, ty 33- 102, — D.C. the Ethnology - the Haida. Amer. Mus. « 8, Pt, FT; iia North Pacific ss aie Yok 5S Ft. — a SO Peete toe and Texts. Aue Ethn. Bull. No. 39. Smithsonian - MINATO, and M. ISHI- KAWA. 1966. tee on paps ss hedron, Supplement No. 7:219-225, TAYLOR, R.L., and B. MACBRYDE. 1977. Vascular Plants of British —* Univ, British Columbia Press, Vancou’ TEHIT,, JA. poet The Shuswap. Mem. Am yi Jesup North Facile . Nat. His Pencdition, Va t. 7. G.E. Stechert and ew York TURNER, NJ. "1970. Ethnobotany of the Queen Charlotte Is- lands. Unpubl ., The Botanical Garden Univ. British pas. Vancouv: L972. illooet Exhnabotay (Fraser ier Dialect). Unpubl. ms., The Botanical Garden, Univ. British Columbia, tices inat™ 973. The ethnobotany of the Bella as Indians of British Columbia ae 193-220. 1979. Plants in British Columbia Indian Technology British Columbia Prov. Mus. Ha ndbook No. 38, Victoria, B.C. d “ern BELL. 1971. The ethnobotany of the Coast Salish Indians of Vancouver Island. Econ. Botany 25(3): 257-310 . and M.AM. BELL. 1973. ethnobotany of the Southern nen Bo Indians of nai Columbia. Econ. Botany 27(3):257-31 = " BOUCHARD, and D.I.D. bia and Washington. British Columbia Prov. Mus. Occ. Paper No. 21, Victoria. ., and B.S. EFRAT. In Press. Ethno- botany of the Hesquiat People of the West Coast of Vancouver Island. British Colum- bia Prov. Mus. Cultural Recovery Series ictoria. DD. LEVINE. 1972a. Haida Plant Names (Skidegate Dialect). Unpu ubl. ms., The Botanical Garden, Univ. Bn ritish Columbia, ot R.D. LEVINE ate Haida Plant =e (Masset Dialect). U publ. ms., The Botanical Garden, Univ. pea Colum- bia, Vancouver. ., J. THOMAS, B.F. CARLSON, and RT. OGILVIE. In Press. Ethnobotany of the Nitinaht Indians of Vancouver Island. riti um P. tish Columbia Prov. cc. Paper aes Victoria. (Co-published by Parks estern Region, Pac Rim oie Park, Tofino, B.C. ie MPSON, coed K Garden, Univ. British Columbia, Vancouver. J. Ethnobiol. 2(1): 39-61 May 1982 VERTEBRATE FAUNA FROM FOUR COASTAL MISSISSIPPIAN SITES ELIZABETH J. REITZ Department of Anthropology, University of Georgia Athens, GA 30602 ABSTRACT.—Samples of vertebrate fauna recovered from three archaeological sites on barrier islands off the Georgia coast and one on the mainland coast indicate a specialized economy emphasizing marine resources. Faunal evidence suggests a use of fishes from the immediate estuarine-salt marsh system. The few highly seasonal resources of the region were not exploited. Vertebrate evidence indicates that occupation was either intermittant or continuous at these coastal sites. The collections total 17 bird, 213 mammal, 941 fish, 114 reptile, and 40 amphibian individuals. INTRODUCTION Along the Atlantic coast between North Island, South Carolina, and Anastasia Island, Florida, lies a series of barrier or sea islands (Fig. 1). Behind these islands is found a rich estuarine-salt marsh system, For many years scholars of diverse interests have examined this unique environment because of its importance to the American fisheries industry. Archaeologists have also been attracted to the coastal islands, and have excavated and tested sites of all temporal phases. Although questions on chronology and settlement patterns have been and continue to be the major focus by archaeologists in this region, questions pertaining to subsistence have been considered by coastal archaeologists for a throughout human occupation on the coast (Larson 1980). In spite of this interest in human use of the estuaries, until recently an adequate vertebrate data base had not been assembled for studies of aboriginal use of marine resources. As of 1979, some 17 sites have been excavated: between Sapelo Island and Anastasia Island, but only five of the faunal assemblages passed the test of adequacy, that is, had over 200 individuals and 1400 total bones (Reitz 1979a; Wing and Brown 1979). Of the five adequate samples, only two were aboriginal: the Archaic St. Simons Shell Ring excavated by Rochelle Marrinan (1975) and the Savannah Kenan Field excavated by Ray Crook (1978). Since 1979 three additional sites have produced large samples. Asa result there are now four adequate late Mississippian vertebrate faunal assemblages available for analysis from the Georgia sea islands. These four collections indicate a prehistoric/historic subsistence strategy based on fishing, supplemented by deer hunting. Deer (Odocoileus virginianus), sea catfishes (A Tiidae), and drums (Sciaenidae) were the major resources to the exclusion of most other species. Species utilization was not the same at the four sites, suggesting that even in a small geographical area where resources are basically similar throughout, there were rari Nek variations to be reflected in the faunal record. Of further interest wee * or importance of highly seasonal resources such as migratory waterfowl, bluctishes PES saltatrix), and herrings (Clupeidae). Based on the faunal evidence it appears that occupation of the coastal islands was not confined to a single season, but may have been sporadic throughout the year, or even continuous. SITE DESCRIPTIONS The sites all date primarily to Savannah or Irene Phases. The Savannah phase is the late Mississippian phase on the Georgia coast, beginning around AD 1000 (Crook 1982) or AD 1150 (Pearson 1979). The Savannah Phase is followed by the Irene Phase, but the time span is subject to debate. Charles Pearson initiates the Irene Phase at AD 1350, REITZ Vol. 2, No. 1 Savannah \ 28 HELENA 5) HILTON HEAD Ogeechee L/ Savanna C _ SKIDAWAY | SSARAW 7) St. | CATHERINE ATLANTIC (Psarero OCEAN Altamaha™Qes t. Brunswi ’SIMONS - ; Satilla } OLONELS (CUMBERLAND St. Marys ame.ta \ \\ St. @ \\ANASTASIA ugust 2 FIG, 1—Sea Islands of the Atlantic Coastal Plain. May 1982 REITZ + a ee ~--7--- Ss < Zz oo. == Rs. ALYMOLDS STATE WKOLWE REFUSE NORTH OF THE HELL RIN NY =n WAS 41 —» Z FIG, 2—Sapelo Island, Georgia (Anonymous 1968). 42 REITZ Vol. 2, No. 1 while Lewis Larson, Ray Crook, and Robin Smith interpret this phase as an early historic aboriginal phase beginning about 1540 (Crook 1982) or 1526 (Milanich 1977). The Irene Phase may have ended as early as 1550 (Pearson 1979), but the excavators of the sites terminate it with the end of the Mission Period at 1680 (Milanich 1977). A Cj 4 date from Irene-San Marcos Phase context at Bourbon Field indicates an occupation within the 1540-1680 period (Crook 1982). Pine Harbor Phase is the Irene Phase equivalent on the Altamaha River and Smith’s San Marcos Phase corresponds with Crook’s Sutherland Bluff Phase. Irene and San Marcos Phases may be contemporaneous (Crook 1982). The temporal designations used below are those provided by the excavators. At contact, Spanish administrators identified the residents of the Kings Bay area as Timucuans and the Sapelo area as Guales (Larson 1978). Three of the four sites discussed here are on Sapelo Island, Georgia (Fig. 2). Kenan Field was excavated by Dr. Ray Crook in 1976-1977 (Crook 1978). It is a multicompo- nent aggregate village covering 60 ha. The site is composed of 589 discrete shell middens measuring 5-10 m in diameter, two mounds, and several structures. Pine trees currently are grown on the site. The site is located on the west side of Sapelo Island, bordered on the northwest by Duplin River and on the south by Barn Creek, near a salt marsh. Six structures, one mound, and 11 shell middens were excavated. Quarter-inch screens were used at each non-feature until except for portions of three test pits in shell middens. A 11 soil sample was taken from columns from each shell midden but not analyzed. Soil from two portions of each shell midden and from features was screened through 1/16 in mesh and chemically floated. Although a Sutherland Bluff component was found at Kenan Field, the occupation was primarily Savannah and Irene Phases. Bourbon Field is a similar multi-component aggregate village approximately seven km northwest of Kenan Field, on the east side of Sapelo Island. It is located behind Black- beard Island facing Blackbeard Creek and a salt marsh (Crook 1982) and was excavated in 1979-1980 by Crook. Bourbon Field covers 14 ha and is composed of 119 discrete shell middens and a small earthern mound. It is currently an open field used by the Georgia Department of Natural Resources as a migratory bird feeding station. During excava- tions, 63 tests were conducted in the off-midden areas and ten shell middens were stud- ied. A % in screen was used at all excavations features and shell-midden volumetric samples. Volume samples of 20 1 each were taken from each natural level and screened through a graduated series of meshes. Faunal samples collected in % in portion were lumped with % in fraction from the rest of the shell midden, therefore “column sample” refers only to fine-screened fractions of bone with the % in fraction removed. Features were also screened in this manner. While the Bourbon Field site contains materials dating from the Sapelo through San Marcos Periods, the Savannah/Irene Phase occupation is primarily represented (Crook 1982). The third Sapelo site is located on the west side of the island, about 2 km west of Bourbon Field and 5 km north of Kenan Field, and by the Mud River anda salt marsh. The excavations were placed 45 m north of a drainage ditch north of the Sapelo Shell Ring, hence the site is known as North of the Shell Ring Drain. The excavations were conducted in 1979 by Dr. Lewis Larson. The faunal sample comes from four contiguous x 2 m units, placed on one undisturbed shell midden. At present, magnolia trees and live oaks primarily grow at this site. During excavations, %4 in screens were used except for a 50 cm2 column sample from the shell matrix which was screened through 1/16 in mesh. The ceramic assemblage contained Irene Burnished Plain, Irene Filfot Stamped, and Savannah Check-Stamped, and an occasional Spanish Olive Jar fragment. It is thought this represents an Irene occupation somewhat later in time than the main occupation at either Kenan or Bourbon Field (Crook, personal communication). The fourth site, the Kings Bay Site (9Cam171), is located on the mainland behind Cumberland Island (Fig. 3) approximately 81 km south of Sapelo Island and was ¢xc@ vated by Robin Smith in 1978-1979 (Smith et al. 1981). The site is a multicomponent discontinuous shell midden covering 91.5 ha and stretching 4.5 km along the western meuaiaiall Peeing ERE NeN: May 1982 REITZ 43 SATILLA FIG. 3—Kings Bay (9Cam171), Georgia RIVER CUMBERLAND ISLAND ST. MARYS RIVER ce Ne N Sc 0 5 \ miles —————— 4 0 10 Ft. veil ner es 44 REITZ Vol. 2, No. 1 edge of Kings Bay, bordered by a salt marsh. The Kings Bay Site is covered by pine plantation and Southern Mixed Hardwoods and is currently being impacted by construc- tion of the Naval Submarine Support Base. During excavations, a series of sheet deposits and features were uncovered, but reference here is primarily to the Savannah Phase fea- tures. Soil from the Savannah Phase features was screened with 1/8 in mesh. Column samples also screened with 1/8 in mesh, were taken from zone deposits, but only two [San Marcos period column samples] were studied. Zone material was screened using 1/4 in mes In most respects the environments of the Sapelo Island sites and the Kings Bay site are similar (Johnson et al. 1974). All are bordered by salt marshes and estuaries con- nected by large sounds leading to the ocean. Tidal creeks of various sizes twist through the marshes. Freshwater biotopes are available within .5 km of all sites but beaches are more distant. ee Aa ni aboriginally was probably Maritime Oak Forests in all cases (Johnson et al. 1974). This is a region of moderate mean annual temperatures of 22 degrees C and average rainfall of 110 cm. Mean tide level at Sapelo Island and at Crooked River, just north of Kings Bay site, is 1.036 m, with a spring tide range of 2.438 m. Mean tidal ranges are 2.072 m (Tide Tables 1981). Sapelo and Cumberland Islands are in the Carolina Region with water temperatures of 15-20 degrees C (Ekman 1953; Briggs 1974) and where wide turbidity, salinity, and temperature fluctuations place unusual stresses on the estuarine fauna (Hoese 1967; Stickney and Miller 1973). Among the species best adapted to these conditions are herrings (Clupeidae), sea catfishes (Ariidae), and drums (Sciaenidae) (Mahood et al. 1974; DEIS 1978). Seasonal variations in species availability are less pronounced in this region than elsewhere. None of the mammals hibernate (Golley 1962). While sea turtles are strictly warm weather visitors, ae turtles are inactive only on very cold days (Carr 1952). Birds increase in density and diversity during the winter, representing one of the most seasonally variable resources, however coastal Georgia is not a major area for migratory flocks (Robertson and Kushlan 1974). Sharks are primarily present during warm weather conditions, but the bony fishes (Osteichthyes) present a variable dispersal pattern de- pending on individual age and species. Many species are oa oan in the estuaries by adults, juveniles, or both throughout the year (DEIS 1978). Some species remain inside the estuary throughout the year, but occupy ea habitats within the estuary in response to water temperatures or salinity levels. Other species visit the estuaries only in the winter either as adults or juveniles, but more species visit estuaries in the summer. The young of most fishes live in the estuary for the first year or two of life and can be found there throughout the year. Adults are less able to tolerate estuarine extremes. Diversity, measured as an index of the number of species and the number of individuals, does not change seasonally as much as relative abundance (Dahlberg and Odum 1970); however, the total number of species is higher in the summer and fall (Dahlberg 1972). Variations in species occurrence are also found between years depending upon rainfall and temperatures. Seasonal behavior of fishes in the estuarine setting is complex based upon several variables which may not be known to the archaeologist. The species list itself provides only a general guide. Ideally it should be possible to determine seasonal patterns of exploitation of bony fishes by examining growth rings of otoliths, vertebrae, or scales (Casteel 1976). METHODS The faunal remains were identified and analyzed using standard zooarchaeological procedures (Wing and Brown 1979), Kenan Field and Kings Bay materials were identi- fied using the comparative skeletal collections at the Zooarchaeology Laboratory, Florida State Museum. Bourbon Field and North of the Shell Ring Drain samples were identified at the University of Georgia’s Zooarchaeology Laboratory, Department of Anthropology. Minimum Numbers of Individuals (MNI) were determined by paired elements and age, and aggregated as MNI (strata) for each non-contiguous unit following Grayson (1 1973). May 1982 REITZ 45 Biomass was calculated using least-squares analysis of logarithmic data (Reitz 1979b; Wing and Brown 1979). Formulae used to obtain biomass estimates are on file at the Zooarchaeology Laboratorys of the University of Georgia and the Florida State Museum. Biomas nan Field was not calculated since the necessary weight information was not readily available. Diversity was calculated using the Shannon-Weaver Diversity Index (Shannon and Weaver 1949) and equitability was calculated using the Sheldon Evenness Index (Sheldon 1969). Species were grouped into classes for analysis. ‘“‘Com- mensal” species include amphibians and small rodents. Snake remains are so rare in the collection that these reptiles were probably commensal animals and are included as such at Kings Bay. Oo na => ° tad wa oO RESULTS Two basic characteristics upon which reliable faunal analysis depends are sample size and screening procedures. These field-related variables affect faunal samples drama- tically. It has been demonstrated that samples of fewer than 200 individuals and 1400 bones are subject to biases which influence both species diversity and equitability (Casteel 1976-1977; Grayson 1978, 1981; Wing and Brown 1979). Using these measures as guide- lines for adequacy, all of the samples discussed here are reliable, with the possible excep- tion of that from North of the Shell Ring Drain, with 5462 bones but only 107 individ- uals. Faunal remains identified from the four sites are listed in Tables 1-4. The Shell Ring Drain faunal list is almost identical to that of the other samples, so that while diver- sity may be depressed the list and proportions of fauna are probably reliable. Equita- bility at North of the Shell Ring Drain may be inaccurate, however; the low MNI may be a result of using a minimal distinction method (MNI site) (Grayson 1973). This method was used because the units were contiguous, representing a single midden component. TABLE 1.—Species List for Kenan Field (Crook, 1978) Species MNI No. % Didelphis virginiana Opossum 1 0.3 Scalopus aquaticus Mole 5 1.3 Sylvilagus sp. Rabbit 13 3.3 Sciurus niger Fox squirrel 1 0.3 Peromyscus sp. Mouse 03 Sigmodon hispidus Hispid cotton rat 3 0.8 Cetacea Dolphin 1 0.3 Procyon lotor Raccoon 16 4.0 Mustela vision in I igi Mephitis mephitis Striped skunk : 0.5 Lutra canadensis Otter 1 0.3 Odocoileus virginianus Deer 38 9.6 Bos taurus Cow : ag Unidentified Bird 4 vig ix sponsa Wood duck 1 0.3 Meleagris gallopavo Turkey 1 0.3 Zenaidura macroura Dove 1 0.3 Passeriformes Song bird 1 0.3 Alligator mississippiensis Alligator 1 0.3 Kinosternon subrubrum Musk turtle 17 4.3 Deirochelys reticularia Chicken turtle 3 0.8 Malaclemys terrapin Diamondback terrapin 18 4.5 Terrapene carolina Box turtle 7 1.8 46 TABLE 1.—(Continued) Chelonia mydas Anolis carolinensis aurus sp Unidentified snake Colubridae Agkistrodon piscivorus Ariopsis felis Bagre marinus Archosargus probatocephalus Cynoscion nebulosus Stellifer lanceolatus Mugil sp. Paralichthyes sp. Total REITZ Green turtle Green anole Glass lizard Colubrid snake Skates & Rays Gar Lady fish Herrin Sea catfishes Hardhead catfish Gafftopsail catfish Sheepshead Spotted sea trout Atlantic croaker Star drum Flounder Vol. 2, No. 1 TABLE 2.—Species List for Savannah Phase Features from the Kings Bay Site, 9Caml71. Species Unidentified Mammal Didelphis virginiana Opossum Scalopus aquaticus Mole Unidentified Rodent Peromyscus sp. Sigmodon hispidus Cotton rat Procyon lotor Raccoon Odocoileus virginianus Deer Unidentified Bird Ct 66 MNI No. % 1 0.4 1 0.4 2 0.8 3 12 1 0.4 3 | Be gms 26.6 os 0.3 0.1 0.2 1 0.3 2 0.5 if O03 t 1.8 5 13 1 0.3 12 3.0 10 25 24 6.0 1 0.3 5 Les 64 16.1 23 5.8 27 6.8 6 1.5 5 Lo 11 2.8 28 7.0 1 0.3 Zz 0.5 19 4.8 5 1.3 398 Biomass kg 0 0.54 5.3 0.01 0.1 0.036 04 0.07 0.7 2.03 19.9 0.07 0.7 May 1982 TABLE 2.—(Continued) Species Anatidae uck Mergus serrator Mergansor Unidentified Turtle Kinosternidae Mud turtles Kinosternon sp. Mud turtle Emydidae Basking turtles Malaclemys terrapin Diamondback terrapin Sceloporus undulatus Fence lizard Colubridae Opheodrys aestivus Rough green snake Amphibian Frog/Toad Bufo sp. Toad Rajiformes Skates & Rays Dasyatis sp. Sting ray Unidentified Fish Amaia calva Bowfin Lepisosteus sp. Gar Ct REITZ 0.4 0.8 gms 47 Biomass kg % 0.007 0.07 0.007 0.07 0.394 3.9 0.1 1.0 0.04 0.4 0.15 1.5 0.46 4.5 0.001 0.01 0.001 0.01 0.008 0.8 0.03 0.3 1.94 19.0 0.015 0.1 48 REITZ Vol. 2, No. 1 TABLE 2.—(Continued) Species Ct MNI gms Biomass No. % kg % Elops saurus 1 1 0.4 0.1 0.005 0.05 Lady fish Brevoortia sp. 19 a 12 0.5 0.026 0.3 Herring Siluriformes 7 — _ 0.7 0.01 0.1 Catfishes Ariidae 293 _ _ 35.1 0.62 6.1 Sea catfishes Ariopsis felis 51 10 3.9 7.9 0.15 1.5 Hardhead catfishes Bagre marinus 217 15 5.8 85.7 1.41 13.8 Gafftopsail Lobotes surinamensis 45 2 0.8 40.7 0.63 6.2 Triple tai Archosargus 5 3 i? L? 0.026 0.3 probatocephalus Sheepshead Sciaenidae 31 = ae 4.95 0.157 1:5 Drums Bairdiella chrysoura 31 11 4.3 1.4 0.06 0.6 Silver perch Cynoscion sp. 28 10 3.9 4.45 0.16 1.6 Sea trout Larimus fasciatus 1 1 0.4 0.6 0.03 0.3 Banded drum Letostomus xanthurus 5 5 1.9 0.3 0.017 0.2 Menticirrhus sp. 1 1 0.4 0.1 0.007 0.07 7 Kingfish , Micropogonias undulatus 7 5 1.9 0.8 0.047 0.5 Pogonias cromis 13 3 1.2 2.55 0.097 1.0 Black drum eae earteeeneeme eeeeinemneendnmnnentmaeneeemneenenntiieedinemmeneaaniamemaedll | | ‘ i May 1982 TABLE 2.—(Continued) Species Stellifer lanceolatus ar drum Bairdiella/Stellifer sp. Mugil sp. Mullet Paralichthyes sp. Flounder Unidentified Bone Total TABLE 3.—Species List of Bourbon Field, 1980-1981. Species Unidentified Mammal Scalopus aquaticus Mole Sylvilagus sp. Rabbit Unidentified Rodent Cricetidae New World mice Sigmodon hispidus Hispid cotton rat Carnivore Ursus americanus Procyon lotor Raccoon Mustelidae ink family Ct Ct 3271 REITZ No. MNI 0.5 gms gms Zaie5l 49 Biomass 0.377 3.7 0.104 1.0 0.094 0.9 0.17 iy 10.195 — Biomass, kg 29.95 0.02 Cy 0.0004 0.0006 0.01 0.02 0.003 0.01 0.03 0.004 TABLE 3.—(Continued) Species Mephitis mephitis Skunk Odocoileus virginiana Deer Unidentified Bird Laridae Gull Family Icteridae Blackbird Family Unidentified Reptile Unidentified Turtle Kinosternon subrubrum Mud turtle Emydidae Basking turtles Deirochelys reticularia Chicken turtle Malaclemys terrapin Diamondback terrapin Trionyx ferox Softshell turtle Unidentified Snake Colubridae Colubrid snakes Elaphe sp. Ratsnakes Viparidae Pit vipers Amphibian Rana/Bufo sp. Frog/Toad Ct 271i 46 REITZ No. 63 (1) % 0.4 112 3.6 0.2 1.4 0.2 1365.57 14.65 121.75 0.62 Vol. 2, No. 1 Biomass, kg Kg % 0.06 0.09 19.06 ae 0:25 0.4 0.01 0.01 0.0006 0.0009 3.91 5.7 0.26 0.4 0.98 LA 0.26 0.4 0.53 0.8 0.2 0.3 0.01 0.01 0.005 0.007 0.001 0.002 0.009 ~=0.01 May 1982 TABLE 3.—(Continued) Species Chondrichthyes Sharks & Rays Sharks Rays Dasyatis sabina Atlantic stingray Unidentified Fish Lepisosteus sp. ar Clupeidae Herrings Ariidae Sea catfishes Artopsis felis Hardhead catfish Bagre marinus Gafftopsail catfish Perciformes Perciform fishes Pomatomus saltatrix Bluefish Archosargus probatocephalus Sheepshead Sciaenidae rums Bairdiella chrysoura Silver perch Cynoscion sp. Sea trout Letostomus xanthurus Ct 5435 191 REITZ No. % 2.1 4.5 0.2 gms 300.47 al $5.14 0.1 Biomass, kg 0.10 4.7 0.007 0.01 0.19 % 0.04 6.9 0.01 0.3 52 REITZ Vol. 2, No. 1 TABLE 3.—(Continued) Species Ct MNI gms Biomass, kg No. % kg Micropogonias undulatus 28 19 3.4 4.38 0.15 0.02 Croaker Pogonias cromis 54 21 | 5.84 0.2 0.3 Black drum Scianops ocellatus 9 7 iz 18.77 0.4 0.6 Stellifer lanceolatus 14 11 2.0 0.16 0.01 0.01 Mugil sp. 679 43 7.6 8.3 0.19 0.3 Mullet cf. Eleotridae 2 2 0.4 0.03 0.002 0.003 Sleepers Paralichthyes sp. 20 5 0.9 1.18 0.03 0.04 Diodontidae 1 1 0.2 0.91 ae om Porcupine fishes Unidentified Bone = a - 1321.22 or = Totals 15331 563 as 6827.83 68.228 — TABLE 4.—Species List for North of the Shell Ring Drain Species Ct MNI Weight Biomass No. % Gm Kg % Ud. Mammal 391 = on 2173.91 26.5 63.1 Sylvilagus sp. 68 6 5.6 41.06 0.7 1.7 Rabbit Cricetidae 3 can oe 0.15 0.005 0.01 New World Mice Peromyscus sp. 2 1 0.93 0.03 0.001 0.003 Mouse cf. Sigmodon hispidus 1 1 0.93 0.10 0.003 0.01 Cotton Rat anaes eS Se May 1982 TABLE 4.—(Continued) Species Delphinidae Dolphin Family Procyon lotor Raccoon Odocoileus virginianus Whitetail Deer Ud. Bird Anas Rallidae Rail Icteridae Blackbird Family Ud. Turtle Emydidae Malaclemys terrapin Diamondback Terrapin Ud. Snake Colubridae Non-Poisonous Snakes Coluber constrictor Black Racer cf. Viparidae Poisonous Snakes Ud. Amphibian Rana/Bufo Frog/Toad Squaliformes Sharks Dasyatis sabina Atlantic Stingray sp. Surface-feeding Duck Ct REITZ Weight Gm 29.39 10.15 270.40 Biomass Kg % 0.5 tee 021 0.5 4.06 9.7 0.30 O75 0.25 0.60 0.01 0.02 0.005 0.01 0.24 0.57 0.27 0.64 0.36 0.86 0.0007 0.002 0.0007 0.002 0.001 0.003 0.003 0.01 0.002 0.005 0.002 0.005 54 TABLE 4.—(Continued) Species Ud. Fish Ariidae Sea Catfishes Ariopsis felis Hardhead Catfish Bagre marinus Gafftopsail Catfish Sciaenidae Drum Family Bairdiella chrysoura Silver Perch Cynoscion regalis Weakfish Leiostomus xanthurus Spot Pogonias cromis Black Drum Scianops ocellatus Red Drum Mugil sp. Mullet Ud. Bone Totals Ct 2872 583 106 155 284 5462 REITZ No. 107 6.5 Weight Gm 279.27 35.31 149.01 32.16 14.68 3346.27 Vol. 2, No. Biomass % 2.83 6.7 1.5 a0 2.52 5.5 0.54 13 0.45 ie! 0.03 0.07 0.03 0.07 0.01 0.02 1.20 2 0.02 0.05 0.002 9.005 42.36 a —_ While the samples from these coastal sites are generally outstanding in terms of size, there remains screening biases. As is the custom where fine-screen recovery methods are om entire units was screened through 4 in mesh, but only a portion of the unit was screened through 1/8 in mesh (Kings Bay) or 1/16 in mesh (Sapelo Island). For analysis, faunal remains collected from the % in zone and fine-screened column samples used, soil fr and features were combined. Obviously it is not accurate to present as a single, unified sample a collection in which roughly 6% of the soil was screened through 1/8 or 1/16 in mesh and the remaining 94% was collected by % in mesh. Asa result of such a combina- tion there is an over representation of large bones, representing species such as deer, and a reduction in small species such as star drum. REG meme tene-srmeetues sn eminent resect i ii sc te May 1982 TABLE 5.—Exploitation Patterns Terrestrial Mammals Cetacea Birds Turtles and Gator Snakes Fish, Sharks, Rays Commensals Total Terrestrial Mammals Cetacea Birds Turtles Snakes Fish, Sharks, Rays Commensals Total Kenan 3 % 73 18.3 1 0.3 8 2.0 47 11.8 13 33 231 58.0 25 6.3 398 _ REITZ MNI Bourbon oO. fo 100 17.8 2 0.4 30 5.3 8 1.4 398 BOS7 25 4.4 563 _~ Biomass Bourbon Kg (0) 21.4 72.8 0.01 0.03 1.25 4.3 0.01 0.03 6.7 7 | 0.04 0.1 29.36 oa NSRD No. fo 9 8.4 1 0.9 5 4.7 2 i 2 1.9 84 785 4 PY 107 ~ NSRD Kg % 4.97 48.4 0.5 4.9 0.27 2.6 0.36 35 0.004 0.04 4.16 40.6 0.004 0.04 10.27 we 55 Kings Bay No. () 5 1.9 2 0.8 10 3.8 228 88.7 iz 4.7 257 _ Kings Bay Kg oO. mal A 33.6 0.01 0.2 0.6 8.9 $.5 56.3 0.05 0.8 6.27 = 56 REITZ Vol. 2, No. 1 This observation becomes critical when evaluating the results in Table 5 which pro- vides a summary of major groups of fauna represented at each site. The primary taxa at all four sites are marine animals, while marsh and terrestrial fauna are generally minor by comparison. The most striking anomaly is observed from the Savannah Phase features at Kings Bay where less than 2% of the individuals are mammals. Recall that this is the col- lection where 100% of the soil was screened through 1/8 in mesh, rather than only 6% at Sapelo ei sites. mportance of this anomaly can be seen in two more examples. The first exam- ple compares three components of the Kings Bay Site (Table 6). The San Marcos zone material was recovered with % in screen while all of the column samples were recovered by 1/8 in screen. While the difference between deer and small drums might be the result of cultural factors, it is more likely to be a function of screen-size. With only 50 individ- uals represented, the San Marcos column sample provides only a relative basis for com- parison, but it does serve to explain the difference between the San Marcos Zone example and the Savannah phase features where the MNI is For the second example, the Bourbon ae an midden tests were reevaluated (Table 6). Each unit was excavated 94% using % in screen and 6% was fine-screened. When the column samples are examined alone, terrestrial mammals provided only 2% of the individuals and fish 89%, a distribution similar to that of the Savannah phase ak and San Marcos column samples at Kings Bay. the purposes of this pie tena it seems egestas to conclude that features and aes samples where 100% of the sample recovered in the same screen size are more reliable than samples collected eeuaet satci screen sizes. DISCUSSION It is common to consider the Mississippian subsistence strategy as more or less uniform based upon horticulture and hunting, primarily of deer (Cleland 1966; Smith 1975). While in many interior areas this may well have been the case, it seems reason- able to predict that there would be variation on this theme in response to locally avail- able food stuffs. The response might be so precise that populations occupying the same environment separated by only 81 km of coast may have practiced different subsistence strategies, none of which emphasized deer. There also appears to have been a subsistence shift between Savannah and Irene phases on Sapelo Island as evidenced by the differences in faunal assemblages from the North of the Shell Ring Drain site and the two villages. When the faunal assemblages from Kenan Field and Bourbon Field are compared there were few striking differences (Table 5). These two sites are primarily Savannah/ Irene phases, although Crook does not think they were occupied simultaneously (personal communication). The only substantial difference is in the amount of turtles consumed at Kenan Field compared with Bourbon and a related reduction in the volume of fish. The Kenan materials are more diverse than those at Bourbon, perhaps because of the more extensive salt marsh bordering Kenan Field compared to that of Bourbon. Another interesting contrast is found between the use of mullets (Mugil spp.), star drum (Stellifer lanceolatus), silver perch (Bairdiella chrysoura), and herrings (Clupeidae) at Bourbon and the virtual absence of these animals at Kenan. These fishes should be equally common on ei sides of ng: ere The mullets at Bourbon Field were small — - arily in 1/8 in mesh samples, the same screen size in which s ms (S. aaa pai perches (B. chrysoura), and herrings (Clupeidae) are most ares to be recovered. It appears that a fine-meshed, mass-capture technique was used exten- sively at Bourbon and seldom used at Kenan Field. If not occupied throughout the year, Bourbon may have been occupied intentionally at specific periods to take advantage of these species. Although they could have been captured off of Kenan Field, it is possible that it was more efficient to catch these fish at Bourbon for reasons not evident today. REITZ May 1982 oLg Lel a1 £ aa £93 L’'v él L°88 826 8's or 80 j 61 g % INW ‘S'q sainjeay yeuuRAes % INW "Sd «SD soorey URS Aeg s8ury $0 I €°9 €I - S02 b'¢ é 6°19 Lal Lol 96 6° 8 0'sT LE % INW PA} au07 sOoIe URS $8 INW dusn ‘S41 4+°SO sajdwuieg uwinyjor soo URS y Vel ra aa j = 9¢T v9 ol 168 6E1 9% ¥ 6T & % INW uoqinog pues] ojades sajdureg uunyjorn uoqinog + pyarparog /42411]99§ 199q] [IOL sjesuawiuM0;) sAvy ‘SyIeYyS ‘YSTA sanday sparg vaorqay SPUR [PINS9L19 7, sajdung pausasrg-ui y puv (Sy) pauaas2g-aury fo uosuvduo)—g ATAVL 58 REITZ Vol. 2, No. 1 It is generally agreed that Sapelo Island and Kings Bay were occupied historically by two distinct groups: Guale at the northern location and Timucuans at Kings Bay and points south (Milanich and Proctor 1978). It is tempting, therefore, to interpret the difference observed between the aboriginal Sapelo Island data and the Kings Bay data as cultural ones. In order to keep recovery technique relatively constant, the column- sample data from Bourbon Field should be compared with the Savannah Phase features at Kings Bay. The animals of choice at Kings Bay were clearly small drums, primarily the star drum (S. lanceolatus). Neither mullets (Mugil spp.) nor herrings (Clupeidae), the dominant species at Bourbon, are common at Kings Bay. The emphasis on star drums at Kings Bay strongly suggests an intentional effort to acquire this fish to the ex- clusion of other animals. Star drums prefer more saline waters than do some other estuarine fishes. More data are needed about natural differences between the two areas and human use of both locations before this difference can be ascribed cultural signi- ficance, but there does appear to be tentative evidence for such a difference. e later coastal strategy, here represented by the Irene component at North of the Shell Ring Drain, was different in some aspects from earlier subsistence patterns. In the first place, deer are a minor component. Since deer bones were more likely to be recovered than the fishes because of recovery techniques, this suggests low numbers of deer taken by the inhabitants. This finding may also be the result of differential distri- bution of refuse as the midden was deposited, since the excavation units were contiguous at the site rather than randomly distributed over a wide area as at the other Sapelo Island sites and at Kings Bay. Placement of excavation units would affect MNI aggregation. The two striking features of this sample are the abundant remains of sea catfishes (Ariidae), primarily the hardheaded catfish (Ariopsis felis), and the somewhat increased number of bird elements. The presence of so many catfishes may indicate a primarily hook-and-line technology in contrast to an earlier net technology. The presence of both spots (Letos- tomus xanthurus) and silver perches (B. chrysoura) niga continued use of nets, weirs, or basketry scoops although both fish will take hooks. of birds during this period is unusual for aboriginal subsistence on the shat oi ease not unusual for historic ea aa (Reitz 1979a, 1979b; Smith et al. 1981). The San Marcos component at Bay also reflected an increased use of birds (Table 6). This may have been the ea of European influence on the aboriginal hunting strate While fishing was clearly a major activity at all of these te: fish were not the only source of animal protein in the diet. Biomass was calculated for three of the sites. As might be expected, terrestrial mammals, primarily deer, contributed a substantial amount of biomass. However, deer contributed over 50% of the biomass only at Bourbon F ield. Based on sampling considerations, the figures from the Savannah Phase features at Kings Bay are thought to give the most accurate picture of the diet. ceili show that fishing and hunting were both important activities, with fishing somewhat more 0.65 to 0.83. In terms of individuals (MNI) the strategy was one in which a few major species and a number of less important ones were used. In terms of biomass, considering only the Savannah Phase features from Kings Bay, once again a limited range of species was used, with a few animals being more important than the others. Diversity range was 1.4 to 2.2 and equitability 0.39 to 0.64. The most important animals were drums, parti- cularly silver perches, and star drums; sea catfishes; deer; diamond-back terrapines; and occasionally spots and croakers. The major biotope exploited appears to have been the tidal creeks. Terrestrial areas were exploited primarily for deer; very few other mammals, freshwater or terrestrial tur- tles, and birds were taken. The fishes, sharks, rays, sea turtles, diamond-back terrapins, bottle-nosed dolphins, and alligators could have been taken in the nearby marshes and tidal creeks. Most fishing could have been done from shore or in shallow waters. | May 1982 REITZ 59 Seasonal indicators are less apparent than could be expected from other geographical areas. The species used are primarily multi-seasonal and those few animals which are restricted seasonally were not exploited to any great extent. Distinctly warm weather species such as sea turtles and sharks attest to warm weather occupation at the Kenan, pian: and North of the Shell Ring Drain sites. Herrings and fingerling mullets at Kenan, Bourbon, North of the Shell Ring Drain, and Kings Bay may document a cold weather occupation (DEIS 1978). Star drums and silver perches are more abundant in the fall and spring and suggest fall and spring occupations at all sites. It appears that occupation of sea-island locations was not confined to a single summer residence, but was either intermittent throughout the year or continuous. This possibility is partially sup- ported by ethnographic and archaeological evidence (Crook 1978) Fishing technology clearly emphasized techniques si ig to the capture of small fishes. Star drums (S. lanceolatus) have an approximate maximum length of 15 cm (Hoese and Moore 1977) and silver perch (B. chrysoura) have a maximum length of about m. The mullets recovered at most sites, particularly at Bourbon Field, were also in this size range. Spots (L. xanthurus) and croakers (Micropogonias undulatus) are small quantity of small drums and the presence of mullets suggest use of impoundment or trapping devices such as nets, scoops, or weirs. Nets could have been placed across tidal creeks, while weirs could have been used where the bottom was firm such as near oyster bars. Even catfishes (Ariidae) and the small sharks found in these collections could have been captured by these devices. Such mass capture techniques would indicate group- subsistence efforts were in use CONCLUSIONS e data presented here suggest resources of the estuarine environment were selec- tively exploited. Sites were oc cupied perhaps during more than one season as these resources became available. Mass-capture techniques may have been employed in secur- ing selected species, most of which were small drums and mullets, while other estuarine species and deer were also taken. Among the sites there is sufficient variations of repre- sented fauna suggest there were strategies specific to each location with some evidence for temporal and cultural variation as well. However, these samples conform to a general coastal pattern which includes the following: use of deer to some extent, varying from site to site but rarely more than 50% of the biomass or 11% of the individuals; low use of birds; occasional use of turtles, both marine and aquatic; heavy use of marine fishes, timarily small drums and sea catfishes. Use of large numbers of small fishes suggests a fishing technology employing nets and weirs rather than hand-lines or trot-lines. Clearly more work needs to be done on coastal subsistence, with appropriate con- cern for field techniques. Based on these collections it can be predicted that systematic column sampling will undoubtedly produce faunal samples over the 200 MNI mark. On that basis it is recommended that future excavators submit for identification and analysis only their column samples and features. It can be argued that 1/4 in mesh is inadequate for sampling coastal aboriginal shell middens. It is also inappropriate to use 1/4 in screen for the bulk of the soil and fine-screen only a portion, but combine the species lists. Archaeologists need to consciously make a decision concerning screen sizes to use in the field as recommended by Thomas (1969). If these guidelines are followed, it may be that the full complexity of coastal subsistence of mainland, marsh-island, and sea-island sites along the Georgia coast for all times periods will be revealed. 