QK | TS I64 JOURNAL (of) ETHNOBIOLOGY An Empirical Assessment of Epazote (Chenopodium ambrosioides L.) as a Flavoring Agent in Cooked Beans - Logan, Gwinn, Richey, Maney, and Faulkner Plantago spp. and Bidens spp.: A Case Study of Change in Hawaiian Herbal Medicine - Palmer Nomenclature of Breadfruit Cultivars in Samoa: Saliency, Ambiguity, and Monomiality - Ragone, Tavana, Stevens, Stewart, Stone, Cox, and Cox The Category of ‘Animal’ in Eastern Indonesia - Forth Ojibway Plant Taxonomy at Lac Seul First Nation, Ontario, Canada - Kenny and Parker Population Trends and Habitat Characteristics of Sweetgrass, Anthoxanthum nitens: Integration of Traditional and Scientific Ecological Knowledge - Shebitz and Kimmerer Structure, Phenology, Fruit Yield, and Future Prospects of Some Prominent Wild Edible Plant Species of the Sikkim Himalaya, India - Sundriyal and Sundriyal Pluralistic Medical Settings and Medicinal Plant Use in Rural Communities, Mato Grosso, Brazil - Amorozo Volume 24, Number 1 Spring/Summer 2004 OURNAL STAFF EDITOR: econ F. Miller, University of Cope Museum-MASCA, 33rd and Spruce Sts., Philadelphia, PA 9104 (nmiller0@sas.upenn.e BOOK atch EDITOR: Darron Collins, Amazon Conservation Team, 4211 N. Fairfax Dr., Arlington, VA 22203 (dcollins@amazonteam.org SOCIETY OFFICERS PRESIDENT: Jan Timbrook, Santa Barbara Museum of Natural History, Santa Barbara, CA PRESIDENT-ELECT: Eugene Hunn, Department of Anthropology, University of Washington, Seattle, W. — REASURER: C. Margaret Scarry, Department of Anthropology, University of North Carolina, Chapel 1, NC (scarry@email.tinc.edu) BOARD OF TRUSTEES John D. Speth, University of Michigan, Ann Arbor, MI Dana Lepofsky, Simon Fraser University, Burnaby, B.C., Canada Enrique Salmé6n, Christensen Fund, Palo Alto, CA SOCIETY COORDINATORS come COORDINATOR: Mollie S. Toll, Museum of New Mexico, Office of Archaeological Studies, Box 7, Santa Fe, NM 87504 WEBSITE pobanisitin: Douglas Trainor (trainor@ethnobiology.org) Ex Offic Past conde Steven A. Weber, Amadeo M. Rea, Elizabeth S. Wing, Paul E. — Cecil Brown, Catherine S. Nancy J. Turner, Deborah M. Pearsall, and Karen R. Adam: ae anion member Steven D. Emslie Editor, conference coordinator, and website co-ordinator EDITORIAL BOARD Eugene N. Anderson, University of California, Riverside, CA: ethnobotany, China, Maya Scott sigs Nae eda France: ee classification, ethnoecology, cognition, evolutionary psychology, history of nce, Maya, M William saiec Tulane Gave, Sas Orleans, LA: historical ecology, ethnobotany, Amazonia Brent Berlin, University of Georgia, Athens, GA: ethnobiological classification, medical eae Maya Robert A. Bye, Jr., Jardin Botanico, Universidad Nacional Aut6noma de México, México, D.F.: ethnobotany, Mexico Roy Ellen, University of Kent, Canterbury UK: ethnobiological classification, cms Indonesia George Estabrook, University of Michigan, Ann Arbor, MI: Mediterranean p re, Portuguese speaking world, sik Hog on technology, computational methods Nina Etkin, University of Hawaii, Honolulu, HI: medical ethnobotany, ethnopl logy, ethnobiology of food, West Africa Pacific Gayle Fritz, Washington University, St. Louis, MO: paleoethnobotany, agricultural origins and developments, North America Brian Hesse, Penn State University, University Park, PA: haeology, animal domestication, pastoral systems, Near East, Andes Eugene oe University of _— Wes WA: ethnobiological classification, ethnoecology, Sahaptin and Zapotec guages, subsistence use of pa Timothy Joka, Macdonald College of ee University, Montreal, Quebec: Africa and the Americas, ethnobotany, McGill University, Quebec, He pate Ethno/human nutrition; nutrition of Indigenous Peoples David L. Lentz, Chicas Botanic Garden, Glencoe esoamerica, North America, Central Asia Brien Meilleur, CNRS, Paris, France: ethnoe ecology, ain knowledge and plant conservation, Alpine Europe, Polynesia Elizabeth Reitz, University of Georgia, Athens: zooarchaeo ology Editorial consultants: Elia San Miguel, Real Jardin Botanico de Madrid, Spain; Alain Cuerrier, Montreal Botanical Garden, Canada; Eraldo Medeiros Costa Neto, Universidade Estadual de Feira de Santana, Brazil The Journal U Ethnobiology is published semi-annually by the Society of Ett th Pecan ie for ' PRE wee ne ppror pay oo Ps CODEN Be es ys inside bee nee of this i issue. © Society of hoiplociongaa 2 ISSN 0278-0771 COVER THT Li soh oe es Market, Mexico City (Logan et al., this issue Figure 1). MISSOURI BOTANICAL JUN 2 2 2004 GARDEN LIBRARY Journal of Ethnobiology VOLUME 24, NUMBER 1 SPRING/SUMMER 2004 CONTENTS ETHNOBIOTICA Vv AN EMPIRICAL ASSESSMENT OF EPAZOTE (Chenopodium ambrosioides L.) AS A FLAVORING AGENT IN COOKED BEANS Michael H. Logan, Kimberly D. Gwinn, Tina Richey, Beth Maney, and Charles T. Faulkner 1 Plantago spp. AND Bidens spp.: A CASE STUDY OF CHANGE IN HAWAIIAN HERBAL MEDICINE Christian Palmer 13 NOMENCLATURE OF BREADFRUIT CULTIVARS IN SAMOA: SALIENCY, AMBIGUITY, AND MONOMIALITY Diane Ragone, Gaugau Tavana, Joan M. Stevens, Patricia Ann Stewart, Rebekka Stone, Paul Matthew Cox, and Paul Alan Cox THE CATEGORY OF ‘ANIMAL’ IN EASTERN INDONESIA Gregory Forth 51 OJIBWAY PLANT TAXONOMY AT LAC SEUL FIRST NATION, ONTARIO, CANADA Mary B. Kenny and William H. Parker 75 POPULATION TRENDS AND HABITAT CHARACTERISTICS OF SWEETGRASS, Anthoxanthum nitens: INTEGRATION OF TRADITIONAL AND SCIENTIFIC ECOLOGICAL KNOWLEDGE Daniela J. Shebitz and Robin W. Kimmerer 93 STRUCTURE, PHENOLOGY, FRUIT YIELD, AND FUTURE PROSPECTS OF SOME PROMINENT WILD EDIBLE PLANT SPECIES OF THE SIKKIM HIMALAYA, INDIA Manju Sundriyal and R.C. Sundriyal 113 PLURALISTIC MEDICAL SETTINGS AND MEDICINAL PLANT USE IN RURAL COMMUNITIES, MATO GROSSO, BRAZIL Maria Christina de Mello Amorozo 139 BOOK REVIEWS 163 Journal of Ethnobiology 24(1): v Spring/Summer 2004 ETHNOBIOTICA In the past few months, there have been several changes in the editorial board of the journal. I would like to take this opportunity to thank Bill Balée, who is stepping down from the editorial board, for all his help and advice; we are for- tunate that Glenn Shepard is willing to take his place. I would also like to thank Elia San Miguel for editing and translating the Spanish abstracts for the past few issues, and Annick Sullivan for providing translations in French. Beginning with this issue, Alain Cuerrier will be the consultant for French and Eraldo Medeiros Costa Neto for Portuguese. Manuel Pardo de Santayana will be the journal’s Span- ish consultant starting with volume 24, number 2. It is thanks to the volunteer labor of many people—the editorial board and consultants, the many anonymous reviewers, and lest we forget, the authors— that we have been able to cover such a variety of ethnobiological topics in our journal. I hope you include the ones in this issue in your summer reading pro- gram! ~~ Vam~ Journal of Ethnobiology 24(1): 1-12 Spring/Summer 2004 AN EMPIRICAL ASSESSMENT OF EPAZOTE (Chenopodium ambrosioides L.) AS A FLAVORING AGENT IN COOKED BEANS MICHAEL H. LOGAN,* KIMBERLY D. GWINN,;? TINA RICHEY, BETH MANEY: and CHARLES T. FAULKNER? * Department of Anthropology (mlogan@utk.edu) » Department of Entomology and Plant Pathology © Undergraduate Programs in Business Administration 4 Department of Comparative Medicine University of Tennessee, Knoxville, Knoxville, TN 37996 ABSTRACT.—A common culinary practice in Mexico and elsewhere in Mesoam- erica is the use of Chenopodium ambrosioides L., a small herbaceous plant known most widely as epazote, to flavor black beans and other dishes. While some people find the taste and odor of this herb to be mildly disagreeable, there are good empirical reasons for its use as a flavoring agent in cooked, unrefrigerated, foods. Through a series of experimental trials we observed that beans prepared with C ambrosioides remained edible, as judged by sight and smell, long after plain beans had begun to spoil. Microbiological tests revealed significant bacteriocidal activity in this species. Epazote has a large and diverse range of potential scientific and commercial applications. Key words: sues Chenopodium ambrosioides, Mesoamerica, culinary practices, food preservativ RESUMEN.—Una practica culinaria frecuentemente utilizada en México y en otras partes de Mesoamérica para darles sabor a los frijoles negros y a otras comidas es el uso de Chenopodium ambrosioides L., una pequefna planta herbacea comtnmente llamada epazote. Aunque algunas personas encuentran el sabor y el olor de este condimento algo desagradable, existen razones empiricas para justi- ficar el empleo de agentes para mejorar el sabor de alimentos cocinados que no son refrigerados. A través de andlisis experimentales, se observ6 que los frijoles preparados con C. ambrosioides permanecian comibles, al juzgar por la apariencia y el olor de éstos, aun cuando los frijoles comunes y corrientes habian empezado a avinagrarse. Analisis microbiolégicos revelaron la actividad bactericida de esta especie. El potencial de utilizacién con fines cientificos y comerciales de esta planta es amplio y diverso. RESUME.—Le Chenopodium ambrosioides L., une petite plante herbacée plus con- nue sous le nom d’ambroisie, est couramment utilisé dans la cuisine du Mexique et, de facon générale, en Méso-Amérique pour aromatiser les plats de haricots noirs ainsi que d’autres plats. Quoique certaines personnes trouvent le godt et l’odeur de cette herbe légerement désagréable, il existe d’excellentes raisons em- piri quant a l'utilisation de l’ambroisie comme aromate dans les plats cuits et non réfrigérés. A la suite d’une série d’essais expérimentaux, nous avons ob- servé que les haricots préparés avec le C. ambrosioides demeuraient comestibles— - LOGAN et al. Vol. 24, No. 1 comme on peut le noter par leur aspect et leur odeur—bien aprés que les haricots préparés sans la plante aient commencé a se gater. Des tests microbiologiques ont révélé une activité bactéricide significative chez cette espéce. L’ambroisie posséde un grand éventail d’applications différentes ayant un potentiel a la fois scienti- fique et commercial. INTRODUCTION Many residents of Mexico and Guatemala subsist on a diet based largely on black beans and corn tortillas. These foods, of course, have served as the corner- stone in the diet of Mesoamerican cultures for thousands of years (Mangelsdorf et al. 1964). An equally ubiquitous, and possibly ancient, culinary practice is the use of epazote (Chenopodium ambrosioides L.) as a flavoring agent in cooked beans." As one authority on Mexican cuisine remarked, ‘To cook black beans without it is unthinkable’ (Kennedy 1978:239). Yet many individuals, especially children and adults unaccustomed to traditional Mesoamerican foods, find the taste and odor of epazote to be mildly disagreeable (Johns 1990:284). Even the plant's common name reveals much about its pungent qualities: epazotl, the Aztec name for this species, is based on epatl, their word for skunk (Coile and Artaud 1997).? There are, however, sound empirical reasons for cooking beans with several leaves and stems of epazote added, reasons that far overshadow any concerns one might have regarding the plant’s strong taste and odor (Figure 1). Three different experiments were conducted to better understand why this malodorous plant is so greatly valued as a flavoring agent in cooked beans. The first and second of these tests were simple in design, yet important in outcome. Moreover, each provided results that warranted a third phase of experimental investigations into the bioactivity and phytochemistry of C. ambrosioides. We now turn to a discussion of our methods and findings. EXPERIMENTAL DATA: PHASE I The idea behind the present research originally surfaced when the senior author (ML), who has considerable fieldwork experience in Guatemala and Mex- ico, suspected there must be some advantageous reason for using C. ambrosioides to flavor cooked black beans and many other traditional foods or dishes (in ad- dition to Kennedy 1978, see Bayless and Bayless 1987; Gerlach and Gerlach 1994; Martinez 1992; Ortiz 1967; Quintana 1986). He was also aware that many other botanicals used as spices thwarted food spoilage by curbing the growth of bac- teria and fungus that rapidly invade cooked, though unrefrigerated, foods (e.g., Sherman and Hash 2001). A simple experiment was conceived that would help resolve the enigmatic question of why it would be “unthinkable,” in Kennedy’s assessment, to cook black beans without epazote. With the assistance of Beth Maney, the experiment was initiated. A supply of black beans was purchased at a local grocery store. Two batches of beans (approximately 226 g [% pound] each)—one containing fresh Mexican epazote (about a dozen whole leaves and attached stems), the other without epazote—were then cooked by vigorous boiling. This was done in the Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 3 See 4 ee FIGURE 1.—A hierbera in Mexico City’s Sonora Market vends numerous fresh and dried plant products used in cooking and health care, including epazote. Photograph by Michael H. Logan. evening. When thoroughly cooked the two batches of beans were drained, then spooned into plastic containers that were clearly labeled: “epazote’ and “‘plain.” No lids were used as the beans sat overnight at room temperature. The following morning the cooked samples were transferred out-of-doors and placed in a shad- ed and covered area. The month was August, with daytime highs in the lower 90s (>32°C), and evening lows in the 70s (>21°C). At 6:00 p.m. (approximately 24 hours after the beans were cooked), the senior author visually inspected the beans and then sniffed each sample. Although they looked the same, the batch without epazote had a slight odor of decay. However, the beans containing epa- zote smelled fresh. The samples remained out-of-doors for another evening. At 10:00 a.m. the following morning the samples were again assessed. The plain sample now exhibited signs of fungal growth. It also had a strong, disagreeable odor. Interestingly, though, the beans cooked with epazote looked and smelled fresh, save for the odor of this additive. That evening at 6:00 the plain sample was reassessed. It was thoroughly rotten, and the odor of the beans was offensive. The beans containing epazote still smelled, after approximately 48 hours, as if they remained edible. They were not eaten, however. The beans in each batch were then discarded. This simple experiment—the “sniff test’’—was repeated several times. And the results were uniformly the same. Black beans cooked with epazote remained 4 LOGAN et al. Vol. 24, No. 1 Ls G SICEN m monary & SPiCEs cs. PRopiy ‘E FIGURE 2.—Dried epazote is sold by a variety of commercial outlets in Mesoamerica and the United States. Photograph by W. Miles Wright. fresh long after plain beans began to rot. But would use of dried epazote, rather than fresh, produce different results? Packets of dried C. ambrosioides were purchased from a retail and mail order firm in Florida that specializes in herbs and spices used in preparing ethnic foods (Figure 2). The amount of epazote added to the beans followed traditional Mex- ican recipes (one tablespoon per half cup of beans). Interestingly, the results with dried epazote paralleled what was observed in our earlier trials. Use of fresh or dried epazote when cooking beans significantly extended the length of time dur- ing which they would remain edible, as judged by sight and, more importantly, smell. The experiment faithfully tried to replicate the domestic environment of cook- ing and food storage found in the past, or among many Mesoamerican peasants today. While ceramic vessels and a wood fire were not used, the factors judged most important in these tests were to leave the cooked beans uncovered in a shaded area and at ambient daytime and evening temperatures. Reheating beans that already contained epazote would undoubtedly further extend the ‘‘shelf-life’’ of this staple food. To gain better insight into the biological activity of C. ambrosioides as a food additive, the senior author enlisted the aid of three colleagues (the other co-au- Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 5 thors), each having considerable laboratory experience. Additional experimenta- tion was the logical next step. EXPERIMENTAL DATA: PHASE II To better control for environmental factors that could have affected the cooked samples in our earlier trials (e.g., wind-borne spores or contact by insects), we decided to place the cooked beans—those with epazote and those without—into Petri dishes, which would then be stored in an incubator at a constant temperature and light level. The black beans were prepared in the same manner as in the previous tests. In all, three different batches of beans were cooked: two with fresh epazote (in one of these the sprigs of epazote were removed after cooking) and one without epazote. These samples were then taken to Charles Faulkner’s laboratory. The beans were then placed into clearly marked, yet coded, Petri dishes. Nine dishes were used (3 with epazote left in [ei]; 3 with epazote removed [er]; and 3 plain [p]). Bacteria (Micrococcus lutea) were then added to two of the dishes in each of the three categories (ei, er, p), leaving one dish in each of the three groupings not spiked with bacteria. It should be noted here that only Logan knew what letters and numbers corresponded with “epazote in,” ““epazote removed,” and “plain,” as well as “bacteria yes” and ‘‘bacteria no.” All dishes were then placed inside a lighted, airtight incubator set at 26°C.° After 24 hours in the incubator (approximately 40 hours had elapsed since the beans had been cooked), the samples were removed and inspected (by CF) for visual signs of spoilage (e.g., any change in color or consistency). All dishes where bacteria had been added showed signs of decay, especially p1 and p2 (the plain beans with bacteria added). However, the samples prepared with epazote, but ones not containing additional bacteria (ei3, er3), showed no visible evidence of spoilage. All lids were then removed and the “sniff test” was conducted on the contents of each dish. Again, this was a blind trial. A disagreeable odor was detected (by CF) in all of the samples, save for two: the samples prepared with epazote that were not spiked with bacteria. These results warranted additional, and more rigorous, experimental trials and microbiological assessment. EXPERIMENTAL DATA: PHASE III Again, black beans were cooked with and without epazote in nonlaboratory conditions. El Charitto (EC) brand epazote was purchased in Mexico. One table- spoon of epazote (ca. 4 g) was added per 0.5 cup dry beans (ca. 200 g). Standard microbiological techniques were followed in order to determine bacteria counts in both samples (Harrington and McCance 1976).* The results were dramatic. The sample without epazote contained approximately 800 colony-forming units per ml. The sample cooked with epazote contained none. It was totally free of bac- teria, yet some 42 hours had passed since the beans were cooked. After 72 hours, samples were again drawn, diluted, plated and counted. The one without epazote had more than 160 million colony-forming units per ml, while the batch with epazote had eleven thousand. These findings strongly confirmed what had been 6 LOGAN et al. Vol. 24, No. 1 observed in our earlier trials. Epazote does extend the length of time that unre- frigerated black beans remain edible, by a magnitude of at least one full day, if not more. Activities of EC and an epazote commercially available in the United States (Penzey’s Spices [PS], Muskego, Wisconsin) were compared. The EC epazote was determined to be heavily contaminated with multiple species of bacteria that could alter bacterial counts in beans cooked with EC; therefore, for further ex- periments, black beans were cooked under controlled laboratory conditions using propylene oxide-sterilized epazote. Propylene oxide did not alter the chemical composition of either EC or PS epazote (data not shown). In subsequent experi- ments, epazote was sterilized by submersion overnight in propylene oxide that was then allowed to evaporate. To each of two sets of sterile cookware, dry black beans (200 g) and tap water (400 ml) were added. The beans were heated to boiling, then cooled to room temperature for two hours. Sterile epazote (4 g) was added to one pot and mixed well. Both batches were simmered for one hour. The next procedure to be pursued was to assess epazote’s potential, if any, to curb the growth of two common food spoilage organisms, Escherichia coli (0157) and Salmonella (8326). Again, more beans were prepared. The cooked samples were inoculated with the bacteria (10° per gram beans). Appropriate controls were used. Colonies were counted at 24 and 48 hours. Epazote had little or no effect on the introduced bacteria. Interestingly, the two control groups (each with epa- zote) had no bacteria, whereas the plain batch was laden with some type of bac- teria. The naturally occurring organisms in the untreated cooked beans were clas- sified as Gram-negative or Gram-positive by differential staining, and Gram-pos- itive organisms were tentatively identified as a species of Bacillus, probably B. cereus (Lund et al. 2000). In order to determine if Bacillus species are sensitive to epazote, both treated and untreated beans were divided into two sterile flasks to which 10° Bacillus BA101/gram beans was added to one flask of each set. The flasks were incubated at 30°C, and samples were plated at 24 and 48 hours as described above to de- termine bacterial counts. EC and PS were evaluated in separate experiments. Ap- proximately ten times more bacteria were isolated from untreated beans than for beans treated with EC epazote. With the addition of Bacillus, there was also a ten- fold reduction in the number of bacteria growing in beans cooked with epazote compared to untreated beans. There was approximately twice the number of bac- teria in untreated beans compared to beans cooked with PS epazote. Apparently these brands of epazote differ in their phytochemical composition, with EC epa- zote exhibiting greater bacteriocidal activity than PS epazote. Dr. Duke’s Phytochemical and Ethnobotanical Database was used to identify commercially-available antibacterial compounds that had previously been isolated from Chenopodium ambrosioides.* Nine compounds were tested for activity against several types of bacteria, including Gram-positive and Gram-negative organisms isolated from beans, Salmonella 8326, and nine strains of Bacillus (Table 1).° Gera- niol and safrole were the only compounds of the nine tested that inhibited the growth of bacteria (Bard et al. 1988). TABLE 1.—Antibacterial properties of epazote compounds. m1t))6CU GCP sOSaB6 «6CGA77 ~~. OBA 101 = E2i E 61 E 65 E 66 E 69 E720 = evar Cymene el — _— a oO | BB rt. welt lobcch: Leeedctrlod od usd — — N ae | am N a ul — w j= N pmb ho Ten pl of undiluted cymene, limonene, myrcene, geraniol, pinene, or a was added to a sterile filter paper disk. Ten jl of ferulic acid (50 mg/ml in 95% ethanol), vanillic acid (50 mg/ml in 95% ethanol), or ascorbic acid (100 mg/ml) was added to a sterile filter paper disk. The treated disk was placed on a lawn . bacteria obtained by sett plating one ml 10° bacteria/ml Sisson on TSA plates. After 24 hours the cleared zone around the filter paper disk was asured. 'G(+) and G(—) organisms were naturally occurring isolates from beans; Sal 8326 = Salmonella; BA and E = isolates of Bacillus. — No effect of compound on ba —— a wth. > Measurements are cleared diamete mm. £007 Jouuns /Sutds XOOTOISONHLA JO TVNYNO[ 8 LOGAN et al. Vol. 24, No. 1 DISCUSSION The traditional use of epazote to flavor cooked beans is an empirically sound practice. Geraniol and safrole effectively retard the growth of Bacillus bacteria, which is a natural component in bean spoilage. This would have been particularly important in the past, as well as in contemporary settings where a sizeable num- ber of persons lack electricity and refrigeration (Valdes-Ramos and Solomons 2002: 149). Because this species safely extends the period of time that unrefrigerated cooked beans remain edible, its potential as a preservative in other dishes, both vegetarian and meat-based, certainly warrants additional study. Preliminary findings on a closely related species—Chenopodium berlandieri Mogq.—suggest that it, too, retards the rate of spoilage in cooked beans due to fungal invasion. Fresh C. berlandieri was acquired from a local farm in Blount County, Tennessee, and three batches of black beans were prepared following the same practices done with the epazote tests described above. Beans cooked without C. berlandieri exhibited significant amounts of mold after 48 hours, while the con- trol sample with C. berlandieri added before cooking appeared to be suitable for consumption. The third batch, prepared with dried C. berlandieri achenes also looked and smelled fresh some two days after they were cooked. These trials, though simplistic, are especially interesting because this species of Chenopodium held an important place in the diet of prehistoric Indians in the eastern woodlands of North America as early as three to four thousand years ago (Fritz 1999; Gremillion 1993; Smith 1984, 1985a, 1985b). Perhaps these peoples, like their counterparts in Mesoamerica, had discovered the food preservation qualities of Chenopodium. Our research on C. berlandieri continues, yet it should be noted here that both C. berlandieri and C. ambrosioides were eventually domesti- cated by aboriginal peoples in the past. Domestication confirms how culturally important these species once were. In Mesoamerica, though, the cultural importance of C. ambrosioides has per- sisted to the present day (e.g., Heiser 1985:82-99). Aside from its role as a flavor- ing agent, and by extension a food preservative, it is eaten fresh as a potherb by Indians in northern Mexico (Bye 1981:116). Leaves and stems of this plant contain important amounts of calcium, phosphorous, and vitamin C (Ortiz de Montellano 1990:240). The fruits or achenes of epazote, which are rich in protein, also hold value as a food (Minnis 2000:223). While most residents of the United States would view chenopods to be lowly weeds, these plants have considerable worth as an alternative crop (Coile and Artaud 1997). Epazote is also culturally valued for its medicinal properties (e.g., Moerman 1998; Morton 1981). It is widely used in the Americas and beyond to combat a large array of health-related problems. Frequently it is the remedy of choice for controlling intestinal worms (Berlin et al. 1996:413-417), yet its efficacy as a ver- mifuge in humans is, to some degree, uncertain (Kliks 1985; but also see Kight- linger et al. 1996).”? While epazote’s role as a curative must be of considerable antiquity, this cannot be determined archaeologically. However, it is described as a useful medicinal herb in early post-Conquest ethnohistorical sources (eg, OF ellana 1987). The volume of literature on C. ambrosioides as a medicinal plant far overshadows what has been published on this species as a dietary item. Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 9 In addition to its bacteriocidal properties, C. ambrosioides has been shown to possess other realms of biological activity. It is a strong allelopathic species, in that its leaves exude compounds that prohibit or delay seed germination of other nearby species (Jimenez-Osornio et al. 1996). Researchers have also learned that C. ambrosioides has insecticidal and repellent properties (Morsy et al. 1998; Su 1991; Tapondjou et al. 2001). Antiviral (Verma and Baranwal 1983), antifungal (Kishore et al. 1989; Montes-Belmont and Carvajal 1998), trypanocidal (Kiuchi et al. 2002), and molluscicidal (Hmamouchi et al. 2000) activities have also been reported for this species. Other studies have shown that C. ambrosioides has hypotensive (Gohar and Elmazar 1997) and antidermatophytic (Kishore et al. 1996) activities as well. It becomes obvious that this species has a vast number of potential uses, and in a variety of industries, ones ranging from agriculture and pest control to human (Lall and Meyer 1999) and veterinary medicine (Perezgrovas et al. 1994). CONCLUSIONS Precisely when and where prehistoric peoples of Mesoamerica discovered the food-preservation qualities of epazote is not known. How this discovery was made, though, is easier to reconstruct. Olfaction was undoubtedly the critical el- ement lying behind this acquired knowledge concerning C. ambrosioides. The Archaic Period of Mesoamerican prehistory (ca. 8000 B.C. to 2000 B.C.) witnessed a huge number of cultural innovations. None was equal in importance to the domestication of plants. Dozens of different species were transformed from their wild states into useful foods (West and Augelli 1989:220). By 2000 B.C., maize and beans held an important place in the diet of these early peoples. They, and no doubt their predecessors, also consumed C. ambrosioides (Manglesdorf et al. 1964:434). There is clear evidence that the seeds of this plant were roasted and eaten. It is also highly probable that fresh leaves and stems were eaten as well. Placing sprigs of this edible potherb into a stone, and later ceramic, vessel that contained beans and water brought to a boil would be a logical method of adding volume, if not variety, to a meal soon to be eaten. Leftover beans would remain in the vessel or a gourd bowl until reheated for the next meal. And humans invariably employ smell, coupled with sight, as a proximate gauge of a food's edibility. The results of our original experimental trials—the “sniff’’ test—revealed, in no uncertain terms, that beans prepared with epazote seemed safe to eat long after the beans lacking epazote smelled foul. The same observation undoubtedly occurred, and repeatedly so, among women in the past. A culinary tradition, one seen today, emerged from this simple, though important, discovery about epazote and its value as a food additive. And the microbiological data identified in the third phase of our experimental trials confirmed the merits of this widely diffused and ancient custom. Kennedy (1978) was indeed correct—there are strong empir- ical reasons why it would be unthinkable to prepare black beans without epazote. Hopefully the findings of this study will stimulate additional research on this small herbaceous plant originally domesticated in Mesoamerica thousands of years ago. The potential scientific and commercial applications of C. ambrosioides are con- siderable. This species certainly warrants further cross-disciplinary inquiry. 10 LOGAN et al. Vol. 24, No. 1 NOTES 1 C. ambrosioides is a small (up to 1 m) herbaceous plant native to tropical and subtropical zones in the Americas, although it has naturalized to many other locales beyond its original range. It has been exported abroad and is now found in several countries in Africa, Asia, and Europe. The strongly scented leaves are alternate, lanceolate or oblanceolate in shape (ca. 4-12 cm long) with toothed, lobed, wavy, or smooth margins. It is known by at least fifty different common names, depending on the region and languages spoken. In English it is frequently called goosefoot or lamb’s-quarters due to the shape of its leaves. The genus dium contains, worldwide, approximately 150 species. Some of these are of consid- erable cultural importance, especially in the Americas. 2 These authors use a variant spelling, one common in Guatemala: “apazote.” 3 This setting on the incubator was chosen because it would simulate ambient daytime (noon) temperatures in much of Mesoamerica. 4 Following an incubation period, three beans were placed in 10 ml sterile 0.75% NaCl, vortexed, serially diluted and plated in duplicate on tryptic soy agar (TSA). Plates were incubated at 30°C. Bacterial colonies were counted after 24 hours. 