60 REITZ Vol. 2, No.1 ACKNOWLEDGEMENTS I am grateful to Dr. Lewis Larson, Dr. Ray Crook, and Robin Smith for the gaia to study nes Eonar from their respective excavations. Funding for analysis of the Kin d by the United States Department of Navy, Contract Number N00025-79-0013, os Bourbon ay pee by a National Geographic Society Research and Exploration Grant, and Kenan Field by a National Science Foundation for Dissertation Improvement Grant Number 77-07565. Support was also provided by the Georgia Department of Natural Resources. Elizabeth S. Wing, Stephen Hale, and George Burgess all Sapa valuable comments o presented at the n the paper. An earlier version of this paper was Annual Meeting of the Southeastern Archaeological Conference, Asheville, 1981, The eee drawn in this paper are, however, the responsibility of the author LITERATURE CITED ANONYMOUS. 1968. General Highway Map: McIntosh County, Georgia. State Highway Dept. Georgia, Atlanta. gore JOHN. 1974. Marine zoogeography. -Hill, New York. aa. ARCHIE. phen re of turtles. Cornell Univ. Press, It CASTEEL, RICHARD 9 "1976, Fish remains eeded community nhasiiias on the Georgia coast. Unpubl. cin Back (Anth- rop.), Univ. Florida, Gainesv —— ms FUG. Space, time, nga subsistence at Bourban Field. Nat. Geo. Res. Reports, 1980, in press. rs pane MICHAEL D. 1972. An ecolog- tudy of Georgia coastal fishes. Fishery Seal. "70(2): 323-353. ————., and E.P. ODUM. 1970. Annual cycles re species occurrences, abundance, and diversity in Geo estuarine popula- tions. Amer. Midland Natur. 83(2):382-392. DEIS. 1978. Draft environmental impact statement for agi hpi location for a fleet ballistic missil (FBM) submarine support base, Kings Sher, Georgia, Dept. Navy Washington, D.C. EKMAN, SVEN. 1953. Re hy of the sea. Sidgwick and Jackson, Lon GOLLEY, FRANK. 1962. poner of Ge orgia. iv. Ge power DONALD. 1973. On the meth- dology of faunal analysis. Amer. Antiquity 38(4) :432-439, ——___.._ 1978. Minimum numbers and sample size in leony faunal analysis. Amer. Antiquity 43:53-6 = es FONE. The Bick of sample size on some ram ed measures in vertebrate faunal alysis. J. nce Sci. 8:77-78. HOESE, H. DICKSO 1967. Effect of Sa rmal salinities on salt marshes. Marine aa 12:249-261 JO HNSON, A. SIDNEY: H.O. HILLESTAD: S.F. SHANHOLTZER, and G.F. SHANHOL- TZER. 1974. An ecological survey of the coastal region of Georgia. Natl. Park Service i. Monog. Series No. 3, Washington, D.C. H. 1978. Historical Guale Univ. Presses of Florida, . 1980. Aboriginal subsistence tech- nology on southeastern coastal plain during es late Faso periods. am P. Bullen Monog. Anthrop. and Hist., No. 2, Univ. Presses of Florida, en MUSIC; B.A. PALMER. 1974, Survey of the fisheries resources in Georgia’s estuarine and inshore ocean waters. Part I. Georgia Dept. Nat. Res. Contrib. Series No. 22. 1975. Ceramics, rgia publ. .D. dissert. (Anthrop.), el salen iste MILANICH, JERALD T. 1977. A chronology for the aboriginal cultures of northern St. Simons Island, Georgia. Florida Anthropol. 30(3):134-142. AMUEL PROCTOR (eds.)- oe ae SS 1978. Tacachale. Ripley P. Bullen Monos. Anthrop. and Hist. No. 1. Univ. Presses of Florida, Gainseville. May 1982 REITZ 61 LITERATURE CITED (continued) PEARSON, CHARLES. soe Patterns of rages sens adapta in coastal eor 2 PED: . (Anthrop.). Univ. SAE aes Ss. REITZ, ELIZABETH J. 1979a. Availability f fish along coastal Georgia an Florida. Paper presented i the Southeastern Arch. Conf., Atlanta, Georgi 1 Pilg ai British sub- sistence niga at St. Au oenetigs nee a, F rica, Georgia, between 1565 a 1783. Catal Ph.D. dissert. haben: : le. ROBERTS LLIAM and JAMES KUSH- LAN 974. The South Florida avifauna Pp. 4 » in Environ Florida: Present and Past. (PJ. Gleason, ed.). Miami Geol. Soc. Mem. No. 2, Miami. SHANNON, CL E. d ARREN WEAVER. 1949, The mathematical theory of communication. Univ. Illinois Press, Urbana SHELDON, A. 1969. Equitability indices: Dependence on the species count. Ecology 50:466 SMITH, BRUCE. 1975, Middle Mississippi exploitation of animal populations. Univ. Michigan, Anthropol. Paper 57, Ann ies SMITH » GO, Mig Ned. ap aeological sites at Kings Bay. Final Report on 3 ago testing at Kings Bay, Camden County, Georgia. Report on file, Univ. sville. STICKNEY, ROBERT R. and D. MILLER. 1973. Chemical and biological sig of the Savannah adjacent to Elba Island. a Marine Sci. Center, Tech. Report yet d and. THOMAS, DAVID HURST. 1969. Great asin hunting pete a quantitative meth- od for treating faunal remains. Amer. Anti- quity 34:392401. TIDE TABLES. 1981. Tide Tables for the East Coast of North and South America. Natl. Ocean Survey, U.S. Dept. Commerce, Rockville, Maryland. WING, ELIZABETH S., and ANTOINETTE B. ROWN. 197 Paleomutidon: method and theory in prehistoric foodways. Acade- mic Press, New York. J. Ethnobiol. 2(1): 63-77 May 1982 BIOLOGICAL CLASSIFICATION FROM A GROOTE EYLANDT ABORIGINE’S POINT OF VIEW1 JULIE WADDY Angurugu Community Bag Service No. 1, via Darwin, N.T., Australia ; ot ABSTRACT.—This paper reviews the contributions of Berlin, Bulmer, Dwyer, Hunn and landt in northern Australia is described using Berlin’s set of terms as a convenient reference point. Examples given provide additional evidence for the notion of a basic unit of percep- tion within folk classification systems though this unit cannot ip a be een de- fined. is suggested that folk taxa may be arranged in contras o purpose and that only those sets required for the purpose would be ane on to senetite a hier- archical ae INTRODUCTION In recent years increasing interest has been taken in the principles of folk classifica- tion. Research workers have been fascinated by the relatively high degree of correspon- dence between folk biological taxa and scientific taxa. Our scientifically trained minds ee see the parallels between folk categories at different levels of inclusiveness ficatory hierarchies are similar in many ways, descaneees have jmnalioes them from different isa which has led to different taxonomic structures accompanied by different sets of t The reasons rey digxe differences I consider to be threefold. Dwyer (1976:425) identified two of these issues in reporting his analysis of Rofaifo mammal taxonomy, viz. 1) ‘To what extent does the folk classifier perceive the same entities as the scientific Zoclogist?” and 2) ‘What is the cognitive status for folk of taxa located at opera’ levels their zoological taxonomy?’ The third issue has been raised by Randall (19 244), i Are hieaienieab « classification systems stored as such in the memory or are i only a result of classification behavior on the part of the folk classifier? THEORETICAL ISSUES Comparing perceptions. pomies er has pointed out that determining the correspon- dence between two systems is a pants of perception. There must be some way of oe their relation. He s ected the scientific species as the objective unit with ich folk taxa, regardless of — must be compared. Berlin advocated comparison a the scientific species with his folk genera, though he has also compared it with his folk species (Berlin et al. 1973:267-8, 1974:102). However I agree with Hunn (1977: 64) that ‘it is not the case that the scientific species must be selected.’ I understand him to be saying that, while the scientific species is indeed a basic objective unit, irrespective of evolutionary theory, we need to take cognizance of the range of scientific species in a given environment before determining the degree of correspondence. Hunn has devised 64 WADDY Vol. 2, No. 1 what he calls a coefficient of dissimilarity, which is calculated after removing any scien- tific taxa that cannot be found in the local environment plus scientific taxa below the level of (labeled) terminal folk taxa. This measure is not affected by the cognitive status of folk taxa. It utilizes scientific taxa as the objective basis of comparison with folk taxa of whatever status. It should be noted that despite the objectivity of the scientific species it necessarily has a cognitive status within the scientific classification hierarchy. Implicit in Dwyer’s first question is the western scientifically oriented view point. The question could equally have been framed: To what extent does the scientific zoolo- gist perceive the same entities as the folk classifier? As there are many folk ae systems but essentially only one scientific classification system comparison would certain ly be easier if the degree of correspondence is determined with reference to the tei system, But a further difficulty is in determining a unit from within the folk classifica- tion system to which general agreement can be given. To my mind no-one has yet been able to suggest a satisfactory objectively defined unit from within the folk system. Berlin, Breedlove and Raven (1973:215) proposed the folk genus, defined largely on the basis of the distinction between primary and secondary lexemes. Thus generic taxa are so called because they are labeled by generic names. Hunn (1977:45) has sug- gested that this association of taxa and names should be verified by specifying indepen- dent criteria for recognizing types of names and types of taxa. I have heard Berlin has now had second thoughts in the light of more recent examples that do not fit easily into his original scheme. Bulmer and Tyler (1968:349) proposed the specieme or folk species which is the lowest level taxon defined in terms of multiple criteria. The specieme is seen as a ‘nat- ural’ category within the environment—‘something crying out to be named,’ as someone has said. But Bulmer too has had second thoughts about this concept in the light of Kalam interchangeable usage of names at apparently different levels of inclusiveness. The rejection of these two apparently objective units brings us to the second issue raised by Dwyer. ognitive status and perception of discontinuity.—Assessing the cognitive status for folk of taxa within their zoological (or botanical) taxonomy revolves around what are perceived as ‘natural’ categories within their environment. Bulmer (1970, 1974) con- sidered that, in the case of locally familiar organisms, the majority of folk taxa corres- ponded to ‘natural’ categories, i.e. to those defined on the basis of multiple criteria. When such folk taxa correspond to recognized scientific taxa it is then tempting to assume that these folk taxa refer to ‘natural’ categories of equivalent cognitive status. It is much easier to match taxa with scientific species, genera, families, etc. than it is to establish with 7 that they are ‘natural kinds’ in the perception of the local people (Bulmer, pers. comm In discussing Lé vi-S trauss’s concept of espéce which has been translated as species, Bulmer (1970:1072) identified one of the assumptions underlying Lévi-Strauss’s argu- ment, viz. ‘that in any total folk-classification of plants and animals there are certain important lower order categories which are seen as “objective” by the users of the clas- sification . . .’ Bulmer (1970:1081) then argued that ‘Karam noalogien! classification, at the lowest level, is concerned with objective discontinuities in nature.’ He considered that the basis of such objectivity is in the observable differences between biological species although he recognized that not all folk taxa will be classified in a biologically realistic manner. Hunn (1977:50) has indicated that the majority of folk taxa can be recognized by characteristic configurations and are defined by significant discontinuities between contrasting categories. Underlying his mathematical treatment of these discontinuities is the assumption that the points at which discontinuities are perceived vary from culture to culture, and between folk and scientific taxa because the perception of discontinuity— in Hunn’s terms, the perceptual salience—varies. Hunn’s approach seems to readily include all folk taxa whereas Bulmer’s approach accepts some and makes exception for others. aimee eRe | May 1982 WADDY 65 The perception of discontinuity may be affected by: 1) identifiable characteristics, 2) cultural significance, and 3) frequency of observation. If the identifiable characteris- tics of two or more scientific taxa are minimal and there is little or no difference in their cultural significance then they may be perceived as one entity even though the differences between them may be recognized. Similarly, if an animal or plant is only rarely encoun- tered it may be included with another scientific taxon, and thus again be perceived as one entity. At this point I would differ with Bulmer and Dwyer who consider terminal unlabeled subdivisions of taxa to have the same status as labeled terminal taxa. To me they appear to have fallen into the trap of assuming that correspondence with scientific taxa implies the same degree of perception of discontinuity despite Bulmer’s awareness of this trap (Bulmer, 1970:1078). Discontinuities are not perceived at the same point. As against Bulmer and Dwyer, I would say that the basic perception of discontinuity, the basic units as seen by folk themselves, must firstly be labeled, i.e. named, and second- ly be undivided, i.e. perceived as a unitary whole which, on occasion, can be further sub- divided. iinguiedeatte labeled subdivided taxa, such as Berlin’s folk - hte represent a different degree of perception. Their identifiable characteristics would not be expected to differ as much as the identifiable parponeteaich separating undivided taxa, though the subdivided taxa may still be defined o asis of multiple criteria. Linguistically un- labeled subdivisions of a taxon pate a np ends degree of perception again. I fail to see how an unnamed subdivision of a taxon can have the same conceptual content as a labeled taxon nor, for that matter, how a secondary lexeme can have the same conceptual content or cognitive status as an undivided primary lexeme. Berlin, Breedlove and Raven (1973:240) go so far as to say that there are different psychological processes involved in distinguishing taxa at different levels of inclusion. I think it is implicit in the work of Berlin, Hunn and Hays that it is only the named taxa, at least at the lowest levels, which truly reflect the perception of discontinuity and thus of ‘natural’ categories. There are differences in the degree of the perception of dis- continuity as indicated above. It is these differences which give rise to differing cognitive Status and thus the different levels of a taxonomy. he problem arises in seeking to assess the cognitive status of those taxa at the lowest levels and likewise of taxa at higher levels of inclusion. It is at this point that Berlin has confounded the questions of perception and of cognitive status as Dwyer (1976:433) claims. Berlin’s folk genus purports to convey both cognitive status and perception of discontinuity—without, however, the degree of discontinuity being satisfactorily defined. Hunn (1977:51) has sought to redefine the status of generic taxa in terms of ‘the width of the gaps isolating taxa and the “width” of heterogeneity, of the taxa them- selves.’ The major difficulty of such a formulation, as Hunn himself has said, is the problem of measurement. The nature of folk taxonomies.—Berlin, Breedlove and Raven (1973:216) have recog- nized folk generic taxa as the basic building blocks of all folk taxonomies, i.e. the most commonly referred to groupings of organisms in the natural environment and the most Salient psychologically. They have then ranked folk taxa by inclusion relationships to produce five levels of inclusiveness (Fig. ite essentially hedhinies Berlin, Breedlove and Raves schema but with minor nadatsennwie 66 WADDY Vol. 2, No. 1 Level 1 Life form If) lf, oj AKAL Level 2 Generic &3 4 85 &6 ---Sm Bn $1 2068} sj \ i Level 4 Varietal v Vo Vm Yn FIG. 1—Schematic presentation of Berlin’s schema. (Adapted from Berlin, Breedlove & Raven, 1974:26). In establishing his hierarchical Pear aad of Kalam vertebrates Bulmer (1968: 622) began with the most inclusive labe taxa, i.e. primary taxa, and worked down- wards, through as many as three a seuccad levels, to the terminal taxa. A schematic interpretation of Bulmer’s data is shown in Figure 2 for comparative purposes. Level 1 Primary Py +-Pp taxa Level 2 ay a Level 3 weedy Level 4 specieme s Sg S83 S4 (ss) Ge) 87 nae ten. ty Sn FIG. 2—Schematic interpretation of Bulmer’s data. May 1982 WADDY 67 Bulmer’s primary taxa vary considerably in their degree of internal variation. In his ex- perience terminal taxa may be at any of four levels but most are at Level 2. In the majority of cases it is the terminal taxa which represent the ‘natural kinds’ or folk species that Bulmer considered the Kalam themselves recognize. In some instances these ‘natural kinds’ are unlabeled subdivisions of a terminal taxon and could be considered as covert species us the rank, or cognitive status, of folk species cannot be fixed within the licrarchy, either by position or by the terminology employed. I find it difficult to accept that the cognitive status of the basic units of perception can be variable Dwyer (1976:435) used Bulmer’s concept of ‘specieme’ or folk species but reversed the levels applied to taxa. In other words he worked from the bottom upwards through categories of increasing inclusiveness. Again a schematic interpretation of Dwyer’s data as been provided for comparative purposes (Fig. 3). AAA A Level 1 Primar tax Level 2 Secondary oe 83 “xn taxa Level 3 Tertiary ty tg tg t4 ee ‘n taxa Level 4 Quarternary q q2 “4m taxa FIG. 3—Schematic interpretation of Dwyer’s data. ( ) designate unlabeled taxa at the lowest level. From the figure it will be seen that unlabeled subdivisions of taxa have been given equiva- lent rank to labeled undivided taxa. As indicated in the previous section, I would ques- tion the validity of Dwyer’s interpretation. Since Dwyer’s study is limited essentially to mammals he gives no examples of taxa equivalent to Berlin’s unaffiliated generics. How- ever he does give examples of taxa not included in any but the highest level of inclu- Siveness. Each of these workers has assumed that there is a valid observable hierarchical system of folk classification. Randall (1976:546) questioned this, suggesting that while the various adjacent levels of a hierarchy may well represent valid relationships, the total hierarchy is something contrived in the mind of the informant, generated by lige aaa questioning. The trouble, as Randall has seen it, is that there may be instances of no transitive seernerea appearing in such hierarchies where, for argument’s sake, a on oak is a kind of oak and an oak is a kind of tree but a scrub oak is not a tree, it is a shrub. andall tie a non-hierarchical classificatory schema, based on an association between categories and their perceptual characteristics stored directly in (Fig. 4 By his mixing of categories from a variety of special purpose Sen systems, e.g. food classification, with the general purpose biological classification system, the complexities of Randall’s system are mind-bogging, especially if there is a high degree of bin Oomialisati 68 WADDY Vol. 2, No. 1 VEGETATION-NESS TREE-NESS FRUIT TREE-NESS CHERRY-NESS “black cherry "pear tree vocado “weeping cherry *birch tree “weeping beech }WEEPING-NESS| OAK-NESS - *black oak ° , *live oak % ~—— pine / weeping willow ae 7r r — Se WILLOW-NESS SHRUB-NESS/ * pussy willow BUSHINESS 3 SCRUB-NESS ° ON BERRY-NESS CANE-NESS * blueberries *blackberries “raspberries fies ee *wild ro Ps PLANT-NESS Noor es *jack-in-the-pulpit FLOWER-NESS St ee FIG. 4—Randall’s model. A memorisation of characteristics model of some English plant categories. (Randall, 1976:551). other words the basic problems still remain; How do we determine what ae een categories within a folk classification system? How do we determine their cognitive status? and, How is the folk classification system derived? I want to return to these questions after first discussing biological classification from the point of view of Groote Eylandt Aborigines. GROOTE EYLANDT CLASSIFICATION Groote Eylandt, which is roughly 40km wide and 60km long, is situated in the - of Carpentaria, (Fig. 5). I have been living at Angurugu on Groote Eylandt since 1975. 69 “eIJEI]SNW JO PULTUTEW 9Y} 0} UOT}PIAl UT spurs! SUIpUNOIINs puke }puR]Aq 910015 Surmoys deW—S ‘OL WADDY 4 ONV ISI NOLYSyNIIG e1yejuedues pao) May 1982 a G 70 WADDY Vol. 2, No. 1 The language spoken by the Aborigines of Groote Eylandt is Anindilyakwa, a langu- age that is confined almost entirely to Groote Eylandt and surrounding islands. This language is characterized by its multiple noun classes, its extensive prefixing and suffixing systems, and by its very long words.3 I have drawn on a comprehensive inventory of some 220 plant taxa obtained largely by Dulcie Levitt (1981), and some 420 animal taxa, many of which were first recorded by Judith Stokes, linguist with the Church Missionary Society at Angurugu. Although the list of folk taxa is virtually complete the number of scientific species represented by obtaining scientific classification of animal specimens. Having gained familiarity with almost all animal and plant kingdom taxa on the island, at least through references if not with the actual specimens, I then turned my attention to the classification system as a whole. Much of the information which follows has been patiently imparted to me by Peter gurama, a man about 55 years old of the Wurrawilya clan with an extraordinary knowledge 7 the plants and animals on the island. He is a recognized local authority. I have also learned a tremendous amount from a number of the old women. It is only the older folk who lived as young adults in the bush who have any extensive knowledge. There have been very few discrepancies in the naming of taxa but there may be slight changes when some of the less clearly defined areas, such as covert categories, are checked with other people. For convenience I shall use Berlin’s set of terms as a point of reference in describing Anindilyakwa taxa The Plant Kingdom: Amarda.—Unlike the majority of languages, Anindilyakwa has terms which are used as unique beginners both for the plant kingdom, viz. amarda, and for the animal kingdom, viz. akwalya.4 The term amarda is also used to refer to one of the two life form taxa. ese taxa are based on binary opposition of woody vs. non- woody. Thus eka refers to all woody plants, viz. trees and shrubs, and amarda includes all non-woody plants, viz. grasses, sedges, rushes, herbs, vines, creepers, ferns, seaweeds and so on (Fig. 6). The only plant which does not fit unambiguously into these two life form taxa is the cycad or burrawang, Cycas angulata. The burrawang stem is soft rather than woody, despite its tree-like form, but is deep-rooted like other trees. Within the woody plants there is a total of 114 generic taxa. Nangurama has grouped them into eight categories, partly on the basis of similarity in form and partly on the basis of shared habitat. Three of these categories are further subdivided into three or four categories. One of these latter categories is named, viz. alyukwurra the paperbarks. The seven taxa included within alyukwurra (Fig. 7) all appear to have the same psychological salience as other generic taxa, such as the examples in — 6. Thus alyukwurra has been interpreted as a labeled intermediate taxon in Berlin’s term The non-woody plants include a total of 79 generic taxa. eaenabia has grouped these taxa into three large covert categories and one small one which includes the six seaweed taxa. An alternative grouping was proposed by another local authority on the basis of root form. He divided each of the two larger categories into two. The existence of these alternative categories suggests that they may not be as well-defined as the covert os reported by Berlin, Breedlove and Raven (1968:294-296). In comparison with the data on plants presented by Berlin, Breedlove and Raven and by cg aa are extraordinarily few labeled taxa of specific rank in Anindilyakwa. One example is in the group of grasses with awned seeds dingarrkwa. Dukwulyadada dingarrk- wa meaning ‘white seeds’ refers to Aristida browniana and dumurrijungwa dingarrkwa meaning ‘black seeds’ refers to Pseudopogonatherum irritans, neither of which has parti- cular cultural significance. Other examples of specific taxa found are big-leaved/ small- leaved (3 generic taxa), good/ bad (of no use) (lgeneric taxon), and the ‘real’ or ‘true’ one, (several instances "cmrmemem en etme : May 1982 WADDY 71 amarda ~plants, foliage eka amarda -woody plants 9 -non-woody plants (114 taxa) (79 taxa) Se | eet alabura mabalba miyarrawa muninga dingarrkwa murungkwurra akwurena ~ Darwin -yellow -red -burrawang -awned -roundyam - wild grape stringybark hibiscus kurrajong grasses Eucalyptus Hibiscus Brachychiton Cycas JEIDISCUS Dioscorea Ampelocissus tetrodonta —_tiliaceus paradoxum —_angula bulbifera acetosa dumurrijungwa dukwulyadada dingarrkwa dingarrkwa dingarrkwa - Pseudopogonatherum - (As - other grass irritans niana with awned a FIG. 6—Biological classification in the plant kingdom from an Anindilyakwa speaker’s point of view. Numbers of taxa are those designated generic by Berlin. alyukwurra - paperbarks Melaleuca spp. 4 “a | \ < ! % , x \, yirarrnganja mamarra | mawilyaburna angwurralya M. leueadendron M. sp. aff. cajaputi M. symphyocarpa M. acacioides i iy | x > | \ / - ayalukwa eee es yilyerrbirradangwa M. leucadendron M. viridiflora Melaleuca sp. FIG. 7—Terminal taxa included within the folk taxon alyukwurra paperbarks. 72 WADDY Vol. 2, No. 1 The Aborigines of Groote Eylandt were hunters and gatherers who relied largely on fish and turtles, some land animals and on bush fruits and roots. They ate few seeds and no leafy vegetable matter. It will be interesting to see if other hunter-gatherer societies also have such a sparsity of folk specifics. If so it would support my contention that folk specifics and folk varietals may have developed largely in societies where agriculture plays a significant role in the economy and there is a subsequent need to make finer distinc- tions within a taxon. This seems to bea corollary of Berlin, Breedlove and Raven’s find- ing that the proportion of folk specifics is much higher among cultivated and protected plants than among other plants (Berlin et al. 1974:99). The Animal Kingdom: Akwalya.—As previously noted the unique beginner for the animal kingdom is akwalya. The first division (Fig. 8) including akwalya ‘animals in the sea’ and yinungungwangba ‘animals on the land’ seems strange in comparison with scienti- fic thinking. akwalya - all animal life akwalya wurrajija yinungungwangba - animals in the sea - winged creatures - animals on the land and others (122 taxa) / / / - fish - marine turtles - snakes and - other land (136 taxa) (5 taxa) legless lizards animals (16 — (27 taxa) / | : i (sea mammals) ieee (coelenterates) (grubs) - earthworms & tas / (14 es (6 taxa) (5 taxa) / \ | \ fe 4 (octopus, squid) (echinoderms) - beachworms (2 taxa) (3 taxa) FIG. 8—Biological classification of ihe animal kingdom from an Anindilyakwa speaker’s point of view. Numbers of taxa are those d d generic by Berlin. May 1982 WADDY 73 However the naming of these two categories reflects the basic dichotomy between life in the sea and on the land that is borne out in other areas of life for the people of Groote Eylandt. The old Anindilyakwa word for women, warningaribumanja, when literally translated, means ‘people of the land’ whereas the men are people of the sea. Although it became apparent that this division was the source of a number of anomalies resulting in nontranstivity, e.g. land snails that are classified with marine molluscs, there was no way in which this division could be deleted, despite my informants’ awareness of the need to stick to purely animal classification without interference from special purpose uses such as food source. To me it seems that habitat must be accepted as a valid factor influencing folk biological classification. At least for the Groote Eylandt Aborigine at this level of inclusiveness, habitat cannot be dismissed as interference from a special purpose classifi- cation, There has been some difficulty as to the relative status of wurrajija ‘winged creatures and others’, Nangurama wanted this taxon to be included within both land animals and sea animals, which would have violated normal taxonomic principles. Another know- ledgeable man has given wurrajija equal status to land and sea animals. The latter view point is followed in Figure 8. This aspect of the classification system needs to be checked further. The primary focus of the taxon wurrajija appears to be birds. When asked for defin- ing features of the taxon, the immediate response given was ‘wings’. It is thus easy to see how most insects, flying foxes and bats are included. Both winged and non-winged forms of green tree ants in particular are easily recognized. Green tree ants crawl on one’s body as do other ants and insects, ticks, spiders and even scorpions and caterpillars. So one can understand how the taxon has been extended to include almost all arthropods. Grubs that live inside trees or in the ground are an exception. gurama arranged sea birds (34 generic taxa) into two large and three small covert categories. He considered land birds (40 generic taxa) as one large covert category in contrast to six covert categories of insects (45 generic taxa) and one covert category of bats and flying foxes (3 generic taxa). Labeled life form taxa included within akwalya ‘animals in the sea’ are akwalya ‘fish’, adidira ‘shellfish’ and setae ‘marine turtles’. Akwalya ‘fish’ divides into aran- Jarra which includes all the cartilaginous fish and akwalya which includes all the bony fish (112 generic taxa) including a small subdivision of freshwater fish (8 generic taxa) (Fig. 9).4 akwalya - all fish oe : Fish - bony fish cartilaginous i is a 3 ) ~ sharks . (stingrays ‘sien - suckerfish (9 taxa) nosed ite (1 taxon) (14 taxa) FIG. 9—Labeled coneaccion within the folk taxon a ne fish. 74 WADDY Vol. 2, No. 1 Aranjarra is further subdivided into mangiyuwanga ‘sharks’ (9 generic taxa), an unlabeled category which includes stingrays (11 taxa) and shovel-nosed rays (3 taxa) and the generic taxon amadengmina ‘suckerfish’. This means that, in Berlin’s terms, there are labeled intermediate categories at two levels, an unusual feature in relation to other languages. e orm taxon adidira includes almost all members of the phylum Mollusca and also hermit crabs. The only exceptions are the octopus and the squid. Nangurama has given five covert categories of five or more taxa and fifteen covert categories of one to three taxa, making a total of 64 generic taxa. Land snails (2 taxa) and freshwater mussels (1 taxon) are included in this life form and are thus examples of nontransitivity. The taxon yimenda ‘marine turtles’ is interesting because of its apparent status as a life form taxon but one which includes only five, or at the most, six generic taxa. Life form taxa tend to include a large number of generic taxa (Hunn 1977:44). In addition to the labeled taxon yimenda ‘marine turtles’ there are three significant but covert taxa which can be glossed as Crustacea, marine mammals and Coelenterates. Scientific cate- gories such as these are very distinctive and yet are limited in species diversity, or at least in readily observable species diversity in a given area such as the seas adjacent to Groote ndt. The existence of generic taxa, of equivalent cognitive status to other generic taxa within the total system, within each of these more inclusive categories suggests that these higher level categories should be considered as life form taxa, whether named or covert. These higher level taxa are in contrast to other labeled life form taxa. Thus neither yimenda ‘marine turtles’ nor any of the generic taxa in question can be dismissed as ora generics. argest covert taxon is the Crustacea. There is a labeled intermediate taxon alkwa one includes all the bait crabs (9 generic taxa) but not the large edible mud crab, Scylla serrata. There are three other taxa of generic rank included in the life form taxon. One of iia taxa is amilyungwurra ‘freshwater yabbies and shrimps’ which is now ex- tended to include prawns. In its original meaning it is another example of nontransitivity. The other two covert taxa are marine mammals (5 generic taxa) and the Coelenter- ates (6 generic taxa in 2 or 3 groups). There are another three to five groups containing one to three generic taxa and totalling seven or eight taxa. Approximately two thirds of all animal taxa are found in or near the sea There is one definite labeled life form taxon within yinungungwangba ‘land animals’. Yingarna includes all snakes as well as nae lizards and the eel Piscodonophis boro. Yingarna is subdivided into dingarna ‘pythons and tree snakes’ (9 generic taxa) and yingarna which includes the remainder but eae y the poisonous snakes (7 generic taxa). Sea snakes are included with pythons and tree snakes and thus provide another example of nontransitivity. There is some debate as to whether the remaining land animals, i.e. 4-footed land reptiles and mammals, should be polysemously labeled or not. Within this grouping there is a covert taxon of marsupials and rodents (9 generic taxa), another of goannas, lizards and the crocodile (11 generic taxa), another of skinks and geckos (4 generic taxa) and three ungrouped taxa. Although the crocodile is the saltwater species, Crocodylus porosus, it is seen as a land animal because it lays its eggs on land. Inclusion of frogs and tadpoles, which were not seen as related by the majority of old people (i.e. pre-contact times), within this group is ambiguous. Otherwise the only unaffiliated generic taxon On land is arrkwara ‘earthworms’. This taxon is also used for beachworms but because it is unaffiliated has been placed within both land and sea animals, thus avoiding problems of nontransitivity. Labeled subgeneric divisions in the animal kingdom are rare. Binomially labeled specific taxa are limited to the ‘real’ one. In most instances where more than one scien- tific species is included in the one Anindilyakwa taxon, the differences between the species are recognized though not labeled. For example, there are three doves all called darrawurukukwa. Geopelia humeralis, the bar-shouldered dove, is larger than the other two species. G. cuneata, the diamond dove, is about the same size as G. striata, the peace- jotianhontnecpeaieniensenian May 1982 WADDY 75 ful dove, but it is not as common as the other two species. The distinctions between the two most common species are clearly recognized and yet there is no indication of any labeled subdivision of the taxon. This folk taxon is one of the best known taxa today. It is also a totemic taxon The use of different names for younger forms of certain taxa, where the young are known to develop into the adult form, is more common. These names have not been included in the numbers of generic taxa quoted above. As far as I can establish thus far, there is no case where the so-called young and adult forms represent different scientific species. DISCUSSION Where differences between species are recognized but not labeled, such as in the case of the doves, I have interpreted these subdivisions to be unlabeled specific taxa, or covert specifics following Berlin’s typology. Bulmer and Dwyer would regard these subdivisions as speciemes. For Groote Eylandters it is generally irrelevant which member of a labeled taxon is considered. As Berlin (1976:392) says, ‘subgeneric taxa are recognized (linguis- tically) primarily because of the close attention they receive as a result of their cultural significance’ The lack of labeled subdivided taxa and the fact that labeled subdivisions are so rarely used, even if they exist, in this folk classification schema makes it relatively easy to determine the ‘natural’ categories at the lower levels. Additional evidence that these named categories are basic to the perception of Anindilyakwa speakers comes from an unusual source. Dwyer (1976:441) hints at, but provides only a very general example of, a possible relationship between social organization and biological classification. Australian Aborigines have a totemic classification system which differs from place to place. On Groote Eylandt each clan has a number of totems which may or may not be folk biological taxa. The relationship with biological taxa in particular is personified so that a man who sees, for example, wurruweba a red-winged parrot flying overhead, might say, ‘There goes my brother-in-law!’ If several scientific taxa are included in the one folk taxon, such as the doves, it wouldn’t matter which of the scientific taxa was sighted, nor whether there were any subdivisions, named or unnamed, the relationship would remain the same. Nor is there any example of any totem which is a taxon ata higher level of inclusiveness. ‘Natural’ categories of this kind are all represented by simple primary lexemes and appear to be of equivalent psychological salience, thus supporting Berlin’s concept of a folk genus but, as I have indicated from my previous arguments, we need to be wary of such agreement. Perhaps this agreement does no more than reflect a widespread Seneral pattern of relationship between nomenclature and taxonomy (Bulmer pers. comm.,), This leads into the second question of how we are to determine cognitive status. Whether the term folk genus or folk species or specieme or anything else is applied to the basic units as perceived by Groote Eylandt Aborigines, there seems to be an equiva- lent cognitive status based on apparently equivalent degrees of cognitive perception. They do not worry about the finer details of discrimination between any subdivisions of the taxon. The unit which they themselves ‘see’ is the labeled category which is not sub- divided. There is one instance in Sav ES of a subdivided taxon where one member of the set is labeled by a simple lexeme, viz. dubudekbuda oystercatchers, dubu- dekbuda dadumamalya Pied Sy and dakwurrinya Sooty sp ialane ea r. The taxon dubudekbuda is a totem of the Warnungwamakwula clan. The name dakwurrinya is used specifically in the songs of that clan. I wonder whether ination of taxa need only be referred to in the context of some special purpose, yet at the same time are available for inclusion in the general purpose biological oactiiation. If so, it would additional substance to Berlin’s concept of folk genus. It is this basic labeled asetee 76 WADDY Vol. 2, No.1 category which I see as the potential unit of agreed perception of discontinuity and cogni- tive status. I think that is what Berlin, Hunn, Hays and I have all been groping towards. Just how we can objectively define it in a manner satisfactory to all remains a problem. Randall (1976:550) has raised the issue of special purpose classification systems. Bulmer (1974:24), in commenting on Berlin’s schema, has noted that folk taxonomies generally seem to be characterized by considerable flexibility and elasticity, contracting or expanding according to context. Dwyer (1976:438) suggests there is a need for flexi- bility in that the same folk taxon can apparently change its status within a folk taxon- omy. Hunn (1976:520) has said that taxonomic structures are inadequate as models of the process of classification. In the light of these comments and the dissatisfaction with systems previously out- lined, I would support Randall’s suggestion that folk taxa are stored in the memory imply as a series of (direct) contrast sets, as defined by Kay (1971:877), but I would pe that the contrast sets remain free to be manipulated as required rather than fixed within a hierarchy. Each contrast set represents perfectly valid relationships. Such sets could readily be ranked by vertically overlapping set inclusion relationships to produce a hierarchical classification. Nontransitive relationships, (for example, a land snail is not a sea animal), are then explained as inclusion of non-typical members of a set by virtue of form or behavior. In the Groote Eylandt Aboringinal biological classification there is only one contrast set, viz. the dichotomy between land and sea animals included within the unique beginner for the animal kingdom, which gives rise to all nontransitive relation- ships. Apparent change in status of a folk taxon would be explained by the formation of an additional contrast set at a higher level. I would take this to include polysemy of folk taxa, pes I don’t think that was what Dwyer inten Th oote Eylandt Aborigines can produce a Seidel food classification system which abe considerably but is by no means identical with the general purpose biolo- gical classification system (Waddy in press). The overlap of terms such as akwalya, which in the food classification systems means all edible flesh, but in the biological classification system means all animal life, highlights the need to be particularly careful that terms in- cluded in a particular hierarchy are rightly included for the purpose of that classification. On the other hand .their totemic classification system results in an entirely different grouping of folk taxa, completely crosscutting higher biological folk taxa in many in- stances and lacking in hierarchical depth If folk taxa are arranged in contrast sets according to purpose, then only those sets required for the purpose would be called on to generate a hierarchical classification. This would appear to me to provide the flexibility and the potential for overlap whic has been observed. Because Kay’s contrast sets are defined upon taxa rather than upon the lexemes that realize them (Kay 1971:874), it seems reasonable that a contrast set may contain un- labeled taxa, i.e. covert categories, as well as, or even in place of, labeled taxa. This does not seem to violate his definition that a contrast set is composed of just those taxa which are mE preceded by the same taxon e problem still remains. Uniostunatedy: as Berlin (1976) says, members of the same contrast set often do not exhibit the same degree of internal variation. This is where I think that Hunn’s idea of monotypic genera or indeed of higher level monotypic taxa can be applied. Certain taxa, termed unaffiliated generics by Berlin Breedlove and Raven (1974:219), do not appear to have the same psychological salience as the more inclusive life form taxa, even though technically they can be included in the same con- trast set. Berlin, Breedlove and Raven rank these taxa as generic on the basis of linguistic criteria which are now being questioned. It would seem either that these unaffiliated contrast set or that the next higher level contrast set is one-membered. For such an inter- pretation to be valid however we still need to be able to define the basic ‘natural’ cate- gory within the folk taxonomy and that to me depends on perception of discontinuity. May 1982 WADDY ab LITERATURE CITED BERLIN, B. 1972. Speculations on the growth of ethnobotanical nomenclature. Lang. Soc. 1:51-86. ——$_ —_, 1976. The concept of rank in inch ase bscmeepin gua hens evi- dence from A folk b . Amer. Ethnol., 3:381 "399. D.E. BREEDLOVE and P.H. RAVE 1968. Covert categories and folk sche Amer. An sent 70:290-299. D.E. BREEDLOVE and P.H. RAVEN, 1973. namie principles of clas- sification and nomenclature in folk biology. Amer. Anthrop., 75: os i 42. DLOVE and P.H. » 2 RAVEN, 1974, plies of Tzeltal plant classification: An introduction to the botan- i aking com- Academic Press, New York, 66 BULMER, R.N. 1968. Worms that croak and other mysteries of Karam natural science. Mankind, 6 :621639. Which came first the chicken or ihe egghead? In J. Pouillion & P. Maranda (Eds.), Echanges et communica- tion, mélanges offerts 4 Claude Lévi-Strauss a Voccasion _ son 60&€me anniversaire. The Hague-Mouton gee a ree Folk biology in the ae Guinea re nay Soc. Sci. Inform., 28. ——_—______.,, and JI. geweere: 1972. Karam eee of marsupials and rodents, Part 1. J. Polynesian ne 0 81: 2-499, ——___.., and J.I. MENZIES, 1973. Karam classificationof marsupials and os Part 2. J. Polynesian Soc., 82: ; EL MENZIES and’ F. KER, 1975. Kalam eee ae of ios and fishes. J. Polynesian Soc., 84:267-308. ah ee ee MJ. TYLER, 1968. Karam classification of frogs. J. Polynesian Soc., 77:333-385. CONKLIN, H.C. 1954. The relation of Hanu- “ a ond world. Unpubl. Ph.D. thesis Yale DWYER, PD 1976. An analysis of Rofaifo mmal taxonomy. Amer. Ethnol., 3:425- 445. HAYS, T.E. 1979. Plant classification and nomenclature in Ndumba, Papua ore Guin- ea highlands. ype 18:253-27 HUNN,E.S. 197 arda conta ‘iedlel of folk bologcl peta ti Amer. Ethnol., 3:508-5 Se Tzeltal ames zoology: The classification of discontinuities in nature. Academic Press, New Y York. KAY,.2, 1971. Ligh erie - semantic con- trast. Language, 41:866-8 LEVITT, D. 1981, sina rik People: Abori- ginal of plan n Groote Eylandt. stein bak ie sy aia Studies, Can MAINED, 18, and R.N.H. BULMER. 1977. my Kalam country. Auckland al al RANDALL, R.A. 1976. How tall is a taxon- omic tree? Some range for dwarfism. ree? Amer. Ethnol., 3:543-55 from * Groote Eylandt A boriginal point of cage and Southeast Asia. Canibeten Univ NOTES 1. The data on ere peach included herein ap he Bota mer, Peter Dwyer, Terence Hay Kenneth and Maddock for their rare fem kee on these ers. 2. The provision of a grant-in-lieu-of-salary b the Avistiaitian Institute > FISH — SNAKE MAMMAL Stages: 0 1-3 4-5 FIG. 1—Revised folk zoological life-form encoding sequence (Brown 1981a). DISCUSSION The encoding sequence of Figure 1 is interpreted as a series of stages in the growth of folk zoological life-form vocabularies with one life-form term being added at each stage. Stage 0 languages totally lack terms for animal life-forms. Stage 1 guages encode one, Stage 2 languages encode two, and Stage 3 languages encode three of the life-forms FISH, BIRD, and SNAKE. At Stage 4 languages add a term for either WUG or MAMMAL. The remaining animal life-form class is encoded at Stage 5. The critical features associated with the five life-forms of the encoding sequence are as follows: BIRD Large creature (relative to creatures such as bugs) possessing wings and usually having feathers and a bill or beak. (This life-form always in- cludes birds. In its greatest extension it includes birds and flying mam- mals such as bats.) FISH Creature possessing a streamlined body and fins, usually having gills. (This life-form always includes true fish. In its greatest extension it includes true fish and fish-shaped mammals such as dolphins and whales). SNAKE Featherless, furless, elongated creature usually ag appendages. (This life-form always includes snakes and/or worms. In its greatest extension it includes snakes, worms, lizards, eels, oe eae, other elongated creatures such as reptile-like insects.) WUG Small creature other than those included in BIRD, FISH, and SNAKE. (This life-form always encompasses bugs, i.e. insects and other very small creatures such as spiders, and frequently is extended to worms. Occasionally the category also includes other creatures such as lizards, tortoises, and frogs if these are small.)2 MAMMAL Large creature other than those included in BIRD, FISH, and SNAKE. (This life-form always includes mammals. It is often extended to other € als such as iguanas and crocodiles and, in addition, to such creatures as tortoises and frogs if these are large.) 3 To a greater or lesser extent each of the above five categories encodes a large, pan-environmental discontinuity in nature. (As noted page the status of WUG and MAMMAL as true discontinuities is problemati In other words, each is a linguistic reflection of a morphologically ‘deaincaians but highly heterogeneous grouping of creatures found in most environments inhabited by humankind. Thus i May 1982 BROWN 97 these categories are distinguished from other general animal classes which do not encode discontinuities in nature but rather are based on criteria other than gross morphology. Such criteria include animal habitat (e.g., house vs. forest), edibility (e.g., poisonous vs. nonpoisonous, tabooed vs. nontabooed), symbolic status (e.g., sacred vs. profane), rei tionship to human beings (e.g., flying vs. crawling vs. trotting vs. burrowing), and s In the literature on folk biological classification those categories based on the ae criteria are identified as ‘special purpose” classes while those encoding discontinuities in nature are called “general purpose” classes. Clearly there are other large zoological discontinuities in addition to the five noted above which are pan-environmental. Examples include ants, spiders, wasps, moths and butterflies, and toads and frogs to mention a few. However, the sell napa treated investigation are singled out for special attention because they appear to be espec- ially significant for humans. This special importance is mirrored neg oar ie tire in folk zoological classification. For example, these discontinuities are consistently encoded by languages. In other words, they are realized as labeled zoological classes over and over again. Most importantly, categories reflecting them tend to be the most polytypic i Four folk in detail wou = animal classes found in Chrau (Thomas 1966), Kyaka Enga (Ralph Bulmer, personal communication), Ndumba (Terence Hays, personal communication) and Tzeltal (Hunn 1977) respectively. TABLE 1. Seven most polytyic general purpose animal classes in Charu (extracted from Thomas 1966). Number of immediately Class included labeled classes stim (BIRD) 17 ca (FISH) i vih (SNAKE) 7 ky6q (“‘frog’”’/“‘toad”’) 5 si (“louse”’) 5 khlang (“bird of prey”)* 5 5 ong (‘‘wasp”’) *khlang is not included in sim (BIRD). oe: Semtiatieieiyg Aceiance: TABLE 2. Eight most polytypic general purpose animal classes in Kyaka Enga (Ralph Bulmer, personal communication) Number of included Class terminal labeled classes yaka (BIRD) id kau (SNAKE) he sa (MAMMAL (“large mammal”)) 32 mugi (“‘frog”/“toad”) = wi (MAMMAL (“small mammal’’)) _ mena (“‘pig”’)* sa maemae (“‘butterfly”/‘‘moth’’) q re (“ant’’) ‘ *mena is not included in either sa (‘large mammal”) or wi (‘‘small mammal’’). eeennetennen, 98 BROWN Vol. 2, Ne. 1 TABLE 3. Nine most polytypic general purpose animal classes in Ndumba (Terence Hays, personal communication). Number of immediately Class included labeled classes kuri (BIRD) 86 to’vendi (WUG) 49 fai (MAMMAL (“large mammal’’)) 15 kaapa’raara (SNAKE) * 11 faahi (MAMMAL (‘small mammal”’)) 10 feqana (‘‘frog”’/‘‘toad’’)** 9 kaapura’rora (“butterfly’’/‘‘moth’’) 8 kaa’pari (“ant’}*** 8 quara (“‘pig’’/‘‘domestic mammal’’) **** 8 *kaapa’raara is immediately included in to’vendi (WUG). **feqana is immediately included in to’vendi (WUG). ***kaapura’rora and kaa’puri are both immediately included in to’vendi (WUG). ****quara is not included in either fai (“large mammal”’) or faahi (‘small mammal”). TABLE 4. Eight most polytypic general purpose animal classes in Tzeltal (extracted from Hunn 1977). Number of immediately Class included labeled classes mut (BIRD) 106 éanbalam (MAMMAL) 36 tan (SNAKE) 23 o (“small rodent”’) * 12 Su (“wasp”) 10 ?am (“spider’’) 10 Cay (FISH) 8 €anul ha? (“water bug’’) *E? o is immediately included in canbalam (MAMMAL). Tables 1-4 show that BIRD, FISH, SNAKE, WUG, and MAMMAL, if encoded, are consistently among the most polytypic animal classes in a taxonomy. Indeed, with the single exception of the Tzeltal FISH category (Table 4), classes encoding the five zoologi- cal discontinuities are the most polytypic animal categories in all four languages. The Tzeltal exception is due to the fact that the language is spoken in a mountainness region of southern Mexico where fish are severely restricted in number and diversity. Similarly, Kyaka Enga and Ndumba are spoken in highland regions of New Guinea where fish are virtually lacking, accounting for the fact that these languages do not encode FISH. Six languages among the 144 recently surveyed treat WUG and MAMMAL in a spec- ial manner. Instead of encoding WUG and MAMMAL in separate categories, creatures of these groupings are lumped together in a single labeled class. These “combined WUG- MAMMAL.” categories typically include bugs and mammals and frequently extend to May 1982 BROWN 99 other creatures which are neither birds, fish, or snakes, such as lizards, turtles, frogs, and so on. Distributional considerations indicate that languages encode combined WUG- MAMMAL only after addition of all three classes of the initial triad, BIRD, FISH, and KE, While zoological life-forms are typically encoded through use of a single label, they are sometimes lexically realized in other ways. For example, languages may lack a term for a class extended to mammals in general, but lexically encode MAMMAL through the binary opposition “large mammal”/‘small mammal” (e.g., Ndumba, see Table 3). Langu- ages doing so are judged as having a MAMMAL life-form in my studies (Brown 1979a, 1981a). In some cases, languages may encode only one-half of a binary opposition, e.g., only “small mammal.” These languages are also judged as having MAMMAL life-forms. Binary opposition is particularly prevalent in the encoding of SNAKE. SNAKE is fre- quently recognized through the binary contrast ‘‘small elongated animal’’/“large elong- ated animal.” The “small elongated animal” category usually ag Soa worms alone while the “large elongated animal” class is usually restricted to true sna Explanatory framework I (Brown 1979a, 1981a) have proposed an explanatory framework to account for the developmental priority of BIRD, FISH, and SNAKE and the late emergence of WUG, MAMMAL, and combined WUG-MAMMAL. Incorporated into this framework are three principles of naming-behavior: (1) criteria clustering, (2) binary opposition, and (3) di- mension salience. riteria clustering occurs when certain features of natural objects correlate or cluster thus producing discontinuities in nature. Criteria clustering underlies encoding of BIRD, FISH, and SNAKE. For instance, Bruner et al. (1956: 47) cite the example of birds, creatures possessing feathers, wings, and a bill or beak. A creature’s possession of feathers is highly predictive of wings and a bill or beak, so much so that an expectancy of all these features being present together is built up. This expectancy can lead to the lexical encod- ing of BIRD. Similarly, FISH and SNAKE have respective sets of defining features show- ing high levels of mutual predictability. The occurrence of fins predicts a streamlined body and gills, and greatly elongated creatures usually lack appendages as well as feathers and fur. In brief, BIRD, FISH, and SNAKE constitute salient discontinuities in nature and, thus, are natural candidates for lexical encoding. (See Hunn [1976, 1977] for a detailed consideration of the influence of discontinuities in nature on folk classification). The late encoding of WUG, MAMMAL, and combined WUG-MAMMAL is in part a function of their relative indistinctiveness as natural discontinuities vis-a-vis the distinc- tiveness of BIRD, FISH, and SNAKE. Each of the former three groupings is excep- tionally heterogeneous and demonstrates little criteria clustering. For example, while most mammals have four appendages used for locomotion and/or object manipulation, so do many other animals including such common creatures as lizards, s amanders, frogs, and turtles. Consequently, possession of four appendages is not particularly pre- dictive of other faunal characteristics such as fur or hair. Exemplars of WUG have even less in common than creatures included in MAMMAL. WUG, for example, encompasses animals having legs and lacking them, having wings and lacking them, having segments and lacking them, and so on. Combined WUG-MAMMAL, of course, aggregates the heterogeneity pertaining to WUG and MAMMAL. bin opposition and dimension salience underlie the encoding of WUG and MAMMAL. joie) sage through = opposition is a common feature of language. Physical and conceptual dimensions are u versally encoded initially through binary contrast, e.g., deep/shallow, sharp/blunt, ena good/bad. Only later are such dimensions recognized by single terms, e.g., depth, sharpness, texture, and value respectively. The priority of binary contrast in dimension encoding is often apparent in the development of terms for whole dimensions. These are frequently derived from one 100 BROWN Vol. 2, No. 1 of the two labels for associated oppositions: for example, depth from deep and sharpness from sharp. Sometimes classification of natural objects involves their “dimensionalization.”’ In other words, they are treated as if they are distributed along a dimension and are encoded through binary contrast. When this occurs, the dimension involved is invariably size. The importance of size in biological classification illustrates the principle of dimension salience. Dimensions are not particularly salient if they only apply to a small number of different objects. Since all biological organisms vary by size, there is a strong tendency for this dimension to underlie encoding of plant and animal classes through binary con- ” r the three major zoological discontinuities are encoded as life-form classes, there remains a large and varied group of creatures which are not affiliated with life- These left over or ‘‘residual” creatures ones include mammals, lizards, frogs, turtles, snails, wo and bugs to mention just the more obvious ones. Life-form en- coding beyond BIRD, FISH, and SNAKE a ae involves lexical recognition of sub- groupings of these animals. owever, among residual creatures distinct discontinuities are not easily discerned since criteria clustering is not much, if at all, in evidence. As a consequence, languages usually resort to a common classificatory strategy that need not necessarily involve distinct discontinuities, that is, binary opposition based on the salient dimension size. Thus the addition of WUG and MAMMAL encodes the contrast ‘‘small residual creature’’/“‘large residual creature.’’4 re is another way of dealing with residual creatures which is occasionally resorted to by languages. Instead of lexically recognizing them through binary opposition based on size, some languages simply regard residual creatures, both large and small, as forming a unified grouping which is encoded by use of a single term. This, of course, creates combined WUG-MAMMAL life-form classes. The relative rarity of combined WUG-MAMMAL among the 144 languages surveyed (Brown 198 1a) suggests that humans are usually disinclined to use classificatory strategies that do not incorporate substantive defining features. Membership in a combined WUG- MAMMAL category does not involve substantive characteristics of creatures but rather their lack of membership in other life-form classes, i.e. their residualness. On the other hand, the binary contrast WUG/MAMMAL does entail a substantive feature, that is, animal size. The relatively high frequency of occurrence of the latter contrast among the world’s languages indicates that humans are somehow more comfortable with life- form classes which are anchored in objective reality, even if only minimally so Folk biological life-forms and societal scale In two studies I report that size of both botanical (Brown 1977) and zoological (Brown 1979a) life-form vocabularies is positively correlated with societal scale. Lang- uages having few biological life-form terms are usually spoken by people living in small- scale societies with little of the political integration, social stratification, and technolog- ical elaboration found in large urban societies where people speak languages usually having —— life-form terms. e special usefulness and aptness of biological life-forms in large-scale societies may relate to the increasing separation of humans from direct reliance and dependence on the natural environment in these societies. The typical individual in a small-scale society can usually name and identify hundreds of separate plant species (Berlin et al. 1974; Conklin 1954; Hays 1976), while typical nonspecialist members of modern urban society might do well to name and identify even one hundred (Dougherty 1978). When people lose detailed knowledge of plants and animals including names for them, less specific terms, such as life-form labels, tend to grow in number and become increasingly salient. Addition of biological life-form classes to languages, then, indexes a general decrease of interest in and concern with the world of plants and animals. May 1982 BROWN 101 Salience of biological classes can be measured through frequency of use of terms for them in ordinary language. The more frequently used words of a language tend to label more salient classes and the less frequently used words, less salient categories. Thus, in languages of modern nation-state societies terms for animal life-form classes generally should be more frequent in use than terms for less general animal categories. Tables 5-7 organize information relating to the salience of animal concepts (classes) as measured by frequency of use of terms for them in three nation-state languages, American English, Arabic, and Peninsular Spanish respectively. As expected, these tables show that animal life-form classes for the most part are ranked among the very most salient animal cate- gories in these languages. TABLE 5, Poles of the 66 most salient animal concepts in American English based on fre- quency of occu of terms for them in written language (extracted from the “‘Lorge-Thorndike Semantic Rane: in Thorndike and Lorge 1944). FREQUENCY ANIMAL CONCEPT(S) Re ee th ne can “‘animal/creature/beast” (849 animal, 324 creature, 202 beast), horse De ch ea CWLWRER hoe Sk ke ER ee ae eee og ee Pe ee ee re es ee ere er a FISH (fish) G00: sc cain Gale fh ee ae BIRD (bird) ROO oe ned yo one eee Wd Ge oe a hn robin Oe S54 2a bt oe eee a ees ee See ee WUG (216 insect, 31 png EEC RE CRT ET ee eee eee SNAKE (127 snake, 54 serpent, 65 wo re Ae 2h ak 624 deed wE ORES SOAS Oe re ree eS cattl a ee ee ere err re ee re eee ae eS tion ee eee er ern re re ee ee er ee ee ee ROR 6. Uk © w ghardrol be Wee ee ai nee ea is Be a sheep SIS £265 4.5.04 886 SNS EAENDAS EEA RO ORR eae ess goose WER cece, on ad Up RRTS s Gatrh ee Loa eee es rabbit i ere eo eer eee rr eT ee ree eae ee deer en en EE ee eee ee Ee ee ee ee ee ee ee cow Ce er re ey een ene es ee ee ee ee eee ee 2 toad eee e ee ee Peer ee eT ere eS Tr ee eT Tee eee ee ee wolf re rae wae nee ean. erase a eer ee eee ee ee ee 2 pig ROE ke eAR Soe ww oie ei 1S Sk eee BREASTS OHS bee, crow, monkey, sie ROO cote» UG hte an ie ee od a aa we ee a ee ee ee Oe EER «eae hs ee oe Ee CRS Ne ha WAS ES CA Nes fe RUD) % aay wea eld Ae Reale 5 eS WN eR ee duck Ds he Ge We as Oe Ce OLR Dee es RR cardin RO. 8 kee ae Oe ee ee a nan ew es oe aes chicken SE dee te techn ch SL OL AS EE ee ee ee ee u Se sy Ya a oe ee ee ee RS lamb, mule OF s-cwse doe we cwh ak hee bee eee Oe ae ea Ke ee es T OF GN Ge ee re a ee ee Ce SY OS He ee ews OP Swe i ae Ee Oe AE Oe a ee es elephant _) ae ere ar ee ke eee ae ee ee ee hound TO oe GIRLS Se CRO eS ee 88 goat, mouse, possum TE ea Re eA a ee ee are ee NS a MCG CO ee i a Oe a ee tee eee ss oe eee ee ee ee cock Ge eR Te ie Se ee ee eee eh eee es shark 102 BROWN Vol. 2, No. 1 TABLE 5 (Continued) FREQUENCY ANIMAL CONCEPT(S) ee ae ee eee ee ee ee ee ee ee ee muskrat re eee ee ee ee ee ee ee ee ee ee ee ee mole eh Fe ao Cie. 8 Wd Ae es A SO ee RO ce oe cricket Oe hives a a ale ie oe Ke She OHA & OE OE Gk oe BSS Oe aos Ste eas oe oe eee calf, dragon, dragonfly, hen, kitten, pony Oe ich be ie Soe ARERR ROS AD HRB AR LR aes fow ee eas eee ed ee en ol Sn a Ae ie oy oe ee 8 buffalo ee ee ee ee ee ee ee eee ee eee ee badger OF. Cede ceeh Sa CAR AN A de Ree A Ok donkey, hare, hog o Rats ee hake eo ee Re S Seow ei a ges eG Owe be abe aes Ox A 4 ha eA oe Ae eR EN squirrel, gorilla, hawk