5 Duke, Jim. n.d. Dr. Duke’s Phytochemical and Ethnobotanical Databases. Agricultural Re- search Service, U.S. Department of Agriculture. [http:/ /www.ars-grin.gov /duke/] (verified 9 December 2003) * Ten pl of each compound was added to a sterile filter paper disk. The treated disk was placed on a lawn of bacteria obtained by spread plating one ml 10° bacteria/ml suspension on TSA plates. Bacteria tested included Gram-positive and Gram-negative organisms iso- lated from beans, Salmonella 8326, and Bacillus sp. isolates BA77, BA101, E21, E61, E65, E66, E69, E726, and E727. Cultures of Bacillus were obtained from the collection of Bonnie Ownley, University of Tennessee. After incubating 24 hours at 30°C, the diameter of cleared bacterial growth was measured. Cymene, limonene, and pinene were obtained from Al- drich Chemical, Milwaukee, WI. Myrcene, ete! safrole, ferulic acid, vanillic acid and ascorbic acid were obtained from Sigma Chemical, St. Louis, MO. Cymene, limonene, myr- cene, geraniol, pinene, and safrole were used as concentrated oils. Ferulic acid and vanillic acid were suspended at 50 mg/ml in 95% ethanol and ascorbic acid was suspended at 100 mg/ml in distilled water. ’ The anthelmintic activity of C. ambrosioides is due to the presence of ascaridole, which is known to be toxic to round worms (Ascaris lumbricoides), as well as a number of other parasitic organisms. Oil of Chenopodium was once widely used to expel intestinal worms, yet the dosage required to be totally effective approached the lethal limit in humans * Co-author Beth Maney, who is an avid horse breeder, found during our epazote tests that a water-based solution of C. ambrosioides, when applied daily as a wash to the soles of her horses’ hooves, effectively eliminated hoof (thrush) infection from her animals. It proved te Se to some common, over-the-counter, remedies designed to control this equine hoof ection. Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 11 REFERENCES CITED Bard, M., M.R. Albert, N. Gupta, CJ. Guynn, and W, Stillwell. 1988. Geraniol interferes with membrane functions in strains of Candida and Saccharomyces. Lipids 23(6):534-538. Bayless, Rick and Deann G. Bayless. 1987. Authentic Mexican Regional Cooking from the Heart of Mexico. William Morrow and Company, York. or Berlin, B., D.E. Breedlove, X. Lozoya, M. Meckes, M.L. Villarreal, and J. Torto- riello. 1996. Medical ethnobotany of cage species used to treat regular In Medical fetta of the Highland Maya of Chiapas, Awrp ed. A. Berlin and B. Berlin, pp. 413-417. Prin ceton University Press, Rieti Bye, Robert A. 1981. anak ne ahaa By of edible greens—past, present, and future. ee, oF Ethnobiology 1: 109. 123. Coile, Nancy C. and Carlos R. Artaud. 1997. Chenopodium ambrosioides L. (Chen- opodiaceae): Mexican-tea, wanted weed? B Circular, Gainesville 33:1-6. Fritz, Gayle J. 1999. Gender and the early omestication of gourds in eastern North America. American Antiquity 64: 417-429. Gerlach, Nancy and Jeffrey Gerlach. 1994. Foods of the Maya: a Taste of the Yucatan. University of New Mexico Press, Albu- querque. Gohar, Ahmed A. and M.M.A. Elmazar. sp 0 ing in Egypt Phytotherapy Research 11: 564-56 actin, Kristen J. 1993. Crop and weed in a rehistoric eastern North America: Chenopodium example. American An- tiquity 58:496-509. Harrington, W.E and Margaret E. Mc- Cance. 1976. Laboratory Methods in Food and at Microbiology. Academic Press, Heiser, hares B., Jr. 1985. Of Plants and People. University of Oklahoma Press, lorman. Hmamouchi, M., M. Lahlou, and A. Agou- mi. 2000. Molluscicidal activity of some Moroccan medicinal plants. Fitoterapia 71:308-314. Jimenez-Osornio, EM.V.Z. Juan, Junji Ku- mamoto, and Christian Wasser. 1996. Allelopathic activity of Chenopodium am- brosioides L. Biochemical Systematics and Ecology 24:195- Johns, Timothy. 1990. With Bitter Herbs They Shall Eat It: Chemical Ecology and the Or- igins of Human Diet and Medicine. Uni- versity of Arizona Press, Tucson. Kennedy, Diana. 1978. Recipes from the Re- gional Cooks of Mexico. Harper and Row Publishers, New York. Kightlinger, Lon K., J. Richard Seed, and a Boodhoo Kightlinger. 1996. As- caris ee eee aggregation in relation to child growth status, delayed cutane- ous hypersensitivity, and plant anthel- mintic use in Mada an Journal of Par- oe 82(1):25- Kishore, N., J.P.N. Janets and N.K. Dubey. 1996. Antidermatophytic action of the essential oil of Chenopodium am- brosioides and an ointment prepared from it. Phytotherapy Research 10:453- 455 Kishore, N., S.N. Dixit, and N.K. Dubey. 1989. Fungitoxic studies with Chenopo- dium mee for control of damp- ing-o eolus aureus caused by Rhizoctonia solani. Tropical Science 29: 171-176. Kiuchi, tan ae Yoshiaki Itano, Nahoko Uchiyama, Gisho Honda, Akiko Tsu- Takashi Aoki. 2002. Monote droperoxides with Trypanocidal activi- rom C, ium ambrosioides. Jour- nal of Natural Products 65:509-512. Kliks, Michael M. 1985. Studies on the tra- ditional herbal anthelmintic Chenopodi- um ambrosioides L.: ethnopharmacologi- cal evaluation and clinical field trials. er. 1999. Inhibition of drug-resistant and drug-sensitive strains of Mycobacterium tuberculosis by Lund, — a“ a, . neaay ape and G W. Gould. 2000. The Mi- casts nore ou mr prs of Food. Aspen Publishers, Gaithersburg, Mary- and. Mangelsdorf, Paul C., Richard S. McNeish, and Gordon R. Willey. 1964. Origins of a2 LOGAN et al. agriculture in Middle America. In Nat- ural Environment and Early Cultures, ed. R.C. West, pp. 427-446. Handbook o Middle American Indians, vol. 1. Univer- sity of Texas Press, Austin. Martinez, Zarela. 1992. Food from My Heart: Cuisines of Mexico Remembered and Re- imagined. McMillan Publishing Compa- ew Yor Make Paul E. 2000. Famine foods of the North American desert borderland in historical context. In Ethnobotany: a Reader, ed. PE. Minnis, pp. 214-239. University of Oklahoma Press, Norman. loerman, Daniel E. 1998. North American Ethnobotany. Timber Press, Portland. Montes-Belmont, R. and M. Carvajal. 1998. Control of Aspergillus flavus in maize with plant essential oils and their com- ponents. Journal of Food Protection 61(5): 6-619. Morsy, Tosson A., Ahmed Shoukry, Said .M. Mazyad, and Karam M. Makled. 1998. The effect of the volatile oils of Chenopodium ambrosioides and Thymus vulgaris sony the larvae of Lucilia ser- icata (Meigen). Journal of the Egyptian So- ciety of Parstlogy 28:503-510. Morton, Julia F Atlas of — Plants of Middle aves Bahamas to Yu- catan. Charles C. Thomas, Sprin, field. Orellana, Sandra L. 1987. Indian Medicine in Highland Guatemala: the Prehispanic and Colonial Periods. SO ison of New Mex- ico Press, Albuquer Ortiz, Elisabeth L. 1967. The Complete Book of Mexican Cooking. M. Evans and Com- pany, Inc., New York. Ortiz de Montellano, Bernard R. 1990. Az- tec Medicine, Health, and Nutrition. Rut- gers University Press, New Brunswick. Perezgrovas, R., A. Parry, M. Peralta, L. Za- ragoza, D. Trow, and P. Pedraza. 1994. Chiapas sheep, wool production and Vol. 24, No. 1 animal health in a unique sheep bree Proceedings of the New Zealand Society - Ant 77-182. spicy. Evolution and Human Behavior 22: 147-163. Smith, ites D. 1984. Chenopodium as a pre- historic domesticate in eastern North A evidence from Russell Cave, a ee ee 226:165-167. a. Chenopodium ie a Hopew Southeastern Archaeology 4:107-133. . 1985b. The role of Chenopodium as a domesticate in the pre-maize garden systems of the eastern United States. Southeastern Archaeology 4:51-72. Su, H. 1991. Toxicity and repellency of Che- ium oil to four species of stored- roduct insects. on of Entomological Science 26(1):178- icon | aes Va ak hee H. Bouda, and Fontem. 2001. Essential oil from pales ambrosioides leaves as post- harvest protectants against six-stored product beetles. pid of Stored Prod- ucts Research 38:395-4 be eg nat Roxana pi Noel W. Solo- 2002. Preventive nutrition: its ing context in Mesoamerica. Nu- trition Research 22:145-152. Verma, H. and V. Baranwal. 1983. Antiviral activity and the physical properties of the leaf extract of Chenopodium ambro- sioides L. Proceedings of the Indian Acad- emy of Sciences (Plant Sciences) 92:461- West, Robert C. and John P. sn ie Sa merica: Its Lands les. Prentice Hall, Englewood Cliffs, New Jersey. Journal of Ethnobiology 24(1): 13-31 Spring/Summer 2004 Plantago spp. AND Bidens spp.: A CASE STUDY OF CHANGE IN HAWAIIAN HERBAL MEDICINE CHRISTIAN PALMER Yale School of Forestry and Environmental Studies, New Haven, CT 06511 (christian.palmer@yale.edu) ABSTRACT.—The substitution of alien species for native plants in Hawaiian tra- ditional medicine is discussed. Substitutions examined are the switch from vari- ous native Bidens spp. to the more ubiquitous Bidens pilosa L. and the switch from the native plantains, Plantago pachyphylla A. Gray, Plantago hawaiensis A. Gray, and Plantago princeps Cham. and Schlechtend, to Plantago major L. Historical, cultural, and scientific data are synthesized to examine these changes in the traditional Hawaiian pharmacopoeia. Major elements that affect the inclusion of new plants in the Hawaiian pharmacopoeia are availability and biological activity. This pro- cess of inclusion also provides insight into how indigenous scientists adapt phar- macological traditions to the changing biological and cultural environment. Key words: Hawaii, la‘au lapa‘au, ethnomedicine, Bidens, Plantago. RESUMO.—A substituicao de espécies introduzidas pelas plantas nativas na med- icina tradicional do Havai é debatida. As substituigdes examinadas sao das es- pécies de Bidens nativas para Bidens pilosa e a substituicéo das Plantago spp. na- tivas, inclusive Plantago pachyphylla, P hawaiensis e P. princeps, para Plantago major. Os aspectos histéricos, culturais e cientificos sao sintetizados para examinar estas mudangas na farmacopéia tradicional do Havai. Os dois elementos mais impor- tantes na inclusao de novas plantas na etnofarmacopéia do Havai sao a facilidade de acesso das plantas e atividade biolégico. Este processo ensina como cientistas indigenas adaptam tradicdes farmacolégicos 4s mudangas nos meios bioldgico e cultural. RESUME.—Cette étude rend compte de I’utilisation d’espéces introduites aux dé- ens des plantes indigénes dans la médecine traditionnelle hawaiienne. La plante ubiquiste, Bidens pilosa L., ainsi que le Plantago major L., ont remplacé respective- ment plusieurs Bidens spp. indigénes et les Plantago pachyphylla A. Gray, P. hawaien- sis A. Gray et P princeps Cham. et Schlecht. Afin de comprendre ces changements dans la pharmacopée traditionnelle hawaiienne, une synthése des données his- toriques, culturelles, et scientifiques a été faite. Les principaux facteurs détermi- nant l’inclusion de nouvelles plantes dans la pharmacopée hawaiienne résident dans leur cos Ame et leur activité biologique. Egalement, ce processus d’inclusion permet de mieux saisir comment les «scientifiques autochtones» ont adapté leurs traditions pharmacologiques selon les changements survenus dans leurs environnements biologique et culturel. INTRODUCTION e purpose of this paper is to synthesize disparate cultural, historical and biological information to examine the factors that influence the adoption of Bidens 14 PALMER Vol. 24, No. 1 pilosa and Plantago major into the traditional Hawaiian pharmacopoeia. Previous botanical, ethnobotanical, demographic, historical, and cultural research provides background on the plants, as well as their ethnobotanical history in Hawaii. Per- sonal interviews let Hawaiians speak for themselves about the plants and the processes of adoption and substitution. A literature search of pharmacological tests as well as antibacterial screening offers insight into the comparative biolog- ical activity of the native and introduced species, an important factor in the in- troduction and use of a new plant species. Substitutions appear to be a result of a variety of cultural factors including empirical testing and observation to deter- mine biological activity, spiritual inspiration, and other ecological considerations such as availability of plant material. The Hawaiian Islands provides an excellent place for a study of the evolution of a pharmacopoeia because both its flora and traditional herbal medicine have been well studied. It is also significant that, even after the introduction of the western medical system in Hawaii, the traditional healing system has continued to be practiced and even expanded in recent years due to a Hawaiian cultural renaissance. Over the years, the traditional herbal medicine system (la‘au lapa‘au) has adapted to the changing biological and cultural environment. The substitution of Bidens pilosa L. and Plantago major L. for native species exemplifies this process. A growing body of ethnobotanical research has begun to examine the origin and evolution of traditional medicine. For example, some recent research suggests that indigenous groups in pre-European contact America had much different pharmacopoeias than they do now, and the introduction of new diseases prompt- ed the discovery of new plant medicines in the relatively short amount of time after the Europeans’ arrival in the New World (Davis 1995). The simultaneous introduction of new diseases and new plants seems to potentially result in the expansion of medicinal plant use through scientific experimentation. Bennet and Prance (2000) found 216 post-European introduced plants that are a part of the indigenous pharmacopoeias of northern South America. Other recent ethnobotanical studies illustrate how indigenous people experi- ment with plants. For example, around 40% of the plants in the local pharma- copoeia in 1995 were unknown to the sixteenth-century Portuguese settlers when they arrived in the Madeiras; they were either introduced from Africa or the Americas or were native to the islands (Rivera and Obon 1995). Similarly, Paul (2002) found that 40% of Haitian medicinal plants belong to genera that have similar medical uses in West Africa. This illustrates the tendency to adopt new species of the same genus when confronted with changed ecological conditions brought about by migration. Some ethnobotanists focus on indigenous adaptations to these changed en- vironmental circumstances. They theorize about the origins of a pharmacopoeia and the experimental processes by which new medicinal plants are selected. Moer- man (1998) hypothesizes that plants used for medicine in traditional society will be those that are available, perennial, and widespread geographically, as well as easily noticed, large and distinctive. Perennials, especially larger tree species, live longer and thus generally have more secondary chemicals, making them more useful as medicinal plants. The other characteristics all have to do with the ease of finding and recognizing medicinal plants. Johns (1990) discusses the selection Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 15 of new medicinal plants focusing on the methods by which biological activity of plants i is discovered—including taste, odor, empirical testing, and the observation of anim Cox (1995) explains how Kuhn's (1965) work on scientific paradigms provides a method of understanding indigenous scientific processes. Indigenous healers operate in paradigms that differ from those of western science and these para- digms, in turn, direct the hypotheses and methodologies of indigenous experi- mentation (Cox 1995). Indigenous healers do, however, follow the scientific meth- od, making observations, developing theories, and then using this knowledge and experience to predict future outcomes. Origins and Evolution of the Hawaiian Pharmacopoeia.—In Polynesia, there is some debate about the origin of its herbal medicine tradition. Whistler (1992) has ar- gued that Polynesians had a limited herbal medicine tradition, especially in re- lation to medicines taken internally. He shows that prior to European arrival, Polynesians had few infectious diseases, and as a result, the medical system was focused around external ailments such as wounds, sprains, rashes, and infections. More complex diseases were treated with spiritual healing, and the plants used in this process were ‘‘fetid herbal preparations applied externally or taken inter- nally that were believed to have the power to repulse malicious spirits and to thereby heal’’ (Whistler 1992:37-38). Consequently, the complex Hawaiian medical system described by early native historians developed in the mid-nineteenth cen- tury in response to the introduced epidemic diseases and contact with medicine from European and other more developed herbal medical traditions, and “the widespread use of medicines for internal ailments that is so prevalent in Polynesia medicine today appears to have developed after, and as a result of, contact with the western world’ (Whistler 1992:38). Cox (1991), however, suggests that herbal medicine in Polynesia is a conser- vative tradition, in that most knowledge is gained from a mentor and not through experimentation (although he recognizes that new plant remedies are sometimes acquired in dreams). Cox extends these arguments to explain how Polynesian herbal medicine developed before western contact. He discounts early missionary histories relied on by Whistler and then supports this claim by noting the simi- larity between different medical traditions within Polynesia, the variation among different family traditions (a European introduced system would look more lo- cally homogenous), Polynesian belief in the endemic nature of their healing tra- ditions, and the fact that a majority of plants used medicinally in Polynesia are not used in other parts of the world. Abbott and Shimazu (1985:220) comment that Hawaiians ‘‘have a tendency to stick to tried and true remedies,” which mainly consist of plants the Polyne- sians carried with them throughout the Pacific. They note however, that many endemic plants are used, although less frequently, and that some plants intro- duced after European contact have also been added. This supports Cox’s notion of the endemic origin of Polynesian herbal medicine based on similarities between different Polynesian pharmacopoeias. Many of the processes observed in the evolution of traditional pharmacopoe- ias in other parts of the world apply to Hawaii. The introduction of infectious 16 PALMER Vol. 24, No. 1 diseases and new plants, the migration of other ethnic groups to Hawaii, and the decreasing availability of native plant species all had an effect on the pharma- copoeia. The Hawaiian pharmacopoeia, like many others, has had rapid devel- opment and change in the number and species of plants used in the relatively short period since European contact. While this study does not directly address the origin of the Hawaiian pharmacopoeia nor the importance of Polynesian in- troduced plants to it, it does reveal some of the processes behind the introduction of new plants to Hawaiian herbal medicine, presenting a dynamic view of the evolution of the Hawaiian ethnomedical tradition. Two methods for the discovery of new medicinal plants are evident in Hawaii: empirical practice and spiritual guidance. Handy describes the origins of medic- inal plant usage in Hawaii, It will be plain to anyone that makes even a superficial study of Hawaiian medicine that the system has arisen mainly through empirical practice. . . . Many Hawaiians will tell you that the system has been established through a trial and error method, but that the original knowledge of the healing qualities of various elements used has always been, and is still, revealed by the ancestral aumakua [spirits] in dreams. Handy (1934:16) Empirical practice is a research method based on experimentation and obser- vation, much like the scientific method. According to Larsen (1946:19), ancient Hawaiians were ‘developing a system of medicine that was built upon observa- tion, deduction, experimentation, and clinical trial,” but that ‘this system was lost as their whole culture succumbed to the pompous tide of European conquest” (Larsen 1946:19). From the observations made in this paper it is clear that the experimentation, observation, and deduction in Hawaiian traditional medicine continued long after European conquest. While this study focuses on the empirical processes behind the discovery of new medicinal plants, the spiritual roots of Hawaiian healing cannot be ignored: in Hawaiian healing traditions, spiritual guidance is an essential element in the discovery of new plant material. In the ethnoscientific paradigm of Ia‘au lapa‘au, dreams, visions, and inspired thoughts provide /a‘au practitioners with hypotheses from which they can begin the empirical testing process for new remedies (Cox 1995). REVIEW OF THE ETHNOBOTANICAL AND BOTANICAL LITERATURE Bidens spp.—The nineteen endemic Bidens species in Hawaii evolved through adaptive radiation from a single primal ancestor (Ganders 2000; Wagner et al. 1999). All the species are interfertile, but most do not occur sympatrically (Gan- ders and Nagata 1984). In modern times four species have been introduced: Bidens pilosa L., B. cynapiifolia Kunth, B. alba (L.) D.C., and B. gardneri Baker. Bidens pilosa, which first arrived sometime before 1845, is native to the tropical Americas (Wag- ner et al. 1999). Bidens cynapiifolia was first collected in 1929, almost seventy years later (Degener 3781 BISH), B. alba in 1958 (Pearsall s.n., BISH), and lastly B. gard- neri in 1983 (Hobdy 1883 BISH). Bidens alba and B. pilosa are both widespread in Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 17 low lying disturbed areas with B. alba becoming increasingly more abundant on Oahu (Wagner et al. 1999). Generally, a distinction is made between native species of Bidens, called ko‘oko‘olau or ko’olau, and the non-native B. pilosa, which is called kinehi. However, there is still some ambiguity and ko’oko’olau can be used to refer to any species of Bidens. For example, Kaiahua identifies ko‘oko’olau as B. pilosa. He also mentions that there are two kinds of ko‘oko‘olau, indicating that the term ko‘oko‘olau can be used to refer to more than one species (Kaiahua 1997). Comparing the use of native and introduced Bidens species is thus complicated because most texts make no attempt to distinguish among them. A source ascribed to Kupunihana, written around 1922, mentions three types of ko’oko’olau used as medicine for stomach ailments and numerous other ailments, indicating that many of the species within the genus were recognized and used medicinally (Chun 1998). None of the early texts, even into the early twentieth century, use the term kinehi (Chun 1994a, 1994b, 1998; Handy et al. 1934; Kaiakamanu and Akina 1922; Kamakau 1964). Nineteenth-century sources on Hawaiian herbal medicine mention that ko‘oko‘olau is useful for asthma, throat and stomach troubles, stimulating the ap- petite, general debility of the body, childhood ailments, as a purgative (with other plants), and as a tonic (Chun 1994a; Kaaiakamanu and Akina 1922). All of the treatments that were common in precontact times are for ailments that are easily recognizeable (sore throat, general debility, stomach troubles). Thus, following Whistler’s logic, they are more likely to have been treated with an herbal remedy during precontact times. The earliest mention on ko’oko’olau in the literature is in 1866 in Kamakau (1964) and the ‘Ahahui report in 1867 (Chun 1994a). After this, ko’oko’olau is men- tioned in a number of other studies on Hawaiian herbal medicines (Chun 1994a, 1994b, 1998; Handy et al. 1934; Judd 1997; Kaiakamanu and Akina 1922; Kamakau 1964; McBride 1979), whereas kinehi is only mentioned in one published sources (Kaiahua 1997). This suggests the native species were more widely used until recently. Ko’oko’olau was also used to treat a wide variety of endemic health con- ditions, suggesting that the native plant has been used since before European contact. It would appear from the widespread use of B. pilosa, its early introduc- tion, as well as the variety of Bidens species that are used, that B. pilosa has been in the pharmacopoeia for some time, even if the name kinehi appears only recently in the ethnobotanical literature. Plantago spp.—The genus Plantago is a similar example of substitution. Multiple native and introduced species have been reported to have medicinal use. The taxonomy of endemic plantains, however, is much simpler. There are three en- demic species: Plantago hawniensis A. Gray, P pachyphylla A. Gray, and P. princeps Cham. and Schlechtend. P. pachyphylla and P hawaiensis appear to be more closely related to each other than to P princeps (Wagner et al. 1999). There are five intro- duced species, including Plantago major and P. lanceolata. P major is widespread throughout the Hawaiian Islands and was first collected in 1864-1865 (Mann and Brigham 423, BISH). P lanceolata was first collected on Kauai in 1895 (Heller 2457, BISH). The other alien species are more recently introduced and less widely dis- tributed (Wagner et al. 1999). 18 PALMER Vol. 24, No. 1 The Hawaiian generic name for Plantago is laukahi. The native species are referred to as laukahi kuahiwi, indicating that they are found in the mountains. The native species also have other names that differentiate between them, manene for PR. pachyphylla, and ale for P princeps (Wagner et al. 1999). Laukahi is used as a purge for mother and babies, to draw the pus out of sores and boils, and to cure the diseases of pa‘ao‘ao and ‘ea (Chun 1994b, 1998) Other authors mention that laukahi is used as a tonic, a laxative, a poultice for boils and sores, and as a tea for diabetes, urinary tract infections, kidney problems, pul- monary disease, and high blood pressure (Gross 1998; Handy et al. 1934; Kaiahua 1997; Krauss 1979; Nagata 1970; Whistler 1992). Interestingly laukahi is used to treat ancient Hawaiian disease categories (pa‘ao‘a0) as well as more modern ail- ments (diabetes, high blood pressure) and external, internal, and ‘spiritual’ dis- ease categories, suggesting that new plants are not just adopted to treat new diseases. There is some disagreement on whether or not the native Plantago species were used in precontact Hawaiian herbal remedies. Krauss (1979, 1993, 2001) and Gross (1998) suggest that the introduced P major replaced the native Plantago species while Whistler concludes that Several authors have reported that these native plantains were the original medicinal species and that the introduced weed Plantago major has re- placed it, presumably because the latter is so common and easy to obtain. It is more likely, however, that Plantago major was the first and perhaps the only species commonly used in Hawaiian medicine. (Whistler 1992: 188) Given the widespread medicinal use of Plantago throughout the world, this is plausible (Basaran 1996; Bayon 2000; Henderson 1994; Johnson 1999; Ramos 1996). A children’s book on Hawaiian herbal medicine mentions that the Japanese, who migrated to Hawaii in large numbers, also use Plantago major in a similar manner to the Hawaiians (Corum 1985). Most of the immigrants who came to Hawaii arrived well before 1922 (Juvik and Juvik 1998), the date of the first study that specifically mentions and describes the medicinal use of laukahi lauli‘i (clearly identifying P major) (Kaaiakamanu and Akina 1922). This allows time for signif- icant interaction between Hawaiians and new immigrants, supporting Whistler’s (1992) claim. However, in support of the use of native Plantago spp., the earliest reference to laukahi is found in Kamakau’s writing about the people of old Hawaii. This work was published in a Hawaiian language newspaper series from 1866-1871, around the time the Chinese first immigrated but before they had moved through- out the Islands. Another early reference is found in a group of interviews con- ducted with a variety of healers in 1867 as part of the establishment of a native Hawaiian healers’ organization (Chun 1994a). Five different healers all used lau- kahi in many different ways. One healer, Pupuka, even specifically mentions the use of laukahi kuahiwi, referring to one of the native species. These sources seem to confirm the early widespread use of Plantago, definitely including the native species and possibly P major. The number of different people using Plantago spp. Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 19 and the early date of use all seem to indicate a switch from the use native Plantago spp. to the introduced Plantago major. After this 1867 mention of laukahi kuahiwi there is no specific mention of the use of the native Plantago species besides Krauss (1979, 1993, 2001) and Gross (1998) (who mention it being replaced with P major, not its contemporary use. Whenever Jaukahi is subsequently mentioned in research on Hawaiian herbal med- icine it is always identified as P major (Gross 1998; Handy et al. 1934; Kaiahua 1997; Kaiakamanu and Akina 1922; Krauss 1979; Nagata 1970; McBride 1979; Whistler 1992). Bushnell et al. (1950) suggest that P lanceolata is also used as medicine, however, the source material he cites—Kaaiakamanu and Akina (1922)—does not support this assertion. This information, plus the conspicuous absence of current references to all other Plantago spp., strongly suggests that P major is currently the only member of genus used in the Hawaiian pharmacopoeia and it probably replaced the use of the native species. Antibacterial Assays for Bidens spp. and Plantago spp—The widespread use of both Bidens spp. and Plantago spp. in traditional medicine has stimulated a significant amount of research on their biological activity (Table 1). These studies suggest a broad range of healing affects from both Bidens and Plantago species. Bidens has been screened for antibacterial activity, a treatment for liver diseases, and an anti- ulcer treatment; Plantago spp. has been tested for anti-inflammatory and wound- healing properties. Thus, the biological activity of both Bidens spp. and Plantago spp. is consistent with their uses in the Hawaiian pharmacopoeia, suggesting that Hawaiian healers possessed a sophisticated knowledge of biological activity. The biological activity of Bidens is due to large classes of compounds: poly- acetylenes and flavonoids.' Marchant et al. (1984) researched polyacetylene con- tent in Hawaiian Bidens species and found a surprising array of different mole- cules in different species. While all contained polyacetylenes in their roots, thir- teen species did not contain any polyacetylenes in their leaves, suggesting highly variable antibacterial activity within the native Bidens species. The native species Bidens macrocarpa (A. Gray) Sherff, Bidens populifolia Sherff, and Bidens campylotheca Schultz-Bip. all contained some level of polyacetylenes in their leaves, which in- dicates that the leaves have antimicrobial properties (Marchant et al. 1984). METHODOLOGY Interviews.—Semi-structured interviews were conducted with both specialists and nonspecialists.2 The purpose of the research was explained and verbal consent granted. The participants were questioned about their knowledge of both intro- duced and native species of Bidens and Plantago and their uses and relative effec- tiveness, as well as possible reasons for the preference of one species over another. When possible, the plants were examined in the field or shown to the participants. In other cases, the plants were described with salient features to distinguish be- tween species. All the species in question had some visible, easily distinguishable features (e.g., yellow versus white flowers, narrow versus rounded leaves). Antibacterial Assays.—The selection the plants species for the antibacterial assays was dependent on locally available species. The introduced B. pilosa, P major, and i) TABLE 1.