SE GG eRe RW Ak OM be he ee aS he Ee Re ORE RAS turtle, lark, tiger Tokens (running words counted) = 4,500,000 TABLE 6. Ranking of the 44 most salient animal concepts in Arabic based on frequency of occurrence of terms for them in written language (extracted from Landau 1959). FREQUENCY ANIMAL CONCEPT(S) Oe ee ie Ae ie ed Oy Oe ER & A os ee animal/creature es oa Os ee ke ER Ee ES a IR MURS eas gi eas Aino We acts & EM gk al ba ae Seg i es hk lanka as SG “dog” Petal as AAS ooo Cy Oa eR Ke hE ee eR “lion” Dds bie Ol aS oe Heh AK Ge oad aed a ak ae “camel” BP astute Me Om care, cas Bh cal Sen Nat alae Mn “ae Sob tps Sh NO wa ae ew to Go de 2 FISH eee et Ae ee ed A oe ee As Gi dws oe aes “horse”’ RRS fie eek ke ree eee es SNAKE (9 “snake,” 1 “worm’’), “locust ica hues cig Se ae ee ede der ee a ead mira y cin kee 4g a ie Re a eS 5.4 a oe oe a ee A ad ig hake oes hie A we ce he “cow” Dstt oe te eee ak 6h oe eae “deer,” “mule,” “elephant,” “bee” ee eh ieee aie ko ee SL CORE Sl eS ww OM “monkey/ape”’ © cain vote waeak a ees “reptile/burden animal,”’ “‘goat,” “spider,” “cat” ee eee ee ee “mosquito,” “‘sheep,” “wolf,” “ostrich,” “cock” ID aia Gi tat me ee eres “swine,” “hen,” “fly, ” “leech,” “vulture/eagle,” “ant”’ Dh Sce gee ates “duck,” “fox,” “ox,” “buffalo,” “dove,” “donkey,” “giraffe,” “hawk,” “hyena,” areata? “lynx/leopar,” i a “tiger” Tokens (running words counted) = 272,178 — a TABLE 7. Ranking of the 65 most salient animal concepts in Peninsular Spanish based on frequency of occurrence of terms for them in written language (extracted from Buchanan 1941). FREQUENCY ANIMAL CONCEPT(S) RA 6 6K fin ee ee ee eee ee animal/creature OE siete 6h Wl Reale WA i ona. aetna rete cee ie ae BI 188 ce ae aw TB oO ew ooo OL io eae ee ye ere ada “horse” BGO ck o's oa db a ee ee a “dog” em ee.) May 1982 BROWN 103 TABLE 7 (Continued) FREQUENCY ANIMAL CONCEPT(S) Fir weet ahs eae ore ee Re ee ee te ee “bull” EAD AGGRO SN LOR EES OA Oe ON RE eee ee ee FIS Ue Gitaw ham Caw bik Roe oo aa ae ee ed eee oa “lion” FE. Ric Aas WA Oe ee ee ee ee ae “cat” Phe 0 Ue Ei Ohh ee ee ER ee a ee “cow” BO Gis ae bo ue he We en ig ee Cee ee eee “nan, “ox” SO ads ede aie Ace ca tae ae ae aw aa eae SNAKE (35 “snake,” 18 “‘worm”’ SP Cease we Sou ad eee Cee Lee ee OL ee **goat,”’ “chicken” OER,» si Seon Sg5a serie kak Sec whateva k etae e e SLe eee ““mouse”’ UD 34.4% 66 ENE eo we BESO ee 4 eee eta rete! “fly (insect)” Pe WRG eo ok eee ek de ae eee oo a ea ee ee ae “pig” Se eee CMB Sr eat ar ay My PARR MIS, Wor eR ees “eagle” at Fisk Ss CLEA Dw VERE ee ee ee eee ee ‘*sardine”’ PE ce eb uses oda w a pee eee eee ees “cattle,” “pigeon,” “donkey” OF yaa Seas Ae Ue kw eee eae eet “mule” DE” aid sake ho ote he a ae ee OE Oe eee eee “wolf” Oh. 6 ew fae eh cea dire CC ee ee eee “fox” ae 2&4 Re CA ee Sa eee ee eee “butterfly” tne eek Ded Ro ON a awe ee a IGS Oo SA Oe ROK ORS “mole” GAS he se Aces Glacial ed “hog,” “‘cardinal,” “‘bear’’ Fe Wie dhe alg Decale ree ee ie oan al Se ieee a we, he “flea,” “falcon” OS tush ca eee ee “turtledove,”’ “reptile,” “lizard,” “woodpecker” a. £0 keg Mok dae Pek ye & ae oe a ees “deer,” “swallow” G bk keene ERE SO ee ee CR ae “whale,” “clam” Tokens (running words counted) = approximately 1,200,000 Salience rankings of animal concepts (classes) presented in Tables 5-7 are based on frequency of occurrence of terms for them in written rather than spoken language.5 nimal concepts ranked in each table constitute a group of the most salient animal classes in a language. For example, those of Table 5 are the 66 most salient animal number of cases different terms of a language label the same animal class. Frequency counts for these are added together to yield an overall count for the class. For example, American English WUG has a frequency score of 247 which is the sum of the occurrences 104 BROWN Vol. 2, No. 1 of insect (216) and bug (31) in a token of approximately 4,500,000 running words. The frequency score for the concept SNAKE in each of the three languages is the sum of counts for “snake” and “‘worm’’ classes. Individual scores for the latter two animal con- cepts are also give ables 5-7 eae that two zoological life-forms, BIRD and FISH are among the six most salient animal concepts in each of the three nation-state languages. In addition, KE is found among the 11 most salient animal concepts in all three. In two langu- ages, American English (Table 5) and Arabic (Table 6), SNAKE and WUG have virtually the same high degree of salience (see frequency scores for these). Only Spanish (Table 7) has a life-form concept, WUG, which shows a relatively low salience ranking. It should also be noted that the most general faunal concept of all, “creatures, beasts, or animals in general,” is ranked first in salience in all three languages. Since people living in small-scale societies have less need for general animal concepts than people of nation-state societies, it is probably the case that life-form classes and Sane general animal categories, if encoded, are not especially salient for them. As it happens, most of the world’s languages apparently lack very broad “unique beginner”’ or in apes encompassing plants in general or animals in general (Berlin 1972; Berlin et al. 1973), indicating that cross-linguistically these are not very salient for speak- ers of nonnation-state languages. In addition, word frequency data from small-scale society languages should show that many, if not most, generic animals classes are ranked higher in salience than animal life-form classes, in sharp contrast to the relative rankings of generics vis-a-vis life-forms presented in Tables 5-7 for nation-state languages. Unfor- tunately, adequate word frequency counts for small-scale society languages are not now available to test this proposition. The association between size of biological life-form inventories and societal scale indicates a tendency for the number of life-form terms to increase with increases in the scale and complexity of societies. Since societal scale has generally increased during the course of human history, especially so during the last several thousand years, it follows that biological life-form vocabularies in the vast majority of cases have grown rather than shrunk in size and, thus, that the animal life-form encoding sequence is basically additive in nature. This conclusion has been borne out in several studies which have used the comparative method of historical linguistics to reconstruct biological life-form growth in the histories of several genetic groups of languages (cf. Fowler 1972; Brown 1979b, 1981b, 1981c; Brown and Witkowski 1982). Life-forms and linguistic marking Cross-language regularities in animal life-form classification are related to linguistic king. The framework of marking has been developed over the years by Jakobson (1941), Greenberg (1966, 1969, 1975), and others. Marking involves all components of language: phonology, grammar, and the lexicon. Marking in the lexicon entails a distinc- tion between marked and unmarked words. The animal life-form encoding sequence is in fact a universal marking sequence or hierarchy. Terms for BIRD, FISH, and SNAKE are regularly unmarked in languages vis-a-vis terms for WUG and MAMMAL which are mar. e are several diagnostic features of marking that tend to co-occur in typical marking relationships. Some of these are as follows: Unmarked Item Marked Item 1. The implied in an implicational rela~ 1. The implier in an implicational rela- tionship. tionship. 2. Earlier acquisition by languages. 2. Later acquisition by languages. | | | ce May 1982 BROWN 105 3. Greater frequency of use (in text or 3. Lesser frequency of use. spoken language). 4. Less complex (phonologically or mor- 4. More complex. Phologically). 5. Earlier child acquisition. 5. Later child acquisition. Marking features 1 and 2 are closely interrelated and entail a cross-language perspec- example, the cross-language data (Brown 1979a, 1981a) show that if a language has a WUG term, it will have terms for FISH, BIRD, and SNAKE. However, if a language has a term for any one of the latter three life-forms, it will not necessarily have a label for WUG. Thus WUG implies FISH, BIRD, and SNAKE, but none of these imply WUG. Similarly both MAMMAL and combined WUG-MAMMAL imply the former three life- forms, but not vice versa. Thus, WUG, MAMMAL, and combined WUG-MAMMAL are marked vis-a-vis BIRD, FISH, and SNAKE which are unmarked (feature 1). Implicational associations involving lexical items are often the synchronic result of cross-language regularities in the order in which these items are acquired languages. Such relationships form the basis for the proposal of an animal life-form encoding sequence (Brown 1979a, 198 la). For example, the fact that WUG implies BIRD, FISH, and SNAKE but not vice versa is understandable if languages regularly encode BIRD, FISH, and SNAKE before encoding WUG. In addition to implicational relationships there is independent evidence which corroborates the acquisitional hypothesis outlined in Figure 1. This is evidence developed through the eo ative approach of historical linguistics showing that languages add animal life-form terms to their vocabularies in the order of the encoding sequence (cf. Brown 1981b; Brown a? Witkowski 1982). In terms of this evidence alone, one could determine that FISH, set and SNAKE are unmarked relative to WUG and MAMMAL which are marked (featur arking features 3, 4, and 5 are closely a but these associations are real- ized in individual basen rather than across languages in the manner of features 1 and 2. For example, Zipf (1935, 1949) has shown that frequency of use (feature 3) correlates strongly with phonological (or orthographic) length of words (feature 4). High frequency is associated with short word length and, thus, with less complexity, and low frequency with long length and more complexity. This correlation is attributable to efficiency of communication factors: efficiency is enhanced when frequently used Ne are short rather than lon ng. Since unmarked items are less complex than marked items and sin they occur more frequently, it is not surprising that they tend to be aioe? by children learning language before marked items (feature 5). Since the animal life-form encoding sequence is also a marking hierarchy, it should reflect other criteria of marking in addition to features 1 and 2. For example, in individ- ual languages we should expect that terms for BIRD, FISH, and SNAKE occur more frequently in ordinary use than terms for WUG and MAMMAL (feature 3). ieee of use Above, data are presented showing that folk zoological life-form names are among the most paar used animal terms in three languages affiliated with nation-state Societies. In Table 8 similar data are compiled for 11 nation-state languages showing that fr €quency of use of the five animal life-forms of the encoding sequence correlates strong- y with the order in which these are added to languages. In other words, an additional 106 BROWN Vol. 2, No.1 ase of marking, aNd of use, attests to the universal marking hierarchy for life-forms which is also evidenced by other marking features such as implicational iia and language acquision order. TABLE 8. Frequency ranking of folk zoological life-forms in eleven nation-state languages. LANGUAGES FREQUENCY RANKING FUT aca a eh ose) seem ee nna) Aue eee ta etre rene ssid nis ale fa ie Tae ea Low Arabic BIRD (29) FISH (16) SNAKE (10) WUG (9) MAMMAL (n.f.) razilian Portuguese BIRD (291) SNAKE (142) FISH (133) WUG (43) MAMMAL <5) Chinese FISH (12) BIRD (9) SNAKE (9) WUG (<9) MAMMAL 9) French BIRD (108) SNAKE (62) — FISH (54) WUG (45) MAMMAL 5) German FISH (1025*) BIRD (612*) SNAKE (598*) © WUG (<100) MAMMAL (<100) Italian FISH (17) BIRD (10) SNAKE (8) WUG (4) MAMMAL (<4) Japanese BIRD (16) SNAKE (11) _ FISH (8) WUG ( 6) MAMMAL (6) Rumanian BIRD (53) SNAKE (28) FISH (19) WUG (<4) MAMMAL (<4) Russian BIRD (114) FISH (84) SNAKE (32) WUG (<13) MAMMAL «13) Spanish BIRD (207) FISH (115) SNAKE (53) WUG (15) MAMMAL « [BIRD] - [WUG] — [MAMMAL] SNAKE FIG. 2. Order in which American children acquire folk zoological life-forms (Chase 1980). This acquisitional order seems to be related in part to American children’s knowledge of generic terms for creatures. The younger children interviewed by Chase were unable to identify individual fish pictures by generic names (e.g., trout, bass, catfish, etc. ) with the exception of the shark (called jaws by some) and names for individual snakes were not known. On the other hand, they were able to assign generic terms to numerous bird pictures (e.g., penguin, seagull, parrot, duck, owl, and so on). Perhaps as a conse- quence, when sorting pictures, all fish were usually put into a single pile and all snakes ir Sorting pictures thusly possessed knowledge of FISH and SNAKE life-form classes, but not BIRD. Children lacking a BIRD life-form are not unfamiliar with the ago bird. akon they simply do not use the term in a way corresponding to adult American usage; other words, they do not use it as if it were a label for a full-fledged i cas ae Rather, they apply bird only to those creatures with feathers, wings, and a bill or beak which are unknown to them and cannot be identified by use of a generic term. Conse- quently, for these children known creatures such as ducks, parrots, owls, and so on are BROWN Vol. 2, No. 1 definitely not birds in their system. Thus younger American children use bird as a label for zoological class having the characteristics of incipient life-form categories described above. Marking and principles of naming-behavior An important question is what factors determine linguistic marking? Specifically, in this context, what generates the marking hierarchy for folk zoological life-forms? Since the relative marking values of BIRD, FISH, SNAKE, WUG, and MAMMAL are uniform across languages, conditions affecting these values must themselves, for the most part, be regular across languages. Probable influences are the principles of naming-behavior proposed earlier an an explanatory framework accounting for uniformities in animal life-form encoding. un entally, the animal life-form marking hierarchy is a linguistic reflection of criteria cslcie in the physical world, that is, it mirrors the indistinctiveness of WUG and MAMMAL as natural discontinuities relative to the distinctiveness of BIRD, FISH, and SNAKE. In other words, terms for BIRD, FISH, and SNAKE are unmarked vis-a-vis terms for WUG and MAMMAL because the physical objects labled by the former three terms figure into highly salient breaks in nature while those labled by the latter two do not. However, criteria clustering alone does not explain these marking distinctions. Such breaks are consistently followed by humans in classifying and naming objects because they are innately inclined to do so. The marking hierarchy, then, is in part attributable to internal constraints on humans in the processing of external stimuli. CONCLUSION The close agreement of physical-perceptual constraints and linguistic marking values for animal life-forms indicates that the former are converted or translated into the latter. The fae age BIR ISH, and SNAKE are naturally salient and are always encoded first in the eens of zoological life-form lexicons. This physical salience is also converted inne lexical encoding into linguistic salience which is manifested through typical marking effects such as high frequency of use, simplicity of form, and early acqui- sition by children learning language. ACKNOWLEDGEMENTS Terence Hays, Albert Heinrich, and Eugene Hunn are to be thanked for reading and commenting on an earlier draft of this paper. LITERATURE CITED rag esa 1972. Speculations on the h of Ethnobotanical Nomenclature. spainheehee in Soc. 1:51-86. BLACK, MARY B. 1969. Eliciting Folk Taxo- Inc., eee BRENT, DENNIS E. BREEDLOVE, D PETER H. RAVEN. 1973. General Pan of aPC SAN and Nomencla- re in Folk B . Amer. Anthropol. 75: eer 1974, Principles of Tzeltal Plant Classifi- cation: An Introduction to the Botanical Ethnography of a Mayan-speaking People of Highland Chiapas. Academic Press, New York. Rinehart and Winston, York. si CECIL H. 1977. Folk Botanical a orms: Their Universality and Gro asa Anthropol. 79:317-342. 1979a. Folk Zoological Life-Forms: Theit Universality and Growth. Amer. Anthropol. 81:791-817 —. 1979b. Growth and Development of Folk Botanical Life-Forms in the Mayan Langu- age Family, Amer. Ethn. 6:366-385. Semen | | | May 1982 BROWN 111 LITERATURE CITED (Continued) BROWN, CECIL H. 1981a. Folk Biological ife-Forms: A Cross-Language Study of Plant and Animal Classification. Unpubl. ms. Dept. of Anthropol., Northern Illinois Univ. ~——- 1981b. Growth and Development of Folk Zoological Life-Forms in as igame Langu- ages. J. Polynesian Soc. 90:83-110. —.-1981c. Growth and Develdeaient of Folk Botanical Life-Forms in Polynesian Langu- ages. Unpubl. ms., Dept. of Anthropol., Iv. More on Folk Zoological Life- Forms. Amer. Anthropol. 83:398401. BROWN, CECIL H. AUL K. CHASE. 981. Animal Classification in Juchitan Zapotec. J. of Anthropol. Res. 37:61-70. BROWN, CECIL H. a STANLEY R. WITKOW- SKI. 1982, wth and Development of Folk Zoological Life Forms in the Mayan a siite Family. Amer. Ethnol., 9:97- minwe, CHARLES B., WESLEY M. CARR, MILTON L. SHANE. 1945. A Graded Word Book of Brazilian Portuguese. New York: Co, ork: Wile BUCHANAN, MILTON A. 1941. A Graded Spanish Word Book.- Toronto: The Univ. of . 1980. Acquisition of Folk Sees sage Life-Forms by American ec ing M.A. thesis, Northern Tine CONKLIN, siRotn C. 1954. The Relation of Hanunoo coasiae to the Plant World. Ph.D. thesis, Yale Uni DOUGHERTY, JANET ¥o. 1978. Salience and Relativity in Classification. Amer. Eth- nol. 5:66-80. FOWLER, CATHERINE LOUISE SWEENEY 1972. Comparative Numic Eihiobinisey. Ph.D. thesis, Univ. of Pittsburgh. GREENBERG, JOSEPH H. 1966. Lan anguage Universals with Special Reference to Feature Hierarchies. The Hague: Mouton. ——- 1969. Language Universals: A Research Frontier. Sci. 166:473-478. ——.- 1975. Research on Language Universals. 4, Annu. HAYS, TERENCE E. 1974. Mauna: Explora- tions in Ndumba milena Ph.D. thesis, Univ. of Washingto i IPTG. Am Empirical Method for the Identi- fication of mes Categories in Ethnobiology. Amer. Ethn. 3:489-507. HUNN, EUGENE 2 1976. Toward a Perceptual Model of pen spain Classification. Amer. Ethnol. 3:5 pa 6 as Se sel Zoology: The Classifi- cation of Discon tinuities in Nature. New p 5 JOSSELSON, HARRY H. — The Russian Word Count and Frequency Analysis of ig MOS Categories of a a Literary . Detroit: Wayne Univ. Pre Pe pores ALPHONSE, P.M.H. EDWAR RDS and ANA JUILLAND. 1965. oi saa — of Rumanian Words. The Hagu Mou seh AA, ALPHONSE and VINCENZO TRA- RA 973. Frequency Dictionary of Italian Words. The Hague: Mouton. LANDAU, JACOB M. 1959. A Word Count of Modern Arabic Prose. ae York: American ouncil of Learned Societie LINDFORS, JUDITH WELLS. ” tite Childre Language and Learning. Englewood ronmia New Jersey: Prentice-Hall Inc LIU, ERIC SHEN. 1973. Frequency Dictionary of Chinese Words. The : Mouton. eran, HIROSHI. fies a Frequency Dic- tionary Japanese Words. Ph.D. Stanford Univ. MORGAN, Ee 1923. German Frequency Word Book Based on Kaeding’s Haufig Seana eed ounce Sprache. New an Com THOMAS, stipe ar "ten Zoology: An Ethnolinguistic Study. mn Papers on Four Vietnamese Lanuguages, Pp. 1-1 (David D. Thomas, ed.) Auckland: Ling. nan of New Zealand. THORNDIKE, EDWARD L. and IRVING LORGE. 1944. The Teacher’s Word Book of 30,000 Words. New York: Bureau of = ep a Teacher’s College, Columbia Uni wANDER BEKE, page E. 1929. French Word Book. ork: The Macmillan ompany. ZIPF, G.K. 1935. The Psycho-Biology of Language. Boston: Houghton Mifflin . 1949. Human Behavior and the Principle of Least Effort. Cambridge, Mass.:Addison Wesley Publishing 112 BROWN Vol. 2,No.1 NOTES 1. In the initial animal life-form study (Brown 1979a) Stage 0 was not recognized. In cena ie o 5 _ < is ~* a i) i=) ° > fom is oO oO =] OQ io} & =) tic] wn oO ro) i=) oO =] fe oO ao] Lond ° | ie) @ _ Q La} ao = B Me uo] = oO Lo) oO Qa Oo wm =: language cases surveyed were extracted from dictionaries (see Brown 1981d for a discussion of this point). 2. WUGisa nic derived from worm and bug. 3. Animal is more commonly used than mammal as a MAMMAL life-form label by speakers of a English. Since animal is also used as a unique beginner term to refer to creatures in general, it is not employed as a life-form gloss to avoid ambiguity of reference. 4. Terence E. Hays working with Ndumba (Tairora) speakers of New Guinea Han Eugene S. Hunn working with Tzeltal speakers of Mexico independently identified “residual” biological classes during the early 1970’s. For discussions of the role of residualness in folk biological grees Obra see Hays (1974) and Hunn (1976, 1977) 5. Frequency of occurrence of terms for these concepts are extracted from word frequency books, respectively Thorndike and Lorge (1944), Landau (1959), and Buchanan (1941). In the case of American English only one frequency count of the several found in Thorndike and Lorge (1944) is used, i.e., the “Lorge-Thorndike Semantic Count.”’ In Tables 6 and 7 Arabic and Spanish animal con- cepts are denoted by English glosses. Associated frequency scores are those of actual animal terms. Actual terms and their frequency scores are given in Table 5 for American English. 6. In American English the term animal is used to designate both “creature in general” and MAM- MAL. Consequently, two distinct usages contribute to the his salience of this term (frequency count = 849, see Table 5). Such a dual application may also pertain to the equivalent Spanish word animal. It is, of course, impossible to determine what proportion of frequency counts for such polysemous items trace to one usage as opposed to the other 7. The following word frequency books were used as sources de data presented in Table 8: Arabic (Landau 1959), Brazilian Portuguese (Brown, Carr, and Shane 1945), Chinese (Liu 1973), French (Vander Beke 1929), German (Morgan 1923), Italian ee A Travera 1973), Japanese (Miyaji aggregated figures from two of these counts, i.e. from the ““Lorge Magazine Count” and the “Lorge- Thorndike Semantic Count.” Several word frequency books consulted break counts down according to genre of written materials surveyed, e.g., drama, essays, newspapers, technical/scientific literature. These include sources for Chinese, Italian, Japanese, Rumanian, and Spanish. In each of th ases frequency figures from technical/scientific literature are excluded from counts presented in Table 8 since these do not reflect “folk” usage. The approximate number of running terms (tokens) pertaining ra each word frequency study is as follows: Arabic (272, 178), Brazilian Portuguese (1,200,000), Chinese (250,000), aia (1,147,748), German (10,910,777), Italian (500,000), Japanese (250,000), Rumanian (500,000), Russian (1,000,000), Spanish (1,200,000), and U.S. English (9,000,000). The considerable differences in ranges of frequency counts for different languages (see Table 8) reflect the fact that counts for these languages are based on tokens which vary considerably in size. 8. The frequency count for U.S, English MAMMAL given in Table 8 is the frequency of occurrence of the word mammal. 9. In counting orthographic segments, all symbols occupying spaces in the horizontal presentation of a word are tallied. This includes symbols indicating vocalic length and symbols indicating glottali- zation of conson ants. For example, pn 6u'm se ono is judged as having four orthographic seg- ments and @’itin “bird” as having six. n a language class, e.g., deities: Paiute with “bird” a “large bird,” gacltcn of all terms are counted and figure into calculations for that life-form 10. Incipient life-form classes are also residual biological categories (cf. Hunn 1976, 1977; Hays 1974). | —" RE RNR crepe aaa ye ’ May 1982 NEWS AND COMMENTS 113 SOCIETY OF ETHNOBIOLOGY, INC. The Society of Ethnobiology, Inc., is now established as a non-profit corporation. The found Board of Directors was covened at its first meeting in San Diego, during the Fifth Ethnobiology Con- ference held in April 1982, by Steven A. Weber, founding president. This Board was enlarged to nee the existing editorial board of the Journal of Ethnobiology. The full Board then acted as ollow 1. Elected Steven Weber to serve as President of en sarin for 1983 and Steven Emslie to serve during that period as Secretary/Treasurer of the Society. 2. Chose Dr. Willard Van Asdall of the University of Arizona to succeed Steven Emslie and Steven Weber as editors of the journal for volumes 3 and 4 (1983-1984). Until further notice, all journal correspondence should continue to be sent to the present editorial office: P.O. Box 1145, Flagstaff, AZ 86002. Authors of manuscripts should refer to the inside back cover of this issue for specific instructions. Chose Dr. Richard S, Felger and Lynn Reitner of the Arizona-Sonora Desert Museum to serve as book review editors. Several journal pages will be devoted to book reviews each number beginning with Volume 3. - Chose Dr. Paul Minnis of the University of Oklahoma as host for the Sixth Ethnogiology Con- ference to be held at Norman, Oklahoma, in 1983 (see notice below). a a SIXTH ANNUAL ETHNOBIOLOGY CONFERENCE The sixth ethnobiology conference will be held in Norman, Oklahoma, on March 18-19, 1985. A ny issued in January Paul Minnis, Department of eer olny University of Oklahoma, 455 West Lindsey, Room 521, Norman, OK 73019. SOCIETY OF ECONOMIC BOTANY MEETINGS The Society of Economic Botany will hold its 23rd annual meeting at the University of Alabama in University, Alabama, June 14-17, 1982. Featured will be a symposium entitled “U.S. OILSEEDS INDUSTRY—GERMPLASM TO UTILIZATION”. Further telecom can be obtained from C. Earle Smith, Jr., Anthropology, Box 6135, University of Alabama, University, AL 354 COMMITTEE FOR NUTRITIONAL ANTHROPOLOGY The Committee for Nutritional Anthropology is an organization within the American aaa ogical Association which fosters communication among scholars interested in issues of nutrition and anthropology. The yearly dues of $5 covers the cost of quarterly newsletters on topics of current interest in the field. To join the organization, send a letter of self-introduction and the dues to the current president: Dr. Cheryl Ritenbaugh Department of Community Medicine niversity of Arizona Tucson, AZ 85724 WEST COAST NETWORK OF NUTRITION AND ANTHROPOLOGY e West Coast Network of Nutrition and Anthropology is an organization of individuals inter- ested in social and biological aspects of food, nutrition and health. Usually, local meetings are h held bi-monthly and an annual meeting is hel Francisco area to share research reports. The organization dues are $6 per year and should ic sent to: Little Department of Nutritional Science rsity of California estireesie CA 94720 Dr. Little can supply information on local California meetings. Information on local meetings in the Vancouver, Canada, are can be obtained from: r. Harriet Kuhnlein Division of Human Nutrition University of British aay Vancouver, B. C. V6T 1W5 suibianishadssmasiusinenbningin ae NOTICE TO AUTHORS The Journal of Ethnobiology accepts papers on original research in ethnotaxonomy and folk classification, ethnobotany, ethnozoology, cultural ecology, plant domestica- tion, zooarchaeology, archaeobotany, palynology, dendrochronology and ethnomedicine. Authors should follow the format for article organization and bibliographies from articles in this issue. All papers should be typed double-spaced with pica or elite type on 8% x 11 inch paper with at least one inch margins on all sides. The ratio of tables and figures to text pages should not exceed 1: 2-3. Tables should not duplicate material in either the text or graphs, All illustrations are considered figures and should be submitted reduced to a size which can be published within a journal page without further reduction. Photo- graphs should be glossy prints of good contrast and sharpness with metric scales included when appropriate. All illustrations should have the author(s) name(s) written on the back with the figure number and a designation for the top of the figure. Legends for figures should be typed on a separate page at the end of the manuscript. Place all tables and figures in consecutive order at the end of the manuscript. Do not place footnotes at the bottom of text pages; list these in order on a separate sheet at the end of the man Metric units should be used in all measurements. Type author(s) name(s) at the ‘op left corner of each manuscript page; designate by handwritten notes in the left margin of manuscript pages where tables and graphs should appear. If native language terminology is used as data, a consistent phonemic orthography should be employed, unless a practical alphabet or a phonetic is justified. A brief characterization of this orthography and of the phonemix inventory of the language(s) described should be given in an initial note. To increase readability native terms should be indicated as bold-face italics to contrast with the normal use of ttalic type for foreign terms, such as latin binomials. If necessary, the distinction be- tween lexical glosses, i.e., English language approximations of a term’s referential mean- ing, and precise English equivalents or definitions should be indicated by enclosing the gloss in single quotation marks. As the editorship of the journal is changing in 1983, authors should submit their Manuscripts beginning in September 1982 to: DR. WILLIARD VAN ASDALL, Editor Tucson, Arizona 86721 Manuscripts submitted before September 1982 should be sent to: P.O. Box 1145, Flagstaff, AZ 86002. Authors must submit two copies of the manuscraipt plus the original copy and original figures. Papers not submitted in the correct format will be returned to the author. NEWS AND COMMENTS Individuals with information for the “News and Comments” section of the journal should submit all appropriate material to Eugene Hunn, Department of Anthropology, DH-05, University of W Washington, Seattle, “Washington 98195. Please note that the former Folk Classification Bulletin SUBSCRIPTIONS Subscriptions to the Journal of Ehnobiolog) should be be godcemesl to P.O. Box 1145, Flagstaff, Arizona — Subscription rates are $22.00, institutional; $15.00 regular Canada. an $6.00. Write checks meémt , for aged 7: oa sen or o of fasirsat oe os Defective copies | or copies ee in nt will be P a, a CONTENTS ANIMAL DOMESTICATION AND OE SE Seer Cas ES, Bitten Hesse. os Ge 1-15 TRADITIONAL USE OF DEVIL’S-CLUB (OPLOPANAX HORRIDUS; ARALIACEAE) BY NATIVE PEOPLES IN WESTERN TPR A Naas fe Paneer. ee se es 17-38 VERTEBRATE FAUNA FROM FOUR COASTAL MISSISSIPPIAN SITES, Elizabeth J. Reitz ...........---20+-eeeees 39-61 BIOLOGICAL CLASSIFICATION FROM A GROOTE EYLANDT ABORIGINE’S POINT OF VIEW], Julie Waddy..........-.-.- 63-77 DIFFERENTIAL GRAIN USE ON THE TITELBERG, LUXEMBOURG, Ralph M. Rowlett & MariaHopf ......... 79-88 UTILITARIAN/ADAPTATIONIST EXPLANATIONS OF FOLK BIOLOGICAL CLASSIFICATION: SOME CAUTIONARY NOTES: Terence BE. Hays coos so oe kk ee ee 89-94 FOLK ZOOLOGICAL LIFE-FORMS AND LINGUISTIC. MAREING. Cece H. Brown. 2-5 6 on oe ee ees 95-112 a ee OO 8) Oe Oe Fe EE eee ee “ : a ‘ E a aR a eI Gee eae ae ournal of Ethnobiology MISSOURI BOTANICAL VOLUME 2, NUMBER 2 DECEMBER S322 Nol GARDEN LIBRARY. JOURNAL ORGANIZATION EDITORS: Steven D. Emslie and Steven A. Weber, Center for Western Studies, Inc., P.O. Box 1145, Flagstaff, Arizona 86002. ASSISTANT TO THE EDITORS: Karin L. Doerr, Center for Western Studies, Inc., P. O. Box 1145, Flagstaff, Arizona 86002. NEWS AND COMMENTS EDITOR: Eugene Hunn, Department of Anthro- pology, DH-05, University of Washington, Seattle, Washington 98195. EDITORIAL BOARD BRENT BERLIN, Department of Anthropology, Columbia University, New York; ethnotaxonomies, linguistics. ROBERT A. BYE, JR., Department of Environmental, Population and Organismic Biology, University of Colorado, Boulder; ethnobotany, ethno- ecology. RICHARD S. FELGER, Senior Research Scientist, Arizona-Sonora Desert Museum, Tucson; arid land ethnobotany, desert ecology. RICHARD I. FORD, Director, Museum of Anthropology, University of Michigan, Ann Arbor; archaeobotany, cultural ecology. B. MILES GILBERT, Adjunct Research Associate, Division of Vertebrate Paleontology, University of Kansas, Lawrence; zooarchaeology. TERENCE E. HAYS, Department of Anthropology and Geography, Rhode Island College, Pr y, ethnotaxonomies. RICHARD] H. HEVLY Department of Biological Sciences, Northern Arizona y,f : y, palynology EUGENE HUNN, Department of Anthropology, aicaiy of Washington, Seattle; ethnotaxonomies, zooarchaeology, cultural ecology. HARRIET V. KUHNLEIN, Division of Human Nutrition, University of British Columbia, Vancouver; ethnonutrition. GARY P. NABHAN , Me 2 fc - we saae Fi %, En Seg , Tucson; cultural ecology , plant domesticatio ion. 2 DARRELL A. POSEY, Center for Latin American Studies, University of Pittsburgh; gy, tropical cultural ecology. AMADEO M. REA, Curator « of Birds and ‘Mammals, San Dieeo Museum of : Natural Hist ny: th ecology. > 2 sh ) : : of eUveE Journal of Ethnobiology Proceedings of the Fifth Ethnobiology Conference 22-23 April 1982 San Diego Museum of Man San Diego Natural History Museum San Diego, California VOLUME 2, NUMBER 2 DECEMBER 1982 : : oa) ge 1 i a : ey ere ed ; arias - ok i 7 . 7 %) / Tor is aa) J. Ethnobiol. 2 (2): 114-122 December 1982 PINE NUTS AS AN ABORIGINAL FOOD SOURCE IN CALIFORNIA AND NEVADA: SOME CONTRASTS GLENN J. FARRIS Department of Anthropology, University of California Davis, CA 95616 ABSTRACT.