—A summary of some of the pharmacological studies done on biological activity of Bidens spp. and Plantago spp. Plant Study Results B. pilosa Bushnell et al. 1950 Antibacterial—moderately effective; anti-enteric pathogens—moderately effective Wat et al. 1979 Polyacetelynes are phototoxic to bacteria and fungi N’Dounga et al. 1983 Antibacterial activity Machado et al. 1988 Antiparasitic activity Geissberger et al. 1991 Antihypertensive Sarg et al. 1991 Anti-ulcer activity Chih et al. 1995 Anti-inflammato Gupta et al. 1996 Tumor inhibition; cytotoxicity; DNA intercalation; brine shrimp toxicity Jager et al. 1996 Prostaglandin-synthesis inhibition Chih et al. 1996 Protect liver injuries from various hepatotoxins and have antihepatic agents Alvarez et al. 1996 Antimicrobial; cytotoxic; insecticidal i Rabe and Van Staden 1997 Antibacterial activity ae Brandao et al. 1997 Antimalarial = Pereira et al. 1999 Immunosuppressant activity ee Dimo et al. 1999 Antimicrobial activi Alvarez et al. 1999 Antimicrobial; not antidiabetic Tan et al. 2000 Possible anti-ulcer effects Ubillas et al. 2000 Antihyperglycemic Dimo et al. 2001 Hypotensive effect Khan et al. 2001 Antimicrobial Chang et al. 2001 Possible treatment for leukemia B. campylotheca Red] et al. 1994 Anti-inflammatory active polyacetylenes B. subalternans Ortega et al. 1998 Anti-inflammatory B. aurea La Case et al. 1995 Anti-ulcer agent Martin et al. 1996 Anti-ulcer agent B. alba Lopez et al. 2001 Does not modify insulin or glucose levels L ON ‘FZ TOA TABLE 1—(continued) Plant Study Results P. major P. lanceolata P. asiatica Bushnell et al. 1950 Ravn and Brimer 1988 Dat et al. 1992 Guillen et al. 1997 Ringborn et al. 1998 Michaelson et al. 2000 Hetland et al. 2000 Ikawati et al. 2001 Marchesan et al. 1998 Deliorman et al. 1999 Schmidgall et al. 2000 Tezuka et al. 2001 Antibacterial—moderately effective; anti-enteric pathogens—slightly effective Anti-inflammatory and analgesic activity No diuretic activi Component plantanajoside has possible antibacterial effects Anti-inflammatory; anti-ulcer by inhibition of COX-2 Possible wound healing effects Polysaccharide fraction PMI protects against pneumococcal infection Inhibited IgE-dependent histamine; possible asthma or allergenic disease uses Anti-inflammatory Hepatoprotective effects Moderate activity on treatment of irritated buccal membranes. Anti-inflammatory FOOT Jeuruns /S3utids ADOTOISONHLA dO TYNANOL 22 PALMER Vol. 24, No. 1 P. lanceolata were readily obtained from the lawns on the Brigham Young Univer- sity-Hawaii campus while P pachyphylla required long hike to the summit of the Ko‘olaus. Similarly, many of the native Bidens species are rare and occur in small, localized populations (Wagner et al. 1999) although they are generally more com- mon than the Plantago species. Voucher specimens were collected.* For Plantago spp. the antibacterial screening portion of this study includes only one native species, P pachyphylla (I was unable to collect Plantago hawaiensis and P princeps because they are endangered species and have very limited distri- butions), and the introduced species of P major and P lanceolata (Wagner et al. 1999). For Bidens spp., the three native species endemic to the northeastern section of the Ko’olau Mountains—Bidens populifolia, B. macrocarpa, and B. campyloptheca— were screened along with the two most common introduced species—B. pilosa and B. alba. An agar diffusion method (Ingraham and Ingraham 2000) was used to ana- lyze the antibacterial properties of native Bidens in comparison with the intro- duced species, and the native Plantago pachyphylla in comparison with the intro- duced P major and Plantago lanceolata. All eight plants were tested against Staphylococcus aureus and Escherichia coli bacteria. These microorganisms were used because they are common skin and digestive tract microorganisms and correspond with Hawaiian uses of Bidens and Plantago spp. They also offer examples of Gram-negative and Gram-positive bac- teria and are common test organisms, allowing for comparison with other anti- bacterial studies. (Bushnell et al. 1950; Locher et al. 1995). Fresh plant material including leaves and stems was blended and squeezed with a garlic press to obtain the plant extract. One or more plants were used depending on the amount of liquid extracted per plant. However, all plants of a given species used for antibacterial assays as well as separate voucher specimen were collected from the same site. No solvent was added. Plant extract (0.75 ml) was applied to a 3-mm diameter paper disk. The paper disk was placed in a petri dish of Mueller-Hinton agar inoculated with either the Escherichia coli (Migula 1895) Castellani and Chalmers 1919 or Staphylococcus aureus Rosenbach 1884. Each dish contained both positive and negative controls—a commercially prepared an- tibiotic disk (Erythromycin) and a blank paper disk. Five plates were prepared for each plant to be tested. The petri dishes were incubated for 24 hours at 37.5°C. The zone of inhibition was then measured in millimeters. It is important to note that an agar diffusion method does not offer a complete picture of the multiplicity of possible medicinal effects of the plants such as the potential interactions of plant compound in a living system or in conjunction with other plant mixtures. It does, however, offer some insight into the comparative biological activity against certain bacteria. RESULTS AND DISCUSSION Interviews.—Personal interviews, including those with both healers and nonheal- ers, provided the most important source of information regarding the introduc- tion of new plants into the Hawaiian pharmacopoeia. Today, kinehi (B. pilosa) and ko‘oko’olau (native Bidens) are used topically to cleanse wounds, heal sunburns and Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 23 insect bites, and to strengthen the body (often in conjunction with more powerful traditional medicine). While the names for the various species of Bidens are some- times used interchangeably, all of the people interviewed, healer and nonhealers, knew about the use of Bidens and recognized the difference between the native and introduced species, even if they had never used the native species. Students of the late Papa Auwae, a well-known Ia‘au lapa‘au practitioner, de- scribed how he taught that kinehi is more potent than the native ko‘oko’olau and is useful in treating a wider variety of illnesses. The ko’oko’olau is used as a general tonic, whereas kinehi is used to treat all manner of internal ailments. He also taught that the potency of the different Bidens populations is also dependent on the location from which the plant was gathered, possibly indicating a much more sophisticated knowledge of bioactivity of Bidens.* Another interview suggests that e plants were mixed to achieve various effects and possibly to control the po- tency. Some people prefer kinehi or ko‘oko’olau because of taste and custom. Taste, however, may also be an indicator of biological activity (Johns 1990). This suggests that people recognize the differences in biological activity among different spe- cies, which is an important factor in favoring the use of one species over another and the adoption of new plants into the pharmacopoeia. Authors on contemporary /a‘au lapa‘au practices most often cite availability as the reason for the introduction of new plants into the herbal medicine tradition (Judd 1997; Krauss 1979). Krauss (1979) specifically mentions that kinehi is used in place of the native ko’oko‘olau because of its availability. People interviewed also indicated that availability was an important factor in the decision to use the in- troduced plants. Many people using traditional medicine are aged and are un- willing to hike into the mountains to find the native Bidens and Plantago species, especially when a similar plant is growing in their front yard. Plants, especially those with healing properties, are seen as gifts from God, and it makes sense that God would make a useful plant readily available. Personal experience while gath- ering plants on Oahu for this project also supports the idea that the introduced plants are much more readily available. This is due, in part, to the degradation of native ecosystems. Native plants are limited to small areas higher up in the mountains (Juvik and Juvik 1998). In sum, in the case of Bidens, there is no simple substitution. The reality is much more complex. Both native and introduced species are still used, sometimes together, and the decision to use one or the other is based on the availability, biological activity, disease to be treated, personal preference (taste), and familiar- ity with the plant Although the use of P major is widespread, there is no use and little knowl- edge of any of the native Plantago species. The literature about contemporary la‘au ‘au practitioners does not mention the native species (Judd 1997; Kaiahua 1997; McBride 1979). While Krauss and Whistler comment on the switch from the native to the introduced Plantago (Krauss 1979, 1993; Whistler 1990), people interviewed for this study did not recognize the medicinal uses of the native Plantago species. The rarity of the native species compared to the ready availability of P major seems to account for the complete substitution of the native Plantago species with P. major. Another important lesson learned from Hawaiian /a‘au practitioners is that 24 PALMER Vol. 24, No. 1 the spiritual component of healing is absolutely essential. Four people mentioned that 80% of healing is “‘spiritual’’ while only 20% comes from the biological ac- tivity of the plants. In addition, plants are seen as gifts from God to aid in the healing process, so it could be said that all healing comes from God. Furthermore, divine inspiration is an essential source of knowledge about new medicinal plants. There are stories of people being inspired to use a new plant that they had never used medicinally before. This seems to be a common local explanation for the addition of new plants into the pharmacopoeia. Antibacterial Assays.—In the case of Bidens, both native and introduced species have antibacterial activities. The most activity was found in B. macrocarpa and B. pilosa. While the native species all exhibit some antibacterial activity, there is quite a bit of variability between the different species (Table 2). A study comparing B. campylotheca to B. pilosa for antimalarial activity found both to have antimalarial activity with B. pilosa being more effective (Brandao et al. 1997). This correlates with native Hawaiian perceptions of comparative biological activity of the native Bidens spp. and B. pilosa. The decline in use of the native species may be a function of the variable antibacterial activity (depending on the species) as well as decreas- ing availability. Bidens alba, although readily available on Oahu (more so than any other Bidens species), is rarely used medicinally. Its moderate antibacterial activity suggests that this maybe a function not only of its recent introduction to Hawaii but also its moderate biological activity. Although this study did not include al- ternative Bidens extracts (e.g., methanol, ethanol, EtOAc, petrol, CH,Cl), it is im- portant to note that other tests that have included them have shown significantly greater antibacterial activity (Jager et al. 1996; Khan et al. 2001; Rabe and van Staden 1997). Initially, it was not possible to detect variable antibacterial activity among the native and introduced Plantago spp. Given that native preparations of Plantago major usually include rock salt and that pharmacological studies of Plantago spp. found significantly more activity with methanol extracts (Guillen et al. 1997; Te- zuka et al. 2001), salt and methanol extracts of PR major were made and tested. However, still no antibacterial activity was found (Table 3). It is possible that the strains of bacteria have developed some kind of resistance to certain antibacterials or that the chemical constituents of P major vary between different populations depending on environmental and/or genetic conditions. Other kinds of biological activity besides antibacterial (anti-inflammatory, wound healing effects, etc.) could also explain the use of Plantago spp. in the Hawaiian pharmacopoeia. The widespread use of a variety of Plantago species in ethnomedical traditions throughout the world (Bayon 2000) and previous pharmacological studies suggest that P major and the Plantago genus (possibly including the endemic Hawaiian species) are biologically active. CONCLUSION Substitutions and additions of Plantago and Bidens species to the Hawaiian pharmacopoeia appear to be the result of myriad intermingled biological and cultural factors. This study indicates that some of these elements could include: TABLE 2.—Comparative antibacterial activity of native and introduced Bidens species. Native Introduced B. macrocarpa B. campylotheca B. populifolia B. alba B. pilosa E. coli Plant 9.8 + 0.2 8.0 + 0.8 — 16.3 + 9.4 8.0 + 1.1 (n = 5) (n = 4) (n = 5) (n = (n = 5) Erythromycin 174 +04 178 = 1.0 16.8 + 0.7 14.0 + 1.0 1622 fl (n = 5) (n = 5) (n = 5) (n = 5) (n = 5) S. aureus Plant 7.0 + 0.4 a2 G2 45215 3 8.8 + 0.4 (n = 5) (n = 4) (n = 2) (n = 1) on = Erythromycin 14.0 + 0.3 452 11 10.4 + 1.7 12.22 G7 11.0 + 0.3 (n = 5) (n = 5) (n = 5) (n = 5) (n = 5) Antibacterial activity indicated by radii of the rings of inhibition measured in millimeters; not detected (—): for all five runs rings of inhibition measured <0.5 mm Five trials were undertaken for each plant. Whenever one or more trial was unsuccessful (ring of inhibition <0.5 mm), that trial was not included in the statistical assessment; rejected data of this kind are indicated by the n value. F007 Jaws /Sutds KOO TOISONHLA JO TVNANOL 26 PALMER Vol. 24, No. 1 TABLE 3.—Comparative antibacterial activity of native and introduced Plantago species. Native Introduced P. pachyphylla P. major P. lanceolata E. coli Plant Eekibs ae Mle (n = 5) (n = 5) (n = 5) Erythromycin 16.0 + 0.5 16.0 + 1.5 17.2 + O04 (n = 5) (n = 5) (n = 5) S. aureus Plant a= — — (n = 5) (n = 5) (n = 5) Erythromycin 162.2704 10.5: > 14 10.8 + 0.9 (n = 5) (n = 4) (n = 5) Antibacterial activity indicated by radii of the ri f inhibiti d in millimet t detected oO (—): all five runs rings of inhibition measured <0.5 mm. : Five trials were undertaken for each plant. Whenever one or more trial was unsuccessful (ring of inhibition <0.5 mm), that trial was not included in the statistical t; rejected data s kind are indicated by the n value. of availability of the introduced and native plants, often as a result of environmental degradation; biological activity of the plant; and a variety of cultural factors, in- cluding the spiritual component of traditional Hawaiian healing, interaction with different immigrant groups, time of introduction, familiarity with a plant, and taste preference. Conversations with local Hawaiians about Bidens confirm these observations, which are also supported by the pharmacological literature, studies of modern Hawaiian ethnobotany, and other ethnobotanical literature that dis- cusses the adoption of introduced plants into a pharmacopoeia. If availability of a species is essential for its inclusion in a pharmacopoeia, then it becomes increasingly important to preserve the environments where in- digenous people live and gather medicinal plants so that cultural knowledge about plant use may be preserved. If indigenous medicinal plants are not avail- able, local healers will replace them with plants that are. These conclusions about availability and biological activity of Bidens and Plan- tago in Hawaii are consistent with existing theories of medicinal plant selection. As Moerman (1998) suggests, the plants adopted are abundant (available), wide- spread geographically, and distinctive. (Although, contrary to Moerman’s theories, neither species is very large nor perennial.) Abundant and widespread geograph- ically, weed species (such as B. pilosa and P major) are often highly competitive because they have high concentrations of active chemicals. This suggests that the factors that allow for their abundance and worldwide distribution also make them biologically active and thus effective medicine. Johns’s (1990) work discussing in- digenous observation and methodological techniques for selecting biologically ac- tive plants is also relevant to the discussion on how the Hawaiians selected plants with biological activity. Interviews with local Hawaiians illustrates that native healers are knowledgeable about comparative biological activity. Whistler (1992), Davis (1995), Rivera and Obon (1995), Bennet and Prance (2000), and Paul (2002) all show the dynamic interaction of ethnomedical systems Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 27 with both the physical and cultural environment—how indigenous scientists ex- periment and adapt to meet the demands of changing social and environmental systems. While all of these studies mentioned the rapid adoption of new plant species when people are introduced into a new environment, Paul’s (2002) work is interesting in that it supports the notion that new species of the same genera as known medicinal plants are more likely to be adopted into the indigenous pharmacopoeia. This research on B. pilosa and P. major supports that hypothesis; both plants could have been added because of the previous use of a similar plant in the precontact Hawaiian pharmacopoeia. Whistler (1992) and Davis (1995) imply that the introduction of new disease, subsequent population decline, and other factors that accompany contact with European cultures encourage the addition of new plants to indigenous pharma- copoeias. This study, by examining two plants apparently adopted around 100- 150 years ago, generally supports this hypothesis. However, it appears that neither P major and B. pilosa were adopted to treat introduced diseases. A more general study looking at all the introduced species in the Hawaiian pharmacopoeia would be more appropriate to evaluate the larger effects of introduced diseases and migration on the pharmacopoeias. While this research does not directly address the origin of the Hawaiian and Polynesian pharmacopoeias, the emphasis on the indigenous scientific process suggests that traditional herbal medicine did not arise solely after western contact. It suggests, rather that traditional knowledge of medicinal plants has a long his- tory in Polynesia but was profoundly affected by the cultural, biological, and ecological changes that occurred with the arrival of Europeans in Polynesia. Thus the discussion started by Cox (1991, 1995) becomes particularly impor- tant. The paradigm provides the pertinent questions and hypotheses that any group of scientists consider relevant. Science becomes the process of inquiry rather than the technological product. Ethnobiological research should, in realization of the importance of scientific paradigms, focus on understanding these processes. Studying the adoption of these introduced plants into the Hawaiian ethnophar- macopoeia, we can clearly observe the spiritual, pragmatic, active, and empirical processes involved in the science of /a‘au lapa‘au. By focusing on the introduction of two specific plants this paper clearly de- scribes some of the processes behind substitutions. By studying plants that have been recently added to indigenous pharmacopoeias, insight is gained into the methodology of indigenous healers. This has significant implications for viewing ethnobotanical knowledge as dynamic and active in the context of cultural change. On a larger scale, examining the nature of change in Hawaii helps to understand the processes that continue to happen throughout the world as eco- systems are degraded and pharmacological traditions adapt to the changing re- alities of the twenty-first century. NOTES 1 Dharmananda, Suhhuti. n.d. A popular remedy escapes notice of Western practitioners. [www.itmonline.org/arts/bidens.htm] (verified September 2003). 2 Interviews were carried out from January to April 2002. Having been raised in Ko’olauloa, 28 PALMER Vol. 24, No. 1 some of the people interviewed were already acquaintances. Through those I already knew I branched out and found others they recommended as having some knowledge about Hawaiian medicinal plants. The only people who considered themselves healers and who have active herbal medical practices were Bula Logan and Alapa‘i Kahu’ena. The other interviews were with people who had widely differing experience with medicinal plants (personal experience while growing up, some formal training, etc.): their recollection and knowledge of the plants and their uses varied greatly. The participants were generally between 30 and 60 years old. Those interviewed were Gladys Ahuna, Millie Enos, Kawika Eskeran, Alapa’i Kahu’ena, Norman Kaluhiokalani, Bula Logan, Henry Na‘awao, Howard K.K. Pali, Harold H. Pukahi, Rueben H. Pukahi, Kapua Sproat, Ipolani Thompson, Kamo’s‘e Walk, and William Wallace. I would like to thank all of the people above who shared their knowledge and time in order help me better understand and appreciate la‘au lapa‘au. * Voucher specimens were collected and identified by Clyde Imada at the Bishop Museum Herbarium and were deposited at the Waimea Arboretum and Botanical Gardens, 59-864 Kamehameha Hwy, Haleiwa, Hawaii 96712. * Several of the people interviewed including Alapa’i Kahuena, Kapua Sproat, and Norman Kaluhiokalani had studied for some time under a well know Hawaiian healer, Papa Auwae. Much of what they knew came from their experiences with him. REFERENCES CITED pharmacopoeias of northern South America. Economic Botany 54:90-102 Brandao, M.G.L., A.U. Krettli, L.S. Soares, .G.C. Nery, and H.C. Marinuzzi. 1997. Antimalarial activity of extracts and fractions from Bidens pilosa and other Bidens species (Asteraceae) correlated with the presence of acetylene and fla- Abbott, Isabella A. and Colleen Shimazu. 1985. The geographic origin of the ee most commonly used for medi- cine by Hawaiians. pay of Ethnophar- macology 14:213-222 Alvares, L., S. Marquina, MLL. Villarreal, D. Alonsa Aranda, 1996. Bioactive polyacetylenes from Bi- dens pilosa. to Medica 62:355-357. M.A. Seville, and MJ. 1999. Y Gaatsie antisecretory and antiulcer activities of an ethanolic extract of Bidens pilosa L. var. radiata Schult. yi Journal of Ethnopharmacology 67:333-34 Basaran, ee EW. Ya, age? Plewa, and D. Anderson. 1996. vestigation of some Turkish herbal wiadiaincs in Sal- monella typhimurium and in the line ania assay in human lymphocytes. Teratogenesis, ere and Muta- genesis 16(2):125- apie N.D., M.N. sok A.M. Arambar- and SZ. Vina. 2000. Anatomia de las edicinales de la provincia Pampeana: ches, L. (Plantaginaceae). ring farmaceutica-bonaerense 19(4):263- Bennet, Bradley C., and Ghillean T. Prance. 2000. Introduced plants in indigenous sg 1950. The antibacterial ae fe of some plants found in Hawaii. Pa- cific ging i (July):167-183. pe iS, hiang, C.C. Chen, L.T. Liu. KC. wae and L.C. Ching. 2001. Antileukemic activity of Bidens pilosa L. var. minor (Blume) Sherff and Houttuy- nia cordata Thunb. American Journal of Chinese Medicine 29(2):303-312. Chih, H.W., C.C. Lin, and K S. Tang. 1995. Anti-inflammatory activity of Taiwan folk medicine ““‘ham-hong-chho” in rats. American Journal of Chinese Medicine 23(3-4):273-278. - 1996. The hepatoprotective effects of Taiwan folk medicine Ham-Hong- Chho in rats. American Journal of Chinese Medicine 24(3-4), 231-240. Chun, Malcolm N., tr. 1994a. Must We Wait Spring/Summer 2004 in wig the 1867 of the Ahahui La‘au au of Wailuku, Maui on Native Hawatian "Health, First iPeopies Produc- tions, Honolulu. , ed. 1994b. Native Hawaiian Medi- cines: Originally Published as Hawaiian Herbs of Medicinal Values Published by the Troi, Board of Health in 1922. First People’s Productions, pire ulu. < aed 1998. ve Hawaiian Medicine, vol Il. First People's Produc- tions, Honolulu. — A.K. 1985. Folk een from Ha- i. Bess Press, Hono Cov, P Paul A. 1991. Retriinaien herbal med- e. In Islands, Plants, and Polynesians: an ekitioiaction to Polynesian Ethnobotany, eds. Paul A. Cox and Sandra A. Banack, pp- 147-168. Dioscorides Press, Port- 995. Shaman as scientists. In Pro- ceedings of the Phytochemical Society in ope: Phytochemistry of Plants Used in Traditional Medicine, eds. Hostett- mann, A. Marston, M. Maillard, and M. Hamburger, pp. 1-15. Clarendon Press, Oxford. ee in N.N. Ham, D.H. Khac, N.T. Lam, T. Son, D.N. Van, M. Grabe, R. Johans- son, G. Lindgren, and N.E. Stjernstrom. 1992. Studies on the individual and combined diuretic effects of four Viet- namese traditional herbal remedies Zea mays, Imperata cylindrica, Plantago major, and Orthosiphon stamineus. Journal of Ethnopharmacology 36:225-231. Davis, E. Wade. 1995. Ethnobotany: an old a Discipline, ed. Richard Evans Schultes, pp. 40-51. sept i Portland. Deliorman, D., E E. Yesilada. 1999. A seliinth ay on hepato- protective ats of some Turkish folk remedies by using pentobarbital in- duced hypnosis anodes mice. Journ of Faculty of Pharmacy of Gazi bideiesity 16(2):77-8 Dimo, T., J. roe PV. Tan, J. Pellecuer, G. ano. 2001. chloride extracts of Bidens pilosa on fruc- tose-hypertensive rats. Journal of Ethno- pharmacology 76:215-221. Dimo, T., T.B. Nguelefack, P. sere E. Dongo, A. Rakotonirina, and S.V. Rakotonirina. 1999. Hypotensive effects JOURNAL OF ETHNOBIOLOGY 29 of a methanol extract from Bidens pilosa Linn. on hypertensive rats. Comptes ren- dus de l’Academie des oleae serie III sci- ences de la vie 322(4):323-329. Ganders, ER., M. Berbee, and M. Pirseyedi. 2000. ITS base sequences in Bidens (As- teraceae): evidence for the continental relatives of Hawaiian and Marquesan Bidens. Systematic Botany 25(1):122-133. Ganders, ER., and Na agata., K.M. 1984. The role of hybridization i in the evol eraceae). In Biosys- tematics, ed. W.F. Baas pp. 179-194. Ontario. Geissberger, P., and U in. 1991. Con- stituents of Bidens pilosa L.: do the com- ponents found so far explain the use of this plant in traditional medicine. Acta Tropica 48(4):251-262. Gross, Annes. 1998. La‘au lapa‘au: an Intro- ductory Guide to Hawaiian Medicinal Plants, rev. ed. Special Publication, no. 8. Center on ing, School o blic Health, University of Hawaii, Honolulu. Guillen, M., J.A. Da Silva Emi ionge, G.A. Kile, A. manian medicinal plants for brine shrimp toxicity, crown gall tumor inhi- a cytotoxicity and DNA intercala- n. International Journal of Pharmacog- rie 34(1):19-27. — — Craighill, Mary K. Pukui, and rmore. 1934. Outline of Hawniian Plapioal Therapeutics. Bishop Museum Bulletin 126. Krauss Reprints, Honolu- I u. Henderson, H.M. 1994. The physicians of Myddfai: the Welsh herbal tradition. Bo- tanical Journal of Scotland 46(4):623-627. Hetland, G., B.A. Samuelsen N, V.M. Los- lash, S.B. peiteet S.I. Aaberge, E. Groeng, and E.T. Michaelsen. 2000. Pro- tective effect of Plantago major L. pectin polysaccharride against systematic Streptococcus pneumoniae infection in mice. ener Journal of Immunology 52(4):348-35 Ikawati, Z., S. ona and K. Maeyama. 2001. Screening of several Indonesian medicinal plants for their inhibitory ef- 30 PALMER wen on histamine release from RBL-2H3 s. Journal of Ethnopharmacology 75: pri Ingraham, 1c, d C.A. Ingraham. 2000. Introduction re Microbiology. Brooks/ Cole Thomson Learning, Stamford. hee A.K., A. Hutchings, and J. a Stad- 1996. Screening of Zulu medicinal nee for prostaglandin-synthesis in- — Journal of Ethnopharmacology 52: 95- Johns, i 1990. The Origin of Human iet and Medicine: Chemical Sasi Uni- sisi of Arizona Press, Tuc Johnson, T., 1999. The CRC E ioe era Desk . CRC Press, Boca Raton. Judd, Nanette Kapulani Mossman. 1997. aau lapaau: a Geography of Hawaiian Herbal Healing. Ph.D. Dissertation (Ge- ography), nea of Hawaii at Man- oa, Honolulu. Juvik, S. P. and J.O. Juvik. 1998. Atlas of Ha- waii, 3rd ed. University of Hawaii Press, Honolulu. Kaaiakamanu, D.K., and J.K. Akina. 1922. =_— Herbs of Medicinal Value Found mong the Mountains and a in the Hawaiian Islands, and Known to the Ha- waiians to Possess Curative and. Palliative Properties Most Effective ing Physical Ailments. Territorial Board of Health, Honolulu. Kaiahua, Kalua. 1997. Hauniian Healing Herbs: a Book of Recipes. Ka‘imi Pono Press, Honolulu. Kamakau, Samuel M. 1964. Ka poe kahiko: the People of Old, tr. Mary K. Pukui. Spe- cial Publication 51. Bishop Museum Press, Honolulu. Khan, M.R., M. Kihara, and A.D. Omoloso. 2001. soe oor paragon “ sae pilosa, Bischofia javanica, Elmerillia pap- uanan, = Sigesbekia cricntalie saci pia 72: Krauss, Beatrice H. 1979. Native Plants Used dicine in Hawaii. Lyon Arboretum, Honolulu. . 1993. Plants in Hawaiian Culture. University of Hawaii Press, Honolulu. . 2001. Plants in Hawaiian Medicine. Bess Press, Honolulu. Kuhn, Thomas. 1965. The Structure of Sci- entific Revolutions. University of Chicago Press, Chicago. La Case, C., CMI Martin, C. Alarcon De La Lasra, V. Motiva, M.J. Ayuso, C.C. Vol. 24, No. 1 Martin, and V.A. Lopez. 1995. Role of 0 Biosciences Sen gh teenie P..19 Historical Society for the Year 1944, pp. 27-44. Hawaiian Printing Co., Honolu- lu. Locher, C.P., M.T. Burch, H.E Mower, J. Be- restecky, H. Davis, B. Van Poel, A. La- sure, D.A. Vanden e, and A.J. sdaed tinck. 1995. Antimicrobial and an complement activity ~ extracts a tained from __ selec Hawaiian medicinal plants. piaea of Ethnophar- macology 49:23-32. Lopez, G.R.L., P.M.C. Ventura, R.M. Ro- extracto hidralcoholico de Bidens alba en rata normales y con diabetes Aloxanica. Acta farmaceutica bonaerense 20(2):89-93. Machado, J.O., E. Dos Santos, and A EV. Lefevre. 1988. Antibacterial activity of Bidens pilosa L. extracts. Revista de cien- cias farmaceuticas 10:55-62. Marchant, Y.Y., ER. Ganders, C. Wat, and G.H.N. Towers. 1984. Polyacetylenes in Hawaiian Bidens. Biochemical Systematics and Ecology 12(2):167-178. Marchesan, M., D.H. Paper, S. Hose, and G. Franz. 1998. Investigation of the anti-in- flammatory activity of c> ai extracts of Plantago sree L. Phytot search 12(1):s3. Martin, C.M., C. La espe V. Motliva, A. Lo- pez, and C. Alarcon de la Lastra. 1996. Healing process induced by a flavonic fraction of Bidens aurea on chronic gas- tric lesions in rats. Role of angiogenesis and neutrophil inhibition. oo fiir Naturforschung 51(7-8):570-577 McBride, L.R. 1979. Practical Folk Medicine of Hawnii. Petroglyph Press, Hilo. Michaelsen, T.E., A. Gilje, A.B. Samuelsen, K. Hogaasen, and B.S. Paulsen. 2000. In- teraction between human complement and a pectin type polysaccharide frac- tion, PMII, from the leaves of rgb major L. ices Journal of Immu- nology 52(5):483-4 Moerman, Daniel E. gore Native North American food and medicinal plants: epistemological considerations. In Spring/Summer 2004 Plants for Food and Medicine, eds. H.D.V. Prendergast, Nina L. Etkin, D.R. Harris, and PJ. Houghton, pp. 69-74. Royal Bo- tanical Gardens, Kew. Nagata, Kenneth M., 1970. Hawaiian me- dicinal plants. Economic Botany 25:245— 254. N’ agg M., G. Balansard, A. Babadja- an, P.T. David, and M. Gasquet. 1983. Contiibition a l’étude de Bidens pilosa L. identification et activité antiparasi- taire de la phényl-1-heptatriyne-1,3,5. Plantes médicinales et phytothérapie 17(2): 75 64-75. Ortega, C.A., A.E. Rotelli, and J.C. Gianel- ad 1998. Chemical components and -inflammatory activity from Bidens olan Planta Medica 64(8):778. Paul, Alexandra. 2002. Origi of the Hai- Ethnopharmacopoeia. Ph.D. Dis- sertation (Ecology, Evolutionary and Conservation Biology), Columbia Uni- New York. ee ait munosuppresive an flammatory effects of menthanolic ex- tract and the es poe tl isolated from Bidens pilosa L. Immunopharmacolgy 43(1):31-37. Rabe T., and J. Van Staden. 1997. Antibac- terial activity of South African plants used for medicinal purposes. Journal of -87. Journal of Ethnopharmacology 52(3):123- 127. Ravn, H. and L. Brimer. 1988. Structure and antibacterial activity of plantamajoside, a caffeic acid sugar ester from Plantago major subsp. major. Phytochemistry 27: 3433-3437. Redl, K., W. Breu, B. i and R. Bauer. 1994. Anti-inflamm i Planta Medica 60(1):58-6 Ringborn, T., L. Segura, Y. Noreen, P. Per- era, and L. Bohlin. ay Wieck acid JOURNAL OF ETHNOBIOLOGY 31 from Plantago major, a selective inhibitor of seme daar catalyzed prosta- gla ewer Journal of Natural Products 61:1212-1215 Rivera, D., and C. Chen. 1995. The ethno- pharmacology of Madeira and Porto Santo Islands, a review. Journal of Eth- nopharmacolgy 46:73-93. Sarg, T.M., A.M. Ateya, N.M. Farrag, and RA. Abbas. 