—Seeds of a large number of western species of the genus Pinus have been used e native peoples of the southwestern and Pacific Coast states as food. A sense of uniformity of the food value across species has been engendered by the use of the common term pifion to describe pine seeds (colloquially, “nuts”), regardless of species. In fact, a wide variation exists in their food values, especially between the major edible species. The difference is particularly striking in comparing the singleleaf pinon (Pinus monophylla) of Great Basin with the gray or digger pine (P. sabiniana) of the Central Valley foothills of California. Each of these species occupies a discrete, non-overlapping territory. However, the level of importance as a food item varies between the Great Basin peoples, for whom P. monophylla was a staple, and the California Indians who considered P. sabiniana the source of a highly desirable, if incidental, food item INTRODUCTION When the Spanish explorers traveled through the south Coast Ranges of California in the late 18th century, they met Indians along the way who offered them a variety of oods. For the most part these foods were new to the Spaniards, however, one was very familiar, pittones, the seeds of the pine trees. One explorer wrote in 1775 concerning these pitiiones, “There are many pine nuts like those of Spain” and, “In the mountains there are seen many pines like those of Spain” (Fages 1937:59, 35). Presumably these Pines were either P. sabiniana Dougl. (the digger or gray pine) or P. coulteri D. Don (the Coulter or big-cone pine). Pinus sabiniana would be the better candidate since its notably hard-shelled “nut” (actually seed) would have been the most similar to the pinon of Spain, the seed of P. pinea Linn., the Italian stone pine. The Spanish encountered pine trees and their seed in a number of other places, Particularly in the American Southwest (Lanner 1981; Long 1941) and so the term pinon has become commonly attached to the seeds of a number of species of pine. Since there is such concern among botanists over the use of the correct term, seed, versus the col- loquial term, nut, the Spanish word pinon forms a nice ambivalent compromise When the word pifion (or pinyon) is used it usually means the seed of P. monophylla Torr. & Frem. (single-leaf pifion), P. edulis Engelm. (New Mexico pifion), or P. cem- broides Zucc. (Mexican pifion). However, the term is also applied to the seed of a num- ber of other species of pine which causes a certain degree of confusion. This is particu- larly true in publications listing nutritional values of American foods. These will generally list pignolia and pifion as the two forms of pine nuts (Watt and Merrill 1963:46; Adams 1975:123; Pennington and Church 1980:108). Pignolia clearly refers to the Mediter- ranean species, P. pinea. When the term pifion is used, it is very unclear, although an examination of the nutrient content seems to narrow it down to P. edulis or possibly F ee (cf., Botkin and Shires Sil 9). edible seeds. The California Indians exploited P. sabiniana, P. lambertiana, P. ponderosa, P. coulteri, P. monophylla, P. quadrifolia, and P. torreyana for their seeds (Yanovsky 1936:5-6). Whereas most of the Indian peoples of California used the oi seed for food (Farris 1982), its importance is largely overshadowed by the acorn. How r, the case is quite different for the Indians of the Great Basin for whom the seed of - peeiones was of 115 FARRIS Vol. 2, No. 2 major importance (Steward 1941:230). Paiute families in the Great Basin would establish their fall encampments near a producing stand of P. monophylla and, if the harvest were particularly good, might actually remain in the vicinity through the winter despite the relatively high elevation favored by this species (Steward 1938:232; Bettinger 1976:83). h it may be supposed that pine seeds of varying species are similar in their nutritional qualities as well as their availability, this is not the case. In fact, it has been stated: This wide and unexpected variation [in nutritional component proportions] between the different species of pine nuts suggests that, if all other commercial nuts were no longer vailable, their nutrients in any proportion could be supplied by some species of pine nut (Botkin and Shires 1948:12). With this in mind I would like to turn attention to two particular species which differ radically in many of their physical and nutritional qualities: the Pinus sabiniana of California and the Pinus monophylla of the Great Basin (Fig. 1). It should be noted that Transverse Ranges of southern California and on down into Baja California (Barrows 1971:310-311; Bean 1972:40; Zigmond 1941:30-32). However, it primacy as a subsis- tence item occurred in the Great Basin. aw FIG. 1—Cones of P. sabiniana (left) and P. monophylla (right). PINUS SABINIANA This species is most often associated with an environment of grassland and/or chapar- ral-covered hillsides typical of the foothill regions of the Sierra Nevada and the Coast Range areas of California. It survives well on poor soils such as the serpentinite soils of the Coast Range (Griffin 1965; Jepson 1910:88). It often shares an environment with manzanita (Arctostaphylos spp.) and blue Oak (Quercus douglasii), both of which were important as food sources for the California Indians. December 1982 JOURNAL OF ETHNOBIOLOGY 116 When mature, P. sabiniana trees may grow from 15-25 m high. The interval between large seed crops is said to be 2-4 years on the average. It takes two years for a cone to produce mature seeds (USDA 1974:609, 611). The cones are large, usually ranging from 10-25 cm in length and often produce over 100 seeds each. It is important to remember that not all of the seeds have developed kernels. The variation from cone to cone can be quite remarkable in terms of size (Griffin 1964). Kernel weight averages 195 mg (Farris 1982; Senkee 1962). r this study seeds were obtained from 26 cones collected by the author in the Sierra ‘Neva foothills and the Coast Range. The cones were split apart and the seeds gathered, counted, and separated into those with developed kernels and those with un- developed kernels. It is fairly easy to separate the seeds with developed kernels. When the seeds are placed in water, the filled seeds sink whereas those with undeveloped ker- nels float (Griffin 1962:135; USDA 1974:621). Table 1 shows the results of this investi- gation. TABLE 1.—Seed production of 26 P. sabiniana cones. * Cone No. Total seeds seeds with seeds with developed kernels undeveloped kernels 1 137 126 11 2 149 146 3 3 61 4 57 4 137 116 21 5 188 88 100 6 166 58 108 7 111 85 26 8 84 76 8 Ss 107 72 35 10 186 182 4 1] 128 109 19 12 101 85 16 13 92 69 23 i? 75 39 36 15 61 41 20 16 74 65 9 17 99 82 17 18 100 87 13 19 79 55 24 20 64 41 23 21 156 111 45 22 162 141 21 23 119 119 1) 24 79 69 10 25 113 78 35 26 114 99 15 X = 113.15 X = 86.26 X = 38.96 S.D. = 37.8 S.D. = 38.4 S.D. = 69.56 Range—61-188 Range—4-182 Range—0-108 *The above cones were obtained from the Sierra Nevada foothills just east of Sacramento and from the Coast Range immediately west of Sacramento on October 13 and 20, 1981. (Herbarium Voucher, Farris 94907, DAV). 117 FARRIS Vol. 2, No. 2 Indian people often maximized their efforts by sampling a few cones from a tree before settling down to collect the cones in earnest. This practice is illustrated in some fore, the averages of 113 seeds per cone and 86 seeds with developed kernels per cone (Table 1) include some cones which would probably have been rejected by the Indians. The difficulty with collecting digger pine cones is that they usually adhere firmly to the branches and often need to be twisted off by hand. This a required climbing the trees and such climbing was normally done by men. By contrast, processing of the cones was undertaken by the women (Willoughby 1963:28- 29). It is not sufficient to wait for the cones to drop because they only do so after opening their scales while still attached to the tree and scattering the seed over a period of several months. The cone may remain on the tree for as much as seven years after shedding the seed (Jepson 1910: 87). Competition with animals made it advisable to pick the cones before they were ready to open on their own. They were then heated to remove the bothersome pitch and also to get the cones to open somewhat to facilitate the removal of the seeds. The seeds could be eaten raw, but were commonly roasted either in the cone or in parching trays. In addition they were often ground up into a meal to be boiled as a pine nut soup or baked into a bread. If stored for more than a year the high fat content would cause the seeds to become rancid. PINUS MONOPHYLLA In many ways P. monophylla forms a striking contrast to P. sabiniana. The cone is comparatively tiny, often only 5-8 cm high (Fig. 1). There are only 10-20 seeds in an average cone. The trees are much smaller and more aaa usually not more than 8m igs Rai are found in southern Idaho, Utah, Arizona, California and Baja Calif- ornia ange of elevation is generally above 1200 m tah 1908 :35-37; Critch- field and ia 1966:9, 48). The interval between large seed crops is 1-2 years (USDA 1974:611). The seeds are often quite large and have a high kernel-to-shell ratio. In samples measured by the author the kernel averaged 72-77% of the total weight of the seed while the shell averaged 23-28%. The average kernel weight was 270 mg. The thin shell meant that it was very easy to hull the seeds. Although water flotation is not effective in separ- ating the seeds with developed kernels from those with undeveloped ones, there is a clue in the coloration of the seeds. The dark seeds tend to have the developed kernels while the undeveloped seeds are usually a tan color (Lanner 1981:48). or people obtained the cones by knocking them down with a stick or shaking the tree. As in the case of P. sabiniana, men would usually knock the cones off the tree and then the women would collect them and process them (Steward 1941:312-313). How- ever, in an eyewitness account from 1891, the women knocked down the cones, collected them, and processed them (Dutcher 1893:378-379). e cones were down they were heated to open them since they were usually collected prior to full ripening. The seeds could be hulled by rolling them on a flat stone (metate) using a handstone (mano). It appears that even the hulls were eaten in some cases since they were present in human coprolites (fossilized feces) found in southern California desert archaeological sites (Wilke 1978:79). For storage the seeds were cleaned of the chaff and dirt through winnowing and then packed in baskets or, in later times, in cloth gunnysacks. NUTRITIONAL COMPARISON Dramatic differences are to be found in nutritional data on the seeds of P. mono- phylla and P. sabiniana. Considering three major constituents: protein, carbohydrates December 1982 JOURNAL OF ETHNOBIOLOGY 118 and fats, these two species differ significantly. Whereas seeds of P. sabiniana have over 25% protein, those of P. monophylla have under 10%. On the other hand, P. monophylla seeds have over 50% carbohydrate as against a figure below 20% for those of P. sabiniana having about 50% fat and P. monophylla having only 23% (Table 2). The amino acid content of the protein found in each species is shown in Table 3. The most limiting amino acid, i.e., the amino acid in least concentration, is determined by means of a scale of “chemical scores.” The amino acids are compared against an ideal protein using the formula: mg of A.A. in 1 g. of test protein Chemical Score = - X 100 mg of A.A. in ideal protein The “ideal protein” figures were developed to replace specific foods such as human milk and whole egg which had been used in previous chemical or amino acid score calculations (FAO/WHO 1973:62-64). It is therefore necessary to determine the source of data for the calculations of chemical scores found in other published materials so as not to make erroneous comparisons (e.g., Benson et al. 1973:146; Kaldy et al. 1980:356). The overall protein score is derived from the score of the most limiting amino acid. This is due to the necessity that “all amino acids must be present at the site of protein synthesis in adequate amounts for protein synthesis to proceed, an equal percentage deficit of any essential amino acid would limit protein synthesis to a comparable degree” (FAO/WHO 1973:62). The protein scores for P. sabiniana and P. monophylla found in Table 3 show that lysine is ae most limiting amino acid in both species. This is a common finding for plant Protein with some exceptions (e.g., legumes). P. monophylla ranks considerably higher than P. sabiniana in each of the other amino acids with the exception of the sulphur-con- taining ones (methionine and cystine). The fat in P. sabiniana is composed of 4.3% saturated fatty acids, 50.5% oleic acid (monounsaturated) and 45.2% linoleic (polyunsaturated) acid (Semb 1935: 610). P. monophylla fat is 85% composed of unsaturated oleate, linoleate, and linolenate acids (Lanner 1981:102;cf. Adams and Holmes 1913). Although the fiber content was not determined for P. sabiniana, it is apparent from Table 2 that crude fiber is generally quite low for pine seeds. The low 1.1% figure for P. monophylla would have been substantially increased on occasions when the shells were eaten along with the kernels. However, the hard shell of the digger pine seed would have precluded such a possibility for this species. The ash content of digger pine is shown to be nearly twice that of P. monophylla, although quite comparable to several other pine species (e.g., P. pinea and P. lamber- tiana). P. monophylla resembles more its Southwest neighbor, P. edulis, in this regard. The mineral content shows higher levels for P. sabiniana in calcium, iron, manganese and zinc (Table 2). The energy value of 100 g of P. sabiniana is substantially higher than P. monophylia, mainly due to the high fat content. The caloric content places P. monophylla more on a par with acorn meal than with any of the other pine species shown (Table 2). CONCLUSION Particularly subject to erroneous reporting in the past (see Farris 1980), although a recent author seems better informed (Lanner 1981). Although P. monophylla has had some local commercial success as the source of a cash crop, this has not occurred in the case of P. sabiniana, despite the great success of Vol. 2, No. 2 FARRIS 119 “ST6T Wey * “OF S96T [M19 Pur eM ‘E961 SNOuAUOUY *Yy SP6T SAzYS pur UNOg “3 “S86T Suey J “(TS61 “Te 39 8821p) poyyau Suryse-Aup Suysn staeq/A “uay PED Aq parensyes ouvysaquiny “g pur yypXKydouow “gy ‘puDrurgns “g 105 yUIIUOD [RIBUTY*9 ‘(O9T: S961 M422 pur eM) [dg 3 OOT 19d parefnsyes] 2 ¢°g x 1ey 3+ LOH x “QuerE+ ppg x“I0Id B= Oy “—p “Pae[Mofeo JOU SUBIUT a]qe} WOIy 2dUasqe ‘1aqy apni *d ‘soqiy Surpnpur amy ayespAyoqien oy, *q ‘(TOT S961 M22 PUr eM SET? TE6T S2uOl) *N 203 s9yYdNnUT ¢-g & SuISN payemMoqes UIDIOIg —“B = > - . OFF 0% ~ hh Tea 2s. se (Tear) 970907 snosang) * 7 a Sb so (Tz) sto S6T 8s SII (Teaut) nIFopay snosa ng) To re £8 2% b69 ot - £8 38 Vu yDUPnsaquiny ‘d ~ = rae OsI 8¢9 ce 0 (rt) «6S CC 0DCOSTCC'SE gl"p? ‘d - sa rb (OFT ag s+ (60) “IT wle TIS 99 yma 'd 6% I 61 86Ol Ost #2 46(TI) «69g =6okSsCid'V z01 gi kydouow q ost FF ys O'g 12g a - Flt eH Oo SF yPurnusqns ‘q uz = UW 24 Cs) —— ‘dad 3 001/3ur) (uonsog 21qIpPy sures8 QQ [/sures8) gi¥aiuoD [essUTW proorrmy WV sad GUD WA yd 121M saqoadg ‘PAW ULOIp PUD spaag auty auog fo sano, pouoryugnn—Z TIAVL 120 JOURNAL OF ETHNOBIOLOGY “ ) = P-) = 5 a *Ayaatjoadsai aurueyeyAusyd pur sutuoryjeur uo uo de Buueds m9y} JO Isned9q PapnyuUI are suUTSOIA} pue suNSAD sploy OUTUTy * _ ‘youd feapt ur yy 3u (€9:€L61 OHM/OV) 001 X ordgsen-3 | ur yy Su = 2209S PPY ouTUTy Jo eommayD “2 (LOT: S961 II9W pure 131e&M SET: [E61 seuof) sinu 10j “301g 3/yy 3u = et ‘ajdures utajo1d jo stsAjorpAy ut jso] ueydoidAIy, *pouruiajap yy [enuassa g jo / AjUO “q Gc ‘N3/vv Bu €9:€L61 OHM/OVA *? 2861 stuey *e ob bP $2109g UID101g O0I< 6G 186 88 Lad 6&6 0S ouyeA 88 G§ 88I 8g 3S €dI OF auluoa1yy, O0I< GL 86E OOT< s9 bE 09 pues) + auTueye;Ausug IZ GZ Ig $l 93 OFT G¢ pPunsdo + autuompay oP €3 Sol bb Zo re | sg autsA’y 00T< GL 00+F €8 8¢ 906 OL aupney 0OoT OF €1Z 89 Lo ha OF auTona]os] 21095 oig3/yy 8u = *N 38/yy 3u 32409 301g B/yy 3m *N 8/yy wu "og 3/yy 3w P) p P gSPPV ourury [enuessy eyAydououw snug BUETUIQES SNUIg 3 204d [PPI py eliAydouow ° d pup vuvrurqes *g fo spaag fo 3urs02g poomayy pup uowsodwoy poy oump—¢ FTAVL 121 FARRIS the very similar Mediterranean species P. pinea, the pignolia (Farris n.d.). had shown great interest in the seeds o Vol. 2, No.2 Indian people P. monophylla and P. sabiniana, although the former took on more the quality of a staple food whereas the latter was used mainly as a special treat. ACKNOWLEDGEMENTS An earlier version of this paper was presented to the 5th Annual Ethnobiology Conference held at San Diego, CA on April 22, 1982. I want Diego Museum of Man for Rendig and T. Steven Inouye of the UC/Davis to express my appreciation to Ken Hedges of the San the opportunity “i arene - paper. Special thanks go also to Victor of Land, Air, and Water Resources who did the nutritional analyses on P. sabiniana and P. eR a Keen of the Department of Nutrition, UC/Davis, for vasa of ash and mineral content; Alan of the UC/Da , Kathy Kanagaki and Barbara Mortimer is Amino Acid Laboratory for analyses of "0 protein amino acids and to G. James West for his seat Son in the preparation of this paper and thoughtful comments in my basic research. The careful editin the improvement of this paper. g and suggestions of the editors and two unknown reviewers contributed greatly to LITERATURE CITED ADAMS, CATHERINE F. 1975. Nutritive value of American foods in common units. cages Agric. Handbook No. 456. Washing- n, D.C. ay MAXWELL, and AUGUST HOLMES. 1913. Pine nut oil. J. Industrial and Engi- neering aig 5 (4):285-287. ANONYM 1963. Food composition t. Amer. Book (R.F. Heizer and M.A. ait eds). Univ. California Press, Berkele BEAN, LOWELL J. 1972. Sedat’ People: The Cahuilla Indians of Southern California. Univ. California Press, Berkeley. geri EVA M., JEAN M. PETERS, MAR- BETTINGER, ROBERT L. 1976. The devel- opment of aay exploitation in central eastern Californ dF ifornia Anthrop. 3(1):81-95. BOTKIN, C.W., and L.B. SHIRES. 1948. The composition and value of pifon nuts. New Mexico Agric. Exp. Sta. Bull. 344. State College, New Mexico CLEGG, MICHAEL S., CARL L. KEEN, BO Dry ashing. Biol. Trace Element . 3: sie CRITCHFIELD, WILLIAM B., and ELBERT L. LE, JR. 1966. Geographic distribution of the pines of the world. U.S.D.A. Forest bl. 991. Washington, D.C. 8. Pinon gathering among the Panamint Indians. Amer. Anthrop. 6:377-380. FAGES, PEDRO. 1937. A Historical, Political, and Natural Description of California by Pedro Fages, ‘crt of Spain. Univ. Califor- nia Press, Berke FAO/WHO. soe ites and protein require- ment: report of a joint FAO/WHO ad hoc expert committee. World Health Organiza- tion Tech. Report Ser. 522. FAO Nutritional Meetings ins Series 52. Geneva. FARRIS, GLENN J. 1980. A reassessment of the nutritional value of Pinus monphyla. J. California and Great Basin Anthrop. 2(1): 132-136. 1982. Aboriginal use of pine nuts i join gl. b Unpubl. Ph.D. (Anthrop.), Univ. California, Davis. California pignolia: seeds of December 1982 JOURNAL OF ETHNOBIOLOGY 122 LITERATURE CITED (continued) ————- 1964. Cone morphology in Pinus sabiniana. J. of the Arnold Arboretum 45:260-273. 1965. Digger pine seedling res- ntinite and non-serpentinite re onse to serpe 19 Silva of Califomia. Memoirs of the Tie, California, Vol. 2. Berkeley JONES, D. ieee. 1931. Factors for con- verting percentages of nitrogen in foods and feeds into pins of proteins. U.S.D.A. Circular 183. on, D.C, KA HNSTON, and bread-root, squaw-root choke. Econ, Botany 34(4):352-357. LANNER, RONALD M. 1981. The Pifion Pine: A Natural and Cultural History. Univ. Nevada Press, Reno, LONG, HANIEL, 1941. Pifion Country. Duell, 1918. The acorn, a possibly neglected source of food. Natl. PENNINGTON, JEAN A.T., and HELEN Food resins H.B. Lippin- 910. Yana texts. Univ. Calitomia Publ. Amer, Arch. and Ethn. 9(1): 1-235. SEMB, JOSEPH. 1935. A phytochemical study of the seed of the Digger Pine. J. Amer. Pharmaceutical Assoc. 24(8):609-613. STEWARD, JULIAN H. 1938. Basin-plateau aboriginal sociopolitical groups. - Bureau Amer. Ethn. Bull. 120, a sn Culture element distribu- tions: XIII Nevada Shoshoni. Univ. California Ant rane ane sg 209-359. SUDWORTH, GEORGE B. 1908. Forest Trees of the or Slope. naeae§ 1967, Dover Publ., New York. U.S.D.A. 1974. Seeds of woody plants in the andbook 450. U.S. WATT, BERNICE K., and as. MERRILL. 1963. Composition of foods. = Hand- book 8. U.S.D.A., bei sey WILKE, PHILIP J. 1978. tone pore human ecology at he Cahuilla hella Contr. Univ. California Archaeol. Res. Facility, Berkel WILLOUGHBY, NONA C. 1965. Division of labor among the Indians of California. Univ. California Archaeol. Survey-Reports 60:70- 79. YANOVSKY, ELIAS. 1936. Food plants of the North American Indians. U.S.D.A. Misc. Publ. 237. Washington, D.C. ZIGMOND, MAURICE LOUIS. 1941. _Ethno- dissert. (Anthrop.), Yale Univ., iudenitiy Microfilms, Ann Arbor, MI 1971 Sseaingin anteemmamn J. Ethnobiol. 2(2): 124-143 December 1982 PAPAGO INFLUENCES ON HABITAT AND BIOTIC DIVERSITY: QUITOVAC OASIS ETHNOECOLOGY ARY P. NABHAN Office of Arid Lands Studies, University of Arizona Tucson, AZ 85719 ADEO M. REA San Diego Natural History Museum San Diego, CA 92112 KAREN L. REICHHARDT Office of Arid Lands Studies, University of Arizona Tucson, AZ 8571 ERIC MELLINK Office of Arid Lands Studies, University of Arizona, and Centro Regional para Estudios de Zonas Aridas y Semiaridas, Colegio de Postgraduados Chapingo, Mexico CHARLES F. HUTCHINSON Office of Arid Lands Studies, University of Arizona Tucson, AZ 85719 ABSTRACT.—Quitovac, Sonora, is an oasis and Papago Indian community in the U.S./ Mexico borderlands, 54 km from an analagous oasis, Quitobaquito, in Organ Pipe Cactus National Monument. Comparison of the two sites provides insight into how traditional Papago subsistence and land use affects habitat and biotic diversity. Quitovac’s springs an modified lagoon have been utilized by Papago farmers for centuries. Around these peren- nial water sources, Papago land and plant management practices created eight large scale and two small scale vegetation associations. These provided habitat for a diversity of plants, birds and mammals, many of which the Papago harvest for utilitarian or religious purposes. Over 138 species of plants, 14 mammals and 108 birds are Lapeer from a 5 ha stud site at the oasis. The concentration of utili bitats clearly affects how these habitats are managed. Since the initistion of the ae. tees a 125 ha area was cleared and levelled for irrigated agriculture. This has dramatically altered life at Quitovac. INTRODUCTION Native American influences on habitats and associated biotic diversity have been the 1 has been hypothesized that the diversification of habitats associated with native agricul- ture has had a beneficial effect on faunal species richness, due to edge effect phenomena, increased insect and seed availability The values of diversified a aad habitats to fauna, and the potential edible or economic return to farmers, were active topics of research among American So . earlier in this century (see Dambach 1948). However, as agriculture has become more mechanized, larger fields of single crops with clean borders have taken the ea a diver- sified family farms where the maintenance of cover crop borders, hedgerows, or wind- breaks was not only practical but advisable (Burger 1978; Sampson 1 Despite the renewed interest in this topic from agricultural ne and ethno- graphers, there are few data with which to compare directly the richness of species (use- ful or otherwise) associated with native subsistence agricultural habitats with that found in nearby, uncultivated or modern cash crop agricultural ecosystems. 125 NABHAN ET AL. Vol. 2, No. 2 Through the Man and the Biosphere program, we have attempted to document qualitatively and quantatively the plant and wildlife diversity associated with various agro-ecosystems and comparable, uncultivated ecosystems in the Sonoran Desert. The habitat complex, and seed plant, bird and mammal diversity were surveyed at the Papago farming oasis of Quitovac, Sonora and at the similar Quitobaquito, Arizona in Organ Pipe Cactus National Monument, where cultivation has not occurred for over 25 years (Fig. 1). There are considerable differences in the biota associated with the sites. Since the two sites differ more in their management history than their physical character, we focus on Papago land use and subsistence practices at Quitovac which influence habitat and biotic diversity. We hope that this ethnoecological perspective on the last Papago oasis will aid in the archaeological and “natural” historical interpretation of other Sonoran Desert oases, as well as in their management. This study is also the most comprehensive treat- ment of the folk biology of the western Papago of Sonora, whose knowledge and uses of desert biota is in many ways different from the central Papago emphasized in Castetter and Underhill’s (1935) classic work. Li eee eae as e BAJA CALIFORNIA if —— NORTE U aw, 2 Ni 2B ~ * ‘~ ~£D 5 156, a ii S Mey te. 47st ° 8 Exig 476g aI: ee -----~- PAPAGO : See lORGAN PIPE- ; & 1 GACTUS INDIAN - ! NATIONAL fee es lh MONUMENT \ RESERVATION Quitobaauitot. Lukeville ~RR ! = 2B 6%, % yta ~ SVS Ee Ae ae ny va ~ CS aeint es P wn -. dunes 2 fields equitovac " GODT OPT ICTS COT CTCOTPPCCOVCOCOS OTTO TS OT IN Srepererere rower were oO COOO TOOTSIE a Puerto Penasco EI Mar de Cortez PRR IIR ogy LEAP EPO ET. OLE OLIN E ME IS PRA OUPP AALSS OPPS ES, To Caborca FIG. 1—Map showing Quitovac in relation to Western Papago Country. THE STUDY AREAS Quitovac is a spring fed oasis, at an elevation of 350 m, in the municipio of Puerto Pefiasco, Sonora. It is found 41 km south-southeast of the Sonoyta-Lukeville border crossing, and 54 km southeast of Quitobaquito. Hastings and Humphrey (1969) reported its mean annual rainfall as 21.9 cm; it lies within the transition between the Lower age and Arizona Upland vegetation subdivisions of the Sonoran Desert (Shreve 51 The presence of water deposited tufa and marl sediments, some of which contain calcified Rancholabrean megafaunal fossils, indicates that the springs of Quitovac have flowed for millenia. When Juan Manje visited the Papago at the site in 1694, calling it San Luis de Bacapa, and Moicaqui (‘Soft Wash’ in Papago), he described it as ‘“‘close to 4 high peaked mountain at whose foot were some springs of water and some lakes” (Bolton 1948). In 1774, Anza described the site as ‘‘one of the best of all the Papagueria, because it has five springs of water . . . which they gather and use to irrigate some small pieces of sania December 1982 JOURNAL OF ETHNOBIOLOGY 126 very sandy land where at most a half a fanega of maize can be planted... ” (Bolton 930). These observations indicate that Papago water control and agricultural manage- ment of Quitovac were well established prior to the introduction of Old World technol- ogy, draft animals, and crops. Kino visited a Papago camp at another set of springs in 1698; his San Serguio is surely a site along the springs of the pre-Cambrian Quitobaquito Hills, on the present day U.S.-Mexico border. Kino did not explicitly mention a pond there, so some historians have assumed that one did not form until Anglos built a dam there in the 1860s. Others disagree, observing that the Papago improved springs and excavated basins elsewhere earlier; the Quitobaquito pupfish shows considerable divergence from Rio Sonoyta pup- fish populations nearby, suggesting the antiquity of Quitobaquito pond habitat (Robert Rush Miller, pers. comm.). From the 1860s on, the presence of Papago pond-irrigated fields and orchards there are well known (Bell et al. 1980;Nabhan 1982). Organ Pipe National Monument was established in the 1930s, but the Papago continued farming and livestock raising there until 1957. The pond was then a 35 cm deep, swampy marsh edged by a grass flat, riparian trees and an orchard; it was ideal pupfish habitat (Robert Rush Miller, pers. comm.). In 1962, was dredged to a 1-2 m depth, and has since been managed as a popular birdwatching ar Not all Papago from the two oases consider themselves to be the distinctive Sand Papago—(Hia C-ed O’odham, ‘In the Sand People’; Hia Tadk Ku:mdam, ‘Sand Root Crush- ers’; or S-’O’obmakam, ‘Apache-like Papago’). However, after an 1851 yellow fever epidemic, some surviving Sand Papago families moved out of the Pinacate region to these nearby oases, or to other western Papago settlements (Bell et al. 1980). At any rate, the Sand Papago regularly visited Quitovac historically, and shared with the people there the use of a number of plants and animals not found elsewhere in Papagueria (Nabhan 1980). Since the 1850s, a rain and cactus harvest ceremony called the Vi‘gita, originally per- formed among Sand Papago in the Pinacate region to the west, has been observed at Quitovac (Davis 1920; Ives 1936; Bell et al. 1980). Within the following notes on the uses of biota, many of the religious uses are those associated with the Vi igita. Linguistically, the Quitovac Papago may be intermediate between the Hia C-ed O’odham and other Tohono O’odham. They regularly use the fricative [v] in gl sound environments where most Papagos make a sound closer to the English [w], an occasionally utilize [t] in place of the more commonly used [c] ([ch] in Shae Both of these allophones are believed to be proto-Piman (Hale, pers. comm.). We are using the Alvarez and Hale (1970) orthography for Tohono O’odham, but are substi- tuting [v] for [w] to reflect the above-mentioned dialect difference. Currently, 16 houses are maintained by Papago and Papago-mestizo families at Quitovac, but not all are lived in year-round. Population has ranged from 27 to 38 in- dividuals since 1960. Papago simply call the place Vak, and Quitovac is rapidly being replaced by Bak as the officially-recognized name for the oasis and the recently estab- lished indigenous land reserve there. METHODS Study of the plant and bird life, and ethnobiology of Quitovac began in November 1979, and has focused on a 10 ha area surrounding the oasis pond. Through July 1982, 12 visits to the area were made for 2-4 day periods, during which a number of data collecting activities were accomplished. In August, 1981, a more formal comparison of Quitobaquito and Quitovac was initiated, using a 5 ha study site centered on the pond. The study methodology also was used for Quitobaquito as well, with exceptions as noted. Agricultural clearing at Quitovac in autumn, 1981, destroyed the habitat on approxi- mately 3 ha of the study site. Consequently, vegetation transects begun in one area were extended to adjacent areas within the same vegetation associations, and only half the original 5 ha area was sampled for mammals. 127 NABHAN ET AL. Vol. 2, No. 2 Habitat mapping utilized February 1982 hand-held aerial photos taken by Peter Kresan, July 1982 vertical aerial photos taken by Vern Palmer, and a sketched map based on paced distances drawn by Nabhan in September 1980. These three data sources were combined in an attempt to reconstruct the extent of habitat areas prior to the autumn 1981 clearing. Each of these mappable habitat units was described in terms of plant species, vege- tative cover, lifeform mixture, soils and land uses. The project’s plant ecologist (K.L.R.) visually discerned discontinua in the vegetative cover of the site. Each unit was sampled for perennials, 75 cm tall or more, via five 30 m line transects placed randomly from a baseline, and via 250 point frame hits for annuals and perennials shorter than 75 cm. Following Karpiscak (1980), cover values from these two methods, including both August 1981 and May 1982 point frame samples, were combined to express a percent of sampled distance with vegetative cover within each habitat. If species were sampled using both methods or during both seasons, the highest value was used. These coverage values were used as indicators of species importance for calculating the diversity (i.e., hetero- se of each habitat’s vegetation, utilizing both the Shannon-Weaver and Simpson indices as described by Peet (1974). To calculate a plant species heterogeneity value for the entire site, species values within each habitat unit were multiplied by the fraction of the total site area occupied by that habitat, and summed. Within each habitat unit, either a soil or water sample was taken from a 30 cm column below the surface, and analyzed by the University of Arizona Soils, Water and Plant Tissue Testing Laboratory, from which methodological details may be obtained. Lifeform descriptions follow Shreve (1951). Land uses were observed during visits, and further documented by interviews with local informants. In addition to plants found on vegetation transects, an inventory was made of all seed plant species found within the 5 ha site. Over 300 voucher specimens were collected over a four year period during all seasons, and most have been deposited in the herbaria of the University of Arizona and San Diego Natural History Museum. Nomenclature follows Lehr (1978), to which non- botanists are referred for common English names. Determination of species native OTF introduced to North America follows Shetler and Skog (1978). thnobotanical interviews were made in Spanish, with Papago frames and lexemes occasionally used to reinforce questions. Although six Quitovac Papago and one Arizona Papago visitor contributed knowledge of plant names and uses, the bulk of the information was derived from the community elder, Luciano Noriego. Over a 3-year period, informa- tion was volunteered by Noriego while walking the site with us or while vouchers were being pressed. Additional plants observed directly in use by other residents were noted. Birds found on the 5 ha site were surveyed during one two-day visit during each of four seasons, and the highest visual or audial count for each species over the two day period taken as the best population estimate. Dawn and dusk surveys of 2-3 hour dura- tions, with extended time spent in dense canopy areas, were sufficient for most idenfi- fications and population estimates. Dove population projections were based on half hour morning counts. It was assumed that dove visitation to the lagoon occurred at a consis- tent rate, throughout the morning, with no more than one extended watering per bird each half day. Thus, the half-hour count was multiplied by 8 to estimate the total dove population. Linnaean taxonomy for birds includes recent revisions by Rea (in press). Notes were taken regarding the habitats in which each bird species spent most of its time, but since few species utilize space for foraging strictly upon the lines of our map- pable habitat units, certain habitats were combined (or collapsed) in our calculations. Simpson and Shannon-Weaver diversity indices were then calculated for these revised habitat groupings more useful in discussing bird foraging, and for each 5 ha site (Quitovac and Quitobaquito) as a whole. Mammal data gathering included the nocturnal setting of Sherman live traps baited with a commercial grain mixture (millet, oat and wheat), the diurnal setting of snap December 1982 JOURNAL OF ETHNOBIOLOGY 128 gopher traps, and the visual counts of larger mammals. The Sherman live traps were set in the evening to capture small nocturnal rodents; they were checked and closed the next morning. Trapping took place one night each in December, 1981, March, 1982, and two nights in May, 1982. Tra aps were set in a grid pattern with 12 m between traps in the same line, and 20 m between lines. At Quitovac, 100 tr. traps were set on the irregularly- shaped undisturbed half of the study area. At Quitobaquito, 200 traps were set each night. For each season, the sites and the habitats within the Quitovac site were compared using the Simpson and Shannon-Weaver indices. The stray pe utilize (a) animal numbers and (b) animal biomass based on individual weights in grams at the time of trapping. Identifications were made in the field utilizing sesibt (1960), with vouchers collected and identifications confirmed for trapping mortalities. Linnaean taxonomy for renters follows Hall (1981), with the exception of Dama, for which we retain Odocoileu Interviews on bird and mammal knowledge and uses were occasional ic kans the study, but also included 3-4 hours of taped interviews in May 1982 with Luciano Noriego and his grandchildren. A scrapbook of photos or drawings of most bird and mammal species potentially present in the vicinity was shown to Noriego, with explanations of calls, behavior or eating habits discussed. Additional data on animal use come from in- specting hunted carcasses gathered by Papago youth, and from accounts of the Papago Viigita ceremony. RESULTS ugh autumn, 1981, Quitovac was a traditional Sonoran Desert farming oasis ry cl ded eight large scale (mappable) vegetation associations, and two small scale vegetational features worthy of note (Fig. 2). The mapped vegetation associations pro- vided one element of our descriptions of habitats. Soils, lifeform and seed plant species diversity, and land uses were also noted (Table 1). The two small scale associations, were (a) man-made ditches running into the orchard and field dominated by Cyperus, Anemop- sis, Heliotropium and Rumex; and (b) living, fieldside fence rows including intentionally planted Salix, Tamarix, Sambucus, Opuntia and Prosopis, which had associated with them piled brush and self-sown Ambrosia, Bebbia, Olneya, and Cercidium These small scale features are best considered part of the diverse ae complex in the south-center of the study site. In both diversity indices based on plan Coverage data, the cultivated field is the most heterogeneous vegetation association, om the orchard the second most. The Shannon-Weaver index is typically most sensitive to changes in the importance of rare species in the sample, and the Simpson index to com- mon species (Peet 1974). These cultivated habitats make up less than 10% of the area of the study site, which is important in the interpretation of whole-site diversity index comparisons of Quitovac and Quitobaquito. Because each habitat’s coverage values “weighted” by the percentage of the 5 ha upon which that habitat exists, and Guitowac Cultivated habitats are so aR small in area, their influence is “diluted” in our whole site calculations. The contrasts between Quitovac’s whole-site plant diversity values (.971, Shannon-Weaver; 813, Simpson), and those for Quitobaquito (.822, Shan- non-Weaver; .764, Simpson) nevertheless suggest that Quitovac has more diverse vegeta- tion. (Note that the higher the diversity index value, the higher the diversity or hetero- teneity), Floristically, there are considerably more plant species, genera and families repre- sented at Quitovac than at Quitobaquito, no matter how large the areas examined are (Table 2). This is due in part to the number of domesticated species (17) einai Cultivated within the Quitovac site, but cultivation contributes more than just tionally sown plants to a flora. There are an additional 59 species of plants Seiad tn in ae field/orchard complex. Many of these can be considered “biologically [as] weeds which are evolutionary and ecological products adapted to survival in habitats disturbed by , No. 2 3 Vol NABHAN ET AL. 129 130 JOURNAL OF ETHNOBIOLOGY December 1982 ai Neh iti BuIzery soyo}Ip Woy Zut1943e8 jueld JPUIDIpsu {Zur y7e8 Joqy 3uizei ‘3urumg Sunsoarey snjovo {3uidde1 Jo 3ununy {8uizern 8utUunY ‘Surnueydsues peru amngy payeanny Burdde} pue 8urjuny {Burj nN poom {But “routes 1INy PIM Rupesiney suso13 PEM ‘ysaarey dos pue ‘8unesiu ‘sur -MOS paas ‘OBETEL uonesiuur IO} 197M JO asn {8ununy prq onenbe ‘8uruwimg 919° a i és" Lte sTeiguayds pur syeru ~uviad Joo! snosoeqiay M9j ‘sassvis Buruts0j -yeur [eTuUd Lag Sosseig puv spao1 [etuuaiad juazisurq sfeisuisyda Maj pur ‘sqniys ‘sqniys snoyAydosoryy SIUIA pue sTerswiaydo ‘sqniys snonprap jeaproig $291} pue sqniys uaa -1y8no1p ‘yu998 -sutds paxtur ‘uadQ syenuue doi snjd ‘speruusiod pue ‘speroutayda Apaam snoaseqiapy aezye Suneoy pure saiAyd -O1IDPUI JUaZIOUIGNS wnoavsspana pjonds syyousiq pponds syyriysig 1Kaujo sndsag sisuasurmop vyd& Saproiso.quy piso.qup Matpunjiag wntae'T vsKBouowm vayouamKpy uo] «onp uopouk;) saployounuco DUUWLAISOILDS DIADI SNILA vyof ~t4afUuoo misonquy 1uossapun WnN1IK-T puynjaa sidoso1g pupkasoy ppapng nyofizsafuor risosqup uo] K90p uopoury) SLISIDG DIyaYyIIUUDT snypusand uojasounjog 1/82 gy Tg Sy Sudd 8L'h ‘d udd Zerg ‘Nn *9OOT ‘gOT X Ow twdd Lov ‘OL ‘siTes a1qnIos ‘1°6 ‘Hd :ureo] Apueg ‘T/bou 1g°z ‘y Suidd ¢°8 ‘Hd :ureoy Aapig T/baut gz*T “y ‘uidd POG ‘siqes >14njos ‘9° “Hd :pues Awroy sdoio yenuur jo PIPY payeanns (¢) s¥ oules wivo] Apues 1/baur L6°6 “Mudd 7g TT d‘udd ¢1°Z¢ ‘N ‘OOT ‘gO1 X Ow Sudd ¥66 ‘SiTes aTqnjos '¢'Z ‘Hd :ureoy Apueg y/bou gt*y Sy sudd 86°I ‘a ‘urdd g/g ‘N TS Tz ‘sites atqnyos ‘9°8 ‘Hd :ureoy Apurg widd ¢9°g ‘y furdd O10 ‘N-FHN ‘Or'T g0I X Og ‘urdd gg ‘s]7e8 arqnyos !g°4 ‘yd :193em Sutidg Bare JO Ape tsdaas SUTTeYTE YELM SIELJ OAT Mopeayy Bare JO YG] sjauueyo ‘a8po uoo8ey eoIR JO %G"} ‘ureydpooyy -19}eM [erautaydy BIE JO %G°G ‘a8pay opisppay quaoelpe pue ‘sdaz] qin payed “LIT JO preyoio peyeanny BaIv JO YG" LG Sururuns esout vpn, “Bae JO %G"9 ‘s8uuds pue puod PPY pawanmny vare JO %OT ‘sdurrds pue uoo8ey jo 133eM uadg ) hes A a as Sosf) pue'y uosduirg :SaoIpuy IDAvIM -uouueys AJISIDAIT aIN}XIW ULIOJasrT (1811J 12409 ysoySty) saiads yuryd queurulog eee ecilin ., dlaigiia: sonstajORIeYO 12]M 10 [IOS Bare [e}0} JO % pue uones07T Ne Ain, i ‘(241s Kpnis vy G) psouog ‘Ivaopn© IV sqouIqQuEJ—"T AIAVL 131 NABHAN ET AL. Vol. 2, No. 2 human activity”? (Bye 1981). We consider 18 of these species to be found at Quitovac only within the cultivated field/orchard complex. A complete flora of Quitovac is near completion, and will list each species by its habitats (Nabhan et al., in preparation). It is not surprising that more than 21 post-Columbian introduced species, in addition to 11 species of Old World domesticates, are part of the Quitovac flora, and are more numerous than at Quitobaquito today. These are primarily ephemerals that for millenia colonized fields, trails and roadsides in the Old World, before rapidly spreading through New World deserts (Naveh 1967; Young et al. 1972). TABLE 2.—Floristic Richness at two Sonoran Desert oases. * 5 hectare 8-10 hectare oasis, plains study site plains around and closest at oasis-pond oasis-pond hills Quitovac, Sonora plant families 45 (41) 49 (42) 55 (44) genera 115 (100) 131 (106) 139 (114) species 139 (122) 158 (131) 172 (143) Quitobaquito, Ariz. plant families 32 (30) 37 (35) 38 (36) genera 71 (69) 92 (90) 101 (99) species 80 (78) 104 (102) 118 (116) — only seed plants. Data for Quitobaquito are from Adams (1971); Bowers (1980); and and Reichhardt, field notes. Data for Quitovac are from Nabhan, Reichhardt and Rea, (in sccm Values in parentheses cman adjusted totals that exclude intentionally planted domesticated species. Table 3 lists the 78 taxa named by Quitovac Papago in their local dialect, as well as the uses of these plants. Over 40 of these utilized species can be found in the field/ orchard are Even recently introduced species such as Brassica tournefortii are utilized in imilar manner to edible greens of considerable antiquity in the region. oh a species a “native” subsistence resource is somewhat of a misnomer. Numero Old World crops and weeds are well-integrated into Papago cuisine even at gricultural margins of Papagueria. A detailed discussion of how particular plants are tie will be included in the Quitovac flora (Nabhan et al., in preparation), but from the data included here it is clear that named and utilized species are largely concentrated in and affect the management of three habitats more than the scsaoagng | field, the or- chard and the adjacent scrubland. These three habitats are “off limits ing animals most of the time. Such plant uses appear to parallel those which ut Papago prac- ticed at Quitobaquito earlier in this century (Bell et al. 1980). Bird life at Quitovac includes 103 species observed on the 5 ha site during our eight days of survey in 1981-1982. Table 4 indicates that during every seasonal visit, species richness was higher at Quitovac than at Quitobaquito. The diversity indices for the two sites do not show such a clear picture; each site had a more heterogeneous avifauna in two of the seasons. Table 5 shows considerable seasonal variation in bird diversity within each habitat at Quitovac. It appears that the field-orchard complex, and the adjacent micro- phyllous shrubs in the wash provide the habitats with the most consistent diversity from season to season. | i re ; ~ a x poo} sNjOe) jaueg $9700 snyov20497 mart p= a x «4POOM,, ysrug oung mKZouow vayrouau py poyem:t, (ease 330) D é x pooj sn32u) BoyrBpry; sngpnnsvf snasa20u1y Iq BIAS:1, a x (en) AETD £, Aa miopfiasnd vapiosoqoag Bny:t, a‘tq x pooy seag APIOUY (a3naf) oysuvsa0yd pyundo reyq?t, a x “Bras ‘pooy wor) skow p97 ung sry “a x "IAN ‘poom poomuoyy pj0s9) vkauy) weypr, oy 1t4 “a x “pow uvog Sudan uesrxayy aanpnoopiq wnidogs wiry tyaY a'a x ysng amg vaount mqqag sn, yey sm “a x "gm ‘pon orenies (sud) vayun8s8 vaSous) uesiey a‘la x x poo} eTOYD Surduinf opssinf undo areu:ey a x pooj ysenbg meysny Ext DILGANIN) pry > aD x 2029NsTY 4282q twn2140/y)9 UospuapuDsoyd pemoyey 8 q x queig r2dedpueg Magany) xKuopiag weyiaspey 5 a x pooy rag Appoug ayding paonjora myundo ross 2) a x poo UOPULIAIE My snypun) snyynagi7y daa a a‘o x ayeurawog wnjounad votung nAvures ° i ‘tq x apeynysin wnsoyftpou wnunjog iden ano E : . press por2. aden pian kyo0p x1U90YG tsof pnyno x pooy uereury eppynyno = 4 suchas ‘ od Mawyod snyjunowy sty ‘a x poom ‘poos nye) adydurkg Maquny} snasa20uaj9 stanono ro) a xX pooj eTo4D vdav20yjuvov vyyundcC wyjop i‘a x Syjox MOTTIM Burppooy nSurppoo8 xyng 1,339 2 sry “a x poo BqTUIg 11170498 snasa20ydo'] fun:20 re qd é x UIA [EWUIUIEWIO DjIaSSIp Muauay Seryq:iq 5 q x ooseqo yy 12289] mp saydo8 1394904 snuiqun sKmouwoyy oyand x x xX daays wioysig sisuapDpUD? s1aQ urosno x = x yqqesryoel papier- xorg snoiusofyvs sndaT] Jaina yno - 4 xoj ITY xojaa sadjn, raeno xX xX X Jarumbs y0yY snyvsatava snpydowsads yoyao p 3 xX X 230407) SUDLYD) S1UD!) ueq x x yunys panods staoqnd ajv3oj0ds eryd:n, (17e,) AqmaN a2IguQg omsewog prim ‘“3T9y poor aWeN uoUTUIOD IUWIEN ITUIIOS awe y o8edeg Suryovay payodey ‘DL4OU0G ‘9D00}1NZ) 4DaU 40 JY spomUnu fo sasn puDv DxD} yOA—" | ATAVL —— December 1982 JOURNAL OF ETHNOBIOLOGY 138 Waterfowl, doves and quail are the major bird foods utilized by Quitovac Papago. These are taken with .22 rifle, slingshot, or a trip-trigger deadfall box trap made of saguaro ribs, called a kakast. Feathers of several bird species are used ceremonially on staffs and prayersticks during the Vi’igita. These surely include Golden Eagle and turkey; probably Red-tailed Hawk and Great Horned Owl, and possibly raven. Unfortunately (for us!), some are painted bright colors, and others are old and misshapen from years of use, so that casual observation during the ceremony was not enough to confirm identifi- cations noted in the literature (Cano-Avila 1979; Davis 1920). The mammals which we consider to be present on Quitovac’s 5 ha study site include the same four small rodent species live-trapped at Quitobaquito (see those marked wit asterisks in Table 7); a trapped gopher; and nine other taxa observed during our visits. Five of these 14 species are domesticated mammals. The Papago report that 13 addi- tional species can be found in nearby mountain ranges and valleys; particularly in times of drought, certain of these mammals may attempt to drink at the lagoon. Yet due to near-continuous human presence, we doubt whether mammals such as deer and javelina drink or browse at Quitovac as frequently as they do at Quitobaquito. Although the same four rodent species were eventually trapped at both sites, trap- ping at Quitobaquito in December and March resulted in more species and individuals than at Quitovac (Table 8). Unfortunately, no trapping was done at Quitovac prior to the clearing; but mammal diversity was obviously less than at Quitobaquito in the first months following this habitat destruction. The May diversity indices based on mammal weights were higher for Quitovac, while those based on mammal numbers were higher for Quitobaquito. This is because packrats (Neotoma) contributed 70% of the weight of trapped mammals at Quitobaquito, but only 30% of the total number of individuals trapped. Table 7 provides ethnozoological data on 31 mammal taxa occurring in the Quitovac vicinity which the Papago there name and/or utilize. Of the 15 taxa utilized for food, most are now shot with .22 rifle; it has been decades since bow hunting and on-foot drives were regularly used. Of religious uses, the tail of the ringtail (Bassaricus) and many parts of the mule deer (Odocoileus hemionus) are apparently still utilized in the Vi’igita. We could neither con- firm nor deny the Vi’igita’s ceremonial enactment of killing other large mammals (such as pronghorn) in addition to mule deer, as Davis (1920) suggested. Finally, dogs, horses, and cattle are ever-present at Quitovac, and in many ways limit the one of other animals. Pigs and chickens as well as other domesticates are occas- sionally kept in the village, but their influence is not so obvious. TABLE 8.—Mammal species richness and diversity at two Sonoran Desert oases based on live-trapping). Diversity Indices No. of Based on Weight Based on Numbers Locality & Season species Simpson Shannon-Weaver Simpson Shannon-Weaver Quitovac, Sonora Dec. 81 0 pad of _ =a March 82 1 0 0 0 0 May 82 4 .686 545 .493 410 Quitobaquito, Ariz. Dec. 81 1 0 0 0 0 March 82 4 427 332 .667 477 May 82 4 469 393 675 532 139 NABHAN ET AL. Vol. 2, No. 2 CONCLUSIONS Recently, human ecologists have hypothesized that native Americans formerly man- aged habitats in ways that encouraged diversity, resulting in benefits in environmental stability or food abundance and reliability (Nabhan and Sheridan 1977; Brush et al. 1981; Emslie 1981). The meaning of diversity, the best ways to measure it, and its rela- tionship to environmental stability are all controversial among theoretical ecologists (Peet 1974, Murdoch 1975). Nevertheless, Altieri (1980) has demonstrated that in agricultural situations, there is clearly a positive correlation between plant diversity in fields, and stability with regard to vulnerability to animal pests. Utilizing several measures of diversity, we have compared two oases: Quitovac, a “traditional” agricultural setting until the autumn, 1981 clearing in preparation for modern mechanized groundwater agriculture; and Quitobaquito, formerly much like Quitovac, but managed as a wildlife sanctuary in a National Monument since the late 1950s. Because of the removal of cattle and certain introduced plants, as well as the earlier cessation of farming, most Park Service managers would consider that Quito- baquito is undergoing secondary succession “back” to a more natural, perhaps more diverse, condition. Yet when compared to Quitobaquito, Quitovac is more diverse in terms of plants, somewhat more diverse in birds, and not nearly as diverse in mammals, despite recent habitat disruption. The richness of biota at Quitovac has provided its inhabitants with a diversity of foods, medicines and ceremonial paraphernalia, over and above any cash crops produced there. At Quitobaquito, only dying figs and pomegranates, a few field weeds, and the outlines of ditches persist to suggest that additional species (and habi- tats?) may have been present a few decades ago. The implications of these differences should be well understood by archaeologists. To fully explain the present differences between the two oases, it is necessary to consider Papago land use activities. Figure 4 illustrates subsistence-related land uses at Quitovac, some of which affect only target species, while others impact upon all species of one life-form, or a food chain based in a particular habitat. Since we feel that these activities account for the differences in biotic diversity between Quitovac and Quito- baquito more than do other historic or contemporary factors, we will discuss each acti- sr in Figure 4 (according to its letters) in the context of both sites. The habitats in hich these activities take place are shown in Figure 3 and described in Table 1. Some activities may take place in more than one habitat. FIG. 3—Oblique map of habitats at Quitovac, reconstructing pre-August 1981 conditions, based on Figure 2. a Fe alam cig December 1982 JOURNAL OF ETHNOBIOLOGY 140 “ ee? FIG. 4—Papago land uses affecting biotic diversity (see text for explanation). Illustration by Paul Mirocha. At Quitovac, wild plant gathering occurs in the field as well as on the pond fringe and in the arroyo (A). Humans compete with birds for saguaro and wolfberry (Lycium) fruit; Davis (1920) reported that 120 gallons (454 1) of cactus wine was consumed a the Vi'igita alone. Since only a small percentage of the seeds produced naturally ii nate in favorable sites, it is unlikely that wild fruit gathering reduces plant population sizes. Likewise, the wild greens (eg., Chenopodium) harvested ase apuniaat ds good years and produce so many propagules that whole plant harvesting probably does not iminish populations. 141 NABHAN ET AL. Vol. 2, No. 2 The mosaic of disturbed soil, low shrub cover (fence rows and pomegranate bushes) and generally the greater availability of fruits and seeds (and presumably insects, which we did not monitor) at Quitovac promote larger numbers of grackles, Northern Cardinals, Pyrrhuloxia, Canyon Towhees, White-crowned Sparrows and certain transients such as Black-headed and Blue Grosbeaks. However, mistletoe and wolfberry fruits are more abundant, in season, at Quitobaquito. These are utilized by mimids, bombycillids, and several other semi-frugivorous groups. though intentional burning could locally-extirpate fire-susceptible species, it is largely practiced on the pond fringe (B). Emergent Scirpus and Typha stands with much accumulated dead standing crop are annually “cleaned out” at low water, in part so that newer tender shoots will be available to livestock. The plants regenerate, but the tempo- rary openings between them provide habitat for rails, herons, and other wading birds. Livestock grazing and browsing probably eliminates certain palatable species from the area altogether (C). Along channels from the springs to the lagaoon, Quitovac lacks the tender Eustoma exaltam and Centaurium calycosum found at Quitobaquito. Live- stock disperse and “‘plant’”’ seeds. They also compete with other mammals. Plowing and other forms of periodic soil disturbance release the wild seed reservoir in the soil for germination (D). Some weed seeds, including Amaranthus, and Probos- cidea have their dormancy broken by light exposure (Wiese and Davis 1967; Anderson 1968); a plow’s superficial covering encourages germination. At Quitobaquito, due to ae of periodic soil disturbance, few ephemeral or weedy annuals germinate. Plowing © exposes invertebrates to blackbirds and grackles, that readily feed in open furrows rae 1974). The planting of living fence rows (E) provides field- and pond-edge borders that fly- catchers (7 spp.) regularly utilize as perches from which to feed. The planting of Salix, Prosopis and Tamarix has provided some of the most intensively utilized habitat at Quitovac; at Quitobaquito, fewer Salix are regernating on their own. The brush woven between fieldside fence rows provides habitat for the few Neotoma at the Quitovac site. Hunting and trapping, primarily of quail and dove, reduce population numbers only slightly today (F). Occasionally other, rarer bird species are killed with slingshots. Cottontails and jack rabbits are hunted around the fields, but their populations do not appear to be threatened. Irrigation of selected areas (G) provides moisture to germinate and bring to seed numerous plant species. Plants such as Anemopsis, Spergularia and Heliotropium thrive in irrigation ditches. Flowing water, and increased humidity attract certain insects, and in turn attract birds (e.g., phoebes). Transplanting and tending of domesticated ot shea such as palms and figs provide Quitovac with its most diverse habitat (H). e shade, and multiple strata are heavily utilized by orioles, woodpeckers, cowbirds ah migrating insectivores (flycatchers, vireos, and wood warblers). At Quitobaquito, the last dozen or so pomegranates and figs are dead or senescing, and palms have been removed. arge carnivorous birds (families or flocks of Black Vultures, Turkey Vultures, Red-tailed Hawks, Harris’ Hawks) were common and conspicuous throughout the day at Quitovac. They were attracted by several large dead or nearly dead cottonwoods former- ly standing in open fields where the birds could drink and bathe. In spite of constant human activities, these large birds were quite at ease at Quitovac. In contrast, hawks and vultures only incidentally flew over the Quitobaquito oasis. The large erm aa at Quitovac also attracted Purple Martins and several other swallow species. However, at Quitobaquito the immediate juxtaposition of open pond and mesquite laid attracted much higher breeding and post-breeding populations of BP opeplas than we found at Quitovac. Finally, intentional seed sowing (I) provides grain, melons, legumes and forage utilized by humans and other animals. The only domesticated annual at Quitobaquito is safflower (Carthamnus), which is feral along roadsides in northern Mexico. dia ee a aeened Te 0 NER ome meme i aeiiititens ae Fe December 1982 JOURNAL OF ETHNOBIOLOGY 142 The dynamic habitats at Quitovac have provided food, water and shelter to humans and other lifeforms for centuries. Recen y, however, much of this habitat was removed when 125 ha of land was cleared for groundwater irrigated agriculture. The project was promoted by governmental agencies to provide economic opportunities for Papagos. While Quitovac residents look forward to increased crop production in the future, to this date the development has not been completed due to political and economic problems. Residents clearly lament the unnecessary destruction of fence rows, abandoned houses, and other historic structures, as well as the e disruption of the springs. Future pumping of groundwater will likely influence flow to the pond. Thus re-establishment of riparian habitat is questionable. As at Quitobaquito in the 1950s, sustainable, traditional agricul- ture, and the “wild” resources associated with it were not evaluated to any extent before a different course of management was initiated (Nabhan 1982). Johnson et al. (1977) have argued that habitat destruction has contributed more to the post-1600 extinctions of 120 bird and mammal species than have hunting, trapping and other “direct causes.”” In doing so, Johnson and colleagues rightfully call for further efforts to protect “endangered” wild habitats. It may be worth considering that diverse agricultural habitats, including certain ones maintained by native American farmers for centuries, are also now endangered. It is unlikely that one could find environments more rare or more vulnerable than those found in desert oases like Quitovac or Quitobaquito. Their loss will affect not only the bird and mammal populations sustained by them, but may impoverish the life of the human community as well. ACKNOWLEDGEMENTS We thank the Papago community of Quitovac, particularly Luciano Noriego, Hector Manuel and David Manuel, gobernador general del tribu Papago in Sonora. Other Papago, including Ofelia Zepeda and Delores Lewis, helped us appreciate and gain perspective on Quitovac, a people and their langu- age. Several itional scientists cooperated with us in gathering field data: Larry Toolin, Janice Bowers, Richard Felger, Bryan Brown, Peter Warshall, Julian Hayden, John Sumner, Peter Kresan, of Quitobaquito. This work was funded by the Consortium for the Study of Man’s Realtionship with the Global Environment, Man and the Biosphere program, and the Tinker Field F oundation. LITERATURE CITED ornia expeditions. Univ. California A flora of Quitoba- Calif Press, Berkeley. AMS, W. G. 1971. quito, Unpubl. Ms. deposited at Organ Pipe 1948. aaian s historic memoir Cactus Natl. Mon : ALTIERI, M. 1980. vais aed of corn of Pimeria Alta. California Press, agroecosystems as a means of re Berkele ; ‘oO populations. Florida ol BOWERS, J. E. 1980. Flora of Organ sl Bist 63(4):450-456, Cactus National Monument. J. Ariz ALVAREZ, A.and K. HALE, 1970. Toward a Nevada Acad. Sci. 15:1-11; 33-47. manual of Papago grammar, Internatl. J. of BRUSH, S. B., H. J. CARNEY and Linguistics 36:33-97, M Dynamics of oes potato ANDERSON, L.C. 1968. Effects of gibberel- agricultur Econ. Botany 35(1):7 lic acid on germination and continued growth BURGER, es Vv. 1978. A nena pie wild- Of Proboscidea lousianica, Phytomorph. life. . 88-107 in Wildlife and America 18(1):166-173, (i. P- Brokaw, ed.). Council of Environ- BELL, R., K. M. ANDERSON and Y. G. mental Quality, Washington, D. ae R. ib Sen Quelities—cthnoecology of tary and its history. es tee Service Wes- ns—past, present and future. J. tern Archaeol. Center, T aaa = 109-123. BOLTON, H. E, (ed. and oe 1930. Anza’s CANO-AVILA, G. STEWARD. 1980. The Quitobaquito ceme- 1979. La fiesta del Cucu, 143 NABHAN ET AL. Vol. 2, No. 2 LITERATURE CITED (continued) ee de la lluvia de los Papagos. Simposio de oria de Sonora, Memoria IV:138-144, Ry ipentitin S. 1974. Unusual feeding habits in two species of blackbird. Wilson Bull. 86:121. CASTETTER, E. F. and R. M. UNDERHILL. 1935. The yagi of the Papago Indians. ew Mexico Bull. 275, Biol. Ser. a een Stud. Amer. South- west 2 COCKRUM, E. L. 1960, The Recent Mammals of Arizona: Their Taxonomy and Distribu- tion. Univ. Arizona Press, Tucso DAMBACH, C. A. 1948. A io of the ecol- ogy and economic value of crop field borders, Grad. School Stud, — Series 2, Ohio State The Papago ceremony of Vikita. Indian oie and Monogr, 3(4). Mus Amer. ‘iden, Heye Found., New York. SLIE, S. D. o> Birds and prehistoric agricultu the New Mexican pueblos. Human Ecol. 4: ager HALL, E. R. 1981. The Mammals of North America. Second Edition. Wiley-Inter- science, New York. R. R. HUMPHREY meterorology and climatology of Arid Regions 19:1-96. IVES, R. L. 1936. Some Papago migrations in the Sonoyta Valley. Masterkey 10:161-167. JOHNSON, R. R., L. T. HAIGHT, and J. M. SIMPSON. 1977. Endangered species vs. cudaiiaiel habitats: a concept. Pp. 68-79, in Importance, preservation and management of riparian habitat: a symposium (R. R. John- son and D. A. Jones, eds.). USDA Forest Ser- vice General Tech. Report RM-43, Fort Col- lins. KARPISCAK, M. M. 1980. Secondary succes- sion of abandoned = vegetation in south- ern Arizona. ge: Ph.D. dissert., Univ. Tucs LEHR, J. HARRY. 1978. A catalogue of the flora of Arizona. Desert Botanical Garden, Phoenix. LINARES, O. F. 1976. ‘Garden hunting” in the American tropics. Human Ecol. 5:97- LE. MURDOCH, W. W. 1975. Diversity pene complexity and pest control. J. Appl. 12:745-807 NABHAN, G. P. and T. E. SHERIDAN. 1977. Living fencerows of io San Miguel, Sonora, Mexico: traditional technology for floodplain management. Human Ecol. 5:97- 18 ae NABHAN, G. P. 1980. Ammobroma sonorae, an pour parasitic plant in extremely arid North America. Desert Plants 2:188-196. ____—. «21982. Where the birds are our friends—the tale of two oases. Pp. 87-98, in The Desert Smells Like Rain. North Point Press, San Francisco. eo G. “— K. L. RIECHHARDT, and A. In prep. Flora of Quitovac, Sonor with aie on vegetation and ethno- botany. Siar, Z. 1967. Mediterranean ecosystems and vegetation types in California and Israel. sop 48 :443-459, PEET, R. K. 1974. The measurement of species diversity. Annu. Rev. Ecol. Syst 5:285-307 REA, A.M. 1979. The a of Pima fields. Environment ee 484:1 n press. Once a eit life and habitat Fries on the middle Gila. Univ. Arizona Press, Tucso SAMPSON, R. N. 1981, Farmland or waste- land—a time to choose. Rodale Press, Em- maus. SHETLER , S. G. and L. E. SKOG. 1978. A ceoriahcual checklist ao species for flora North America. Monogr. in systematic botany. Missouri Botanical Garden, St. Louis. SHREVE, F. 1951. Vegetation of the Sonoran Desert. Carnegie Inst. Wash. Publ. 5 m tures, ar pga and depths. Weeds 15 (1):118-121 YOUNG, J. ‘. R. A. EVANS and J. MAJOR. 1972. Alien plants in the Great Basin. J: Range Manag.:194-199. J. Ethnobiol. 2(2): 144-153 December 1982 PSYCHOACTIVE PAINTED PERUVIAN PLANTS THE SHAMANISM TEXTILE ALANA CORDY-COLLINS Department of Anthropology, University of San Diego Alcala Park, San Diego, CA 92110 Museum of Man Balboa Park, San Diego, CA 92101 ABSTRACT.—From a cache of over 200 Chavin textiles found in the Ica Valley on the xhibited in the San Diego Museum of Man’s 1980 analysis of South American shamanism, this 2000 year old cloth is painted with images of transformation and transcendence. Of special interest is the repre- sentation of three plants shown in connection with a jaguar, winged deer, hummingbirds, shamans, and a deity. While this textile has been described before in the context of the others in the group, the present paper addresses the plants specifically. One plant is very likely to be the hallucinogenic San Pedro cactus (Trichocereus pachanot); another is more tentatively suggested to be seed pods of the ceca acacia (Anadenanthera pere, or A. colubrina). The third plant, while still eluding precise classification, must be con- sidered as a possible narcotic as well. INTRODUCTION Over 2000 years ago a Chavin artist of ancient Peru painted a cotton textile with a religious message which is dramatically clear today. The message is shamanism and the textile is remarkable because all of shamanism’s basic elements are represented onit: the shaman, his animal familiar, his means for entering a trance to contact the supernatural world, and a deity from that world sli 1). This cloth is hereafter referred to as The Shamanism Textile. Now fragmentary and measuring 54.61 x 68.58 cm, the cloth is a plain weave, 1/1 (1 warp thread by 1 oe thread), of S-spun cotton (Gossypium barba- dense) and is painted with tan and brown colors (pigments as yet untested) over the ecru FIG, 1—The Siaienaibes Textile: a. shaman grasping easicenihers pods, b. sige Pedro Cactus, c. jaguar, d. hummingbird, e. Staff God, f. winged deer, g. mystery plan - % (Ken Hedges photo) 145 CORDY-COLLINS Vol. 2, No. 2 tone of the natural cotton. A total of 65 extant motifs is assignable to 12 categories (Table 1). DISCUSSION The Archaeological Context The Shamanism Textile, along with the others in the cache, was reportedly found at the site of Carhua in the Ica Valley on the south coast of Peru by huaqueros (grave rob- bers)!, However, on stylistic grounds, the textiles belong to a quite different locale; they are most comparable to the art of the Chavin type site, Chavin de Hu4ntar, situated in the eastern Andes at an elevation of 3135 m above sea level, and 644 km distant from Carhua (Fig. 2). Therefore, it is most likely that the textiles were carried from the locus of man- acture, possibly Chavin de Hu4ntar, to the south coast (Cordy-Collins 1976: a2; Conklin 1978:7). Chavin Shamanism Elsewhere I have argued that, as a group, the textiles functioned as a catechism which brought a religious message from one non-literate society to another in pictographic form, and that the message concerned the ascendency of a new Chavin deity (Cordy-Collins 1976). Furthermore, I have contended that the Chavin peoples’ religion was shamanistic and was based on plant hallucinogens (Cordy-Collins 1977, 1980). Simply stated, shamanism is a means by which order and balance are maintained within a society. The shaman is the focal character in the system and through his inter- cession with the supernatural world, homeostasis is sustained. To achieve homeostasis the shaman (1) enters a trance (frequently induced by plant hallucinogens), (2) trans- forms into his animal familiar (usually the jaguar), and (3) then flies upward to the spirit world where he intercedes with the supernaturals residing there. The hypothesis proposed for the textiles’ presence on the south coast is that they served as a medium of a proselytizing movement from a Chavin central locus. Further- more, Chavin religious iconography has been shown to consistently include hallucinogenic references. Curls emanating from the nostrils of supernatural images seems to represent the mucus discharge which results from snuffing hallucinogenic powder (Chagnon 1968: TABLE 1. yuh Suns frequency table of motifs on The Shamanism Textile in present, fragmentary stat Element Number of appearances Stale an ai sia oe os eae a oe Wks HK 1, very fragmentary CE SOE es he a ee oe A ee ks 1 section above Staff God Tee OTS sis ae ee a ed ero a ea a ve heen 1 pavtial plant Gini Peale eed si eds Sec bek boss 1, less than half remaining mugebery ame as od oa ee) iw eee ee ek 2, both in bloom; 1 fragmentary Geet, WHERE: 5 ois ne ee ee Oe a a wade ose ee weet. ASEM CN Oo Se dk oko ee oe Pa we ae 4, all fragmentary SUATIMNS os <6 6 kee eR Re Ne ek oe ee RE tw ee Rees San Pedro cactus © oo 8 hoe a 6: 4 in bloom, 1 not, 1 unclear acacia seed pods .. os ks be oe 6: 5 held by shamans, 1 fragmentary hummingbirds. 