1991. Constituents and bi- ras nee eooued of Bidens pilo ypt. Acta Pharmaceutica Hungarica ie sei - J., E. Schnetz, 2000. Evidence for eae effects of polysaccharides an containing h hesion assay on eal membranes. Planta Medica Pa 48-53. = and E. Dongo. 2000. Ef- fects of fates ine cyclohexane, and methylene chloride extracts of Bidens pi- losa on various gastric ulcer models in rats. Journal of Ethnopharmacology 73: 15-421. Tezuka Y., S. Irikawa, T. Kaneko, A.H. Ban- nese herbal drug extracts for inhibitory activity in nitric oxide production and identification of an active compound of Zanthoxylum sigh pelt —— of Eth- 09-21 Jolda, J. .M. nic glucosides from Bidens pilosa. Planta Medica 66:82-83. _— W.L., D.R. Herbst, and S.H. Soh- 999. Manual of the Flowering Plants of Satie vols. 1 and 2, rev. ed. Bishop useum Press and University of Ha- waii Press, cere lu. Wat, C.K., R.K. Biswas, E.A. Gaham, L. , G.H. Towers, and E.R. Waygood. 1979. Ultraviolet-mediated cytotoxic ac- tivity of phenlyheptatriyne from Bidens pilosa L. Journal of Natural Products 42(1): 103-111. Whistler, W. Arthur. 1992. Polynesian Herbal Medicine. National Tropical Botanical Garden, Lawai noe Journal of Ethnobiology 24(1): 33-49 Spring/Summer 2004 NOMENCLATURE OF BREADFRUIT CULTIVARS IN SAMOA: SALIENCY, AMBIGUITY, AND MONOMIALITY IANE RAGONE,? GAUGAU TAVANA,? JOAN M. STEVENS,» pains ANN peeiin REBEKKA STONE,? PAUL MATTHEW COX: PAUL ALAN COX @ National Tropical Botanical Garden, 3530 Papalina Road, Kalaheo, HI 96741 » California State University, Fullerton, Fullerton, CA 92834 © West Dermatology, Santa Barbara, CA 93110 4 New York Botanical Garden, Bronx, NY 10458-5126 © Brigham Young University, Provo, UT 84602 ABSTRACT.—Breadfruit is an important subsistence crop in the Samoan archi- pelago, where numerous cultivars are grown and used. The diversity of breadfruit in Samoa is indicative of its antiquity and value to this society. The purpose of our study was to document and compare knowledge of breadfruit names by Sa- moans of a wide range of ages in both rural villages and towns and to test the relationship between saliency and binomiality. A total of 354 people were inter- viewed and 46 cultivar names were recorded. A binomial is used to name a breadfruit—the generic term ‘ulu is given first and a second word is added to describe that particular cultivar—when the second word used alone could refer to something other than breadfruit. A monomial is used only when this term does not refer to anything else or has no other meaning. There was no significant relationship between saliency and binomiality of breadfruit names and a signifi- cant relationship between binomiality and linguistic ambiguity. A useful outcome of this study was defining 60 Samoans as “experts” with statistical measures that we will use in continuing ethnobotanical studies in Oceania and that may have broader application. Key words: breadfruit, Artocarpus altilis, Samoa, tropical crop cultivars, ethnotax- onomy. RESUMEN.—E] Arbol del pan es un cultivo importante para la subsistencia en el archipiélago de Samoa, donde se cultivan y utilizan numerosas variedades. La diversidad de Arboles del pan en Samoa es indicativa de su antigiiedad y valor para esta sociedad. El propésito de nuestro estudio era documentar y comparar el conocimiento de los nombres de variedades de los arboles del pan entre Sa- moanos de una amplia gama de edades en aldeas y pueblos rurales y examinar la relacién entre la importancia cultural y utilizacion de binomios. Se entrevisté un total de 354 personas y se registran 46 nombres de variedades. Se utiliza un binomio para denominar un cultivar de arbol del pan—al término genérico ‘ulu y se le agrega una segunda palabra para describir ese cultivar particular—cuando al utilizar la segunda palabra sola podria entenderse algo distinto a los arboles del pan. Se utiliza un monomio solamente cuando este término no se refiere a ninguna otra cosa ni tiene ningun otro significado. No encontramos ninguna re- lacién significativa entre la importancia cultural y la utilizacion de binomios y la ambigiiedad lingiifstica. Durante este estudio utilizamos medidas estadisticas 34 RAGONE et al. Vol. 24, No. 1 para definir a 60 Samoanos como “expertos.” Esta técnica servira para la contin- uacién denuestros estudios ethnobotanicos en Oceania y pueden tener una apli- cacién mas amplia. RESUME.—L’arbre 4 pain demeure une espéce importante en agriculture de subsistance dans l’archipel de Samoa, ot. de nombreux cultivars sont utilisés et cultivés. La diversité de l’arbre 4 pain en Samoa est un indice de son antiquité et de son importance pour cette société. Le but de notre étude était de documenter et de comparer la connaissance portant sur les noms de I’arbre a pain parmi les Samoans provenant d’un large éventail d’ages et issus autant des villes que des villages ruraux. Le rapport entre l’importance culturelle et la bindmialité a éte vérifié. Un total de 354 personnes ont été interviewées et 46 noms de cultivars ont été enregistrés. Un binéme est employé pour désigner un arbre a pain: le terme générique «’ulu» est donné d’abord, puis un deuxiéme mot est ajouté pour décrire ce cultivar particulier. Le deuxiéme mot utilisé seul pourrait cependant se référer a autre chose que l’arbre 4 pain. Un monéme est employé seulement lorsque ce terme ne se rapporte pas a autre chose ou n’a aucune autre acception. Nous n’avons trouvé aucun rapport significatif entre l’importance culturelle et la binémialité des noms de I’arbre a pain, mais il existe un rapport significatif entre la binémialité et l’ambiguité linguistique. Des mesures statistiques ont été utilisées afin de qualifier «experts» 60 Samoans. Cette approche nous sera utile lors de nos prochaines études ethnobotaniques en Océanie. Elle pourrait avoir de plus larges applications. INTRODUCTION Agricultural people throughout the world typically recognize and name nu- merous forms or varietals of important domesticated plant species. These folk specific taxa are typically distinguished by subtle morphological differences such as color, relative size, shape, habit of growth, etc. (Berlin 1992). Culturally salient plants—those species and cultivars that are well known throughout a culture and are easily recognizable—have been the subject of much discussion and debate (e.g., Atran et al. 1997; Berlin 1986, 1992; Brown 1985, 1986, 1987). It has been suggested that highly salient taxa should be named with a monomial (Berlin 1992), while the greater specificity possible in a binomial should be used to dis- tinguish closely related taxa, especially cultivars of domesticated plants (Hays in Brown 1985). This idea is roughly analogous to the use of short telephone num- bers in Western societies to refer to highly salient services, such as 911 for the Police or 411 for Information, while longer numbers are used to distinguish be- tween the numerous Jones families that appear in the telephone directory. In this paper, a data set of 350 interviews with Samoans concerning knowl- edge of breadfruit names is statistically analyzed. We wanted to see if there are general patterns in the names applied to breadfruit cultivars, such patterns being a component of folk taxonomy in general (Berlin 1992), in an effort to determine both consistency and hierarchical diversity in the folk nomenclature of breadfruit The Samoan archipelago lies in the central south Pacific Ocean. It is divided into two political entities: the independent nation of Samoa (formerly Western Samoa, which changed its name in 1997) with the principal islands of ‘Upolu and Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 35 _ : »% J 2 : ? ate = Pe ox ——_ a. ~ ras ‘Oe FIGURE 1.—Ma’afala, a common Samoan breadfruit cultivar, growing by a residence in Saipipi Village, Savai’i. Photograph by Diane Ragone. Savai’i, as well as two smaller inhabited islands, Manono and Apolima. The east- ernmost islands are part of American Samoa, an unincorporated territory of the United States, comprised of five inhabited volcanic islands (Tutuila, Aunu’u, and the Manu’a Group of Ofu, Olosega, and Ta’u), and two coral atolls (Swains Island and the uninhabited Rose Atoll). Breadfruit, Artocarpus altilis (Parkinson) Fosberg (Moraceae), is an important subsistence food crop in Samoa and trees are grown around residences (Figure 1) in all villages and in the towns (Ragone 1997; Whistler 2000). A census in 1989 estimated that 89 percent of agricultural households grew breadfruit and an es- timated single crop equivalent area of 1000 ha of land was in cultivation (Ward and Ashcroft 1998). An aboriginal introduction, breadfruit has been an important component of Samoan subsistence agriculture for more than three millennia as part of a suite of crops that includes coconuts (Cocos nucifera L.), bananas (Musa sp.), taro (Colocasia esculenta (L.) Schott and Alocasia macrorrhiza (L.) D. Don), yams (Dioscorea sp.), ‘ava (Piper methysticum Forst. f.), and sugarcane (Saccharum offici- narum L.). Names, and in some cases descriptions, for as many as 30 Samoan breadfruit cultivars have been recorded by various visitors to Samoa since the 1840s (Ragone 1995). Cultivars are recognized and distinguished based on various morphological characters such as fruit shape and size, skin texture, flesh color, presence of seeds, leaf shape (especially degree of dissection or lobing), and tree form, or by fruit attributes related to cooking or storage qualities (Ragone 1997). The purpose of this study was to document and compare knowledge of breadfruit names by 36 RAGONE et al. Vol. 24, No. 1 females and males of a wide range of ages in both rural villages and towns in Samoa. Our intent is to contribute not only to an understanding of breadfruit names in Samoa, but also to test the relationship between cultivar saliency and binomiality. METHODS Interview Techniques.—In July 2000, 354 Samoans in Samoa and American Samoa were interviewed about their knowledge of breadfruit names. Seven villages or towns were chosen for study: in independent Samoa, Saipipi and Falealupo on Savai’i and Apia (the capital city) on ‘Upolu; and in American Samoa, Olosega and Ofu, Manu’a Group, Pago Pago (the capital city) and Afono on Tutuila Island. The interviews were conducted in the Samoan language by two-person teams and the responses were recorded on a standard form. In each village the teams walked to dwellings and work areas, interviewing any person who agreed to be interviewed. Interviews were conducted in homes, markets, and other areas of work and transit in the towns. In addition to residents, several expatriate Samoans from New Zealand and the United States who were visiting their families were interviewed. The age, date and place of birth, gender, occupation, place of resi- dence, and marital status of each person interviewed were recorded. Each person was asked to name as many different cultivars of breadfruit as they could, to- gether with information about local availability of each cultivar. The names were then read back to the respondent to ensure accuracy and to provide them with the opportunity to add any additional names. One group interview with 43 Samoan chiefs (matai) was conducted during a chief’s council meeting after an ‘ava ceremony in Falealupo, Savai’i. In addition, 16 individuals in Apia with conservation management responsibilities in govern- ment or NGOs (nongovernmental organizations) were interviewed. The latter 16 interviews were not pooled with the rest of the data to allow a comparison be- tween the two data sets. Several of the respondents, particularly those deemed by our statistical pro- cedures to be “‘experts’’ (see below), were interviewed at length to elicit detailed information about uses, cultivation practices, descriptions, and naming rules or patterns of rules used to name breadfruit cultivars, but these data are not reported here. Voucher specimens of breadfruit cultivars were collected and deposited at the National Tropical Botanical Garden (PTBG).! Recording of Data; Definitions of Idiosyncratic and Expert Respondents.—Interview data were entered into a spreadsheet on a portable computer in the field and grouped according to village. The breadfruit cultivar names recorded in the interviews were ranked by order of frequency of mention (Table 1). Breadfruit names in Samoa consist of either a binomial composed of a generic level term ‘ulu modified by a specific epithet or a monomial in which only the specific level epithet is used and ‘ulu is understood. If there was variation in the binomial or monomial form of a breadfruit cultivar name, e.g., ma’afala and ‘ulu ma’afala, the form used by the majority of the respondents was selected. Where there were slight differences in spellings or pronunciations, a standardized spelling / pronunciation used by the TABLE 1.—Frequency of breadfruit cultivar names recorded during interviews with 350 Samoans. Cultivar Number of Type of name respondents* % Rank? name‘ Translation? ma’ afala (308) 315 90 1 UM — ‘ulu ma’afala (7) puou (283) 286 81 fs UM — ‘ulu puou (3) aveloloa 238 68 3 UM _ maopo (214) 218 62 4 UM i ‘ulu maopo (4) ‘ulu ma’‘a 195 56 5 AB rock, hard? ‘ulu ea (185) 194 55 6 AB Uvea Island? ‘ulu wea (9) ‘ulu manu’a 131 oF 7 AB Manu’a Islands? momoleg 116 33 8 UM egg yolk! ‘ulu sina (79) 80 22 9 AB white! ‘ulu asina (1) sagosago (55) 59 17 10 UM — ‘ulu sagosago (4) peti (42) 56 16 11 AM fat! ‘ulu peti (14) ‘ulu tala (51) 54 15 12 AB spiny’ ‘ulu talatala (3) a puou (34) 38 11 13 UM wants to be a puou* fa’a fia puou (3) ‘ulu fia puou (1) ‘ulu fefelo (22) | 9 14 UB at *felo (11) ‘ulu initia 29 8 15 AB India breadfruit? FOOT Jeuruns /Butids ADOTOISONHLA JO TVNINOL TABLE 1—(continued) Cultivar Type of name % Rank? name* Translation* “ulu fau 26 7 16 AB fibrous® mase’e (21) 23 z 17 UM — ‘ulu mase’e (2) ‘ulu se’e 19 3 18 AB sliding* ‘ulu kiripati/kilipati 11 3 19.5 AB Gilbert Islands? gutufagu (9) 11 3 19.5 UM neck of the bottle’ ‘ulu gutufagu (2) puou fatu 10 S$ 21 UM seedy puou' ‘ulu falaoa 8 2 22 AB loaf of bread! vasivasi (5) 6 1 23 UM — ‘ulu vasivasi (1) puou tala 5 1 24 UM spiny puou' ‘ulu fiti 4 1 26 AB Fiji? tui tu 4 1 26 AM spiny! fia maopo 4 1 26 UM wants to be m ‘: puou maopo 3 1 28.5 UM puou that looks like maopo* maualuga 3 1 28.5 AM high® ‘ulu faga 2 1 31 AB eel trap! malali 2 1 31 AM smooth! matatetele (1) 7 1 31 AM big eye! ‘ulu matatetele (1) ma‘afala tala 1 0 39.5 UM spiny ma‘afala' puou tutunu 1 0 399.5 UM roasting puou® ‘ulu toso 1 0 39.5 AB ‘ulu to’elau 1 0 995 AB Tokelau Islands? ‘ulu tau 1 0 95 AB pluck® Te #2? ANOOVA I ‘ON “BZ “IPA TABLE 1—(continued) Cultivar Number of Type of name respondents* %o Rank? Translation® ‘ulu sasalapa 1 0 39.5 AB custard apple!’ puou fefelo 1 0 39.5 AM puou that looks like fefelo* ‘ulu fagaloa 1 0 39.5 AB Fagaloa village? avesasa'a 1 0 39.5 UM a ‘ulu pase’e 1 0 395 AB lazy® ‘ulu mama 1 0 39.5 AB light weight® segatoa 1 0 5 UM aa po’eloa 1 0 39.5 UM -- fia ta 1 0 39.5 AM wants to be slashed® * Number of respondents (in parentheses) who listed a binomial or — variant » Ties are scored by using the average of the ranks of tied agp eg., (19 + 20)/2 = yi 8 5, (33: 46)/14 = 39.5. © Binomial / —— names as identified by Samoans. UM = mbiguous monomial, AM = ambiguous monomial, AB = ambiguous binomial, UB = unambiguous bin 4 Definition of the different names of breadfruit cultivars. 1 = appearance, 2 = —_ origin, 3 = culinary properties, 4 = comparative, 5 = respect term, 6 = descriptive action involving breadfruit. See text for explanation of categories #007 Jeuruns /8utids ADOTOISONHLA JO TVNYNO! 40 RAGONE et al. Vol. 24, No. 1 TABLE 2.—Examples of Samoan breadfruit cultivar monomials and binomials. Generic term Specific modifier Binomial ‘ulu sina sn Pe with respect term fa’atau sina Monomial ma’opo Sonosbial with two words in specific epithet oO fia puou @—generic term is understood. majority of the respondents, e.g., ‘ulu ea and ‘ulu uea, was adopted.’ In these cases, the names were scored together for statistical purposes. Before analyzing the data, idiosyncratic responses and interviews were re- moved and expert respondents were identified. A breadfruit name was regarded as idiosyncratic if it was mentioned by only one respondent, unless that respon- dent was an expert as defined below. The interview of any respondent who men- tioned two or more idiosyncratic taxa was also defined as idiosyncratic. Idiosyn- cratic names and interviews were excluded from the statistical analyses. An expert was defined as any individual who reported a number of breadfruit cultivars equal to or greater than one standard deviation above the mean number of names reported by all respondents, and whose reported names included 90% of the cultivars that were known by at least half of all respondents Each breadfruit name was scored as an ambiguous monomial (AM), unam- biguous monomial (UM), ambiguous binomial (AB), or unambiguous binomial (UB). This was accomplished by comparing the name to two comprehensive dic- tionaries of the Samoan language (Milner 1966; Pratt 1911) and by checking with two bilingual speakers of English and Samoan. A name was regarded as ambig- uous if it conceivably could refer to an object other than breadfruit. This concept of using the term ‘ulu to prevent ambiguity or misunderstanding was posited by one of the matai® when asked to explain how breadfruit cultivars are named and why some include the term ‘ulu and others do not. The example he gave to make this clear was, “If I ask one of the young men to ‘Go get a ma’a’ he'll probably bring back a stone, but if I say ‘Go get an ‘ulu ma’a’ he knows exactly what I’m asking for, whereas if I say ‘Go get a ma’afala’, it is absolutely clear that I want a certain type of breadfruit. It wouldn't be pcos! to say ‘Go get an ‘ulu ma’ afala’’”’ We here use the terms ‘‘ambiguous” and ‘ Breseous as contrast terms rather than Berlin’s (1992) terms of “analyzable/ unanalyzable’ ’ for the sake of simplicity, and because ’ ambiguous’ and ‘‘unambiguous” are direct transla- tions of the Samoan terms “manino” and “le manino’’ respectively. RESULTS Interviews.—Breadfruit cultivar names in Samoa consist of either a binomial com- posed of the term “‘ulu modified by a descriptive term (Table 2), or a monomial in which only a descriptive term is used and ‘ulu is understood. This understand- ing was made explicit to us by several respondents, who, if questioned intensely or if they thought we were naive, would add the term ‘ulu to the description to emphasize that they were indeed referring to a cultivar of breadfruit. Samoan, as Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 41 is the case with many languages that did not have an indigenous orthography, often uses a series of two or more words to express a single concept. Consequent- ly, monomials can also sometimes be expressed as several words that form a coherent epithet or descriptive phrase. All persons interviewed used the generic level term ‘ulu for breadfruit, with the single exception of the village of Tafua, Savai’i. Because the Samoan language of politeness taboos the use of a word if it is a chief’s name, it may not be used in the presence of the chief holding the name as a title (Pratt 1911). Out of deference for the paramount chief ‘Ulu Taufa’‘asisina, the respect word for breadfruit, fa’atau, is always substituted for the term ‘ulu in this village. There are several different categories of breadfruit names regardless of bi- nomiality or monomiality. For example, one kind of name reflects the appearance of the breadfruit, such as ‘ulu sina ‘white breadfruit’, or momolega, whose name evokes the yolk of an egg in reference to its very yellow flesh. Other names are geographical, reflecting the putative origin of the cultivar such as in ‘ulu manu’a ‘Manuan breadfruit’. Another kind of breadfruit name reflects culinary properties, as in ‘ulu ma’a ‘hard breadfruit’, which takes a long time to cook. Yet other breadfruit names are comparative in the sense that they reflect overall similarities to another cultivar such as in puou maopo, a ‘puou that looks like maopo’. Two minor categories are names that are respect terms such as maualuga, which means high, or descriptive actions such as ‘ulu tau ‘to pluck’. Lastly, eleven breadfruit names, such as aveloloa, are irreducible in the sense that they either cannot be translated or their meaning has been forgotten by contemporary Samoans. The Samoan dictionary (Pratt 1911) defines these simply as “a variety or type of bread- fruit.” Recording of Data; Definitions of Idiosyncratic and Expert Respondents.—Using the re- dacted data set (determined by excluding all idiosyncratic names and all four idiosyncratic interviews) and by combining monomial/binomial variants (using a majority rule) and cognates, a total of 46 different names for breadfruit cultivars were recorded during individual interviews with 350 Samoans. The effect of ex- cluding these four interviews had only a small effect in the mean number of taxa reported (6.3 redacted, 6.4 unredacted) and no effect on the median number re- ported (6 names), with the number of breadfruit cultivars reported ranging from 0 to 20 names. Of the 354 individuals who were interviewed, 63 respondents reported 10 or more names, which is one standard deviation above the mean number of names known to all informants. Three of these individuals were excluded as experts because they did not meet the second expert criterion: they did not know 90% of the cultivar names known to more than half of all respondents. Therefore, 60 individuals were defined as “experts.” This statistical definition of expertise is compatible with Samoan folk perceptions of expertise. Tofa mamao, which glosses as ‘deep understanding’, is not found in everyone, but all villagers know who has such ‘deep understanding’. In several instances, respondents suggested that we speak with certain villagers because those individuals would know a lot about breadfruit, and this was borne out during in-depth interviews. These individuals had extensive knowledge of other practices concerning breadfruit such as crop 42 RAGONE et al. Vol. 24, No. 1 LE ee: Number of respondents i O-- 1 22. -4°°5 62.7 8) 79) 10, TE 12 13614) 15.1617): 18619220 Number of breadfruit cultivar names known FIGURE 2.—Number of breadfruit cultivar names known by 350 Samoans. husbandry, how to identify different cultivars, preferred uses, etc. The data are not normally distributed (Figure 2) and do not come from a random sample, so non-parametric statistics were used in the data analysis. Binomiality, Saliency, and Linguistic Ambiguity.—Using the redacted data set (i.e., the entire data set less idiosyncratic interviews), we sought to study the possible relationship between the saliency of breadfruit names and their binomiality. The cultivar names listed in Table 1 were analyzed for prominence of binomial versus monomial ethnotaxa to see if monomials tended to be more salient. H, = there is no relationship between binomiality and saliency H, = there is a relationship between binomiality and saliency were tested using a Wilcoxon rank sum test (Remington and Schork 1985; Sne- decor and Cochran 1989) and testing at the 0.05 level for significance. Ties were scored by using the average of the ranks of tied numbers. The test statistic (z = —0.83) was not significant at the 0.05 level so hypothesis H, was rejected: there is no relationship between binomiality and saliency. To limit the influence of in- frequent names, the two hypotheses were again tested with the Wilcoxon Rank- Sum using only the 66% most salient taxa. In this second test, H, was again rejected (z = 0.74). A third test, comparing only the top ten most salient names, Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 43 TABLE 3.—Summary of ambiguous and unambiguous binomial and monomial breadfruit cultivar names. Binomial Monomial Totals Ambiguous 19 (11.3) 7 (14.7) 26 Unambiguous 1 (8.7) 19 (11.3) 20 Totals 20 26 46 Note: cells have number observed and (expected). was performed, using a W statistic rather than Z because of the small sample size. In this third test, H, was rejected a third time, so we can unequivocally state that there is no significant relationship between saliency and binomiality in breadfruit names reported by 350 Samoans. The hypotheses H, = there is no relationship between ambiguity and binomiality H, = there is a relationship between ambiguity and binomiality were tested by constructing a 2 X 2 contingency table, with the columns repre- senting monomial and binomial names and the rows representing linguistically ambiguous and unambiguous names (Table 3). A x? statistic was calculated (x? = 22.2, p < 0.001) and tested at the 0.05 level for significance using Yates correction for continuity (Snedecor and Cochran 1989) and H, was rejected. Binomiality is significantly related to linguistic ambiguity among our 350 respondents. When Table 1 is analyzed for a relationship between linguistic ambiguity and saliency, the relationship is even stronger: 82% of all breadfruit names are either unam- biguous monomials (UM) or ambiguous binomials (AB); e.g., binomials whose specific epithet alone, out of context, could conceivably refer to another object than breadfruit. These results do not support the rather reasonable assertion by Berlin (1992) that monomials should be used to label highly salient taxa—indeed there is no relationship between saliency and monomiality—but our results do support the indigenous hypothesis that monomials should be used only when the terms are completely unambiguous. Age, Gender, Westernization and Cultural Competency.—To determine if there was a relationship between age of the respondent and number of breadfruit names re- ported (Table 4), the median number of taxa reported by respondents in each of nine age classes was calculated (Class 1 = ages 0-9; Class 2 = ages 10-19; etc.). For statistical continuity age cohorts 0-9 and 80-89 were included; they indicate the age decade but do not imply that a 0 age child or an 89 year old adult were interviewed. The three youngest people interviewed were between two and four years old, all others were six years or older. The oldest person was 84 years old. These terms are merely labels for the cohorts. These data were used to test the following hypotheses: H, = there is no relationship between age class and number of breadfruit names reported H, = there is a relationship between age class and number of breadfruit names reported 44 RAGONE et al. Vol. 24, No. 1 TABLE 4.—Knowledge of breadfruit cultivar names based on age class. Age class Respondents Difference (years) - ean Rank (D) Ea 1 (0-9) 14 14 9 “75 57.76 2 (10-19) 78 4.1 8 3.9 15.21 3 (20-29) 78 6.1 7 -0.9 0.81 4 (30-39) 57 6.7 6 0.7 0.49 5 (40-49) 31 7.4 > 2.4 5.76 6 (50-59) 40 8.0 a 5.0 25.00 7 (60-69) oN 9% 1 8.2 67.24 8 (70-79) 19 8.6 2 6.6 43.56 9 (80-89) 2 72 4 re 12.25 *Sum of D? = 228.08; r = —0.90. by calculating a Spearman's rank correlation coefficient and testing for signifi- cance at the 0.05 level. Since at 7 (n — 2) degrees of freedom, the two-tailed significance level for the correlation coefficient r at the 0.01 probability is 0.798, H, is rejected, showing a strong relationship between age class and mean number of breadfruit taxa reported. We wished to determine if the location of one’s residence had any influence on the number of breadfruit names (Table 5) that were known as well as whether gender played a role in such knowledge. Ordinarily an analysis of variance would be used to see if such differences are important. Since these data are not randomly collected independent samples with normal distribution, and since sample vari- ances were not equal for the subsamples, such an ANOVA analysis with the para- metric F statistic would be inappropriate. Therefore the nonparametric Kruskal- Wallis test was used which generates a statistic comparable to that of an ANOVA to test using the x’ distribution at the 0.05 level of significance for the following hypotheses H, = there is no difference between villages in the number of breadfruit names known H, = villages differ in the number of breadfruit names known. any This test yielded an H statistic of 123.6. Since multiple ties occur in the data set, this statistic was corrected by dividing by the correction factor (1 — (T? — T)/N? — N = 0.99) where T is the number of ties for each observation and N is the sample size. Our corrected statistic H,,,, = 124.7. At seven degrees of freedom, since H exceeds 203, H, at p < 0.005 was rejected, hence place of residence is highly significant in influencing number of breadfruit names known. TABLE 5.—Knowledge of breadfruit cultivar names based on place of residence. Number of names known Ofu Olosega Afono Pago Apia Falealupo Saipipi Expat. Mean 5 5.3 5.5 “, 5.6 ff: 8.7 1.6 Median 5 is 6 S Da 7 8 2 Maximum 10 i 11 ii 11 14 20 4 Respondents (n =) 51 52 28 og 34 33 88 i, Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 45 TABLE 6.