6.6.5 «0:6 #3's% 6OW OO we ee eG, Boer SS ee floating circles Sn ae = Er SE egg, December 1982 JOURNAL OF ETHNOBIOLOGY 146 ee FIG. 2—Map of Peru locating Carhua, oe site of textile cache, and Chavin de Huantar, site of the Chavin art style’s definitio 5; Cordy-Collins 1980). Chavin artists’ use of hallucinogenic snuff has also been sug- gested by Donald Lathrap (1973:96). In addition to psychoactive snuff, the hallucino- genic San Pedro cactus has also been identified in Chavin religious art (Cordy-Collins 1977:360; Lumbreras 1977:23; Sharon and Donnan 1977:377-379). Therefore, any attempts to decipher Chavin iconography should test for hallucinogenic and shamanistic references. A refinement of the aforementioned general hypothesis accounting for the Presence of Chavin textiles on the south coast, specifically aimed at the iconography of the textile under discussion here, argues that all the motifs on the textile in Figure 1 refer directly to the general proselytizing message: plant hallucinogen-based shamanism sup- 147 CORDY-COLLINS Vol. 2, No. 2 ported a deity new to Chavin religion. The following discussion will concern itself with the possible representation of (1) hallucinogenic plants, (2) a shamanistic complex and, (3) the new Chavin deity. Analysis of The Shamanism Textile’s Components To unravel the textile’s message, ethnographic studies of South American shamanism made over the last century are invaluable. Shamans’ transforming agents in South Amer- ica today are commonly plant hallucinogens, and the most readily identifiable design on the textile appears to be one of these, the San Pedro cactus (Trichocereus pachanoi) (Fig. 3). A columnar, ribbed plant, San Pedro is used by coastal Peruvian shamans today to achieve a trance state whereby the supernatural world is opened to them; San Pedro’s active alkaloid is mescaline. Traditionally, shamanic curing sessions employing the cac- tus occur at night when the flower blooms. Apparently the act of blooming is particular- ly important because the language of the curing session makes continued use of the blooming metaphor (Sharon 1978:107). Therefore, it seems especially significant that in four of the five cases where identification is possible, the painted San Pedros are in bloom. It is pertinent to note that the proposed San Pedro cacti, as represented on the textile, have no more than four ribs. This is contradictory to fact: T. pachanoi has between six and eight ribs. However, modern Peruvian shamans believe that four-ribbed San Pedro cacti do exist and are especially potent because of the four ribs (Sharon and Donnan 1977:376). The number four is a magical, ritual one in modern shamanism. Therefore, it is entirely possible that chang San Pedros are entirely mythical. This is an important point to which I shall re The second motif on The Shae Textile which may be a plant hallucinogen appears as linear clusters of circular elements. It is possible that this motif represents Anadenanthera peregrina (acacia) seed pods (Fig. 4). A. peregrina has a documented use —————_$ =$ P)\_p_—_ : od 7 ee wart ee (L) Speg ING 7 > eo jag? LO ~ es “er ee \cawfling, FIG. 3—San Pedro cactus growing at Chavin FIG. 4—Anadenanthera peregrina (after de Huantar, Peru. (Jack L. Riesland photo) Schultes 1976). —— \ December 1982 JOURNAL OF ETHNOBIOLOGY 148 in South America extending back to 1496 (Schultes and Hofmann 1979:116). As sham- anic trance-inducing agents, the seeds are removed from the pods and ground into powder for snuffing or, in some cases, it is reported that the seeds were simply chewed. Whereas actual A. peregrina pods average about 20 cm in length, the motifs on the textile appear to be much larger relative to the individuals who seem to be holding them. However, in defense of the argument, the size of the hummingbirds relative to everything else indi- cates that true scale was not a particular concern of the Chavin artist. Yet, there is a second objective which might be leveled against interpreting the proposed plant as A. peregrina. Actual acacia pods split open along their sides rather than interdigitally between the seeds as seems to be indicated on the textile. Nonetheless, some A. peregrina Pods evidence severe constriction on the pod between iridividual seeds (cf., Schultes and Hofmann 1979:117, lower photograph). Finally, though, the point must be made that the geographical zone occupied by the Chavin artist who painted the textile was not that where the psychoactive acacia grows. spracerneanige! peregrina is apparently native to the tropical lowlands drained by the Orinoco River. Nevertheless, it is documented that there was extensive trade of the drug into the ee (Schultes and Hofmann 1979: 117). What makes the discussion of the motif particularly intriguing is that we cannot be certain of the form the drug was in when traded: whether in whole ict individual seeds, or as ground powder. If it ; was brought into the highlands in either of the latter two forms, then we would not expect the Chavin artist to necessarily be aware of its appearance while growing, except by heresay from the traders. Though the identifi- 3 cation as A. peregrina is tentative, ae it should be noted that some of ae the animals on the textile appear with muzzle emanations which could well represent nasal dis- charge which results from inhala- tion of hallucinogenic snuff. The third proposed plant on The Shamanism Textile presents a conundrum. Only two images of the motif appear, one complete depicts cotton plants (Cordy- FIG. hed cai of | The i leaiiaie Textile ek ae Collins 1979: Figs. 3, 7-9). the complete representation of the mystery plant Furthermore, a flower-like ele- (Alana Cordy-Collins photo). 149 CORDY-COLLINS Vol. 2, No. 2 ment appears atop the motif in question. Therefore, I think it likely that the mysterious image is that of a plant. But what plant? Given the context, one might expect it would be a plant with hallucinogenic properties. One of the problems in identification is that of proper scale, a problem discussed in reference to the supposed acacia pods. Another problem is the uncertainty of the viewer’s vantage point; Chavin art occasionally makes use of simultaneous Picasso-like views. One more difficulty is the abstract quality of the painting itself; and, finally, many hallucinogenic plants used by shamans today are still unclassified. The plant’s diagonal lines are intriguing; they could be gashes to allow for the draining and collecting of sap as with rubber trees. If this is the meaning of the lines, perhaps the plant can be compared with the virola tree (Virola spp.), the resinous sap of which is gathered by contemporary Amazonians to be employed as a snuff ingredient. However, currently shamans obtain virola resin by scraping the inner bark, not by gashing the tree. alternative interpretation is that the mystery plant is really a flower, since the upper portion is very flower-like. Finally, two more possibilities must be entertained: either the plant might by mythical as the four-ribbed San Pedro cactus discussed earlier seems to be, or it might simply have been unknown to the Chavin artists in its live, grow- ing form, as has been suggested for A. peregrina. Here too, it could have been that the processed hallucinogenic substance alone was imported by Chavin people. Therefore, they would have had to depend on foreign descriptions or on their own imaginations to create a visual image of the plant. Taken by themselves, the three proposed plants may not seem to make a particularly strong case for a definition of the painted Chavin cloth as ‘The Shamanism Textile. = However, like any archaeological data, these three iconographic motifs must be studied in their context in order to arrive at a meaningful interpretation. As stated at the beginning of this discussion, shamanism is a complex, the trance-in- ducing agents being but one element in the system. Other elements include the shaman, his animal familiars, spirits and supernaturals from the netherworld. It can be demon- strated that most, if not all, these elements are present on the textile, thus providing both a context for the identification of trance-inducing a. plants—and for the overall interpretation of the painted cloth as The Shamanism Texti The figures associated with the presumed acacia pods may be shamans. While the bodies of these individuals are anthropomorphic, their faces are decidedly zoomorphic (compare these faces with those of the deer and feline below). This could indicate that these individuals are shamans in the process of transformation. Composite faces in Chavin ceramic art seem to represent states of transformation (Figs. 6, 7). Alternatively, these creatures may be spirits, specifically spirits of the acacia. Among the Waika Indians of southern Venezuela and Brazil who regularly make extensive use of A. peregrina snuff, spirits called Hekula are believed to communicate with them during their snuff-induced ecstatic trances (Schultes and Hofmann 1979:118-119). According to widely-held Amazonian beliefs, once the transforming agent is ingested, the shaman is no longer human in form, but j jaguarian. One spotted feline appears wi his = resting on an opened cactus bloom. Emanations exude from his muzzle. I have previ uggested that such emanations in Chavin art refer to the use of hallucinogenic eae (Cordy-Collins 1980). Therefore, a reasonable interpretation of the motifs’ juxta- Position is Boe: the jaguar is the transformed shaman, intimately associated with the trans- forming media The > hens are also appropriate inclusions because, due to their ability to draw nectar from flowers by sucking, the birds are equated with the shaman who, in curing, sucks the pathogens from his patients’ bodies. Here, however, the bird/shaman association extends even further: in eight instances on The Shamanism Textile the hum- mingbirds are depicted with their beaks abutting the cacti as if to draw out the trans- forming juices. Furthermore, invariably all birds in a shamanic context symbolize the shaman’s magical flight to the realm of the deities. December 1982 JOURNAL OF ETHNOBIOLOGY 150 One deity, even though only partially extant, can be precisely identified. This is the Staff God, the new Chavin supernatural around whom revolved the proselytizing move- ment which brought the textile cache to the south coast. The deity has been somewhat reconstructed in Fig.8. Comparable Staff Gods evidence the same headdress form, an inverted aca fanged mouth with serpents (Figs. 9, 10). € apparently very old shamanic symbols, extending back at least to the Upper Paleolithic 4 period in Europe (Furst 1976). Today in Peru the deer acts as a metaphor for e swiftness and elusiveness of the shaman. The wing on its back reinforces the shaman’s power of flight. Additionally, at least one of the deer on the textile is shown with muzzle emanations, probably relating to the use of hallucinogenic snuff. The deer’s association with hallucinogens is corroborated by a Chavin ceramic bottle showing the deer in direct association with a San Pedro cactus (Cordy-Collins 1976: Fig. 110). Therefore, it can be seen that, even without the identification of the three proposed hallucinogenic plants, the message this textile carries is shamanism as a complex of inter- related elements. Yet, because shamanism as it is known today throughout South Amer- ica and as it has been documented since the 15th century, has made consistent use of plant hallucinogens to achieve the desired state of ecstacy and communication with the spirit world, it is reasonable to assume that such plants also played an important part in the shamanism of South America’s prehistoric past. It is reasonable to suggest that the three otherwise unidentified motifs on The Shamanism Textile are meant to represent plant hallucinogens. CONCLUSION A Chavin painted textile from Precolumbian Peru has been, by means of icono- graphic analysis and ethnographic analogy, shown to be a very important document. FIG. 6—Chavin stirrup spout bottle show- FIG. 7—Chavin stirrup spout bottle show- ing a face in process of transformation. ing a face in process of transformation. (Junius B. Bird photo) Ja ck L. Riesland photo) 151 CORDY-COLLINS Vol. 2, No. 2 FIG. 8 —Partially reconstructed Staff God face from The Shamanism Textile. (Alana Cordy-Collins drawing) FIG. 9—Staff God on Chavin textile show- FIG. 10—Staff God on border of Chavin ing same headdress type as on The Shama-_ textile showing same headdress as on The nism Textile. (F.E. Landman photo) Shamanism Textile. (Alana Cordy-Collins photo) cenit aan i ge E t December 1982 JOURNAL OF ETHNOBIOLOGY 152 Not only does it corroborate earlier independent interpretations that Chavin society’s religion was shamanistic, but it suggests crip plants used in its shamanistic practices. b Of the three plants shown, one seems to Pedro cactus, another may be acacia seed pods, but the third image eludes deniaton Nevertheless, because the cactus and outh Americ acacia are both hallucinogens used in an shamanism today, it seems ex- While tremely likely that this image with ce features may be similarly interpreted. it may be that future research will reveal the exact nature of the mysterious motif, it should be born in mind that Chavin art is highly conventionalized with a strong mytho- logical component. That the Chavin artists consistently chose to represent San Pedro cactus with four ribs instead of the actual six to eight suggests SS concerns were in a sense more real than the everyday world about them. Therefore, it would be a mistake to look upon Chavin botanical representations as mirrors of the B.C. actual ecology in Peru in the second millenium NOTES 1 Because the textiles were not excavated arch- aeologically, the locale of their discovery—sup- Posedly Carhua—is undocumented. However, because conditions for optimum preservation of organic materials exist on the south coast, it is se probable that the textiles were discovered ar Carhua, if not precisely ther mpceniss Britton and Rose (1920) report San Pedro growing at elevations from 2000 to 3000 m, Sharon and Donnan (1977:375) have repeatedly observed it growing at sea level along € north coast of Peru. 3 “Smaller ic appendages [in Chavin art] are usually compared to snakes, and in this way they usually i fee directly from the body...” (Rowe 1967: :79). LITERATURE BRITTON, N.L. and J.N. ROSE. 1920. The Carnegie Inst., Washington, D.C. CHAGNON, NAPOLEON. 1968. Yanomamo: the Fierce People. Holt, Rinehart and Wins The revolu- the Early Columbian catechism. Univ. Microfilms, Ann Arbor, Michigan. _— 77. Chavin Art: Its Sha- a eee Origins. Jn Pre-Colum- Hist an Art : Selected Readings ( age cg uaa Jean Stern, eds.). Peek Palo Al nication LTO, Cotton and the Staff God: Analysis of an Ancient Chavin Textile. 4 While there might be some objection to iden- types. For instance, Theodor Koch-Grinberg reported during 1917-1928 that the Taulipang shamans of Venezuela refered to themselves as Risk black jaguar . . .the tapir jaguar ... the ma jagu the multi-colored jaguar . (cf, Furst 1968: 158). These animals, eee outward appearances are a were, none- theless, all categorized as jaguars CITED In The Junius B. Bird Pre-Columbian Textile Conference (Ann Poll i th Benson, and The Textile Mus. and Dumbart tees for Harvard Univ., aan 1980. An nitintie ome of pallnciiakoey experience. The Masterkey ae FURST, PETER T. 1968. The Olmec Were- the Olmec (Elizabeth P. Benson, ed.). barton Oaks > tages for Harvard Washington, D FURST, PETER T. 1976. Hallucinogens and Culture. Chandler and Sharp, San Francisco. LATHRAP, DONALD W. 1973. Some Thoughts on the nneiaet Basis of Chavin. in Variation in Anthropology (Donald W. Lathrap and Jody at eds.). Tllinois Archaeol. Survey, Urbana. 153 CORDY-COLLINS Vol. 2, No. 2 LITERATURE CITED (continued) LUMBRERAS, LUIS GUILLERMO. 1977. Excavaciones en el Templo Antiguo de Chavin (Sector R); informe de la sexta campafia. Nawpa Pacha 15:1-38. ROWE, tose acta 1967. Form and i . In Peruvian Archae PO Selected eee (John H. Rowe woe Dorothy Menzel, eds.). Peek Publ., Palo Alto. SCHULTES, RICHARD EVANS. 1976. Hal- lucinogenic Plants. Golden Press, New York. and ALBERT HOFMANN. 1979, Plants of the Gods: Origins of Hallu- cinogenic Use. McGraw-Hill, New York. noarchaeological continuity in Peru. Archae- ology 30(6):374-381. SHARON, DOUGLAS G. 1978. The Wizard of the Four Winds: A Shaman’s Story. The Free Press, New York. J. Ethnobiol. 2(2): 154-161 December 1982 OOLIGAN GREASE: A NUTRITIOUS FAT USED BY NATIVE PEOPLE OF COASTAL BRITISH COLUMBIA HARRIET V. KUHNLEIN Division of Human Nutrition, University of British Columbia Vancouver, British Columbia V6T 1W5 ALVIN C. CHAN Department of Biochemistry, etapa of Ottawa Ottawa, Ontario KIN 9B4 J. NEVILLE THOMPSON Nutrition Research Division, Health and Welfare Canada ttawa, Ontario K1A OL2 SHURYO NAKAI Department of Food Science, University of British Columbia Vancouver, British Columbia V6T 1 ABSTRACT.—Marine fat, derived from several sources, ste formerly used to great extent by Northwest Coast Indians as a — enhancer of many foods as well as for medicinal and ceremonial purposes. The most prominent pees fat used by British Columbia native people has been from om ooligan ( teetedhss pacificus Richardson, Osmeridae) a small fish which is harvested in bulk in early spring, allowed to ripen = large bins, and then rendered to give a pungent, golden, thick oil called ‘“‘ooligan grease”. preparations of this fat were made in 1981 from the Nuxalk naaepereed of Bella Coola, B.C. and several nutrient analyses were done. Fatty acids, and range o percent methyl esters were: saturated at 32.2 (30-33), mn at 64.5 (63-66), and polyunsaturated at 0.9 (0.8-1.1). The principle fatty acid was oleic acid. In addition, there are ree amounts of Kjeldahl N (16-19 ug/g), Ca 27-206 ug/g) and P(70-100 ug/g). Analyses with high pressure liquid chromatography yielded the following mean and values for a fat soluble vitamins: vitamin A-20 ug/g (18-29); vitamin rE. 220 ug/g (148- 279); vitamin K - 10 ug/g (4-13). It is concluded that ooligan grease is a nutrient-rich food fat that is currently consumed much less frequently than it was formerly. INTRODUCTION Food fat derived from several marine species was used to a great extent in the past by Northwest Coast Indians. Uses of fat from seal and porpoise (Suttles 1951), whale (Druc- ker 1951) and salmon (K’san, People of 1981) have been described; however, the most prominent food fat for British Columbia native people is recognized as that rendered from the small fish, Thaleichthys pacificus (Macnair 1971). This fish is commonly noted as the ooligan, eulachon, eulachen, olachen, olachon, or oolachan, depending on the Pronunciation of the various B.C. native groups, and the interpretation of the writer. In this paper, the spelling “ooligan” is used, since this closely approximates the term commonly used by the Nuxalk People of Bella Coola who graciously provided samples of “ooligan grease” for the analyses reported here. Thaleichthys pacificus grows to a maximum length of 30 cm, and returns in early spring to fresh water rivers for spawning after spending 2-3 years at sea. The important rivers for harvesting these fish are the Stikine, the Nass, the Skeena, the Kitimat, the Bella Coola, the Kingcome, the Klinaklini, and the Fraser (Macnair 1971). In addition to use of the fish as a source of rendered fat, native people have used the ooligan as a popular flesh food. The fish is eaten fresh after one of several possible methods of pre- paration (boiling, baking, grilling, etc.), and it can also be preserved by smoking, drying, salting or freezing—or in combinations of these processes. 155 KUHNLEIN ET AL. Val. 2, Na, 2 The fat rendered from T. pacificus is commonly called “grease”. It is widely used as a condiment with many foods, such as dried fish, potatoes, native root foods and vegetables (McIlwraith 1948; Kuhnlein et al. 1982; Niblack 1970; Rohner 1967; Turner 1975). It is also used as an ingredient in the preparation of bread, stews or salads (Haw- thorne et al. 1960; Edwards 1978). In addition, ooligan grease was formerly used exten- sively as a preservative in that cakes of dried berries were submerged in boxes containing the fat, thus protecting the fruit from oxidation and pests. As well as being a prominent traditional food, ooligan grease is used as a native medicine for skin rashes or for the treatment of various internal ailments (McGregor 1981; Garfield and Wingate 1966; Edwards 1978). It was also used as an all-purpose lubricant for wood and leather items (Edwards 1978). The cultural significance of ooligan grease cannot be underestimated, as it was (and continues to be) a prominent food and gift during feasts and potlatch ceremonies. Early ethnographers among the Nuxalk and Kwakiutl people noted that it was a sign of poverty for a family to be without ooligan grease (MclIlwraith 1948; Curtis 1915). There is documentation for preparation and use of ooligan grease by the Tlingit (Oberg 1973), Tsimshian (Stewart 1975; Boas 1916; Garfield and Wingate 1966) Kwa- kiutl (Macnair 1971; Curtis 1915; Rohner 1967), Kitimat Haisla (Hawthorne et al. 1960), Gitksan (K’san, People of 1981) and Coast Salish (Barnett 1955). This list is not a com- plete one, since many native groups obtained ooligan grease by travelling to the rivers to help residents with fish harvest and preparation, or by trade The process of preparing the rendered fat, and the consequent flavor of the final product, differ among the groups of native people. Usually, the fish are “ripened” in bulk to develop the flavor and to permit decomposition of the carcas for easier release of the fat during cooking. The exact changes in the fish carcass due to microbial action are not known. Ethnographic accounts describe cooking the fish in water heated with hot rocks or over an open fire using a bin with a metal bottom. In addition to procedural variations among native groups, there are preferences and opinions of families on how to make the best ooligan grease. PREPARATION OF OOLIGAN GREASE BY THE NUXALK PEOPLE The people of the Nuxalk community in Bella Coola, British Columbia, age ooligan grease from fish entering the Bella Coola River in early spring. The time of arrival of the ooligans ultimately depends on weather conditions, but usually the first fish appear in late March. In early April, when the seagulls hover over the river which is “black with ooligans,” the fish are seine-netted into boats and hauled in buckets onto shore and packed into bins built on the river bank. The size of the fish bins, called “stink boxes” by the Nuxalk people, varies. An average-sized box will hold 6300 kg of ooligans, and is constructed from cedar planks to be 2-3 m square and 1.5 m high. The floor of the bin is customarily lined with cedar (Thuja plicata Donn) boughs to permit adequate drainage. When full, the box is covered, and the contents left to ripen. De- pending on the weather, this process will take from 4-14 days. Each family has their own way of telling when the ripening is complete—either by smell, or by feeling the texture of the decomposing fish. Some Nuxalk people preferred, in the past, to use only the fattier, female fish for the grease-making process. Today it is the custom to use the entire catch. Usually, it is the men who net the fish, and haul the catch into bins. At the time of cooking the grease, the whole family spends the day, or the weekend, at the riverbank, with women supervising the cooking process. If the fish are properly ripened, a 6300 kg bin will yield around 380 | of prepare’ grease. The ripened fish are placed into boiling water contained in a cooking box wi metal bottom that is placed on supports over an open fire. The contents are slowly sim- mered to extract the fish oil. It is essential to cook slowly, to prevent boiling and froth- December 1982 JOURNAL OF ETHNOBIOLOGY 156 ing, so that the oil from the fish will “melt away” and rise to the surface of the water layer. When cooking is complete more water is carefully added to the bin to make a dis- tinct water/oil interface. The oil is then scooped off into metal pots and the fish residue is either released to the river via a trough from the cooking box, or taken for garden er The oil is then reheated to the frothing point to skim off any particulates. This is semen cae tin by some families with the addition of red-hot stones to the metal pot. families who use this method say they do it to get the “hot rock flavor” they like so nea Others reheat the grease slowly on a portable kerosene stove to complete the skimming process. Still others will reheat it two or three times on a portable stove to ensure that the grease is “safe” and will not get “strong tasting” during storage. When the final cooking is finished, the resulting fat is a golden, Peat thick oil which is poured into gallon jugs and is usually stored in a cool part of the home. A few families also store grease in a freezer. Everyone maintains that grease ee is properly cooked and bottled will keep for several years at room en ial Those who store it frozen say they do so to keep the flavor from getting “stro: An interview study of Nuxalk families (Kuhnlein 1981), revealed that slightly more than 50% of families still use ooligan grease to some degree. The quantity used per family varied from 7 to 38 1 of grease per year. Although there were five fermenting bins yielding upwards of 2000 1 of grease for the village in 1981, this was distributed within the community of about 600 people. Nuxalk elders still use ooligan grease as a medicine, and it still has a prominent role in cultural activities. The following foods have been observed (by author HK) to be eaten with ooligan grease in the Nuxalk community: dried fish, smoked and cooked fish, potato, herring roe, salmon roe, seaweed (Porphyra perforata Agardh.), bannock, and homemade bread. Spring greens eaten with ooligan grease include the shoots of salmonberry (Rubus spectabilis Pursh), thimbleberry (Rubus parviflorus Nutt.) and cow parsnip (Heracleum lanatum Michx Today, marketed fats such as lard, hydrogenated fat, corn oil and margarine are com- monly used by native people, and these have replaced the use of ooligan grease as a regular meal-time food. An essential step in the documentation of nutritional conse- quences of this adaptation is the identification of the nutrient components of ooligan grease. Although ooligan grease has always been considered as a generally healthful food by native people, no reports of its nutrient composition have yet appeared in the scienti- fic literaure. METHODS Samples of ooligan grease were taken in 1981 from five different preparations made in the Nuxalk community of Bella Coola, British Columbia. The samples were poured from the family container into acid-washed 200 ml teflon bottles which were then frozen and stored at -15° C until later analysis. Analyses were completed for proximate compo- sition, minerals, fatty acids, and vitamins A, E and K. Proximate composition was assessed with standard techniques. Moisture was deter- mined in Senne . bec at 70° C overnight and then to constant weight at 60° C in a vacuum o As terminations were made on the dried samples at 550° C in a muffle fen Total ee were determined with method 16.052 of the A.O.A.C. Methods Manual (1970). Total Kjeldahl nitrogen was determined with standard proce- dures (McQuaker 1976). Mineral elements were assessed with a nitric-perchloric acid digest of the sample on an me coupled plasma-atomic emission spectrometer (McQuaker et al. 1979a, 1979b). rminations were made for Al, As, Ba, Ca, Cd, Co, Cr, Cu, Fe, Mg, Mn, Mo, Ni, P, Pb, rt te Ti, V and Zn. Fatty acids were determined as percent methyl esters using standard gas chromato- graph procedures after the samples were saponified and methylated with boron trifluo- ride. 157 KUHNLEIN ET AL. Vol. 2, No. 2 Vitamin A was assayed fluorimetrically as retinol according to the method of Thompson et al. (1978) using 100 mg of oil, or with high pressure liquid chromato- graphy (HPLC) on a silica column. Vitamin K was determined with HPLC with a c18 reverse phase column (Vydac 201 TP 0.32 x 25 cm) installed in a Spectra-Physics HPLC system (SP 8700 solvent delivery system, SP 8400 UV/vis detector, SP 4100 computing integrator), Vitamin Kj (Sigma V-3501) was used as the standard. Duplicate analyses within a laboratory were within 10.5% of each other. Cholesterol was detectable at <5 ng/mg oil using a mobile phase of 1:1 acetonitrile- isopropanol and reading at 200 nm. Vitamin E was measured with HPLC using the method of Thompson and Hatina (1979). A spectrofluorometer set at 290 nm excitation and 330 nm emission was used as a detector. RESULTS AND DISCUSSION The results of analyses of ooligan grease for proximate composition, fatty acids, calcium and phosphorus are given in Table 1. As anticipated, the primary energy com- TABLE 1. Composition of ooligan grease, Component Mean* Range Fat, % >99 co Moisture, % 0.16 0.09-0.24 Ash, % < 0.02 0.005-0.02 N (Kjeldahl), ug/g 18 16-21 Ca, ug/g 68 27-206 P, ug/g 85 70-100 RE en a aes ee % methyl esters — — — — — — 14:0 6.4 5.7-7.2 14:1 0.3 0.3-0.4 16:0 17.9 17.0-18.3 16:1 7a 6.3-8.0 18:0 4.1 3.9-4.4 18:1 54.6 52.2-57.8 18:2 0.8 0.7-1.0 18:3 0.1 0.1-0.1 20:0 0:3 0.3-0.4 20:1 1:2 1.2-1.2 20:5 or 22:1 i3 1.3-1.5 4Mean of 5 samples Eipihlintia, |S sia tcinih | December 1982 JOURNAL OF ETHNOBIOLOGY 158 ponent was fat with less than 1%, by weight, of the oil being the combined components of moisture, ash, and nitrogen. Extrapolating these small amounts of nitrogen to protein with a factor of 6.25, there was less than 0.01% protein present. Carbohydrate content, determined by difference (the sum of fat, moisture, protein, and ash subtracted from 100) was negligible. Calcium and phosphorus were the major minerals found, although at these levels, (6.8 mg/100 g calcium and 8.5 mg/100 g phosphorus) ooligan grease would make a very small contribution to daily dietary needs, even if consumed in large quanti- ties. Of the minerals determined, iron, magnesium and zinc were the only other minerals present. These occurred inconsistently in the samples and in small quantities (5-10 ug/g) and also would not have contributed significantly to dietary needs. etermination of the fatty acid content revealed that the primary fatty acid was monounsaturated oleic acid (18:1). There was a mean content of 54.6% of this fatty acid. The second most prominent was the saturated fat, palmitic acid (16:0), with a mean content of about 18%. There were small amounts of polyunsaturated acids (18:2 and 18:3), so that the ratio of polyunsaturates to saturates was very low. However, the total unsaturated fat content is approximately 65%, which exceeds that of commonly used animal fats, such as butter, lard, beef fat and mutton fat. It is similar in unsaturated fat content to that of poultry fat (Brignoli et al. 1976: Reeves and Weihrauch 1979). The resolution of the longest chain acids was not complete, so that it was not known whether this small component (1.3-1.5%) was eicosapentaenoic acid (20:5) or the longer chained cetoleic acid (22:1), or a combination of these two. Eicosapentaenoic acid has been identified in diets containing fish oils which are consumed by Greenland Eskimos. The fatty acid, at 2.6% of fatty acids in the total diet, is though to contribute to alow incidence of ischemic heart disease in these people (Bang et al. 1980). The results of the analysis for three fat-soluble vitamins are given in Table 2. The samples are rich in vitamin A. At the mean level of the samples reported here, it would take less than 50 g (2.5-3.5 tablespoons) to provide the adult daily need for vitamin A, which is 800-1000 retinol equivalents, or ug of retinol. It is often thought that the diets of native people are low in vitamin A if they do not consume carotene-rich vegetables TABLE 2. Three fat-soluble vitamins in ooligan grease Sample Vitamin A? Vitamin E> - Vitamin K® See Nan ic iota eee 1 18.0, 18.7 232 7.5 2 29.2, 26.5 279 12.5 3 18.2, 18.3 148 13.5 4 16.5, 10.7 183 4.0 5 23.0, 19.5 259 11.0 Mean 21.0, 18.7 220 9.7 “Vitamin A was determined as retinol in two laboratories >Vitamin E was determined as @-tocopherol and vitamin K as Ky (see text) 159 KUHNLEIN ET AL. Vol. 2, No. 2 and fruits or vitamin A containing fats such as butter or fortified margarine. It is clear from the data presented here that the use of ooligan grease as food or medicine, even if irregularly consumed, could provide most of the retinol needed. Since vitamin A is ef- ficiently stored hepatically (Food and Nutrition Board 1980), there is every reason to believe that vitamin A deficiency would be unlikely to occur among people who regu- larly use ooligan grease. The vitamin E levels found in these samples (148-279 ug/g) are considerably higher than those reported for fats from several uncooked finfish (McLaughlin and Weihrauch 1979). Since the entire body of the ooligan is used in the preparation process, higher amounts of tocopherol anticipated to be contained in the fish liver and other organs would also contribute to the tocopherol content of the final product. At a mean level of 22 mg of @tocopherol per 100 g of oil, it would take less than 3 tablespoons to meet dietary standards for adults. Ethnographic accounts indicate that one cup or more was consumed medicinally or at special ceremonies. Among those families who consume ooligan grease in Bella Coola today, it is the usual practice to have a few tablespoons at one meal during the day. Although vitamin E deficiency has been documented in B.C. Indians (Desai and Lee 1974), greater deficiency among inland-resi- dents than coastal residents was reported. Our results indicate use of ooligan grease would be protective against vitamin E deficiency. There is no dietary standard for vitamin K, although it is known that this nutrient is essential for efficient blood coagulation in humans (Food and Nutrition Board 1980). It is thought that most of that needed by humans is synthesized by intestinal bacteria; min K daily, primarily from plant sources (Olson 1973). Therefore, eg vitamin Kj reported here for ooligan grease (4-13.5 ug/g) would have a very minor influence on nutritional status for this nutrient. Vitamin Kj, phylloquinone, is aie found in plants, but is also stored in the liver of animals who consume plants. The origin of this nutrient in the ooligans is probably from sea algae, and it is rendered into the final pro- duct having been present in the liver of the fish or in intestinal algal residues. itamin D content of these samples of ooligan grease has yet to be confirmed. To date, this vitamin was undetectable in the samples using the method of Thompson et al. (1982). The composition of ooligan oil in comparison to other fats which are commonly used by native people in the Nuxalk community is given in Table 3. The saturated fats of TABLE 3.