—Knowledge of breadfruit cultivar names based on gender. Number of names known All females All males Expert females Expert males Mean 5.6 7.1 10.9 ei Median 6 i 10 11 Maximum 19 20 19 20 Respondents (n =) 192 158 23 37 Gender differences were also tested using the Kruskal-Wallis test. The hy- potheses were: H, = there is no difference between genders in the number of breadfruit names known H, = men and women differ in the number of breadfruit names known. o For the total of 192 women and 158 men in our sample (Table 6), the corrected H statistic was 21.7, allowing us to reject H, at the p < 0.005 level: men know sig- nificantly more breadfruit names than women. DISCUSSION We were impressed by the diversity of breadfruit cultivars recognized by the Samoans in our sample. We are unaware of any major supermarket in the United States that stocks anything approaching this selection of crop diversity, which the Samoans claim grows in and about their villages. Our data set of 350 interviews allows us to do more than to document the richness of Samoan breadfruit diver- sity; it allows us to test several hypotheses about knowledge of breadfruit. The five criteria of Brown (1985) are used in determining which Samoan breadfruit names are binomials or monomials: 1) a composite term is considered to be binomial if one constituent of the label stands on its own as the name of the class (e.g., ‘ulu); 2) one constituent is not a major life-form (e.g., breadfruit ‘tree’); 3) morphologically dissimilar (i.e., sea horse is not a type of binomial); 4) shared generic constituent; and 5) composite terms ‘mate of’, ‘like’, ‘similar’ are not binomials. The suggestion that monomials are used in folk taxonomy to label highly salient folk taxa has been asserted for the simple reason that binomial names for lower-salience referents are overall less salient and more easily remem- bered than monomials (Brown 1985, 1986, 1987). Berlin (1986) argued that it is erroneous to state that the increase in binomial taxa results from an overall de- crease in saliency. Rather it is due to direct biological manipulation by humans in the process of domestication and that folk genera with the largest numbers of folk species are always cultivated plants. Folk specific taxa may finely subdivide a single biological species with cultivated plants that have been highly modified under domestication (Berlin 1992). Hays (in Brown 1985) argued that “binomi- alization might be most common in sets of taxa that are highly salient; i.e., do- mesticated plants or animals of which varieties or species (binomially labeled) would have resulted from domestication.’” What is needed to test these competing theses has been a direct indicator of saliency in a folk setting. Frequency of mention in a standard interview as an index of saliency of a 46 RAGONE et al. Vol. 24, No. 1 folk taxon is adopted here. Using that measure, no support was found for either Brown or Hays: there is no statistical association between saliency and binomiality for names of Samoan breadfruit cultivars. It could be argued that the restricted taxonomic focus (i.e., only breadfruit cultivars are considered) makes our study an inadequate analysis of the broader theory. However, by restricting ourselves to a single crop, “intensity of cultural use’ (Turner 1988) is held constant, so our data on comparative saliency are strictly comparable. Studies on other crop cul- tivars in various places would add considerable power to this basic approach. Our data also gave partial support for the argument that there is often con- siderable disagreement among indigenous societies on folk names. Working with the Wola, an agricultural people in New Guinea, Sillitoe (1980) found that they hold in common a set of cultivar names, but when forced to apply these names to actual plants, they only agreed about 50% of the time about which name goes with which plant. He surmised that disagreement over naming plants most likely occurs at this taxonomic level since such identifications frequently depend on fine details of morphological variation (Sillitoe 1995). Although some slight differences in the use of monomial or full binomial names were noted (as might be expected when attempting to clarify a plant name for a foreign investigator), there was surprising little variation in plant names, once cognates with superficial differences were clumped together in the analysis. What surprised us further was not the differences in names, but the overall con- sensus in names which were recorded on islands over 400 km apart. The number of idiosyncratic responses, including those we surmise were invented on the spot to please a persistent investigator, was very low. Fewer than one percent of our interviews were excluded from analysis because of idiosyncrasy. In all settings, however, two broadly different realms of ethnobotanical knowledge were found: common knowledge and expert knowledge. A useful outcome of our study was that we were able to statistically quantify what makes an individual an expert. Future ethnobotanical fieldwork in Samoa will be greatly facilitated by our having defined a large group of experts with whom we can work and conduct in-depth interviews about breadfruit. For ex- ample, we will work with some of these experts to ascertain the conservation status of breadfruit cultivars in Samoa, especially those that were only known by one or a few individuals. We surmise that cultivars such as ma’afala and puou, known by 90% and 81%, respectively, of the Samoans interviewed are common in cultivation and therefore conserved in situ, whereas the more uncommon cul- tivars may be at risk and require special conservation strategies. A rigorous comparison between folk and statistical measures of expertise is beyond the scope of this paper, but we believe that our statistical definition of expert could benefit investigators conducting ethnobotanical projects elsewhere. It is possible in a fairly short time to interview a large number of people about a specific topic and from that group quickly and accurately identify those who possess expert knowledge about the subject at hand. Working primarily with expert individuals is a useful, and timely, strategy to maximize obtaining reliable, specialized, and verifiable information. In our sample, 17 of the experts were in their 60s and six were over 70 years old. It is critical that the traditional cultural Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 47 knowledge of these elderly experts, several of whom were in very poor health, be documented before it is lost. Place of residency has a strong impact on the amount of knowledge about breadfruit names that an individual possesses (Table 5). As might be expected, the traditional villages of Saipipi and Falealupo on the remote island of Savai’i score highest in breadfruit knowledge. We were surprised to find that Ofu and Olosega villages in the remote Manu’a archipelago of American Samoa scored at about the same rate as the residents of the capital cities of Pago Pago and Apia. This may be due to the prevalence of sending high school age boys and girls off- island for education, where they are removed from participating in daily cultural activities and hence do not have the opportunity to learn traditional knowledge and practices from their elders. There is, in effect, a brain drain as adults leave the Manu’a islands for Tutuila, Hawaii, or the U.S. mainland. For example, many families maintain residences both in Ofu or Olosega and on the island of Tutuila. Mid-life adults are working in the wage economy on Tutuila, providing a home for their high-school-attending children, or caring for their elderly parents who have moved to Pago Pago for medical care and long-term convalescence. The mayor of Olosega suggested that since the residents of Ofu and Olosega rely primarily on earned income and family remittances rather than subsistence ag- riculture, there is little need to keep such breadfruit knowledge alive. In any case, it appears that ethnotaxonomic knowledge is exceedingly fragile, and can quickly disappear, even from apparently remote areas. This is evident by the low rate of knowledge possessed by expatriate Samoans. Gender differences in breadfruit knowledge can be supposed to reflect the gender-based divisions of labor inherent in Samoan society. Men are more likely to work in the plantations, plant and harvest breadfruit, and prepare them in the umu or stone ovens. It is important to note that these gender differences, while reflected in the mean and median number of names known by men and women, do not reflect expert knowledge. The second most knowledgeable person about breadfruit names was a woman. In the group of 60 individuals we defined as having expert knowledge, 23 were women and 37 were men. Knowledge about breadfruit among this group of women can primarily be attributed to transmis- sion of knowledge by family traditions. For example, some of these women held matai titles that are conferred by their families and typically recognize those individuals who are knowledgeable about and practice fa’asamoa—i.e., who maintain traditional Samoan customs and knowledge. Several women were relat- ed by marriage and/or birth to matai. If there are no sons living in the household a daughter, out of respect for her father, will become familiar with and learn her family’s traditions, including areas of expertise that are normally associated with men. Wives of matai bring to the marriage their own family traditions and often learn those of their husbands. Upon the death of a matai, wives are the repository of this shared knowledge and ensure that both families’ traditions are perpetu- ated. In interviews with individuals who work in administration of government and NGO conservation programs in Apia, a modest level of breadfruit knowledge was recorded; certainly above the median (8.5 vs. 6.0 names) for all of our re- spondents, but below that of the expert level. One individual reported 15 names, 48 RAGONE et al. Vol. 24, No. 1 placing him well within the expert range, while a long-term expatriate knew only five names. The expert noted that his knowledge was acquired principally from his residence in and subsequent visits to remote villages. This suggests that vil- lagers who possess strong ethnotaxonomic knowledge should be recruited to oversee conservation and agrobiodiversity programs in Samoa and elsewhere. In- digenous knowledge of crop diversity is crucial to guiding conservation and ag- ricultural development projects to ensure that traditional cultivars, cultivation practices, and cultural practices and knowledge are preserved rather than eroded. NOTES ‘ Herbarium specimens were deposited at PTBG, National Tropical Botanical Garden, Ka- laheo, Hawaii. * The standard orthography for Samoan includes a glottal stop or break, indicated here by an apostrophe before the vowel. * Interview with Vaiga Uaealesi, in Saipipi Village, Savai'i, 19 July 2000. ACKNOWLEDGMENTS We are grateful to the many Samoans who welcomed us into their homes and with humor and patience allowed us to ask them personal questions and query them about breadfruit. We especially appreciate the kindness of the matai who welcomed us to the villages of Saipipi and Falealupo with traditional ‘ava ceremonies. We were honored to meet paramount chief ‘Ulu Taufa’asisina during our visit to Tafua. Special thanks to the Tavana family for providing their home in Saipipi as our base of operations while on Savai’i island; to Elizabeth Laolagi who housed us during our stay in the Manu’a Islands; and the Tavana and Gurr families who housed some of us during our work on Upolu and Tutuila, respectively. David Lorence kindly provided the Spanish and French abstracts. REFERENCES CITED an, Scott, Paul Estin, = Joi ey, aa Milner, G. B. 1966. Samoan Dictionary. Ox- ance in folk Sieh Rusia of Ethno. biology 17:17-43. Berlin, ey "1086. Comment on “The growth of ethnobotanical nomencla- ture” by ~ Brown. Current Anthro- pology 27:12 ee a Ethnobiologica Classification. nceton University Press, ceton. H. 1985. Mode of subsistence and folk ae oe Current An- thropology 26: . 1986. The ge growth o of ethnobiologi- cal nomenclature. Current Anthropology 27:1-9. The folk subgenus: a cthnobiological rank. Journal of ‘Baa: biology 7:181-19 ford University Press, London. whines George. 1911. A Samoan copenye J English and Samoan, and Samo glish; With a Short Grammar of the ‘kanioai Dialect. London Missionary Society Press, Apia, Samoa. Ragone, Diane. 1995. Description of Pacific Island breadfruit cultivars. Acta Horti- culturae 413:93-98. . 1997. Breadfruit. Artocarpus altilis (Parkinson) ee 5 Promoting the Con- servation and Use of Underutilized and Ne- glected on 10. Institute of Plant Ge- netics and Crop Plant Research, Gter- sleben and International Plant Genetic Resources Institute, Rome. Remington, Richard D. and M. Anthony Schork. 1985. Statistics With Applications Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 49 to the Biological and Health Sciences, 2nd Turner, Nancy J. 1988. ‘The importance of e ic -Hall, Inc., Englewood a rose” : evaluating the cultural signifi- Cliffs, New Jersey. cance of plants in Thompson and Lil- Snedecor, George W. W.and William Cochran. iooet Interior Salish. American Anthro- 1989. Statistical Methods, 8th ed. Iowa pology 90:272-290. State University Press, Ames. _ Ward, R. Gerard and Paul Ashcroft. 1998. Samoa: Mapping the Diversity. Institute of Pacific Studies, University of the South agg iataeie?! chose Saeuint of Pacific, Suva and National University of the plant resources of the asin Region, Samoa, Apia. southern Highlands Pro Papua Whistler, W. Arthur. 2000. Plants in Samoan New Guinea. raPna of Efhnobiology 15: Culture: the Ethnobotany of Samoa. Isla 201-235. Botanica, Honolulu. us Pr Journal of Ethnobiology 24(1): 51-73 Spring/Summer 2004 THE CATEGORY OF ANIMAL’ IN EASTERN INDONESIA GORY FORTH Department of eee University of Alberta, Edmonton, Alberta, T6G 2H4 Canada ABSTRACT.—It is a generally accepted idea among ethnobiologists that most non- western languages lack a term for ‘animal’. Evidence from eastern Indonesia re- veals that, understood as labels for an ethnotaxon comparable to vernacular En- glish ‘animal’, such terms are by no means rare in this part of the Austronesian- speaking world. At the same time, the lexical resources employed to name a gen- eral ‘animal’ category reveal a notable diversity that corresponds to the variety documented by K. Alexander Adelaar in regard to Austronesian languages as a whole. In this article, I review terms translatable as ‘animal’ in several eastern Indonesian languages. I conclude by addressing issues illuminated by the eastern Indonesian evidence, including the perceptual salience of the ‘animal’ taxon and Berlin's evolutionary thesis concerning the lexical recognition of categories be- longing to different ethnotaxonomic levels. Key words: Eastern Indonesia, Austronesian languages, ethnotaxonomy, ethno- zoological nomenclature, terms for ‘animal’. RESUMEN.—La idea de que la mayoria de las lenguas no occidentales carecen de un término que signifique ‘animal’ esta generalmente aceptada entre los et- nobidlogos. Los datos de Indonesia oriental muestran que estos términos, enten- didos como etiquetas para un etnotaxon comparable al de ‘animal’ en espafiol verndculo, no son en absoluto escasos en esta parte del mundo de habla austro- nesia. Al mismo tiempo, los recursos léxicos empleados para nombrar una cate- goria general de ‘animal’ revelan una notable diversidad que corresponde a la variedad del conjunto lenguas austronésicas. En este articulo reviso los términos traducibles como ‘animal’ en varias lenguas de Indonesia oriental. Finalizo pro- poniendo ideas, basadas en la evidencia indonesa, sobre la prominencia perceptual del tax6n ‘animal’ y la tesis evolutiva de Berlin en lo que concierne al reconoci- miento léxico de categorias de diferentes niveles taxondémicos. RESUME.—Parmi les ethnobiologistes, il est généralement admis qu'il n’existe pas d’équivalent au terme «animal» dans la plupart des langues non occidentales. Cependant, dans les régions ou I’on parle malayo-polynésien, de pareils termes ne sont pas rares et des faits provenant de I’est de I’Indonésie indiquent que ces termes pris en tant qu’étiquettes pour un ethnotaxon comparable au terme anglais vernaculaire «animal» existent. Aussi, de facon paralléle, les ressources lexicales utilisées afin de nommer une catégorie «animal» générale montrent une diversité remarquable qui correspond a celle documentée dans l’ensemble des langues ma- layo-polynésiennes. Dans cet article, je fournis une synthése des termes se trad- uisant par «animal» parmi plusieurs langues de l’est de I’Indonésie. Je termine cette synthése en soulevant différents points a la lumiére des faits tirés de l’est de I’Indonésie, incluant la perception du taxon «animal» ainsi que la thése évo- lutive de Bertin qui a trait 4 la reconnaissance lexicale des catégories appartenant a des niveaux ethnotaxonomiques différents. Sz FORTH Vol. 24, No. 1 INTRODUCTION In the study of ethnobiological classification, it has become a virtual maxim that terms in nonwestern languages denoting a category corresponding to English ‘animal’ are uncommon—even “normally” absent (Berlin 1992:15, 27, 190; cf. Ber- lin et al. 1973:215; Brown 1984:4; Lévi-Strauss 1966:1). Among the Austronesian languages of Indonesia, however, such terms are not nearly so rare as this gen- eralization would suggest. At the same time, as Adelaar (1994:12-13) has noted, Proto-Austronesian, the hypothetical ancestor of all Austronesian languages, ap- pears to have lacked a general term for ‘animal’. Accordingly, the lexical means employed by modern Austronesian speakers to refer to ‘animal’ are remarkably various.’ The purpose of this paper is to demonstrate a comparable variety among general terms for ‘animal’ encountered in several eastern Indonesian languages spoken on the islands of Flores, Sumba, Roti, Timor, and Seram. I further consider the implications of this variety for ethnobiological theory pertaining to folk zoo- logical classification. One interest in this connection is evidence indicating that ‘animal’ exists, at least as a covert category, even among speakers of languages that lack a term unequivocally denoting the taxon. Especially relevant here is the widespread incidence of numeral coefficients (or classifiers) cognate with Malay ekor ‘tail’ (cf. Proto-Austronesian *‘ikuy ‘tail’, Dempwolff 1938:68), which are em- ployed when counting or enumerating any kind of animal (see Berlin et al. 1974: 30; also Taylor 1984:107, 1990:44). In his review of ‘animal’ terms, Adelaar (1994:13) lists four general ways in which the folk taxon appears to be labelled in Austronesian languages. These include: naming with a descriptive phrase (or paraphrase) such as ‘living creature’ or ‘animate thing’; with a word denoting a particular animal kind; with a term referring to ‘domestic animal’; or with a loan word (often deriving from Malay binatang, Sanskrit sattva, or Arabic haywan). As 1 demonstrate below, all of these methods are reflected within a much more restricted group of eastern Indonesian languages. This variety is discernible within clusters of the most closely related languages or dialects, and in some instances even possibly within one and the same language. LANGUAGES OF FLORES, SUMBA, AND TIMOR All the languages I survey here have been identified by Blust (1980) as mem- bers of a Central-Malayo-Polynesian grouping within the Malayo-Polynesian fam- ily of Austronesian languages. Included in this grouping are two subgroupings identified by Esser (1938) as the Bima-Sumba and Ambon-Timor groups. More recently, Wurm and Hattori (1981) have proposed a more detailed classification of languages included in the second group, but this need not overly concern us here.* Bima-Sumba languages include those spoken on Sumba, Savu, western and central Flores, and Komodo, as well as the Bimanese language of eastern Sum- bawa. Of the languages treated in this article, Esser’s Ambon-Timor group in- cludes those spoken in more easterly parts of Flores—including Sika, the Lama- holot languages of East Flores (Flores Timur) and the smaller islands immediately Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 53 to the east (Solor, Adonara, Lembata/Lomblen)—as well as Rotinese, the Tetum (or Tetun) language of Timor, and the Nuaulu language of Seram. [ begin by reviewing Bima-Sumba languages, partly because their ethno- zoological lexicons are rather better documented than those of Ambon-Timor lan- guages, and I begin with Nage and closely related dialects of western Keo, since ethnozoologically this is the case that I know best (see Forth 1995, 1999, 2004). An alternative procedure might have been to frame the lexical data with regard to the four methods of labelling ‘animal’ isolated by Adelaar. However, since some languages exhibit more than one of the four ways of referring to animals in gen- eral, this is less convenient. BIMA-SUMBA LANGUAGES Nage (and Western Keo), Central Flores—The Nage term ana wa labels a category of living things that closely corresponds to the English vernacular sense of ‘ani- mal’ where it contrasts with ‘human’. By the same token, the expression corre- sponds to modern Indonesian (and Malay) binatang ‘animal’. Accordingly, Nage recognize the taxon as comprising a number of labelled and unlabelled (or covert) life-form taxa, including nipa ‘snakes’, ika ‘fish’, and ana wa ta’a co ‘flying crea- tures’ or ‘birds’ (coinciding mostly with the zoological class Aves), even though the focus of ana wa is large mammals and then especially domesticated varieties (Forth 1995:47-48).° Instancing an apparently universal feature of folk taxonomy, Nage ana wa definitely excludes human beings (kita ata), although, as I discuss presently, the term can be applied metaphorically to a certain category of human eings. its most common usage, ana means ‘child, children’ or ‘child of’. In a broader sense, the term can further refer to a member of any human collectivity or social unity (see, for example, ana loka ‘participant in a ritual assembly or other activity’; ana one ‘insider’, cf. one ‘inside’). Since wa means ‘wind,’ ana wa might thus be glossed as ‘children, people of the wind’ Entailing a figurative usage (insofar as Nage contrast ‘animals’ with ‘people’), this interpretation is rec- ognized by Nage themselves, who rationalize it with reference to the idea that, like the wind but unlike humans, animals are uncontrolled and unpredictable in their behavior (Forth 1989, 1995:47). Consistent with this representation, Nage further apply ana wa to small children (ana éno; Forth 1995:47-48), who—as one informant explained—do not yet understand speech and cannot be constrained by verbal commands or admonition. (In this connection, the informant noted how toddlers will heedlessly grab at everything in sight.)* Contrariwise, in some contexts Nage use ana ‘child’ alone with reference to animals, although mostly it appears with reference to birds. In attempting to identify a particular kind of bird, for example, one might thus inquire ana apa ke? ‘what (animal, bird) is that?’5 In this context, ana might simply be construed as an abbreviation of ana wa; alternatively, it can be understood as specifying an instance of a larger collectivity, a ‘member of’ the larger group of “flying animals’ (ana wa ta’a co). Interestingly, in the Wangka dialect of Rembong (northwestern Manggarai), the cognate anak similarly occurs in anak reman (reman refers to wild vegetation, see note 16), identified by Verheijen (1977 s.v. anak) as a general 54 FORTH Vol. 24, No. 1 term for ‘bird’. Also relevant here is the mostly optional use of ana in Nage names for many kinds of birds (e.g., koka and ana koka—Helmeted friarbird, Philemon buceroides), although ana also occurs in the names of other sorts of small animals (e.g., ana gu—house lizard, Hemidactylus frenatus; ana fe—tadpole; ana bo and ana tebhu—two kinds of freshwater fish). Interestingly, a clan resident in villages near the Nage center of Bo’a Wae is named ‘Ana Wa’. As shown by their alternative naming simply as ‘woe Wa’ (clan Wa), however, the name in this context does not necessarily translate as ‘animal’, but is usually understood to mean ‘Wind people’. According to another local interpretation, it can be construed as ‘animal’, but only in the metaphorical sense of ‘small children’ However ana wa is precisely to be understood, the Nage term clearly instanc- es the use of a descriptive phrase to express the general sense of ‘animal’—or to label an ethnotaxon at the level of the ‘kingdom’ (or ‘unique beginner’, Berlin 1992:15). As a general term for ‘animal’, ana wa is also known in western Keo, where it was defined as referring to all four-footed animals, livestock, birds, and snakes. Two other Keo terms, both elicited when asking about local terms for animal’, are ngawu nitu and bugu lara. Meaning ‘possessions (goods, wealth) of spirits’, ngawu nitu more precisely denotes wild animals, and reflects the idea, also found in Nage (Forth 1998:70-72), that various wild creatures are the do- mestic animals of free spirits (nitu). The endemic Flores giant rat (Papagomys ar- mandvillei, bétu) is thus considered the water buffalo of these spirits, Green jun- glefowl (Gallus varius) are their chickens, and so on. As these specific equations are restricted in number (if only by virtue of the fact that humans possess limited kinds of domestic animals), it is equivocal how far ngawu nitu can be understood as including all wild creatures. Nevertheless, in response to questioning, I was assured that nipa (snakes), for example—which are more often identified as man- ifestations of nitu spirits themselves rather than as some particular kind of animal belonging to the spirits—are also included in this catego The second Keo term, bugu lara, refers specifically to livestock (owned by humans). A synonymous expression recorded in Nage is bugu beti. For the most part equivalent to ngawu (‘wealth, possessions’; also, in context, specifically ‘bridewealth’), the relevant sense of bugu is ‘thing, possession, good(s)’ (cf. bugu ngawu, wealth, including both livestock and inanimate objects). No one I ques- tioned could explain either lara or beti in these contexts. The usual sense of Keo lara (cf. Nage laza), however, is ‘ill, illness’, while in neighboring Ngadha, beti (cf. Nage bugu beti) also means ‘ill’ (Arndt 1961). One possibility, therefore, is that the phrases distinguish domestic animals from other possessions as things which are subject to illness, and which thus may decrease through sickness and death. I introduce these expressions in order to demonstrate that, while ana wa includes both domestic and wild animals in Keo as well as Nage, there are also special terms distinguishing wild and domesticated kinds. Like the general term, moreover, the latter are descriptive phrases designating essentially utilitarian clas- ses of animals as the ‘property’ of spirits and humans respectively, though a peculiarity of ngawu nitu is that, by virtue of a cosmological principle of ‘““recip- Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 55 rocal inversion’ (Forth 1998), it is used to specify not human livestock but wild creatures. One interest of the western Keo use of ana wa relates to the fact that while wa is the Nage word for ‘wind’, in western Keo ‘wind’ is waya. Yet one does not hear ana waya. This circumstance, then, suggests the use of a loan word for ‘animal’, though one adopted from a neighboring dialect rather than from Malay or another quite different language. Lio and Endenese, Central Flores.—Located to the east of the Nage, inhabitants of the Lio and Ende regions speak dialects that are closely related to those of Nage, Keo, and Ngadha. Indeed, they form a single grouping with these, distinct from both the language of Sika (spoken immediately to the east of Lio) and Manggarai (the language of western Flores; see Wurm and Hattori 1981:map 40). In what remains the major source for the Lio lexicon, Arndt’s dictionary (1933) lists two terms that may be glossed as ‘animal’. One is binata, clearly a loan from Malay (see binatang). The same term is given for ‘animal’ in Endenese (Stokhof 1983; Suchtelen 1921:330, for the ‘Ja’o’ dialect). For Lio, Arndt defines binata more specifically as ‘large animal, especially four-legged animals’. However, according to Takashi Sugishima,’ an anthropologist who has recently conducted extensive research among Lio, the term is further employed in the general sense. (Sugishima also states that binata is often used in contradistinction to a term for ‘human being’—for example when abusing people by comparing them with animals.) Although the Lio term is obviously borrowed from Malay binatang, for it to appear in a dictionary published as early as 1933—and in regard to Suchtelen’s Endenese word lists, in a publication dated 1921—it must have been adopted in this part of Flores before the earliest years of the twentieth century. In fact, the adoption likely occurred prior to the era of effective colonial administration, a circumstance entailing that its introduction in Lio and Endenese was not a function of wide- spread bilingualism or the establishment by the Church of elementary education in Malay. Another possible candidate for ‘animal’ in Lio is ule (Arndt 1933). As in Nage and other Flores languages, the primary meaning of Lio ule is ‘worm, maggot, grub’ (see Appendix 1). The term is thus comparable to Malay/Bahasa Indonesia ulat ‘caterpillar, worm, insect’, and, like the latter, evidently reflects a Proto-Aus- tronesian or Proto-Malayo-Polynesian form that referred, at least primarily, to worms and similar creatures.’ In regard to the variety of small creatures named by the term, ule (like Malay ulat) appears largely to correspond to the sort of widespread folk taxon generally designated ‘wug’ (a neologism formed from ‘worm’ and ‘bug’; see Brown 1984:16). Yet ule further occurs in compound names of several Lio folk generics that denote birds.’ Among these are ule a ‘crow’, ule mesi ‘heron’, Stokhof 1983, ule mi’u ‘a bird that shrieks mi’u’, ule si ‘a small bird’, ule molo, and ule polo.? A particular connection of Lio ule with birds is further indicated by the term haba ule ‘bird’s nest’ (Arndt 1933:132, s.v. haba ‘(bird’s) nest’, cf. haba manu ‘hen’s nest’). Other Lio compounds with ule listed by Arndt (1933) denote folk generics including worms, grubs, and insects. In regard to the application of the term to birds, it is interesting that of five insect terms, at least four refer to flying insects 56 FORTH Vol. 24, No. 1 (see ule ae ‘various sorts of dragonflies’, ae ‘water’; ule api ‘wasp with a red lower body’, api ‘fire’; ule hetu ‘moth’; ule n’gake ‘butterfly’; ule si ‘a sort of beetle, chafer’ but also a small bird). Apparently nonzoological applications of Lio ule include ule re’e (re’e ‘bad, mean, ugly’; cf. Nage ’e’e) and ule ola, both of which Arndt translates as ‘evil spirit’ (German bdser Geist). The first term, however, Arndt additionally glosses as ‘all poisonous snakes’ (see note 3, regarding Nage ana wa ta’a ’e’e). It therefore appears that in Lio, a term originally denoting worms and similar small creatures has become extended so as to encompass a far more inclusive category of living things. In other words, one is evidently dealing with an instance of a word denoting a particular animal kind being applied, if not to animals in general, then to a significantly wider variety of creatures than those originally labelled by the term. Interestingly, a remarkably similar extension appears to have occurred in the Tetum language of Timor, as indicated by Hull’s (2001) gloss of ular as both ‘worm, caterpillar; crawling insect’ and ‘creature, animal’. Further evidence for the Lio term is provided by Arndt (1933), who translates ule as ‘creature, worm, grub (larva), maggot, bird’ (German: Getier, Wurm, Larve, Made, Vogel). Arndt also lists the compound ule age as ‘all kinds of animals, worms, reptiles, and birds’ (Getier, Wiirmer, Reptilen, Vogel). On the other hand, according to more recent evidence provided by Sugishima (see note 6), ule age refers exclu- sively to birds, serving as ‘‘a general term for birds, except chickens.” In view of the meaning of ule (and cognates) in other languages, it may be significant that, in addition to birds, the majority of Lio compounds in which ule occurs severally denote small creatures (dragonflies, wasps, larvae, moths, cater- pillars, butterflies, worms). According to Sugishima (see note 6), Lio do not apply ule to mammals or fish, although they do refer to some poisonous snakes as ule bani (bani ‘angry’, ‘aggressive, bold’). Also noteworthy in this connection is the fact that German Getier, Arndt’s first gloss of ule, not only has the collective sense of ‘creatures’, but also applies especially to insects (see Tyrell et al., s.v. Getier). There is thus a suggestion that Lio ule refers only to certain kinds of animals, mostly smaller ones, so that the term may accurately be glossed as ‘animal’ (or ‘bird’, ‘snake’, and so on) only in the context of compound expressions, where the word is modified by another, or in expressions referring collectively to a variety of creatures, where the inclusion of particular kinds is ambiguous. Insofar as ule can refer to snakes, it should be noted that the Lio term cannot be interpreted as a retention of Proto-Austronesian *ulaR (or *qulej) ‘snake, worm’ (Zorc 1994:593, 550). Not only had ‘snake’ become separated at the Proto-Malayo- Polynesian level, as *nipay (Zorc 1994:550), and perhaps earlier (in Proto-Hespe- ronesian-Formosan = Western Austronesian and Formosan, Zorc 1994:550) as *buLay, but the evidence of other Flores languages reveals cognates restricted to worms, maggots, and other similar small animals. It would appear, therefore, that the Lio usage represents a special development, not simply a reversion to a more generalized meaning but a shift to one evidently more inclusive than that of the Proto-Austronesian form. Obviously, the suggestion that ule serves as a general term for ‘animal’, like Nage ana wa, requires considerable qualification. Nevertheless, it is clear that, in Lio, the term has acquired an ethnozoological sense that is far more inclusive than Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY oF ‘worm, maggot’, and which moreover subsumes creatures belonging to more than one life-form—notably, birds and snakes. Yet, by the same token, one cannot def- initely conclude that ule unequivocally denotes an ethnotaxon encompassing all zoological life-form taxa, or figures as a component of productive expressions (such as Nage ana wa ta’a co ‘flying animals’ or ‘birds’). The Lio compound ule age does not necessarily contradict this characterization. Since age appears to have no separate meaning, it cannot decisively be interpreted as a modifier specifying a particular segment of animal kinds. Nor does it clearly function adjectivally, in this context or in any other. On the other hand, another Lio term generally de- noting wild birds, ule bene (see note 6), can be analyzed as ‘wild ule’ (see bene ‘grass, weeds, bush’, Arndt 1933; also note 16). Hence in this instance at least, ule does appear to approximate the general sense of ‘creature’, even if the term cannot be used alone to mean ‘animal’. While in the absence of further evidence regarding Lio usage one cannot definitely conclude that ule designates an ‘animal’ taxon, a fascinating comparison may be found in Chinese chong (or chung). Like ule, the commonest gloss of chong is worm, but other senses of the word include ‘insect’, ‘caterpillar’, ‘larva’, and ‘vermin’ (A Pocket Chinese-English Dictionary 1978). In addition, various kinds of evidence indicate that, in the past, chong has functioned as a general term for ‘animal’. According to the etymologist Xu Hao, in sixteenth-century China chong was used for ‘animal’ regardless of the method of locomotion or physical form of the creature referred to (Chinese Etymological Dictionary 1981). Accordingly, chong further occurs in the names of a variety of particular animal kinds, includ- ing ‘tiger’ (da-chong, literally ‘big worm’) and ‘snake’ (chang chong ‘long worm’). At present, however, all of these categories possess alternative names. Also, in modern Chinese, the general term for ‘animal’ is dong wu.'° If there is an explanation for this similarity between Chinese and Lio, it might be found in a widespread, and probably universal, conception of animals as things that move (or are animated). Thus, as the smallest and morphologically simplest of moving things, and perhaps as creatures which, for humans, display a partic- ularly salient kind of movement (wriggling or crawling), worms, or perhaps better said ‘wugs’, might be regarded as something like ‘atoms’ of animation.'' Also worth noting in this connection is makayidi-yadaku, the eastern Sumbanese term for ‘animal’, which, as I describe more fully below, includes the component yada ‘to team, swarm, wriggle, fidget’. Eastern Sumbanese.—As recently discussed in another article (Forth 2000), eastern Sumbanese possesses at least one expression that functions as a general name for ‘animal’. This is makayidi-yadaku ‘things that move’, a sense that reveals another instance of the use of a descriptive phrase to label ‘animal’. The basis of the expression is the compound yidi-yada, comprising two roughly synonymous terms meaning ‘to move’, and producing an alliterative sound symbolism com- parable to English ‘topsy-turvy’ or ‘twist and turn"? Both Onvlee (1984) and Kapita (1982, s.v. kayidiku) further gloss the expression as ‘the whole of creation or ‘all creatures’. (Like Nage ana wa, however, the category definitely excludes human beings.) As these glosses might suggest, makayidi-yadaku is used mostly when speaking of ‘animals’ in general, rather than referring to single individuals 58 FORTH Vol. 24, No. 1 or single kinds. Nevertheless, not only is the term regularly applied to a variety of animals, but it is recognized by Sumbanese speakers as denoting a category that subsumes less inclusive categories, particularly mahawurungu ‘flying things’ (mostly birds) and mabei ‘creeping, crawling things’, a large and internally di- verse category that includes insects, arachnids, reptiles, amphibians, and even fish. Although makayidi-yadaku can denote all nonhuman animals, its focus ap- pears to be undomesticated kinds. Consistent with this, yada can mean ‘wild, untamed, difficult to tame’, as well as ‘to move, be capable of movement’ (Kapita 1982; Onvlee 1984). According to Onvlee, yada refers more specifically to a quick movement; thus he further translates the word as ‘to teem, swarm’ and ‘to wrig- gle, fidget’. Somewhat curiously (since one might expect the contrast to be with yada), he also describes yidiku as denoting a movement slower than yidi. Similar to Nage and Keo, eastern Sumbanese possesses a special term for domestic animals. This is banda, the main sense of which is ‘goods, possessions, wealth’ (cf. Bahasa Indonesia benda; also Nage and Keo bugu, ngawu). As this derivation may suggest, the term refers particularly to large livestock, a mainstay of the Sumbanese traditional economy. Informants in the eastern Sumbanese do- main of Rindi stated that banda could be understood in the wider sense of ‘ani- mal’ (Bahasa Indonesia binatang), and that wild animals could then be distin- guished as banda matamba ‘wild banda’. Yet neither Kapita (1982) nor Onvlee (1984), the principal lexicographers of Sumbanese languages, records the latter phrase, and I suspect that, even at present, it is not a widespread or standard usage. Whatever the extent of their semantic overlap, makayidi-yadaku and banda are not obviously related by taxonomic inclusion. By the same token, banda sug- gests a utilitarian category, referring mostly, if not entirely, to a class of economic values. Mostly in the sense of ‘wealth’, variants of banda appear in other eastern Indonesian languages. A case where the more inclusive meaning has become restricted, not just to ‘domestic animal, livestock’, but to a particular domesticate, is Nage, where the cognate bhada is the name of the water buffalo, the most valuable animal in Nage traditional economy. Manggarai, Western Flores.—As a general term for ‘animal’, Manggarai kaka in some ways presents a more complex case than any of the usages reviewed above. To a greater extent than Lio ule, the lexeme appears in a large variety of Mang- garai bird names (e.g., kaka ketok, Sunda pygmy woodpecker), all of which ap- parently label folk generics (see Appendix 2). It also occurs in generic names for other kinds of animals, mostly snakes and insects (e.g., kaka ta’a, Green tree viper, Trimeresurus albolabris), as well as in the life-form terms for ‘bird’ and ‘snake’, kaka lélap (lélap ‘to fly’) and kaka léwe (léwe ‘long’).° For purposes of internal comparison, it should be noted that, in place of kaka, several Manggarai animal terms comparably incorporate kala (see kala mango, a kind of crab; kala wara, a kind of small red ant; and kala wura ‘watercock’; Verheijen 1963:686; 1967). According to Verheijen, kala derives from kaka by dissimilation (1963:685 n. 68).'* Whether this also applies to kara, a component of the names of just two birds (kara kuak and kara kua wié, the White-breasted waterhen and the Night heron) is not indicated.'5 Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 59 In all of these usages, kaka and variant forms resemble Lio ule insofar as the resultant compounds apply primarily to insects, birds, and snakes. Yet kaka dif- fers from ule (mostly in the sense of ‘maggot, worm’) in that, by itself, it appears not to designate simultaneously any folk generic, intermediate, or life-form taxon. This circumstance lends support to Verheijen’s (1963, 1967) interpretation of kaka as a general term for ‘animal’; hence an expression like kaka léwe ‘snake’ might be straightforwardly translated as ‘long animal,’ and kaka lélap ‘bird’ as ‘flying animal.’ To illustrate the general sense of animal, Verheijen further cites the phrase tjala oné kaka (1967 s.v. kaka 1) ‘perhaps some animal has entered’. This, he notes, can refer, for example, to a wild pig that may have invaded a cultivated field or an ant that has crawled into a placenta (kept after the birth of a child)— usages which affirm that kaka can refer to quite various zoological kinds.’ Other usages with the same import include akit le kaka ‘(to be) bitten by an animal’ (Verheijen 1967, s.v. soro II) and ngo bang kaka ‘to go hunting’, which incorpo- rates ngo ‘to go’ and bang ‘to bring’, and more specifically means ‘to bring dogs in order to hunt’ (ibid. 1967:186, s.v. kaka; see also bang motang ‘to hunt wild pigs’, motang ‘wild pig’, ibid.: 29, 337). The character of the Manggarai term, however, is complicated by the appear- ance of kaka in Nage and Ngadha names for quite diverse natural kinds, includ- ing, in a couple of instances, plants. In these languages, kaka occurs as a reference to living things only in a limited number of binary names for what are apparently folk generic categories. Nage contains six such names. While similarly few in number, the Ngadha compounds refer partly to creatures different from those designated by the Nage terms. Further variety is revealed by ethnozoological categories named with kaka which Verheijen records for Komodo, a language closely related to Manggarai (see Appendix 2). Some explanation for this diversity is available from evidence suggesting that, in at least some of the Nage terms, kaka reflects homonymous usages. For ex- ample, kaka in the Nage name of the Dollarbird is locally construed as an ono- matopoeic imitation of the bird’s harsh cry, whereas in kaka kea, the more elab- orate name of the Yellow-crested cockatoo (also simply called kea), kaka may be understood as a cognate of words with the same or similar referent in other Malayo-Polynesian languages (see Ngadha and Manggarai kéka, eastern Sum- banese kaka, Malay/Bahasa Indonesia kakatua ‘cockatoo’; Proto-Polynesian *ka(a)kaa or *kakaa ‘parrot species’, Wurm and Wilson 1975:147). By further contrast, kaka watu, the Nage name for a fish that characteristically inhabits the rocky bottoms of bodies of water, can be interpreted as incorporating kaka in the sense of ‘to stick, adhere, be attached to’ and watu ‘stone, rock’. (It is conceivable that kaka also has this meaning in the name of the Praying mantis, kaka koda.) The sense of ‘to adhere, be attached to’, which applies in Ngadha as well as Nage, would also explain the occurrence of kaka in Florenese names for life-forms other than animals. Thus, the two Ngadha terms, kaka bheto and kaka kaju, denoting an unidentified edible plant and species of Ficus, ferns, or vines (Verheijen 1990: 26), can be translated respectively as ‘what attaches to bheto bamboo’ and ‘what clings to trees’.” This evidence tends to rule out the possibility of Nage and Ngadha com- pounds representing remnants of an earlier classification in which kaka consis- 60 FORTH Vol. 24, No. 1 tently denoted a far more inclusive category of living things, and ultimately an ‘animal’ taxon as, according to Verheijen, it does at present in Manggarai. It is similarly difficult to see how kaka, either in Manggarai or central Flores lan- guages, could represent a semantic expansion of a term that formerly possessed a more restricted range of reference (as, hypothetically, Lio ule once did). For the Manggarai usage, a more likely interpretation can be found in further glosses of kaka listed by Verheijen (1967). These include ‘thing, object, article’ and nomin- alizing functions of kaka, in particles translatable as ‘that which’, ‘the thing which’, ‘one who’ (cf. Bahasa Indonesia yang). Rather than ‘flying animal’, there- fore, the Manggarai term for ‘bird’ (kaka lélap) might be glossed as ‘that which flies’ (cf. eastern Sumbanese mahawurungu, where ma is the nominalizer) or ‘fly- ing thing’. Similarly, kaka langu, the one nonzoological Manggarai name incor- porating kaka, which denotes a toxic mushroom (Verheijen 1967:186 s.v. kaka), can be translated as ‘that which intoxicates’ (see langu ‘to intoxicate’, ‘to act as though drunk’). Further supporting this interpretation, the large majority of Manggarai kaka compounds referring to living things do indeed translate as ‘that which (has a certain appearance)’ or ‘the one that (behaves in a certain way, makes a certain sound)’ (see Appendix 2).’* The point applies equally to compounds with kala. Thus kala wura (watercock), for example, may be interpreted as ‘one which is wura (a dead spirit)’; in fact, Verheijen provisionally glosses the name as ‘animal of the spirits of the dead’ (1963:868, n. 87). In view of Verheijen’s knowledge of the Manggarai language and of Mang- garai culture and natural history, one can hardly doubt his interpretation of kaka as a general term for ‘animal’. Nevertheless, the usage is likely to have developed as a synecdoche, whereby a word meaning ‘thing, entity’ has come to denote something more specific, namely, ‘living, animate thing’. Yet there remains the question of which, if any, of Adelaar’s four methods of designating ‘animal’ Mang- garai kaka exemplifies. If my interpretation is correct, kaka ‘thing’ may have its ultimate source in a hypothetical compound, *kaka X ‘thing that X’, where X was a word designating movement or the quality of animate life. Thus we may ulti- mately be dealing with a descriptive phrase comparable to eastern Sumbanese makayidi-yadaku ‘things that move’. Yet it is also possible that kaka ‘animal’ simply represents a generalization from the variety of compounds referring to particular animal kinds in which the term occurs—that is, as a kind of fictive etymology. Although Verheijen (1967) gives ‘livestock’ as one gloss of kaka, there is no reason to believe that this is the primary meaning, or that this meaning is the derivation of the more general sense of ‘animal’. AMBON-TIMOR LANGUAGES Rotinese.—The Rotinese term for ‘animal’, bana (dialectal banda, Jonker 1908), pro- vides an instance of a term denoting domestic animal having come to be u: the more general sense. Although bana is obviously cognate with Bahasa Indo- nesia/ Malay benda (see previously) and eastern Sumbanese banda, it is unclear whether the term retains ‘domestic animal’ as its primary sense. Jonker glosses the word first as ‘animal, especially a four-footed animal’, and lists bana fuik and bana aek as compounds specifying ‘wild animal’ and ‘tame, domestic animal’ Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 61 respectively. At the same time, he translates bana manu as ‘all sorts of animals, four-footed animals and birds, livestock and poultry’. Insofar as Rotinese manu refers specifically to the domestic fowl, this might suggest that the phrase applies, if not exclusively, then in the first instance to domestic kinds. It also suggests a distinction between ‘animal’ and ‘bird’ comparable to one sense of English ‘ani- ma Tetum (Tetun), Timor.—Closely related to Rotinese, and also classified by Wurm and Hattori (1981) as a member of a Timor and Islands subgroup within a larger Timor Area group (see note 2), the Tetum language of Timor contains at least two words for ‘animal’. One is binatan (Morris 1984), obviously borrowed from Malay (i.e., binatang); the other is balada ‘animal, beast’ (Hull 2001; cf. balada si‘ak ‘wild beast’), which is not explained. In addition to these, another, possibly older way of referring to animals in general is the expression buat na’in, glossed by Morris (1984) as ‘living things, any unspecified animal’. Tetum buat means ‘thing, object’ (cf. Manggarai kaka). Na’in functions as a title of respect and a numeral coefficient for persons, and is further described as referring to things that possess agency, or some particular power or skill; thus liras na’in, for ex- ample, means ‘things that have the capacity to fly’ (Morris 1984:146-147). Also noteworthy in this connection is the form na’i ‘lord, master’ (Hull 2001). Evidently an instance of the honorific use of the term, na’i occurs in the compounds na’i- bei ‘grandfather, ancestor; crocodile’, and na’i-boku ‘species of large kite’. Tetum buat na’in provides a further example of the use of a descriptive phrase to designate ‘animal’. The essential qualification is evidently provided by na’‘in, alluding to agency and the possession of (a specific) physical power. Semantically, therefore, the expression is most comparable to Sumbanese makayidi-yadaku ‘things that move’. Nuaulu, Seram.—Although included in Esser’s Ambon-Timor group, the Nuaulu language, spoken on the Moluccan island of Seram, is a fairly distant relative of Tetum and Rotinese. Wurm and Hattori (1981) place it in a Central Maluku group, separate from the languages of eastern Flores and Timor. Nevertheless, thanks to the work of Roy Ellen, Nuaulu is one of the few eastern Indonesian languages for which we possess detailed evidence with respect to ethnozoological classification, and for this reason alone it is worthy of comparative consideration. According to Ellen (1993a:96), Nuaulu ipai serves as a general term for ‘ani- mal’, but does not clearly include all life-forms that one might expect to find under this rubric. This equivocality appears largely to reflect disagreement or indifference among Nuaulu themselves. At the same time, ipai can be used in exclusive contrast to ‘human’ (mansia), in which context, Ellen (1993a:97) states, “it appears to be used to refer to all non-human animals.” Otherwise, the term may have as its primary sense “terrestrial animals, contrasted with those of sea and air” (Ellen 1993a:96). Consistent with the first specification, Ellen also de- scribes the Nuaulu term as somewhat resembling the polysemous use of ‘animal in English. He does not state whether or not Nuaulu explicitly consider named life-form categories (such as ‘bird’, manue, or ‘snakes and allied forms , tekene) to be included within ipai, nor does he discuss the possible derivation of the term. Nevertheless, the ethnographer’s statements on the whole suggest that ipai func- 62 FORTH Vol. 24, No. 1 tions as a label for a general category of ‘animal’ to about the same extent as does Nage ana wa or Manggarai kaka.” CONCLUSIONS: LEXICAL VARIETY AND SEMANTIC UNIFORMITY As the foregoing discussion has demonstrated, general terms for ‘animal’ found in eastern Indonesian languages exemplify all of the four ways of denoting this taxon identified by Adelaar. Naming with a descriptive phrase is illustrated by the Nage, Sumbanese, and Tetum usages. The use of a term referring to a more exclusive animal taxon is exemplified by Lio ule. A term that originally referred to domestic animals is represented by Rotinese bana (and, in a qualified sense, by Sumbanese banda). Finally, the use of loan words (in all instances from Malay binatang) is instanced by Lio binata and Tetum binatan, and also in Nuaulu (see note 20, regarding binatan).”' As this distribution illustrates, one method is not confined to the Bima-Sumba group of languages, nor to the Ambon-Timor group. In fact, as the Lio, Sumbanese, Tetum, and Nuaulu usages suggest, speakers of a single language may use more than one kind of term to express the general idea of * al With the possible exception of Manggarai and Nuaulu, none of the languages discussed above includes a single unanalyzable lexeme serving as a general term for ‘animal’, as exemplified by Malay binatang.” In this respect, the usages contrast with terms for particular life-forms, such as Nage nipa ‘snake’. Yet this does not mean that eastern Indonesians, or a significant portion of them, lack a well-de- fined concept of ‘animal’ As noted earlier, that they do possess such a concept is demonstrated by the widespread Austronesian grammatical feature of employing a single numeral coefficient when enumerating animals belonging to diverse life- forms (cf. Berlin et al. 1974:40, who describe the obligatory use of numeral clas- sifiers in Tzeltal as distinguishing ‘“unambiguously bounded” unique beginner taxa comprising ‘plants’ and ‘animals’). All utilizing the word for ‘tail’ (see also Malay ekor), instances drawn from languages surveyed in this article include Manggarai iko; eastern Sumbanese ngiu, from kiku ‘tail’; and Nage, Keo, and Lio éko (see e.g., Nage ja éko telu ‘three horses’, nipa éko wutu ‘four snakes’, hale éko lima ‘five flies’). A comprehensive ‘animal’ category is also implicit in such representations as the Nage taboo on speaking to animals, a prohibition whose consequential breach is described in oral tradition as involving such diverse crea- tures as snakes, crayfish, and goats (Forth 1989, 1998). In addition, as I hope to show in a future paper, the Nage category of ‘animal’ is indicated by the use of sex terms—comparable, for example, to English ‘bull’ and ‘cow’ and ‘buck’ and ‘doe’—which among living things are assigned only to zoological folk generics and not to plants (see Taylor 1990:117, who describes how, among the non-Aus- tronesian speaking Tobelo, plants as well have both male and female forms, even though in the majority of cases Tobelo are unable to identify these). Among Nage, sex terms are assigned to all categories of animals (ana wa), including reptiles, amphibians, fish, and insects as well as mammals and birds, and all are thought to engage in sexual intercourse, a behavior which Nage are not in every case able to verify empirically. Yet even if one accepts that all eastern Indonesians possess a category of Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 63 ‘animal’, it may not always be clear how far particular terms—whether analyzable or not—actually name the concept. As shown, usages that are equivocal in this regard include Lio ule and, probably, Nuaulu ipai. What the evidence does show, however, is that these, like the other eastern Indonesian terms described above, denote folk taxa which include two or more life-forms (such as ‘bird’, ‘snake’, or fish’). That they do not definitely subsume all life-forms that a modern English speaker might wish to classify as ‘animals’ is a dubious criterion for rejection. Moreover, it is arguably typical of all folk categories, pertaining to so inclusive a taxonomic level, including of course English vernacular ‘animal’, that they are inherently indefinite and subject to “prototype effects’’ (Lakoff 1987), and that what speakers and culture participants will recognize as included will be situa- tional, marked by ambivalence, and subject to individual variation.” All of the foregoing bears upon Berlin’s well-known thesis concerning the evolution of ethnobiological classifications (1992). According to Berlin, in the de- velopment of a language, (folk) generic taxa (local categories mostly coinciding with scientific species or genera) will be named, or “lexically recognized,” before higher order taxa, that is, life-form categories (such as ‘snake’, ‘bird’, ‘fish’, and so on) and ‘intermediate’ classes (categories comprising a limited number of sim- ilar generics included in a life-form, e.g., ‘birds of prey’). Later still, according to this theory, names will be assigned to ‘subgeneric taxa’ (ones comprising ‘folk species’ and ‘varietals’), while lexical recognition is finally given to the ‘kingdom’, of which ‘animal’ and ‘plant’ are of course the prime examples (Berlin 1992:274— 75). How many of these taxonomic levels are distinguished by name, in Berlin’s view, reflects the level of technological development of the society in question. In spite of ambiguity surrounding the question of what constitutes a ‘name’, the evidence of eastern Indonesian languages appears generally to support Ber- lin’s thesis. It almost goes without saying that the large majority of standard names for animals in these languages denote folk generics. In addition, usually two or more life-forms are labelled, and such labels often reflect reconstructed forms at the level of Proto-Austronesian or Proto-Malayo-Polynesian (see, for ex- ample, Nage, Ngadha, Lio, Endenese nipa ‘snake’; central Flores ika, Sikanese i’ang, eastern Sumbanese iyangu ‘fish’; and Tetum manu, Nuaulu manue, and Rotinese manupui ‘bird’). On the other hand, the degree to which eastern Indo- nesians label ‘intermediate categories’ is difficult to determine and defies any succinct summary—a situation which appears largely to follow from an inherent ambiguity reflected in the very designation ‘intermediate’. But even if life-form taxa (and perhaps some intermediates as well) are more consistently named than is the ‘animal’ taxon, this does not mean that early Austronesians (speakers of ancestral languages corresponding to Proto-Malayo-Polynesian or Proto-Austro- nesian) did not have ways of denoting ‘animal (in general)’. Indeed, the fact that the several eastern Indonesian languages surveyed here reveal precisely the same limited number of nomenclatural methods as do Austronesian languages in gen- eral tends to suggest that they did. In other words, these various ways of naming ‘animal’ may have developed no later (to retain the diachronic idiom) than did those for these other ‘higher order’, or supergeneric, taxa. Although the point cannot be fully developed here (but see Forth 1995, 2000, 2004), it may also be noted that names for several life-forms—e.g., Nage ana wa ta’a co and eastern 64 FORTH Vol. 24, No. 1 Sumbanese mahawurungu ‘bird’ (see also Sikanese kenaha horong ‘flying thing’, Pareira and Lewis 1998)—consist of descriptive phrases and so are formally iden- tical to terms for ‘animal’ in the same languages. The same may apply to Mang- garai terms for ‘bird’ and ‘snake’, if as hypothetically suggested, kaka ‘animal’, derives from a similar compound translatable as ‘living thin Two further points should be made regarding Berlin's evolutionary theory. First, if the driving force is technological development, then differences in lexical recognition of different taxonomic levels are evidently a matter of culture rather than human cognition per se. Secondly, if ethnobiological classification is seen to be grounded in universal factors of perception (which is Berlin's position, and one that I basically accept), then it is not clear how it can be subject to any sort of cultural evolution. Only in this light may one usefully raise the question of the ‘naturalness’ or perceptual salience of the taxon ‘animal’. It is by now well ac- cepted that ‘generic’ categories—also called ‘basic’ categories, and in psychology and logic, ‘basic-level’ kinds or ‘individuals’, and ‘basic level sortals’—are those which present themselves in perception as the most obviously discrete, and hence lend themselves most readily to lexical differentiation. By the same token, it is the representation of these categories that appears to be the most independent of the practices and values of particular cultures. Yet it should be considered that a category like ‘animal’ possesses almost equal salience, especially in regard to the property of movement (or animation), which as it were naturally distinguishes animals of all kinds as objects unlike all other objects, including ones that may be recognized as equally possessing the property of life (most notably, plants).” By contrast, intermediate categories (for example, groupings of birds encompass- ing several similar folk generic categories), and even some life-form taxa (for ex- ample, smaller creatures sometimes subsumed in named ‘wug’ categories), are arguably less psychologically salient, which is to say that their recognition, lexical or otherwise, may be as much dependent on particular cultural interests. Of course, one may ask why, if ‘animal’ possesses such salience, are names for this category apparently so uncommon? One response might be, again, that recog tion of a taxon does not always result in monolexemic naming. However, if ‘name’ is understood in an inclusive sense, with reference to the evidence of eastern Indonesian languages I would also suggest that such names may not in fact be as uncommon as has hitherto been supposed. NOTES ' Adelaar bases this assessment on data from 80 languages, belonging to four main branch- es of the Austronesian family, which are recorded by Tryon (1994). ? Wurm and Hattori (1981) retain Esser’s Bima-Sumba group (noting its ultimate derivation from the work of J.C.G. Jonker), but place the Ambon-Timor languages of eastern Flores and the islands of Solor, Adonara, and Lembata in a ‘Flores-Lembata subgroup’, which they then classify within a ‘Timor Area group’. Ambonese and other Moluccan languages are then placed in a ‘Central Muluku group’. * Formally comparable to ana wa ta‘a co is ana wa ta’a laka ‘crawling, creeping animals’, a term I first encountered in the Keo region. The category, however, encompasses snakes Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 65 (nipa) as well as a wide variety of other fauna, including insects, worms, grubs, crusta- ceans, amphibians, large reptiles like monitor lizards and marine crocodiles, and even rats and mice (dhéke). Subsuming or cross-cutting two and possibly three named or unnamed life-form taxa, it is difficult to see how the category could itself constitute a taxon. As Nage informants pointed out, moreover, the term can situationally include creatures that nor- mally fly (co) or swim (nangu), such as eels and crayfish when they find themselves on dry land, and flying insects like locusts and butterflies which otherwise creep or craw]; human infants also crawl. Consistent with this, ana wa ta’a laka appears not to be reg- ularly employed as a standard category, in which respect informants contrasted it with ana wa ta’a co. With regard to the application of the latter term specifically to birds (including bats), and not to flying insects, it is noteworthy as well that all insects that fly (co) also creep or crawl (laka). A similar category, also initially recorded in Keo and ap- parently less familiar to Nage, is ana wa ta’a ’e’e ‘ugly, disgusting animals’, which is identified with snakes—or more particularly dangerous snakes (such as the Russell’s viper, nipa ba), and certain kinds of grubs. *For Térong-Mawong, one dialect of Rembong, a language of northeastern Manggarai, Verheijen similarly records the cognate anak wara (wara = Nage wa ‘wind’) in the sense of ‘baby, infant’. In Rembong, the expression does not simultaneously serve as a general term for ‘animal’, although, interestingly enough, in another Rembong dialect (Wangka), anak wera is listed as a euphemism for ‘wild pig’. Wera ‘spirit, spiritual being’ is cognate with Ngadha wera and Nage wa—thus apparently a homonym of Nage wa ‘wind’—both of which refer to the malevolent spirit of a witch. Arndt’s dictionary (1961) does not indicate a Ngadha term for ‘animal’ (cana wara, corresponding lexically to Nage ana wa, is glossed as ‘snare for catching birds’), but this of course does not mean that none exists. 5 The fact that ana is used in this way more often with reference to birds may be accounted for by the fact that, as Nage themselves recognize, for creatures identified with other named life-form taxa, notably nipa ‘snakes’ and ika ‘fish’, the life-form name can be used instead, at least when this much of an animal's identity is known. ° Takashi Sugishima, Kyoto University, personal communication 2000. 7 Proto-Austronesian reconstructions include *uleg’ and *udaj ‘worm’ (listed by Wurm and Wilson, 1975 under ‘maggot’ and ‘worm’); *qulej, glossed with Bahasa Indonesia ulat (Fer- nandez 1996:158); and *ulaR ‘snake, worm’ (Zorc 1994:593). Fernandez (1996) has also reconstructed a ‘Proto-Flores’ form, *uler (equated with Bahasa Indonesia ‘ulat, see Ap- pendix 1). 81 follow Berlin’s practice of employing “folk generic” (or simply “generic’) to refer to ethnotaxa that comprise particular kinds mostly coinciding with scientific species or gen- era. ° Arndt glosses the last two terms, somewhat inexactly, as ‘Sparrowhawk’ and ‘Eagle owl’ (German Uhu). According to Verheijen (n.d.), ule polo refers to the Common koel (Eudyn- amys scolopacea). Evidence from Arndt’s dictionary that ule can be used alone in the sense of ‘bird’ is the phrase ule léla dzére ‘the bird flies suspended, hovers (1933:86, s.v. dhére; léla ‘to fly’). © With regard to senses of chong, I am most grateful for assistance kindly provided by Dr. Lin Jenn-Shann of the Department of East Asian Studies, University of Alberta, and Dr. Wu Xu, a former doctoral student in the university’s Department of Anthropology. 66 FORTH Vol. 24, No. 1 "Cecil Brown, who coined ‘‘wug” to refer to a life-form category comprising small crea- tures like ‘bugs’ and, frequently, ‘worms’ (1984:16), lists Mandarin chung as a ‘wug’ term, glossing it more specifically as ‘insect+worm+nonsnake reptile’ (Brown 1984:237). 