—Composition of ooligan grease in comparison to other fats (per 100 g)° ooligan pork corn margarine grease lard oil Saturated Fat, % 32.5 39.2 12.7 13.2 Monounsaturated Fat, % 64.5 45.1 24.2 45.8 Polyunsaturated Fat, % 0.9 112 58.7 18.0 Vitamin A, RE 1,985 me or 993> Vitamin E as a-tocopherol, mg 22.0 12 14,2 12.9 Vitamin K as K,, mg 1.0 — oe — *values for marketed fats are primarily from Reeves and Weihrauch (1979). >from Health and Welfare Canada (1979). December 1982 JOURNAL OF ETHNOBIOLOGY 160 ooligan oil are similar to lard and higher than that p t in corn oil and corn oil marga- rine. The total unsaturated fat, that is the eoeibied monounsaturated and polyunsatur- ated fats, of ooligan grease is similar to that of corn oil. There is no doubt about the heey of ooligan grease in providing vitamin A, E, and K in comparison to the other ree Today native families often use marketed fats such as lard and margarine in food prepa- ration, and it is reasonable to assume that these have replaced ooligan grease as the major dietary fat. This is further implicated by the extent of village ooligan grease preparation in comparison to former days. As a stated, in 1981 there were five ripening fish bins yielding in the vicinity of 500 of oil for use by a community of about 600 people. In 1982, there were four ies ripening bins on the north bank of the Bella Coola River. In contrast, the contemporary Nuxalk elders recall that it was usual to have eight or ten different family preparations of ooligan grease each spring, and that everyone would use it as a regular food. Unfortunately, percapita consumption cannot be calcu- lated since population —. were not kept, in addition, it is common knowledge that the grease was a favorite trade it to neighboring groups in exchange for other foods or household shee aaa it is quite clear that current practice is that only about half of the Nuxalk people use ooligan grease today, and that marketed fats are used frequently. In conclusion, ooligan grease is a eee e? food fat that is used less today than it was former It continues to be obtain th ingenious low-cost native technology from local resources. Although the labor aeotivce for fish harvesting and grease prepara- tion is considerable, this food has been the object of much cultural activity and is still highly appreciated by many native people of Coastal British Columbia. ACKNOWLEDGMENTS e would like to thank the Nuxalk families who prepared the ooligan grease and who kindly provided samples for this study: the Siwallace family, the Hans family, the Saunders family, the King family, the Moody family and the Nelson family. The following people have graciously shared pe ane on anges grease, and some have assist- ed with preparation of this manuscript: Margaret, Siwallace, Felicity Walkus, Sandy Moody, Edward Moody, Willie Hans, Sarah Saunders, Alice Tallio, Elsie — Katie Nelson, Rose Hans, Eliza Saun- ders, Andy Siwallace, Clayton Mack and Lillian Siwallace. Appreciation is also cordially extended to Dr. Nancy Turner for discussion of the manuscript. Our gratitude is also extended for technical assistance to Althea Townsend and Linda Jung of the Department of Food Science and to Anthea Kennelly of the Division of Human Nutrition, University of British Columbia. The research was partially funded by support to Harriet Kuhnlein from the Natural Science and Engineering Research Council of Canada (A-7148 LITERATURE CITED ASSOCIATION OF OFFICIAL ANALYTICAL BRIGNOLI, C.A. KINSELLA, J.E. and J.L. MI 1970. Official Methods of WEIRAUCH. 1976. Comprehensive evalu- ation of fa ids in foo dro- Amer, J. Clin, Nutr. 33:2657-2661. Bureau Amer. Ethn. Annu. Rep. 1909-10. U.S, Government Printing Office. genated fats and oils, J. Amer. Dietet. Assoc. 68:224-229. om Sii E.S. 1915. The North American i fo) Johnson Reprint Corp., ae York. peas 1970). DESAI, I.D . LEE. 1974. Vitamin E status of sae ans a Western Canada. Can. J. of Public Health. 65(3):191-196. DRUCKER, P. 1951. The Northern and Cen- 161 KUHNLEIN ET AL. Vol. 2, No. 2 LITERATURE CITED (continued) tral Nootkan Tribes. Bull. 144, Washington, D EDWARDS, G. 1978. anne time in Bella Coola. The Beaver, Autumn Issu: FOOD AND NUTRITION BOARD. av Rec- ommended Dietary Allowances, 9th Ed. Natl. Res. Council, Natl. Acad. Sci. Washington, D.C. GARFIELD, V.E. and P.S. WINGATE. 1966. The Tsimshian Indians and their arts. Univ. “ie Amer. Ethn. attle, HEALTH AND WELFARE CANADA. 1979. Nutrient values of some common foods. Ottawa. HAWTHORNE, H.B., BELSHAW, C.S. and S.M. JAMIESON. 1960. The Indians of British Columbia. A study of contemporary social adjustment. Univ. of Toronto Press. K’SAN, PEOPLE OF 1981. Gathering what Douglas and 1981. Nutrient composi- tion and use of Native Indian traditional food. Project Report 1216-9-169, Health and Wel- fare Summer Canada, Ottawa. KUHNLEIN, H.V., TURNER, N.J. and P.D. KLUCKNER. 1982. Nutritional significance of two important “root” foods (Springbank clover and Pacific silverweed) used by native people on the coast of British Columbi Ecol. of Food and Nutr. 12:89-95. McGREGOR, M. 1981. Native medicine in southeast : Tsimshian, Tlingit, Haida. Alaska Med. 23(6):65-69, parser T.F. 1948. The Bella Coola dians. Vol. 2 Univ. Toronto Press, Toronto. eaves: PJ. and J.L. WEIRAUCH. 1979, Vitamin E content of foods. J. A Dietet. Assoc. 75:647-665 MACNAIR, P.L. 1971. Descriptive notes on the Kwakiutl manufacture of eulachon oil. Syesis 4:169-177. McQUAKER, N.R. 1976. Chemical analysis of waters, waste-waters, sediments and biological materials. 2nd ed. Province of British Colum- ia, Ministry of Environment, Environmental apna Vancouver, McQUAKER, N.R., BROWN, D.F. and P.D. KLUCKNER.. 1979a. Digestion of environ- mental mat for asi by inductively aR plasma-atomic emission spectrom- . Anal. Chem, 51(7): 1082-1084, McQUAKER, N.R., KLUCKNER, P.D. and G.N. CHANG. 1979b. Calibration of an udu coupled aeeregietis emission spectrometer for the analysis of environmen- tal materials. Anal. Chem. 51(7): 888-895. NIBLACK, A.P. 1970. The Coast Indians of Southern Alaska and ong B.C. Johnson eprint Co., Ltd., New York. OBERG, K. 1973. The social economy of the ies Indians. Univ. Washington Press, Seat seas, R.E. 1973. Vitamin K. In Modern nutrition in health and disease. (R.S. Good- hart and M.E. Shils, eds.). Lea and Febiger, shear eer gain REEVES, J.B. oni J.L. WEIHRAUCH. 1979. ilies 8 of foods. Fats and oils. Agric. Handbook No. 8-4. U.S. Dept. of Agricul- ture. png R.P. 1967. The people of the Gil- ord: A contemporary Kwakiutl village. Na Mus. Canada Bull. 225, Anthropol. Ser. o. 83. Ott Pa ai = 1975. The seasonal availa- bility of fish species used by the Coast Tsim- rthern British Columbia. Syesis ee life = the Coast Salish of H Rosario Straits. Unpubl. Ph.D. comin phases ) a Washington. THOMPSON, J.N., ERDODY, P., MAXWELL, W.B. and T.K. MURRAY. 78. The fluo- rometric determination of vitamin A in dairy products. Health and Welfare Canada Amen- ded Bull. VT-5. THOMPSON, J.N. and HATINA, G. 1979. Determination of tocopherols and tocotrie- nols in foods and tissues by high pressure liquid gee J. Liqu. Chromatog. 2:32 PiciNeoey J.N., HATINA, G., MAXWELL, W.B. and S. DUVAL. 1982. High perfor- mance liquid chromatographic determination of vitamin D in fortified milks, margarine and infant formulas. J. Assoc. Off. Anal. Chem. 65:624-631. TURNER, N.J. 1975. Food plants of British Columbia Indians. Part I. Coastal Peoples. B.C. Provincial Museum. J. Ethnobiol. 2(2): 162-176 December 1982 USE OF WILD CHERRY PITS AS FOOD BY THE CALIFORNIA INDIANS JAN TIMBROOK Associate Curator, Anthropology, Santa Barbara Museum of Natural History Santa Barbara, CA 93105 ABSTRACT.—Central and southern California Indians prized the fruit of Holly-Leaved Cherry, Prunus ilicifolia (Nutt.) Walp. [Rosaceae], much more for its seed kernel or pit than for the surrounding fleshy pulp. A survey is made of Prunus pit consumption by tain poisonous hydrocyanic acid, it is shown that native people recognized the danger and, through proper treatment, were able to produce a safe and desirable food. Subsistence, social, economic and ritual roles of Prunus pits in California Indian life are discussed, fol- lowed by comments on changes in recent times. INTRODUCTION genus Prunus, a member of the Rose Family [Rosaceae] , contains a number of our cultivated fruit trees: plum, peach, apricot, almond and cherry. There are several wild shrubs in the same genus which are native to California, and it is not surprising that the fruits of these were eaten by the indigenous peoples. What is surprising is that in some areas the pits were eaten, for they contain potentially dangerous levels of cyanide. This paper discusses these wild Prunus pit foods in California, the methods of dealing with their toxicity, and the roles they played in California Indian life. NATIVE SPECIES AND THEIR USE Prunus in California re are seven species of Prunus indigenous to California. The following brief descriptions, summarized from Munz and Keck (1959), emphasize distribution, plant size and nature of fruit; season of ripening is not given but can be estimated from flowering es, P. emarginata. Bitter Cherry. Mountains throughout California from San Diego County north, elevations up to 2700 m; in Yellow Pine and Red Fir Forests and Chaparral. Shrub to small tree, 1-6 igh. Drupe red, bitter, 6-8 mm diameter. Flowers April to au P. subcordata. Sierra Plum. Coast Ranges from Santa Cruz Mountains north, and Sierra Nevada from Kern and Tulare Counties north to Modoc County, aon below 1800 m; Yellow Pine Forest. Shrub to small tree, 1-6 m high. Fruit 1.5-2 cm long, red-purple, edible. Flowers March to May. P. fremontii. Desert Apricot. Colorado Desert from Palm Springs south to Baja California. Shrub to small tree, 1.5-4 m tall. Fruit yellowish, hairy, 8-14 mm long, dry. Flowers February to March, onii. Desert Peach. East side of the Sierra Nevada from Kern and Inyo Counties north, 1002700 m m ey RH Shrub 1-2 m tall. Fruit 12 mm long, hairy, with thin, dry pulp. Flowers March to f pleat Desert Almond. Mojave and Colorado Deserts at 750-2000 m elevation; variety Punctata occurs in Santa Barbara and San Luis Obispo Counties. Shrub 1-3 m tall. Fruit dry, hairy, 8-12 mm long. Flowers March to May. P. virginiana var. demissa. Western pore Coast Ranges and Sierra Nevada, below 2500 m; Yellow Pine Forest, Ch: Foothill W: and. Shrub or small tree, 1-5 m high. Fruit 5-6 mm across, dark red, bitter but edible, ak late in the season. Black-fruited variety melanocarpa occurs in northernmost California. Flowers May to June. P. licifolia, Holly-Leaved Cherry, Islay. Coast Ranges from Napa County south to Baja Califor- Santa San Clemente Islands. Similar-appearing Catalina Cherry, subsp. /yonii, is 163 TIMBROOK Vol. 2, No.2 restricted to the Channel Islands of Anacapa, Santa Cruz, Santa Rosa, Santa Catalina and San Cle- mente (Smith 1976:164); below 1500 m in Chaparral and Foothill Woodland. Shrub or small tree, 1-8 m tall. Fruit red (rarely yellow), 12-15 mm long, with thin, sweetish pulp. Catalina Cherry fruit is darker and larger, 12-24 mm long. Flowers April to May. Distribution of Prunus Foods Fruit. — The fruit of all these wild Prunus species is characterized by having a large stone and relatively little pulp. What pulp there is is often rather dry. Nonetheless, Indian peoples throughout the state did make use of this fruit pulp, and many are stated to have prized it as a food source. The Cahuilla were fond of Desert Peach (Bean and Saubel 1972:119); the Mendocino County Indians relished Sierra Plum and made long trips to get it since it was not common in their region (Chesnut 1902:356). The species most used for fruit appear to have been Sierra Plum, Western Chokecherry and Islay. it was usually eaten fresh from the trees, but it was sometimes dried and stored for later use (Bean and Saubel 1972:119-121; Chesnut 1902:356; Barrett and Gifford 1933:162; Zigmond 1981:54; Harrington, n.d.). The Luisefo preferred to let the fruit sit for a few days before eating it, to improve the taste (Sparkman 1908:194, 232). Pits. — Some groups, particularly those in central and southern California, preferred to throw the fruit pulp away and eat the pits. To find out which groups made significant use of Prunus pits for food, a survey was made of Culture Element Distribution lists (C.E.D.) compiled for California in the 1930s and 1940s. These lists vary widely in the degree of detail they provide. Usually, plants are merely described as “eaten”, without specifying the exact part used or the method of preparation. A further difficulty is that botanical names are rarely included in this type of anthropological literature, and common names are often incorrectly applied. Nonetheless, C.E.D. lists can still give some indications of traits important in a cultural group. Other ethnographic sources were also checked where they were available. Though Prunus fruit was eaten in many areas, there is virtually no evidence of use of pits in northeastern California (Voegelin 1942; Beals 1933; Gifford and Klimek 1936), central Sierra (Aginsky 1941; Barrett and Gifford 1933; Gayton 1948a, 1948b), southern Sierra (Driver 1937a; Gayton 1948a; 1948b; Voegelin 1938; Zigmond 1981), north- western California (Driver 1937b; Curtin 1957; Foster 1944; Schenck and Gifford 1952), or Round Valley (Essene 1942; Chesnut 1902; Goodrich et al. 1980). Some groups, such as the Maidu (Hill 1972), Shastan (Silver 1978), Northern Paiute (Kelly 1932), Atsugewi (Garth 1953), and Yana (Elsasser 1981), may have made flour of the fruit pulp or mashed whole Prunus fruit, seeds and all, but most of this information is ambiguous. Great Basin and Plateau groups did make a pemmican-like food of mashed Prunus fruit and seeds (Stewart 1940; Ray 1942), and this practice may have extended slightly into northeastern California. In addition, one northern Pomo consultant said “wild plums” — presumably Prunus pits — were stored in grass-lined pits indoors, but this was not mentioned by other Pomo, and no further information was given (Gifford and Kroeber 1937:181). In contrast to the rest of the state where evidence is scanty, in the central coast and southern California every tribe is listed as preparing ‘‘wild plum seed meal” (Harrington 1942; Drucker 1937). This concentration of Prunus pit foods seems rather curious until it is compared with the distribution of the seven Prunus species in California. The practice of making “‘wild plum seed meal” coincides quite closely with the range of Holly-Leaved Cherry or Islay, Prunus ‘ilicifolia, as shown in Figure 1. The name “Is- lay”’ is a Hispanicized version of slay’, the Salinan word for the plant, and is the common name historically used by most Indian people to refer to the plant, the fruit, and the food made from the pits (Harrington 1944:38). This species was used by all peoples within its range south of San Francisco Bay. It is the only one which was consistently specified as being sought for its kernel or pit everywhere it occurs. Although some groups like the Cahuilla (Bean 1972; Bean and Saubel 1972) used other species as well, P. ilicifolia was aii ee) = ST , i 5 TT hieadiniiiiadal December 1982 JOURNAL OF ETHNOBIOLOGY 164 “iy Gy OL}. Northern BiPaiute a Wy tsugewiGWYH Wa a i WAY distribution of Islay qj \ [Prunus ilicifolia] i ee SN i v tS BY groups reported as using ij Prunus pits as food 263 or eS Udtld py i es Gabrie ss Cahuilla A : : wuiiidiiiyy a ZCupenoGyws | G Luis Zz ‘, 2 FZ (0) 100 mi & i tj, a 0 100 km FIG. 1—Comparison of Prunus ilicifolia range with consumption of Prunus spp. pits b California Indian groups. Northern California tribes outside the range of P. ilicifolia may have used P, subcordata and P. virginiana var. demissa. Other species were available but not used. the most important one for pit exploitation. Thus, “wild cherry pit meal” would be a more appropriate designation than ‘‘wild plum seed meal.” Prunus Pit Preparation Methods of preparing foods from the seeds of Prunus will now be described as they were practiced by each tribe within the range of P. tlictfolia. These methods are briefly 165 TIMBROOK Vol. 2, No. 2 TABLE 1.—Patterns of Prunus pit preparation on the central coast and in southern California. Tribe Leached Roasted Mush Cakes* Costanoan (—) + (+) Salinan + Esselen (+) Chumash (—) + Tataviam Kitanemuk + si Serrano + Gabrielino (+) (+) Luiseno + + Cahuilla + (+) + + Cupeno - Ipai + ~ (+) + Tipai + + (+) *“Cakes”’ includes balls, patties, ‘“‘tamales,” ‘‘tortillas”’ + stated as present by at least one source (+) probably present (—) probably not present blank space indicates lack of information summarized in Table 1. No information was encountered for Patwin, Wappo, Pomo or Coast Miwok, all at the northernmost end of the range of Islay. Costanoan. — Seeds of the Islay, or Holly-leaf Cherry (Prunus ilicifolia) were ground to produce a meal that was eaten (Levy 1978b:491). Details are provided in the ethno- graphic notes of John P. Harrington (n.d.), now being prepared for publication by Barbara Bocek of Stanford University. The Islay pits were heated in warm water to remove the clinging fruit pulp, following which they were dried for a while, then opened by hitting with a stone. At this point the two consultants disagreed: one said the shelled pits were further dried and then eaten with no additional preparation; the other said the kernels were put in water, sugar added to sweeten, then roasted overnight in a grass-lined hole. They could then be eaten, or one could “take them to people.” The taste was com- pared to that of beans or chestnuts. No mention was made of the form of the final product—bread, mush, or cakes—or of storage, although the fruit was apparently gathered in some quantity. Isabel Meadows, Harrington’s primary Rumsen consultant, told him that the term “Islay” did not include Western Chokecherry, P. virginiana var. demissa as identified from specimens. She also said that the fruit of this latter species was eaten but that the pits were not (Bocek, pers. comm.). Esselen. — “Cherry stones” were given as an item of the diet (Hester 1978:497). No further details were found, but the method was probably similar to that of the Salinans. Salinan. — As noted, the common name Islay is a Salinan loan word. According to early Spanish explorer Fages, an item of Salinan diet was “‘a fruit like a red plum oF cherry, from the seed or pit of which, with its surrounding substance they make good tamales. They call it yslay, and they eat the little meat which the pit contains” (Priest- ley 1937:59). Mason (1912:121) refers to “chuckberries” being eaten; this may refer to chokecherries, but no preparation method is given . December 1982 JOURNAL OF ETHNOBIOLOGY 166 Chumash, — The diarist of the 1542 Cabrillo voyage recorded that the Indians of the Santa Barbara Channel made “‘tamales” of a white seed the size of maize; these were good food (Bolton 1925:30). When the fruit pulp is removed, the shells of the Islay pits fit this description. Other observers indicated that Islay was still being used as a food as late as the 1890s; they said it was ground, cooked, and made into balls which were esteemed highly (Caballeria y Collell 1892:17-18; Bard 1894:4). While other authors merely mention that the Chumash ate Islay pits (Orr 1943; Grant 1978), Harrington’s unpublished field notes provide considerable detail about the exact preparation methods they followed. These notes form the substance of a major mash ethnobotanical study by the present author, to be published at a later time. Harrington’s principal Venturefio Chumash consultant, Fernando Librado, c mented that the Chumash ate the fruit, but that the kernel was the “really aeatanaes part of the Islay. The fruit was gathered by hand, picking into a large bag hung around the neck. It was piled up until the pulp rotted, then the pulp and skin were rubbed off and the shells cracked open. Another consultant said that the pits were first boiled “until done,” then allowed to sit overnight before cracking. After the kernels were removed, they could be used right away or stored for later use in a big basket in the hou ing was mentioned by only two consultants in the Chumash area, er they Ne on the method used. Fernando Cardenas, a non-Chumash man of Santa Ynez, said the whole kernels were placed in a sack and repeatedly dipped into hot water, then ground into meal (Saunders 1934:59). Harrington’s last Barbaretio Chumash consultant, Mary Yee, said the Islay was leached after it was “mashed” [=ground?] by letting water run through it in a basket in the creek. Neither of these methods is e as that used by the Chumash for leaching acorn meal, which was usually done in a twined tray. None of Harrington’s older consultants, who had been living in an earlier time when Islay was being prepared regularly, mentioned any leaching process. If the Chumash did leach Islay, they oaohaiiy followed the method used by the Kitanemuk, described below To cook, the Islay was boiled for a long time—one person said three aon ae a stone olla. Acorn mush, by contrast, was cooked in baskets with heated stones. After it was cooked, the Islay was mashed like beans using a wooden paddle-like implement, and molded into cakes or balls which were sometimes rolled in pinole flour (Timbrook 1980:277). All of the Chumash consultants mentioned these balls; apparently Islay was eaten only in this form, never as mush. If a person had been gathering Islay a long way from home, the entire preparation process could be done in the field, but it was more usual to do it at home. The finished Islay balls were arranged on a tray ready for eating; they could be kept for a week or more. Chumash consultants compared the taste to beans, as the Costan- oans did, and commented that they liked it very much. Most poco that Islay was usually eaten as an accompaniment to meat such as baked gopher or squirr Of the plant specimens collected by Harrington’s Chumash consultants, all of those labeled as Islay were Prunus ilicifolia (Fig. 2). This is by far the most common species in Chumash territory. Obispenio Chumash Rosario Cooper mentioned gathering two i Chumash used more than one species in the manner described. Cooper may have meant green and ripe Islay fruit, which were segregated by the Kitanemuk (see below). Tataviam. — Very little published information is available about this group. “Berries of Islay (Prunus ilicifolia)’’ have been listed as the fifth most important vegetable food for the Tataviam, after yucca, acorns, sage seeds and juniper berries (King and Blackburn 1978:536). This inference was drawn from Harrington’s (n.d.) notes on the Kitanemuk, Gabrielino, Chumash and San Bernardino Mountain Serrano; presumably the Tataviam were like these neighboring groups in making more use of the Islay pits than of the fruit itself. No information is given on preparation. Kitanemuk, — Once the major source of information is Harrington’s (n.d.) 167 TIMBROOK Vol. 2, No.2 FIG. 2—Specimen of Prunus ilicifolia collected and labeled by Harrington’s Barbarenio Chumash consultant Lucrecia Garcia, ca. 1928. Photo courtesy of the National Anthro- pological Archives, Smithsonian Institution. fieid notes. Tejon area consultants said that only the ripe and sweet islay fruits were chosen for eating fresh, as some were not sweet. Usually they threw the pulp away, si the kernel was the really esteemed part of the Islay [a phrase also used in Harrington Chumash notes]. They would begin picking the Islay when it was still green and keep on until after it ripened, and even pick it up off the ground. They picked the fruit by hand but never beat the tree or hit it with a stick, for it was considered to be delicate. They kept the green and the ripe, including fallen, fruit in separate piles; the shells of the green Islay were used at a later stage of preparation. The fruit was brought home and piled on 4 December 1982 JOURNAL OF ETHNOBIOLOGY 168 swept dirt floor in the house for several va until the pulp rotted. It was then washed off in the creek by rubbing it between the hands. For the initial preparation stage, water was heated in baskets with hot stones and either poured over the Islay or the Islay put into the water. The water was not boiling, and the Islay remained in it only a short time, about ten minutes. Then someone would break open a pit and pinch the kernel to see if it was “done.” The water was poured off and the pits spread in the sun to dry for two or three days. The shells were cracked by rolling on a metate with a stone held horizontally in both hands2 and the hulls removed from the kernels, At this point the shells of the green Islay, which had been kept separate, were burned and the ashes moistened to make a dough which was molded into cakes like soap. The kernels of both the red and green Islay were mixed together after this. Only the green fruit shells were used to make the ash cakes. The shelled kernels were stored in sacks or in big storage baskets to keep them for winter use. The Islay kernels were cooked in an olla with water, which was changed two or three times during the cooking process. One ultant said the kernels were boiled for awhile before the water was poured off, but another said the water was poured off before it oiled. New cold or lukewarm water was added, heated and changed twice when the kernels were old, and three times when they were new because the latter were more bitter. This changing of water was thus both a cooking and a leaching process. Following this leaching stage, hot or cold water was added and the kernels boiled until done. This was said to take a long time, from morning till afternoon, for the ker- nels were very hard. While the pot of Islay was boiling, an amount of the prepared ash cake equal to the last joint of two fingers was added to the pot so the Islay would not be bitter; this was added only to Islay, not to any other food3. One consultant said they added no salt or anything to the Islay, indicating that different cooks had somewhat different recipes. When the Islay was done and the water had almost all cooked away, it was mashed in the cooking olla and molded into little balls the size of biscuits. They were reddish colored like beans. The balls were put into a tray and would keep about three days without souring, They were considered fine to eat with roasted meat. itanemuk consultants knew nothing of there being two kinds of Islay, which had been mentioned by the Obispefto Chumash woman; “when ripe there are many kinds, white, black, purplish, etc., but it is all Islay’ (Harrington, n.d.). It is not known whether Harrington obtained voucher specimens for this area, or whether other species in addition to P. ilicifolia were prepared in a similar way. Serrano. — The Culture Element Distribution indicates that the Serrano made “wild plum seed meal” and that the rag were leached whole (Drucker 1937:9). Further details are no doubt to be found in Harrington’s notes, now being studied by Michael Lerch of the San Bernardino County sa Gabrielino. — Both Gabrielino and Fernandefio are listed as having made wild plum seed meal (Harrington 1942: 8). Johnston (1962:33) notes that “the pits of the wild plum bushes yielded a good seed for grinding into meal. In fact the native fruits were more useful in this fashion than for their pulp, which was often rather sour and At present the species of this “‘wild tice is unknown, although perhaps later investi- gation of Harrington’s notes and specimens will be revealing. Luisevio. — Hollyleaf Cherry or “Islaya” (P. ilicifolia) was one of many types of seeds used by the Luiseno, and was formerly an important article of diet in some parts of their territory (Bean and Shipek 1978:552; Sparkman 1908:194). The pulp of the fruit was eaten, but the kernel was the principal part used. The fruit was spread in the sun until thoroughly dried, when the shells were cracked and the kernels — These were ground into flour which was leached and cooked in ccna the same man- ner as acorn meal: leached in a basket or sand basin with hot or warm water, coo aad a pottery vessel into a mush which was eaten cold a 1908:193-194). Some 169 TIMBROOK Vol. 2, No. 2 Luiseno groups leached the seeds whole (Drucker 1937:9). Chokecherry fruits were eaten, but the pits were apprarently not used (Sparkman 1908:194, 232). Cahuilla. — The kernel of Islay, P. ilicifolia, was used much more than the fruit pulp. These “plums” were gathered in large quantities in August and spread in the sun until the pulp was thoroughly dried. The shells of the pits were then broken open and the kernels extracted. These were crushed in mortars, leached in sand basins, and boiled into mush (Barrows 1900:60-61). At least one Cahuilla group was said to leach the seeds whole, before preparing the meal (Drucker 1937:9). The ground meal was some- times made into a tortilla-like food (Bean and Saubel 1972:120), reminiscent of the “tamales” of the Salinans or the Chumash and Kitanemuk Islay balls. The pits of Western Chokecherry, and possibly also Desert Apricot and Desert Peach, were ground into a meal and prepared in the same way (Bean 1972:43; Bean and Saubel 1972:119-121). This is one of a very few instances where the use of Prunus pits is ieee Ngai to any species other than P. ilicifolia, though all species were used for their fru These ean 4 were gathered by women. The plants are often found near villages and acorn gathering sites. It has been speculated that Prunus fruits may have been sought during the acorn harvest (Bean and Saubel 1972:120), but Islay ripens as much as two months earlier, so this is doubtful. Cupetio. — The Cupefio made “wild plum seed meal” and leached the seeds whole (Drucker 1937:9). They were observed soaking acorns and “plum seeds” in Warner's Hot Springs to leach them (Bean and Saubel 1972: 198). No further information was found. Ipai/Tipai (Dieguewo). — Two species of “plum” and three of “cherry’”’ were gathered (Luomala 1978:600). These may have been, respectively, Desert Peach and Desert Apri- cot [plums] and Chokecherry, Bitter Cherry and Islay [cherries] ; but there is no ethno- graphic documentation of Kumeyaay [=Tipai] use of pits of species other than P. ilici- folia (Hedges 1980:132). Bitter seeds like those of plums were treated like acorns: pounded in a bedrock mortar, sifted, and leached. Seed flours in general were made into mush, cakes, and stews with vegetables (Luomala 1978:600). All Diegueno groups were listed as leaching the seeds of “wild plum” after grinding (Drucker 1937:9). The Santa Ysabel Ipai ate the fresh fruit of P. ilicifolia. The large seed was then cracked, the kernel extracted and pounded in a mortar, and the meal made into patties and roasted (Hedges 1967:34). Leaching is not mentioned here. The roasted patties may be similar to the Cahuilla tortilla-like food made from Islay. The Southern Dieguefio [=Tipai] also used ‘wild plum” seeds, probably Te . Cherry (Hedges 1980:131). These were cracked with a mano and metate. The mea were spread in the sun to dry, then rubbed between the hands and tossed in a nee basket to remove the hulls. After grinding in a rock mortar, they were leached in a basket like acorns, but with only cold water. They were then cooked into a mush in a pot directly on the fire (Spier 1923:334-335). Clan ownership of patches of wild plum trees has been reported in Tipai territory (Spier 1923:307). The Tipai of Baja California continue to use P. ilicifolia seeds even to the present y. The outside pulp is eaten and then the seeds are broken and ground up, and the inner meat is leached [and cooked?] to make a mush. Consultants say it is a harsh food, does not taste very good, and gives them a stomach ache if they have to eat it regularly. But it forms a staple when families are too poor to buy food (Hinton 1975: 217-218). Historical Sources. — It was noted that Spanish explorers described Salinan and Chumash as making “tamales” of a seed which could only be Islay, P. ilicifolia. Natural historian Longinos Martinez wrote in 1792: “The [seeds] most commonly consumed by the gentiles [unconverted Indians] of New California [include] a large seed they call silao which, although it is somewhat bitter, they wash, dry, and roast; it is one of their most important foods” (Simpson 1961:46). Surely this also refers to Islay. ee NL. I = 8