2 Tn combination, the affixes ku- and -ku lend a repetitive or continuous quality to the basic verbal compound, while ma- renders the nominal sense. STt is a point of some interest, although one which cannot be fully developed here, that Verheijen (1967) lists Manggarai ular, clearly a cognate of Malay/Bahasa Indonesia ular ‘snake’, as the name of a particular kind of snake and also as a component of six com- pounds (u.-mandar, u.-mbani = u.-mbangi, u.-paka, u.-walok) specifying other kinds of snakes. 4 Blust (1983, “A Linguistic Key to the Early Austronesian Spirit World,” unpublished manuscript), who does not cite this interpretation, treats kala wara and kala mango as reflexes of Proto-Austronesian reconstructions he collectively designates as “’+qali/kali- forms.” In a complex analysis, he argues that these forms, prefixed to other morphemes, once marked a variety of biological kinds and other natural entities as things associated with spiritual danger, or more generally as ‘‘referents, states or actions that were believed to be connected with the supernatural world’ (Blust 1983:2). Whatever the merits of this argument, which is far too detailed to assess here, Blust evidently does not adduce the numerous Manggarai kaka compounds. 'S Another ethnobiological instance of kala is as a general term for ‘betel’ (Piper betle), in which sense it further appears in compounds denoting varieties of betel as well as several other plants, including some that are considered to resemble betel (Verheijen 1967). How- ever, it is not at all clear that kala in this context has the same derivation as the morpheme that appears in animal names. ‘6 For Rembong, a language, or cluster of dialects, spoken to the northeast of Manggarai (and within the northwestern part of the present administrative region of Manggarai), Verheijen (1977) lists kokaq reman as a general term for ‘wild animal’, and in one dialect as a specific reference to a wild pig. (A comparable double meaning is found in kokaq kazu—kazu ‘forest, wood’—glossed both as ‘monkey’ and ‘animal’.) Further occurring in a variety of compounds referring to particular kinds of mammals, birds, insects, and snakes, kokaq—glossed by Verheijen (1977) as ‘animal; thing, object; person; unidentified object or person (Bahasa Indonesia anu)’—is evidently cognate with Manggarai kaka. On the other hand, he translates reman as ‘leaf (leaves); grass, weeds; undergrowth, scrub; forest’. Relevant here are words with similar meanings used in other languages, including Nage and Sumbanese, to refer to wild varieties of animals that also occur as domesticates (see, for example, Nage wawi witu and eastern Sumbanese wei rumba ‘wild pig’). It is curious, however, that Verheijen glosses kaka remang, the Manggarai cognate of Rembong kokaq reman, not as wild animal but as ‘livestock’ (exemplified by horses and water buf- falo). The Manggarai term specifying wild animals is kaka puar, incorporating puar ‘forest, jungle’. ” The ferns denoted by kaka kaju are epiphytic (see Appendix 2). The only comparable plant name recorded for Endenese is kaka rawa (Dysoxylum, Verheijen 1990). Lio includes no ethnobotanical compound terms which include kaka, although in this language, also, the word has the sense of ‘to wrap around, cling, adhere to’ (Arndt 1933). Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 67 ** Also consistent with an interpretation of kaka as, essentially, a nominalizing particle are kaka dagang or kaka wagang ‘unidentified person; thing; genitalia’ (apparently as a eu- phemism), as well as kaka tana ‘earth spirit’ (tana ‘earth’), assuming the first component is not a variant of another lexeme, kakar (see the synonymous dialectal kakar tana). "’ At present, kaka does not occur as a nominalizing particle in Nage or Ngadha. However, as already noted, most if not all of the central Flores compounds incorporating kaka can be accounted for in quite different ways. >In a personal communication (22 February 2002), Ellen states that, at present, Nuaulu ipai is rarely used for ‘animal’ and is “increasingly replaced with binatan’ (cf. Malay binatang and the usages described above for Lio and Endenese). He also reports makapana as another general term for ‘animal’ (cf. Ellen 1993a:96, where this term is attributed to Rosemary Bolton, 1990). However, Bolton (pers. comm. 9 March 2003) states that makapana (from maka, a nominal prefix, and pana ‘to feed’) refers specifically to domestic animals. Citing a Nuaulu informant whom she questioned in 2003 in Bandung (in Java), she has subsequently claimed (pers. comm. 27 March 2003) that ipai is not a Nuaulu word, or at least is not a general term for ‘animal’. This apparent disagreement with Ellen is probably accounted for by the replacement of ipai with the loan word binatan, which is noted by Ellen himself. An obvious cognate of binatan, pinatane, is reported by Margaret Florey (pers. comm. 4 December 2002) as the only term for ‘animal’ in the Alune language of western Seram. 21 According to Adelaar (1994:13), a method comparable to employing a descriptive phrase is the use of a word meaning ‘game’ or ‘meat’ to denote ‘animal’. Although it does not name animals in general, it is a point of interest that nake (usually ‘meat’ or ‘game’ in central Flores languages) is listed as a general term for ‘bird’ in Endenese (Aoki and Nak- agawa 1993; Suchtelen 1921:340, 389). * Although my discussion has been restricted to Austronesian languages, it is noteworthy that Taylor (1990:49, 50, 67) reports a term for ‘animal’ in the non-Austronesian Tobelo language, spoken on the eastern Indonesian island of Halmahera. This is aewani. Since Taylor provides no interpretation of the term, it is presumably unanalyzable. 31 use ‘unanalyzable’ in the general sense. In contrast, Berlin et al. (1974:28) employ ‘an- alyzable’ and ‘unanalyzable’ in a way largely restricted to taxonomic relations. Thus, in their typology of lexemes, Nage ana wa ‘animal’ would be classified as an unproductive analyzable primary lexeme’, since the second element (wa ‘wind’) does not specify the term as labelling a taxon subordinate to one designated by the first element (ana child, person’; cf. Taylor 1990:40, for a critical discussion of Berlin’s typology). ee extent that he employs this typology in his 1992 book, Berlin (1992:27-28) speaks of names rather than “lexemes,” while he replaces “‘unanalyzable” and ‘“analyzable” with “‘simple” and “complex.” 2 As has often been recognized, English ‘animal’ can contrast for example with ‘bird’ or (according to Wierzbicka 1985) ‘snake’ (see Forth 1995:66, n. 2). 25 Although the matter cannot be explored in this paper, there is perhaps also a question of whether widespread life-form terms, such as those reflecting Proto-Malayo-Polynesian *manuk ‘chicken, bird, fowl’ (Zorc 1994:583; see also Proto-Austronesian *manuk ‘bird’, Bellwood 1997:102, Table 4.1), or indeed the protoform itself, are, or were, as consistently 68 FORTH Vol. 24, No. 1 inclusive as English glosses, such as ‘bird’, would suggest. Since Bima-Sumba reflexes of *manuk (such as Nage and Sumbanese manu), when used without modification, refer only to the domestic fowl, one is also led to ask, with regard to Berlin’s evolutionary thesis, whether the apparent loss of this lexeme as the name of a life-form taxon should be un- derstood as an instance of regression, or devolution. 26 Sexual and reproductive behavior is another feature that sets animals apart from other living and nonliving things. Yet, for Nage and other folk biologists, this is not so evident or observable as is movement and, indeed, for animal kinds that are rarely or never ob- served mating, is mostly attributed on the basis of inference. Nage, Sumbanese, and other eastern Indonesians apply terms for ‘living’ and ‘dead’ equally to plants and animals. Indeed, the idea that plants are ‘living things’ is probably universal, and, as Bloch (1998) has recently pointed out, is arguably part of the reason that plants (including trees) are, like animals, widely employed as human metaphors. This is not to say, however, that this common quality is a sufficient basis for the recognition of plants and animals—or, indeed, human beings (usually, and in a sense universally, distin- guished from animals)—as members of a superordinate taxon of ‘living things’, as is im- plicit in the western scientific concept of a ‘biology’ equally subsuming ‘botany’ and ‘zo- ology’. REFERENCES CITED A Pocket Chinese-English Dictionary. 1978. eae an oe to the Botanical Zhong Hua Shu Ju, Hong Kong. a Mayan-speaking People of Adelaar, K. Alexander. 1994. orhe history of thing, animal, person and related con- cepts in Malay. In Austronesian Termi- nologies: Continuity and Change, eds. A.K. Pal and M.D. Ross, pp. 1-20. Pa- cific Linguistics Series C-127. Depart- ment of Linguistics, Research School of Pacific and Asian Studies, The Austra- lian National University, Canberra. Highlond Chiapas Academic Press, New York and London. Bloch, Maurice. 1998. Why trees, too, are ood to think with: towards an anthro- pology of the meaning of life. In The Social Life of Trees: Anthropological Per- spectives on Tree Symbolism, ed. Laura stig pp. 39-55. Berg, Oxford, New Aoki, E. and S. Nakagawa. 1993. Endenese- Pisa Dictionary. Privately published. Arndt, P. 1933. Li’onesisch-Deutsches Wérter- Nike “Arnoldus-Druckerei, Ende-Flores. . 1961. Worterbuch der Ngadhasprache. Studia Instituti Anthropos 15. Anthro- pos-Institut, Posieux, Fribourg, Suisse. Bellwood, Peter. 1997. Prehistory of the Indo- Malaysian Archipelago, rev. ed. Universi- ty of Hawai'i Press, Honolulu. Berlin, Brent. 1992. Ethnobiological Classifi- cation: Principles of Categorization hd Plants and Animals in Traditional Soc ties. Princeton University Press, sah ton, New Jersey. anna Brent, D.E. Breedlove, and PH. Ra- ogy. American Anthropologist 75:214- - 1974. Principles of Tzeltal Plant Clas- Bist Robert, 1980. Early Austronesian so- zation: the evidence of lan- ng ge. Cond Anthropology 21:205-247. ee R.A. 1990. A preliminary descrip- on of Nuaulu phonology and gram- mar. M.A. Thesis, University of Texas at Arlington Brown, C. H. 1979. Folk pois life- forms: thei i d growth. American Anthropologist 81:791-817. - 1984. Language and Living Things: Uniformities in Folk Classification and Naming. Rutgers University Press, runswick, Chinese Etymological Dictionary. 1981. Cheng Chung Book Company, Taipei. Clark, Ross. 1994. Evolution, migration and extinction of Oceanic bird names. In Austronesian Terminologies: Continuity and Change, eds. A.K. Pawley and M.D. Ross, pp. 73-86. Pacific Linguistics Se- Spring/Summer 2004 ries C-127. Department of Linguistics, Research School of Pacific and Asian Studies, The Australian National Uni- versity, Canberra Dempwolff, Otto. 1938. Vergleichende Laut- lehre des austronesischen Wortschatzes. 3. a Austronesisches Worterverzeichnis. lin. Dietrich Reim Echols, J. and H. Shadily. 1963. An Indone- sian-English Dictionary, 2nd ed. Cornell University Press, Ithaca, New York. . 1989, An Indonesian-English Dictio- nary, 3rd oa. revised and edited by J. U. Wolff and J. T. Collins with the as- sistance of H. Shadily. Cornell Univer- sity Press, Ithaca, New Yor Ellen, Roy. 1993a. The Cultural Reldions of Classification: an Analysis of Nuaulu Ani- mal Categories from Central Seram. Cam- hi-r University Press, ss i 993b. Nuaulu Ethnozoology: a Sys- versi at C ry. Esser, S.J. 1938. Talen; Blad 9b. In Atlas van tropisch Nederland. Martinus Nijhoff, The Hague. Fernandez, Inyo Yos. 1996. Relasi historis kekerabatan bahasa Flores: kajian — historis komparatif terhadap sembilan ba- hasa di Flores. Penerbit Nusa Indah, Ende. Forth, ee 1989. Animals, witches a rm Indonesian variations o the prea complex’. Anthropos 84: 39. 1995. Ethnozoological classification and classificatory language among ¥ Nage of caper Indonesia. Journal Ethnobiology 15:4. ——-., 1998. ‘Beneath ie Volcano: Religion, Cosmology and rl snc among the Nage of Easter Verhande- lingen van het Koninklijk ae voor Taal-, Land- en Volkenkunde No. 177. s! Press, Leiden. Supplementary notes on ia bird classification and ethno-or- nithology. Anthropos 94:568-574. et astern Sumbanese bird clas- sification. Journal of Ethnobiology 20:161- a Nage Birds: Classification and Symboltens among an Eastern Indonesian JOURNAL OF ETHNOBIOLOGY 69 People. Routledge, London and New k ork. Hull, Geoffrey. 2001. Standard Tetum-En- glish Dictionary, 2nd ed., revised and ex- panded. Allen & Unwin in association with the University of Sydney, Crows Nest, NSW, Australia. gees J.C.G. 1908. Rottineesch-Hollandsch ordenboek. E.J. Brill, Leiden Lakoff, George. 1987. Women, Fire, and Dan- gerous Things: What Categories Reveal ane the Mind. University of Chicago s, Chica pos (Falibiey Pas 1966. The Savage Mind. Weidenfeld and Nicolson, Lon- don Kapita, Oe. H. 1982. Kamus Sumba/Kambera- Indonesia. Panitia Penerbit Naskah- ~ kah Kebudayaan Daerah Sumba De Penata Layanan Gereja Kristen ent a Waingapu. Morris, Cliff. 1984. Tetun-English -~wsieries Pacific Linguistics Series C-83. The ro tralian National University, Canberr Onvlee, L. 1984. Kamberaas (Oost- hous): Nederlands woordenboek. Foris Publica- tions Holland/U.S.A., Dordrecht. Pareira, M. Mandalangi and E. Douglas Lewis. 1998. Kamus Sara Sikka Bahasa In- csiigge Penerbit Nusa Indah, Ende, Flo- Stokhof, W.A.L., ed. 1983. Holle Lists: _— ularies in Languages of Indonesia, vol. 6: the Lesser Sunda Islands (Nusa ieatoena) Pacific Linguistics Series D-59. The Aus- tralian National eis th ore Suchtelen, B. C. C. M. 1921. Endeh (Flores). A iesicane van het Bureau voor de Bestuurszaken der Buitengew- esten bewerkt door het Encyclopaedisch Bureau, Aflevering 26. Papyrus, Welte- vreden Taylor, Paul Michael. 1984. “Covert cate- gories”’ — identifying unla- beled classes obelo folk biological classification. oe of Ethnobiology 4: 105-122 _—. 1990. The Folk Biology of the Tobelo People: a Study in Folk Classification. Smithsonian Institution Press, Washing- ton, D.C. Tryon, D. T., ed. 1994. Comparative Austro- nesian Dictionary. Mouton de Gruyter, rlin. Tyrell, P., V. Schnorr, W.V.A. Morris, and R. Breitsprecher, eds. 1991. Collins German- 70 FORTH English English-German Dictionary, 2nd d. Harper Collins Shape a Verheijen, J.A.J. 1963. Bird names in Mang garai, Flores, Indonesia. Aithropos 58: 8. . 1967. Kamus Manggarai, I: Manggar- ai-Indonesia. Martinus Nijhoff, The H. ague. . 1977. Bahasa Rembong di Flores Bar at. I: Kamus Rembong-Indonesia. o S.V.D., Ruteng, Flores. Mimeo- d. - 1982. Komodo: het eiland, het volk en de taal. Verhandelingen van het - lijk Instituut voor Taal-, Land- en Volk- enkunde no. 96. Martinus Nijhoff, The Hague. . n.d. Vernacular bird names in the Lesser Sunda Islands. Collection Ver- heijen, Historical Archive of the Kon- Institute voor Taal-, Land- en thar ec Leiden. Unpublished script. Wierzbicka, Anna. 1985. Lexicography and Vol. 24, No. 1 Conceptual Analysis. Karoma Publishers, Ann Arbor, Michigan. Wurm, S.A. and Shiro Hattori, eds. 1981. Language Atlas of the Pacific Area. Austra- lian Academy of the Humanities in col- laboration with the Japan Academy, Canberra. Wurm, S.A. and B. Wilson. 1975. ae Finderlist of Reconstructions in Austrone sian Languages (Post-Brandstetter). Pacific Linguistics Series artment of Linguistics, Reseach School of Pacific Studies, The Australian National Uni- versity, Canberra. Zorc, R. D. P. 1994, Austronesian culture Australian National University, Canber- ra. APPENDIX 1.—Cognates of Lio ule in other Flores and eastern Indonesian languages. ‘Proto-Flores’ *uler = Bahasa Indonesia ‘ulat’ (Fernandez 1996: 158; cf. Proto- stronesian *qulej, ibid.; cf. 7?PAN *ulaR ‘snake, worm’, Au: Zorc 1994: 593 Manggarai uli (dialectal ules, ulos; Verheijen 1982: ae ‘maggot’ (and ap- parently similar creatures, Verheijen 1 Komodo Ngadha é ma Endenese (Ja’o dialect hoes wo Sika wi uler ‘maggot, type of worm’ ‘erejen 1982: 131) maggot, ay caterpillar’ (Arndt 1961) Nage ule m, grub’ (Forth, field notes) Endenese (Nga’‘o dialect) ule (=oelech) enlerptioe (van Suchtelen ) 1921) rm’ (Aoki and Nakagawa 1993: 92) le (= Behaie Indonesia ulat; see also ule lale, ule nale ‘sea- cio ege ule klobat ‘cocoon’; ule tana ‘worms that eat maize roots’ Pereira & Lewis 1998: 203) ular ‘worm, caterpillar, larva’ (Morris 1984: 193; “fly maggot’ is labelled with another term); ‘worm, caterpillar; crawling in- sect’ (Hull 2001) Note: Several dictionaries which employ Bahasa Indonesia (the Malay-based esian national lan- Indon age) as the target language simply gloss the local word as ulat. Echols and Shadily (second edition, thir 1963) list ulat as: i caterpillar 2. worm, insect’ (cf. ulat ser and editi erangga, insects; cf. serangea, insec nm of their dictionary (1989) gives ‘1. caterpillar, worm (in compounds) 2. maggot, , Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 71 APPENDIX 2.—’KAKA’ compounds denoting living kinds in Manggarai and other languages of Flores. a) KAKA compounds in Manggarai (Verheijen 1963, 1967) Note: Not all names incorporating kaka appear in all Manggarai dialects. Also, in some dialects, the same zoological kinds are named with terms which do not incorporate kaka. Glosses of second components are from Verheijen 1967 and 1963: 716-717 (see “Summary ad hoc translation of Manggarai words”); (ON) indicates that, according to Verheijen (1968), the second term is onomatopoeic. (Onomatopoeic terms can be understood either as names for the sound or as verbs meaning to produce the sound in question.) All terms follow Verheijen’s orthog- raphy. For the sake of comparison, however, it should be noted that /dj/ corre- sponds to /j/ (cf. /j/ in English ‘jaw’) in the transcription of other Indonesian languages discussed in this paper, and /tj/, similarly, to the sound written as /c/ (cf. English /ch/ as in ‘chat’). kaka ando aék waé, kind of dragonfly (provisional identification) (ando aék ‘to bend over, bow’; waé ‘water, river’) kaka awa, a kind of spider (also simply awa) kaka bégol, a kind of poisonous snake (bégol ‘to throw, hurl’. According to Verheijen, under the synonym metjo, this snake is said to be able to spring or jump; the name therefore probably refers to Russell’s viper, see Forth 1995:52-53, s.v. ba bago) kaka dangka ‘earwig, locust’ (referents unclear; cf. Komodo kaka dangka, below) (= iko dangka; iko ‘tail’; dangka ‘branch, fork; hook; branch off, diverge’, evidently referring to the shape of the tail) kaka éa, Flores crow (Corvus florensis) (ON) kaka djurit, Bushlark (Mirafra javanica) (djurit ‘to run’; the bird in question characteristically runs along the ground) kaka kedéngké or kaka koé koé, Pitta (Pitta brachyura) (kedéngké ‘to hop’; koé ‘small’) kaka kék, White-breasted wood-swallow (Artamus leucorhynchos) (ON) kaka kéntu, a species of falcon and a species of hawk (Accipiter). (kéntu, har- vesting knife for rice; to cut, sever; cf. Nage bele teka ‘sharp wing’ as a name for a falcon; Forth 1996) kaka ketok, Sunda pygmy woodpecker (Dendrocopos moluccensis); Great tit (Parus major, Méngé dialect) (ketok ‘to knock, tap’) kaka kiong, Bare-throated whistler (Pachycephala nudigula; also called simply kiong) (ON kaka kuik, Cisticola spp. (small birds) (ON) kaka langu, toxic mushroom (cf. langu ‘to intoxicate’, ‘to act as though drunk’) kaka langu waé, sort of freshwater insect (waé ‘water, river’) kaka lawar, Apodidae and Hirundinidae (swifts and swallows; also simply called lawar; cf. Malay /Bahasa Indonesia kalewnr, bat; eastern Sumbanese kalewaru, swiftlet, Collocalia spp. Forth 2000) FORTH Vol. 24, No. 1 kaka leka, kind of poisonous snake (also simply leka, described as a speckled snake; cf. leka, palm bough, dried palm leaf used as decoration) kaka léros, ‘birds with cup-shaped nests’, generally Zosteropidae (white- eyes). (léros ‘yellow’) kaka lunteng, kind of grayish black snake that eats frogs and rodents (cf. lunteng ‘large piece of firewood’) kaka meésé, literally ‘large creature’, eagles (general term), also ‘water buffalo’ (mésé ‘big’) kaka muntung, dark phase of Spizaetus cirrhatus or other dark eagles (muntung ‘burned, dark-colored’ kaka nanong, kind of small insect resembling a spider; (dialectal) water strid- er, Gerridae (nanong ‘to go up and down’ kaka ndurut, kind of insect (ndurut ‘to hang, be suspended; (of a tree) packed with fruit’) kaka ngé’ok, kind of worm (ngé’ok ‘to move the body repeatedly’) kaka nteleng, kind of insect similar to a wasp and the size of a fly (nteleng ‘still, motionless’) kaka pémpang, kind of flying insect resembling a mosquito (pémpang ‘fever, malaria’) kaka petju, sort of malodorous insect, Pherosophus sp. (petju ‘to fart’) kaka rae, Red cuckoo-dove (Macropygia phasianella; also simply called rae or rae-rae) (rae ‘reddish color, brown’) kaka rawuk, kinds of hawks (Accipiter spp.; synonymous or overlapping with kaka kéntu) (rawuk ‘ash, gray’) kaka sara, centipede, Geophilidae (sara, kind of creeper growing in under- bush) kaka ta’a, Green tree viper (ta’a ‘half-ripe, green’) kaka téi or kaka tik, Brush cuckoo (Cacomantis variolosus) (ON) kaka teret, Bee-eater (Merops superciliosus) (ON) kaka toak, Common koel (Eudynamis scolopacea) (ON) kaka wadja, crocodile (cf. wadja = Malay/Bahasa Indonesia baja ‘steel, ar- mor; hard iron’) KAKA compounds in Nage kaka daza, Dollarbird (Eurystomus orientalis) kaka hika, Flying lizard (Draco sp.; Van Suchtelen 1921 records kaka héka for the Nga’o dialect of Endenese, while Arndt 1961 lists héka, transcribed as xéka, as ‘to have arms or wings’) kaka kea, Yellow-crested cockatoo (Cacatua sulphurea), also called simply kea kaka koda, Praying mantis kaka meo, one or more species of large spiders (cf. meo ‘cat’) kaka watu, kind of freshwater fish (watu ‘stone’) Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 73 c) KAKA compounds in Ngadha (from Arndt 1961, except where otherwise indicated) Note: I employ the same orthography as I use for Nage. Where Arndt’s usage differs from this, his transcription is placed in brackets. kaka, edible crab; ringworm (kaka also occurs as a reference to a skin disease in the Ja’o dialect of Endenese, Aoki and Nakagawa 1993) kaka bheto, edible plant ‘with with sourish leaves’ (Verheijen 1990; thus Dysoxylum sp.; cf. Endenese kaka rawa, Dysoxylum, ibid.) kaka daza, kind of bird (cf. Nage kaka daza) kaka kaju (kaka kadju), vine(s), fern(s) of the genus Asplenium, tree(s) of the genus Ficus (Verheijen 1990) kaka kuwe (kaka kuve), heron (kuve ‘speckled black and white’) kaka meo (kaka méco), large spider d) KAKA compounds in Komodo (from Verheijen 1982) kaka dangka, earwig kaka kéaq, Barn — (Tyto alba) kaka po, Large-billed crow (Corvus macrorhynchus) kaka rao, Glossy swiftlet (Collocalia esculenta); possibly also Drongo (Dicrurus sp. kaka wetog, Sunda pygmy woodpecker (Dendrocopos moluccensis) kaka koaq, Helmeted friarbird (Philemon buceroides) ait i Bek oot seat Journal of Ethnobiology 24(1): 75-91 Spring/Summer 2004 OJIBWAY PLANT TAXONOMY AT LAC SEUL FIRST NATION, ONTARIO, CANADA MARY B. KENNY and WILLIAM H. PARKER Faculty of Forestry and the Forest Environment, Lakehead University, Thunder Bay, Ontario, Canada P7B 5E1 ABSTRACT.—Ojibway (Anishinaabe) Elders of Lac Seul First Nation who are flu- ent in the Oji-Cree dialect and knowledgeable about plant names were inter- viewed. They provided names of 38 plant taxa, with some synonyms. The study also identified 65 Oji-Cree plant-related phrases, nouns, suffixes, and prefixes. There is no word for ‘unique beginner’ that corresponds with the plant kingdom. Plant classification at Lac Seul is based on gross morphology, but also reflects traditional utilization of the plants. The Lac Seul people do have names for two overlapping all-inclusive categories representing trees and all deciduous plants. There are five named and covert life-form taxa that are clearly morphologically defined: ‘conifer’ (covert), ‘angiosperm tree’ (covert), ‘shrub’, ‘herb, fern or fern ally’ and ‘moss’. Two additional plant groupings, ‘medicinal root’ and ‘berry’, reflect cultural utility and overlap with the five principal life forms. Although unnamed, one additional grouping, ‘bark plants’, was identified that is also based on cultural utility. It was found that the names of certain ubiquitous species are the same in Lac Seul Oji-Cree and in ee oe ne languages ranging from northern Saskatchewan to western Quebe Key words: ethnobotany, folk taxonomy, Ojibway, Anishinaabe, boreal forest. RESUMEN.—Se entrevist6 a varios ancianos Ojibway (Anishinaabe) de la etnia indigena Lac Seul, que hablan con fluidez el dialecto Oji-Cree y son buenos con- ocedores de los nombres de las plantas. Ellos facilitaron los nombres de 38 taxones vegetales, con algunos sinénimos. El estudio identific6 ademas 65 dichos, nom- bres, sufijos y prefijos relativos a las plantas en Oji-Cree. No hay ninguna palabra que corresponda al ‘reino vegetal’. La clasificaci6n vegetal entre los Lac Seul se basa en la morfologia general, pero también refleja la utilidad tradicional de las plantas. Las gentes Lac Seul tienen nombres para dos categorias superpuestas que incluyen todos los elementos dentro de ‘Arboles’ y ‘plantas de hoja caduca’. Hay cinco taxones de biotipos bien definidos morfolégicamente; algunos se nombran y otros no: ‘conifera’ (no nombrado), ‘angiosperma arbérea’ (no nombrado), ‘ar- busto’, ‘herba, helecho o similar’ y ‘musgo’. Dos grupos adicionales de plantas, ‘raiz medicinal’ y ‘frutillo silvestre,’ reflejan una utilidad cultural y se superponen a los cinco biotipos principales. Existe ademas una agrupaci6n adicional, no nom- brada, de las ‘plantas que producen corteza’, basada también en utilidad cultural. Se constat6 que los nombres de ciertas especies ubicuas son iguales en Oji-Cree de Lac Seul y en otras lenguas Algonquian extendidas desde el norte de Saskatch- ewan al oeste de Quebec RESUME.—Les ainés ojibwés (Anishinaabe) de la Premiére Nation Lac-Seul qui parlaient couramment le dialecte oji-cri et qui connaissaient bien les noms des plantes ont été interviewés. Ils nous ont communiqué 38 noms de plantes, ainsi 76 KENNY and PARKER Vol. 24, No. 1 que quelques synonymes. Cette étude a également permis de déterminer 65 phras- es, noms, suffixes et préfixes ojibwés liés aux plantes. II n’existe aucun terme pour «royaume populaire» correspondant au régne végétal. Au Lac-Seul, la classifica- tion végétale est fondée sur la morphologie grossiére; elle est également le reflet de l'utilisation traditionnelle des plantes. Les habitants de Lac-Seul possédent des noms pour deux grandes catégories inclusives qui se chevauchent, dont I’une re- présente les arbres, l’autre englobe toutes les plantes a feuillage décidu. II existe cing taxons ayant des noms oji-cris et qui portent sur des formes de vie cachées. Ils sont clairement définis sur le plan morphologique: «conifére» (caché), «arbre angiospermien» (caché), «buisson», «herbacée», «fougére» ou «plante affine» et «mousse». «Baie» et «plantes médicinales» forment deux autres groupes de plan- tes. Ces groupes sont fondés sur l'utilisation culturelle et chevauchent les cing principales formes de vie. Quoique ne portant pas de nom, un groupe additionnel a pu étre identifié qui correspond aux «plantes a écorce». Il s‘agit aussi d’un groupe fondé sur les utilisations culturelles. Finalement, il ressort de cette étude que le nom de certaines espéces ubiquistes ne différe pas entre le dialecte de la communauté du Lac-Seul et celui des autres groupes algonquiens dispersés entre le nord de Saskatchewan et l’ouest du Québec. INTRODUCTION This study focuses on folk plant classification of the Oji-Cree speaking Ojib- way (Anishinaabe) people in northwestern Ontario at Lac Seul First Nation, Lac Seul, Ontario, Canada. While considerable information regarding Ojibway and Cree plant names is known (Densmore 1928; Leighton 1985; Meeker et al. 1993; Smith 1932), there has been no study of the plant classification system of the Anishinaabe in northwestern Ontario. The purposes of this preliminary and mod- estly funded study were to document plant names and plant category terms in the Oji-Cree language at Lac Seul First Nation in northwestern Ontario, Canada; to determine how the plants are classified; and to compare the Lac Seul plant nomenclature with other nonagrarian North American Algonquian folk systems. Only a small percentage of the on-reserve population retains the Oji-Cree lan- guage. Most of these people are Elders over 60 years old. Geographic Setting and Vegetation.—Lac Seul First Nation is situated on the south shore of Lac Seul at approximately 50°09’ north latitude and 92°12’ west longitude. The total land area is inexact due to fluctuating water levels from the dam at Ear Falls, Ontario, but it is estimated to be 27,287 ha (67,375 acres). Lac Seul lies within the Upper English River section of the boreal forest region (Rowe 1972). This area represents a transition zone between the Great Lakes-Saint Lawrence forest to the south and the boreal forest to the north. The vegetational regimes at Lac Seul show typical boreal diversity due to a continuum of patch disturbances by fire, blowdown, and insect infestation. In addition, human activ- ity has created long-term modifications in the forest. To this day people create habitation sites, trails, and portages. The Ojibway (Anishinaabe) of Lac Seul.—From archaeological remains, it is known that aboriginal people have occupied the region of Lac Seul and the English River watershed for about 9000 years (Dawson 1983). Anishinaabe people resided on Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 77 Lac Seul when the first European fur traders came into the waterways. The lan- guage spoken at Lac Seul is a dialect of Oji-Cree, a blend of two Algonquian languages. From precontact times to first contact with European traders in the 1750s and well into the latter half of the twentieth century the Lac Seul Anishinabe have maintained a fishing, hunting, and gathering culture of the Boreal and Great Lakes-St. Lawrence ecosystems. This transition zone was rich and varied in food resources before Lac Seul was flooded to create a reservoir by a joint agreement between Canada and the United States (first in the early 1930s and again in the 1950s). The once extensive wetlands supported wild rice beds and high waterfow] populations, extensive fish spawning grounds, muskrat and beaver habitat, and they supported many wetland plants used for food and medicine. MATERIALS AND METHODS Knowledge of many of the Oji-Cree names for plants and their uses persists only among a few older residents at Lac Seul. Interviews were conducted sporad- ically from 1997 to 1999 with Elders known for their knowledge of the local plants. Plant identification was based on pressed or fresh specimens presented to the respondents.’ Questions were presented in Oji-Cree with the help of a trans- lator who was a resident of the community. In addition to the traditional names of specific plants and plant groups, respondents were asked about traditional economic uses and the relationships of plants to each other. Details of the plant collection and interview process are presented in Kenny (2000). Berlin’s (1992) modified Venn diagrams are used to present the plant classi- fication at different hierarchical levels. All ethnobotanical taxa are represented by gray circles and labeled in bold italics. Botanical taxa are represented by black circles, usually at the genus or species level, and labeled in italics. Hierarchical relationships are shown by nesting the circles representing taxa. Where taxa are interpreted to converge (i.e., they are not mutually exclusive), the circles overlap. Where taxa are covert (unnamed), the circle is a dashed line. RESULTS As has been found in other folk taxonomies, there apparently is no all-encom- Passing word for ‘plant’ at Lac Seul; the covert kingdom Plantae is inferred from descriptive vocabulary (Berlin 1992). This study identified a total of 65 Oji-Cree plant-related phrases, nouns, suffixes and prefixes that were collected from re- spondents at Lac Seul First Nation.2 These 65 names and phrases identify 38 botanical taxa (Table 1) and 19 morphological characteristics of plants (Table 2). A list of the plant taxa is provided in Table 1 with Oji-Cree translations. Table 1 also lists the synonyms encountered when respondents offered dif- ferent names for the same plant. When synonyms were revealed, one name ex- pressed an attribute of the plant and the other described a use of that plant. For example, Cornus stolonifera is named paashkoaatig ‘scraping (inner bark) for use and miskaabemig ‘visible (from a distance) because of its redness’; Diervilla loni- cera is named ozhaawaapimaatig ‘green bark shrub’ and ozhaawaakiimiinaatig TABLE 1.—Plant taxa from Lac Seul First Nation, Ontario, Canada. Taxon! Oji-Cree English translation Common English name Abies balsamea (L.) Mill. shingob bigiw fir/spruce with gum alsam fir Acer spicatum Lam. zshaashaagopemaagaatig (n.) ‘chewing wood (shrub)’ mountain maple zshaashaagopemigoon (v.) ‘chewe d upon this wood has Achillea yer: L. waabigooniinzens ‘flower which i is little’ yarrow Acorus calamus achiitemoo azoo ‘squirrel tai sweet flag wiikensh no translation Alnus crispa ba Pursh. moozpaag oe the) ee moose bell’ green alder Amelanchier spp. zhigaagomiinen = saskato y’ saskatoon berry zhigaagomiinaatig aot e = saskatoon berry wood (sh foes margaritacea (L.) Benth agawaapamakiin pearly everlasting an androsaemifolium L. Aralia nudicaulis L. Arctostaphylos uva-ursi (L.) preng. Betula papyrifera Marsh. Clintonia borealis (Ait.) Raf. Coptis trifolia (L.) Salisb. ornus canadensis L. Cornus stolonifera Michx. Corylus cornuta Marsh Diervilla lonicera Mill. Epilobium angustifolium L. Eupatorium perfoliatum L. mahkwa ochiibig 7) waabooz ochiibig menozhaatig wiigwaas zhaashaagomiinaatig (n.) shaashgobuteh (v.) osawa ochiibigens ozhaawaapimaatig ozhaawaakiimiinaatig wiikaasenseywin shiingibiishkag aazhaabaakesiing ‘when it comes dee in the day- light’ ‘bear's root’ ‘for be ‘rabbit's ‘healing Cad (shrub)’ no translation ‘chewable berry wood (shrub)’ ‘(it is) going through (the body)’ yellow root oa ‘chewing ber ‘scraping it (bark) hs use shrub’ ‘being red it is vis ‘nut wood (s hruby’ ‘green bark shrub’ reen ae berry shrub’ ‘being peelabl ‘like the fish duck’ (red grebe) ‘(stem) going right through’ spreading dogbane wild sarsaparilla bearberry, kinnikinnik birch blue bead lily goldthread bunchberry red osier dogwood beaked hazel bush honey-suckle fireweed boneset daNaAVd pue ANNA T ‘ON ‘FZ “TOA TABLE 1—(continued) Taxon! Oji-Cree English translation Common English name Fraxinus nigra Mars gemag no translatio black ash Gaultheria hispidula ) Muhl. amiinaadekag ‘leaf/be uk that smells good’ creeping snowberry Inonotus obliquus (Fr.) Pil. egun ‘in the li Linnaea borealis L. Picea spp. Pinus spp. Pinus banksiana Lamb. Populus spp. Pteridium aquilinum (L.) Kuhn Sambucus pubens Michx. Sorbus decora (Sarg.) Schneid. Sphagnum spp. Class Bryidae (’ a: mosses’) Thuja occidentalis Usnea cavernosa te Vaccinium spp. Vaccinium silos Sage Ait. Vaccinium myrtilloi saagaategu paapiishaagakiig shingob zhingwaak kik azaadii ginebigoon wiimbashkwaatig mahkwaomiinaatig miishiigan michaa miin —- come in later in the umm. no translation no translation ‘snake ce’ ‘bursts off the stem wood’ (bark) ‘bear’s berry wood’ tion ‘hairy thing’ ah for lichens hanging, from tr ‘bluebe ‘little blueberry’ ‘large blueberry’ twinflower white or black spruce red or white pi jack pine trembling aspen or balsam bracken fern red elderberry showy mountain ash sphagnum or peat moss feather moss eastern white cedar old man’s beard blueberry low sweet blueberry velvet leaf blueberry ‘ Nomenclature follows Baldwin and Sims (1989) and Gleason and Cronquist (1963). F007 Jowuung /Sutidg ADOTOISONHLA JO TVNINOL 80 KENNY and PARKER Vol. 24, No. 1 Mitig . “Conifers > “a a 2 ' meaaga \ : /&ngiosperm trees’ pe fer & fern allies ee f f -angiospe: -herbs, fone & fer allies Ochiibig “\ smedicinal roots/ Miin -berries -aatig -angiosperm shrubs Gaamig Cultural Life-form -mosses FIGURE 1.—Lac Seul life-form taxa and overlapping culturally-based groups. ‘green inner bark, berry shrub’. Apocynum androsaemifolium is named mahkwa ochiibig ‘bear's root’; osheysep ‘for twine’ refers to its root or stem. Clintonia bo- realis is named both shaashgobuteh ‘(it is) going through the body’, which refers to its medicinal activity as a laxative, and zhaashaagomiinaatig ’chewable berry shrub’. The highest ranking plant groups are represented in Figure 1. The two some- what overlapping life-form groups which include vascular plants are mitig ‘tree’, which includes all trees, and aasaagakiik sensu lato (s.1., in the broad sense) which contains nonconiferous woody plants and herbs, together with ferns and fern allies. While the noun mitig denotes all trees, the focus of this term is primarily conifer species reflecting the composition of the boreal forest vegetation at Lac Seul (Rowe 1972). The noun aasaagakiik s.1. defines a large part of the domain of plants. According to one respondent, aasaagakiik s.|. is explained as ‘‘all things which grow up out of the ground, change and die.’”” The deciduous conifer, Larix laricina (Du Roi) K. Koch, is probably excluded from this group. Five additional life-form taxa are ‘conifer’, ‘deciduous tree’, ‘shrub’, ‘herba- ceous plant’ (which includes ferns and fern allies), and ‘moss’. Each of these groups is polytypic since each encompasses more restricted named taxa. Coniferous and angiosperm trees are recognized as distinct entities by the Lac Seul people. Both types are included in the designation mitig ‘tree’, but only giosperm trees are also included within aasaagakiik s.l—angiosperms and fern allies (Figure 1). ‘Conifer’ may be thought of as a covert intermediate life- form taxon inferred from the vocabulary describing parts of coniferous trees (Ta- ble 2): opii ‘conifer needle’, ozhiigoopiin ‘needled branch of conifer’ (usually spruce and fir) and bigiw ‘spruce or fir ‘Deciduous tree’ is also a covert intermediate life-form taxon being described by a respondent as aasaagakiik (i.e., separate from the conifers). The covert life-form groups ‘conifer’ and ‘angiosperm trees’ are shown with Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 81 TABLE 2.—Morphological vocabulary describing plants in Lac Seul Oji-Cree. English Oji-Cree Nouns berry miin niibish inner bark (with sap and latex) oshiiban root ochiibig needled branches of pine ozhtigoopiin needled branches of spruce/fir bigiwoopiin gum bigiw nut bigaan flower waabigwan tree miti plant—that grows, changes and dies aasaagakiik (herbaceous and/or woody deciduous plant) Suffixes wood/shrub -aatig root -ochiibig erry -omiinen conifer needle -opii moss, lichens -gaamig Prefixes chewin zhaashaa- yellow osaa- green ozhaa- red miska- associated genera (Figure 2). The covert ‘conifer’ life form contains four genera. Shingob contains both Picea glauca (Moench.) Voss and Picea mariana (Mill.) B.S.P. as well as Abies balsamea. The two undifferentiated botanical Picea species are an example of the botanical ‘‘type concept” for the folk genus shingob. Shingob is the type species for the genus while shingob bigiw is a binomial referring to the folk species ‘gummy shingob’ (Abies balsamea). Zhingwak contains both Pinus re- sinosa Ait. and Pinus strobus L. but excludes Pinus banksiana, which is called kik. The covert life form ‘angiosperm trees’ contains four folk genera: aagemag (Fraxinus nigra), azaadii (Populus spp.), moozpaagoon (Alnus crispa), and wiig- waas (Betula papyrifera). Azaadii includes both Populus balsamifera and Populus tre- muloides. Alnus crispa (moozpaagoon) has an intermediate stature between a tree and a shrub at Lac Seul. Since it is not modified by the “shrub” suffix -aatig, so it is grouped with ‘angiosperm trees’. Plants with names having the suffix -aatig constitute the life form ‘shrub’ (also translated as ‘wood’), and are included in the designation aasaagakiik s.1. (Figure 1). We have been unable to determine if -aatig also refers to coniferous shrubs such as Juniperus communis L. and Taxus canadensis Marsh., which are pres- ent at Lac Seul. The names of nine folk genera, eight of which are considered shrubs, have the suffix -aatig (Figure 3). The eight characteristic members of the group -aatig are zshaashaagopemaagaatig (Acer spicatum) zhigaagomiinaatig (Amelanchier spp.), menozhaatig (Arctostaphylos uva-ursi), paashkoaatig (Cornus sto- 82 KENNY and PARKER Vol. 24, No. 1 ’ a Gili. rd ‘i ae how ob Ori ca (Moench) A.” aagoon oe (@ek (Mill.) ia, aag WG) oy wy (Ait.) Pursh \NA ee nie i x i \o- balsamea (L.) st | or ‘ —— I (5 an | | (OC) Betula pu Be oS Ore ap ' een, ‘ja jn SEP 6 \ \ ve On = oa aa J (One sont L eee Hh ‘ \ \—O Pinus resinele Ait’ / kik k/ a Oforulus trem loides Michx. hs O us banksiana Lamb. \ wea) ~ ‘Populus spp. , tie, Cle lguet "ete we “Conifers “/ (Covert) “ Angiosperm trees” / (Covert) FIGURE 2.—Lac Seul ees life forms: conifers and angiosperm trees with nested folk generic and folk specific t lonifera), bigaanaatig (Corylus cornuta), ozhaawaapimaatig (Diervilla lonicera), wiimbashkwaatig (Sambucus pubens), an ee mahkwaomiinaatig (Sorbus decora) (see Table 1 for English translation of Oji-Cree). The name of one aberrant nonwood folk genus also includes the -aatig site zhaashaagomiinaatig (Clintonia borealis) is a low, herbaceous member of the Liliaceae. It has not been determine this anomaly exists. e form aasaagakiik is polysemous, being applied at more than one taxonomic rank. Besides referring to all deciduous plants including trees when used in the broad sense, the same word is also used in a more restricted context, -aatig at is eT : “woody /shrub” i Amelanchier “ pee ozhaawaapimaatig Pa ES ia O \ la lonicera Mill. o~ yrren E uva-ursi (L.) Spreng. ; ‘ koaatig net eimins Michx. Pipremantig ae Me cornuta Marsh. A (0) - inaatig ‘onia borealis (Ait.) Raf. FIGURE 3.—The Lac Seul life form -aatig, (shrub) with nested folk generic taxa. Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 83 ochiibig ‘root’ aasaagakiik een : ie Bibi oe (medicinal) \ Pa P Phin, \ Pa : fo ian Set \ 7 achiitemoo azooAwiikensh) | > / Acorus calamus (L.)Mill. ie a } » f | é / ginebigoon f i \ / Pteridium aquilinum#(L.) Kuhn. (O) \ a \ : paapiishaagakiig ©) } Linnea borealis L. WY i | wien “berry : | : | aazhaabaakesiing | f oO) / j Eupatorium perfoliatuma L. \NY 7 aes ‘ ail Fj “4: \ / Bark Utility ‘, : j % 2 Sf waabigoontinzens moe Achillea millefolium L. { CO) — f ( f ©) _sgawaapaamakiin ~~“ Anaphalis margaritacea (L.) Benth & Hook ra shiingib kag ©) Epilobium an lium ee net FIGURE 4.—The Lac Seul life form aasaagakiik s.s. (herbaceous plants, ferns and fern allies) with nested folk generic taxa and overlapping culturally-based groups ochiibig (me- dicinal root), miin (berry) and bark plants (covert). aasaagakiik sensu stricto (s.s., in the narrow sense), to denote only nonwoody herbaceous plants, including ferns, and probably fern allies. The meaning is “things that grow, change and die.” Thirteen folk genera were identified within this life form, two of which have synonyms (Figure 4): waabigooniinzens (Achillea millefolium), achiitemoo azoo (syn. wiikensh; Acorus calamus (L.) Mill.), agawaa- paamakiin (Anaphalis margaritacea), mahkwa ochiibig (syn. osheysep; Apocynum androsaemifolium), waabooz ochiibig (Aralia nudicaulis), osawa ochiibigens (Coptis trifolia), zhaashaagomiinen (Cornus canadensis), shiingibiishkag (Epilobium angus- tifolium), aazhaabaakesiing (Eupatorium perfoliatum), amtinaadekag (Gaultheria his- pidula), paapiishaagakiig (Linnaea borealis), ginebigoon (Pteridium aquilinum), and miin (Vaccinium spp.). The folk genus miin contains two binomial folk specifics that correspond to botanical species: michaa miin (Vaccinium myrtilloides) and mi- inens (Vaccinium angustifolium). Grasses did not emerge as a separate life form or even folk genus category. Although plants in the flowering stage were presented to the respondents when- ever possible, they were included within aasaagakiik s.]. together with other flow- ering herbaceous plants. Two additional groups identified on the basis of cultural utility are contained within aasaagakiik s.s., and include members of the shrub and nonwoody her- baceous plant groups. The names of these two groups reflect a useful morpho- logical feature or medicinal value (Figure 5). The first of these groups obtains its name from the root word miin ‘berry’ and contains seven folk generic names encompassing shrubs and herbaceous taxa listed above. The type is the folk genus miin (Vaccinium spp.) and includes only the ‘blueberries’. The folk genera desig- nated by miin are characterized by their edible berries with only one exception. 84 KENNY and PARKER Vol. 24, No. 1 ochiibig seth. iy Lae ais - “root”, (medicinal paashkoaatig ~. _.— Bark Utility ae ee * / ) rep antes Michx, “< en Sie 2 ozhaatig 7. poner ochiibig hy (O) ©; ere uva-ursi (L.) Spreng. / ©) Apocynum androsaemifolium L. \ Not __.. waabooz dpheii ibig | O | Aralia ss ae L. " \ - fas NS # Pa Pece eee ( ) { @, \ Acer spicatum Lam. wikenih ; z osaawa ochiib ba 4 feecun ee Aconys saa (L) Mil. O cores a c) Salis vine ne \ O pt ee er ae ‘ Sambucus pubens Mich — a a ae min ‘ sree -"berry” ve sep * (O) \ | win miinens IN ~~ ¥ = Ss Vaccini iPS ~~. — — L amiinaadekag oe ad (@) Vaccinium angustifolium Ait. ~ ltheria procumbens L. Me | ocheaahiiinaaig/ Di i \ j /f \ vege % rt neccoretie ie © { & } { O \ \O% nse ies Mi zhaa shaa agomiinad itty a ME iA %; / Clintonia borealis (Ait.) Raf. A — oeN —— chaashaagominen a iigac agomiir iment Cornus canadensis Amelanchier spp. GURE 5.—The Lac Seul culturally-based groups ochiibig (medicinal root), miin (berry) and ‘bark plants’ (covert) with nested folk generic and folk specific taxa. Ozhaawaakiimiinaatig (Diervilla lonicera) has oblong seed capsules, which are described by the word miin ‘berry’. At Lac Seul, the word miin also refers to any tiny, rounded thing The name of the second group is derived from the base word ochiibig ‘root’ and includes herbaceous plants with medicinal roots, all in the life form aasaa- gakiik s.s. (Figure 5). The word ‘root’ is separate from the modifier (e.g., mahkwa ochiibig ‘bear's root’). These are not food plants; but in the case of mahkwa ochii- big there is a synonym, osheysep ‘twine’, which labels another economic use for the species and refers to the stem of the plant. Medicinal roots are of high cultural significance. Although neither of the synonyms (achiitemoo azoo and wiikensh) for sweet flag (Acorus calamus) contain the base word ochiibig, this taxon is also included in the ‘medicinal root’ grouping because it is used extensively in medicine and ceremony. Wiikensh is used by respondents when referring to the ceremonial and medicinal use of Acorus calamus. The name achiitemoo azoo ‘squirrel tail’ de- scribes the spadix, which was used when our consultants identified fresh or pressed specimens. Wiikensh may represent an additional folk genus since the term is also applied to Iris versicolor L., which has a different medical use. There are indications of a covert third group ‘bark plant’ mainly consisting of shrubs having bark used for various purposes. Although there is not a single word for this concept, there are six folk genera names related to the scraping or peeling of the bark to be used for smoking, medicine, and twine making (Figure 6). The Oji-Cree names describe how the bark is harvested or used: zhaashaa- gopemaagaatig ‘chewing wood’ (Acer spicatum); menozhaatig ‘healing shrub’ (Arc- tostaphylos uva-ursi); paashkoaatig ‘scraping (inner bark) for use’ (a synonym for Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 85 Gaamig ae a Bryophyta Pt in. \ 4 waabangaamig agki ’ ff ~“earth’s white ground Cever” \/ i \ / ; j / \ j : \ H ikwe waabangaamig i ~"woman’s white ground cover” i “ : : j es gaamig | Sphagnidae j \ a f / gp ozhaagaamig f c . Bryidae # ne ott FIGURE 6.—The Lac Seul life form gaamig (ground cover; ‘moss’) with nested folk generic and folk specific taxa. Cornus stolonifera); ozhaawaakiimiinaatig ‘green inner bark shrub with a berry’ (Diervilla lonicera)—miin is used to identify the capsule of this shrub; and wiim- bashkwaatig ‘bursts off the stem (when it is peeled)’ (Sambucus pubens). The last folk genus, osheysep, is a robust herb (Apocynum androsaemifolium) whose name means “‘twisted for twine.” This last folk genus also falls into the ochiibig, or ‘root’ group, having the synonym mahkwa ochiibig. The life form gaamig ‘moss’ is mutually exclusive of the other life forms. This term means ‘ground cover” and includes all species of Bryophyta with a spread- ing unbroken growth pattern (Figure 6). We do not know if the ground lichens are included in this category. The life form gaamig contains two folk genera, waabangaamig (Sphagnidae) and ozhaagaamig (Bryidae). It is not known how many additional genera are recognized in this life-form group, either named or covert. Numerous unidentified Sphagnum spp. were grouped together by a re- spondent as waabangaamig and include two folk species, waabangaamig aaki ‘earth white ground cover’ and ikwe waabangaamig ‘woman white ground cover’. Each of these names are trinomials. Respondents at Lac Seul identified two additional folk genera, a lichen and a fungus, that could not be categorized under any life form. Unaffiliated genera are usually morphologically different from the genera of other life forms, and Berlin (1992:172~-181) describes these as “ambiguous.” The arboreal lichen, Usnea caver- nosa is named miishiigan. This may actually represent a life-form taxon since all Usnea spp. were identified as miishiigan, a term which may designate all hanging arboreal lichens. The other unaffiliated genus is saagaategan (Inonotus obliquus), the clinker or cinder polypore conk, which looks like a mass of burned coal (a clinker) stuck on a birch tree. This conk is morphologically discrete from any other life form and even from any other wood decay conks which may also con- stitute another life form. 86 KENNY and PARKER Vol. 24, No. 1 Similar Plant Names in Other Ojibway and Cree Dialects—The Ojibway and Cree languages are part of the larger Algonquian language family which extends from the Atlantic coast of North America across Canada and the United States to lowa, eastern North Dakota and as far west as Alberta (Ningewance 1993). A compar- ison of similar plant names in other Ojibway and Cree dialects is shown in Table 3. The eight species shown are ubiquitous in the boreal forest ecosystem where northern Ojibway and Cree reside and in the Great Lakes-Saint Lawrence forest ecosystem where the more southern Ojibway are located. Numerous ethnobotan- ical researchers working with Ojibway and Cree communities have collected names of plants along with the uses of the species without analyzing the folk taxonomies. Although the researchers used slightly different spelling systems for the Ojibway and Cree dialects, the names of plants show a marked correspon- dence across widely separated geographical areas. Aralia nudicaulis (wild sarsaparilla) is named waaboozojiibik by the Great Lakes Ojibway, wabos’odjibik by the Ojibway in northern Minnesota and On- tario’s Lake of the Woods area, waposocipihk by the Woods Cree of east-central Saskatchewan and waaboozochiibig at Lake Seul (Table 3). The consistency of these names covers a distance of several thousand kilometers. The names for Aralia nudicaulis are translated as ‘‘rabbit’s root’’ in all dialects except the Min- nesota/Wisconsin Ojibway (Smith 1932) where the Ojibway word may mean “rabbit woman's medicine.” Smith (1932:356) gives the Pillager Ojibway name wabo's tiskwe and a description of women’s use of it as a “remedy for blood purification during pregnancy.” Abies balsamea (balsam fir), is similar in the four Ojibway dialects (Table 3), with overlapping among the names and the synonyms (i.e., shingob bigiw, zhin- gob bigiwaandag, and aninaandag). The Lac Seul dialect leaves off the -aandag suffix used in the three other Ojibway dialects. This ending may correspond to the suffix -aatig ‘woody’ which people at Lac Seul use for denoting shrubs. The northern Woods Cree name seems dissimilar but there may be a remnant of similarity in napak. The names for Acer spicatum (mountain maple) were not collected by Dens- more (1928) or Leighton (1985); the species is found in Densmore’s region of study but not in Leighton’s. The names are similar in the three Ojibway dialects. The Lac Seul name includes the ‘shrub’ suffix -aatig. The differences in spelling are more a feature of variation in orthographic conventions than in actual pronunci- ation. The suffixes -gobiimag and -gobi’mtik are comparable. The ‘‘c’”’ in caca- is pronounced “s” in both the Densmore (1928) and Smith (1932) phonetic pronun- ciation systems. Zshaashaa- is pronounced the way it appears. Betula papyrifera (white birch) was a staple product for manufacturing con- tainers, canoe sheathing, and wiigwaamen (home or lodge covered with birch bark). The names for white birch are closely related in all dialects. The Ojibway whom Densmore (1928) interviewed included the -atig suffix ‘wood’ in their di- J Smith’s (1932) word, ode’imindji’bik, for Cornus canadensis (bunchberry) is the only one significantly different from the others. There may have been some con- fusion in Smith’s data (1932) because he also has Fragaria virginiana Duchesne translated as ode’imindji’bik ‘heart berry root’. The wild strawberry had a signif- TABLE 3.—A comparison of plant names in other Ojibway and Cree dialects. Minnesota/ Minnesota / Great Lakes Woods Cree Wisconsin NW Ontario Lac Seul Oji-Cree Ojibway! Nihithawak? Ojibway Ojibway‘ Scientific binomial shingob bigiw zhingob bigiwaandag napak a(h)siht jingo’b pikewa’ndag a’ninandak’ Abies balsamea (L.) syn. aninaandag Mill. zshaashaagopa-maa- zhaashaagobiimag n/a cacagobi'mtik n/a Acer spicatum Lam. gaatig waabooz ochiibig waaboozojiibik waposocipihk wabo’s tiskwe wabos’ odji’bik Aralia nudicaulis L. wiigwaas wiigwaas waskway wigwas wi wasa tig Be . : a papyrifera rsh. zhaashaagomiinen zhaashaagominens sasakomin ode’imidji’bik caca’gomin Conte canadensis L. miskaabemig miskwaabiimizh miskwapimak, syn. — meskwabi’mic miskwabi mic Cornus stolonifera ithkwapimak ichx. bigaanaatig bagaanimizh pakan bapa’ bagan’ Corylus cornuta Marsh. azaadii azaadii mathamitos manasa’ di manasa’ di Populus balsamifera L. azaadii azaadii i asadi ‘di Populus tremuloides Michx. ' Meeker et al. 1993. Be cage 1985. Smith 1932. * Densmore 1928. P00z JeuuINg /ButIdg ADOTIOISONHLA JO TVNYNOL 28 88 KENNY and PARKER Vol. 24, No. 1 icant place in southern Ojibway sacred stories, and it seems unlikely informants would have confused it with bunchberry. It could, be a synonym, however (N. Turner, pers. comm.). The other four dialects contain the morpheme miin or min ‘berry’ and probably all translate as “chewable or edible (little) berry.” The names for Cornus stolonifera (red osier dogwood) show the closest ety- mological relationship in this group of plant names. The extraordinarily visible red color of its late-summer bark is marked in the dialects over a wide geograph- ical area. All names include a descriptor for red: miska, meskwa, miskwa, and the Lac Seul name miskaabemig ‘being red it is visible’. The shrubs are visible in the fall and winter along the lakeshore from a long distance. Corylus cornuta (beaked hazel) is the only nut-producing shrub in the boreal forest regions where the Ojibway and Cree reside. Bigaan, bagaan, and pakan all mean ‘nut’; at Lac Seul the suffix -aatig ‘shrub’ is attached to the end of the word. The word “nut” implies food (Europeans adopted the word “‘pecan’’). The names for Corylus cornuta refer directly to the nut, but also imply cultural use as food. The last two species, Populus balsamifera and P. tremuloides, are not differen- tiated by the Ojibway dialects, both being termed azaadii, asadi, and asa‘di; but, the two are different in Woods Cree. The observed consistency in names of certain plants over a very large geo- graphical area where these groups of people have resided for thousands of years may be the result of three factors. The east-west trade routes brought people together for millennia prior to European contact; these later became part of the fur trade routes. People spoke the same language, if not the same dialect, and would have been able to converse with each other about plants for food and healing. In the boreal forest ecosystem of the north, many of the same species and plant associations are found over large expanses of the region and people would have been able to recognize, talk about, and use the plants in new localities. DISCUSSION Northern Ojibway Plant Classification—This ethnobotanical study is the first re- search concerning the plant classification system of a northern Oji-Cree speaking community in northwestern Ontario. While this work is only preliminary and does not cover all of the vascular plant species found in the region, the results indicate that a complex, hierarchical folk classification system does exist in the traditional culture and language of the Lac Seul Ojibway. The plant taxonomic system of the Lac Seul people is based on the traditional perception and use of the ecosystem and mode of subsistence. The Lac Seul Ojib- way and Cree lived off their traditional lands and waters in the transition zone between the Boreal and Great Lakes-Saint Lawrence forest ecosystems by har- vesting the natural resources, and their plant names exhibit relational compari- sons of the plants to animals and landscapes of their environment (e.g., ginebi- goon ‘snake place’ (bracken fern) and miskaabemig ‘being red it is visible from a distance along shore (implied)’ (red osier dogwood). This study found no name for unique beginner in the kingdom Plantae at Lac Seul, which is typical of most folk taxonomies (Berlin 1992). The rank of covert kingdom Plantae is inferred from a vocabulary of botanical terms (Table 2). Some Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 89 of this vocabulary is genus specific, meaning that a word will describe the plant part of only one genus, such as ozhiigoopiin, which refers to the boughs of the genus zhingwak (Pinus spp.). The plant taxonomy at Lac Seul shows a number of features that have been found in the botanical taxonomies of other hunter-gathering people (Hunn 1982; Turner 1973). While the Lac Seul ethnobotanical taxa are intrinsically hierarchical, they do not fit a perfectly hierarchical classification model nor are they mutually exclusive as in scientific botanical classification. Group overlap occurs both be- tween and within ranks. Subsistence and Folk Species—Mode of subsistence is an important factor in how groups of people conceptually organize their biophysical world. Most ethnobo- tanical studies of plant classification have involved horticultural peoples in trop- ical or subtropical regions of the world (Berlin 1992). The prevailing theory has been that folk species are found primarily in agrarian cultures and only rarely, or in fewer numbers, in foraging or hunter-gathering groups. Foraging people usually average two percent polytypic genera (Berlin 1992:275). The Lac Seul Ojib- way are also considered hunter-gatherers and the folk taxonomic data gathered so far, while preliminary, shows nine percent (3 of 34) polytypic folk genera, a value between those of horticulturalists and foragers. Lac Seul Ojibway have a history of manipulating and propagating plants for special uses (e.g., wild rice, blueberries, sweet flag, and certain other medicinal species) and these numbers may reflect that activity. The six identified folk species are all very important in traditional usage at Lac Seul. Utilitarian-based Taxa in Folk Taxonomies.—Berlin (1992) does not recognize rank based on cultural utility, although he does recognize an intermediate rank based on cultural mythology of utilitarian cultivars in Tzeltal plant taxonomy. Other authors (Hunn 1982; Turner 1973, 1987, 1989; Turner et al. 1990) recognize cul- tural utility as a rank. Lac Seul botanical taxonomy resembles the systems of other foraging peoples in that taxa apparently have been constituted based on cultural utility. Turner (1973), in her study of the Haida, Bella Coola (Nuxalk) and Lillooet (Stl’atl’imx), Pacific Northwest people, found that the folk taxonomies of these three First Nation groups included ‘root’ and ‘berry’ life-form categories although the root category was unnamed. Root plants were either grouped together in discussions about the plant species (i.e., the genus names identify the plants by their roots) or there is a suffix which denotes the root part. In Lac Seul Oji-Cree, the base word ochiibig is applied to certain herbaceous plants whose roots are medicinal. In Haida, Bella Coola, and Lillooet, the ‘berry’ group overlaps with the ‘shrub’, ‘tree’, and/or ‘herbaceous plant’ life forms (Turner 1973). In the Lac Seul system, the ‘berry’ grouping of plants overlaps with the ‘shrub’ and ‘herbaceous’ life forms (Figures 4 and 5). Turner et al. (1990) found that the Thompson (Nlaka pamux), a Salishan-speaking group in British Columbia, included an economic life form of ‘berry/fruit-bearing plants/bushes’. A recent study by Johnson- Gottesfeld and Hargus (1998) with the Witsuwit’en, Athapaskan-speakers in northern British Columbia, also found that life-form taxa may be partly utilitarian and not mutually exclusive. 90 KENNY and PARKER Vol. 24, No. 1 A final life form with considerable cultural utility is ‘moss’. While in man regions recognition of a moss life-form group is unusual (Atran 1985:301), at Lac Seul the mosses are conspicuous as a vegetational entity carpeting the floor of many forest stand types. The life form gaamig contains two folk genera, waa- bangaamig and ozhaagaamig, the former containing two folk species, and both with high utility (Figure 6). While Turner (1973) reported the ‘moss’ life form was named but empty in the taxonomic system of the First Nation interviewed, Johnson-Gottesfeld and Hargus (1998) found that Sphagnum spp. and feather moss are considered yin ‘moss’. The prototype of the life form in this case appears to be Sphagnum magellanicum Brid. The Montaignaise people's concept and namin of mosses most closely parallels the situation at Lac Seul. Clément (1990) observed that the suffix -kamuk, referring to ‘‘plants without roots,” terminates the names for various Sphagnum spp., caribou lichens, and some feather mosses. The name -kamuk also means “surface” or ‘crust of the earth,” which is comparable to Lac Seul’s references to waabangaamig ‘white ground cover’. SUMMARY AND CONCLUSIONS There has been no previous research into the folk taxonomic systems of the Oji-Cree and Cree speaking people of northwestern Ontario. The goal of this preliminary study was to record plant names and botanical vocabulary of the Oji- Cree speaking Lac Seul First Nation and determine the method of plant classifi- cation. The names of 38 folk genera and folk species were elicited with some synonyms and the words for 18 morphological characteristics, together totaling 65 botanical names or phrases. Although the data set was small, it was apparent that the folk genera correspond to morphologically based life-form categories rep- resenting a hierarchical taxonomic system for ordering the plant world. NOTES ‘Voucher specimens are in the collection of Mary B. Kenny, Lac Seul First Nation, via Box 317, Hudson, Ontario, Canada. * The pronunciation and spelling of Lac Seul Oji-Cree words and phrases follow the con- ventions of Ningewance (1993): Consonants Pronounced as in English: m, n, y, h, and w. Ojibway Consonants: k, p, ch, t, and sh are similar to English. 8, b, j, d, and zh are softer than English. Aspirated (sometimes) in pronunciation: k (hk), p (hp), and t (ht). Consonants at the end of a word are voiced: -iw, -ng, -nd, -nzh, -tw, -dw, -shkw, owels Short vowels: (i) as in bit 0) as in look (a) as in cup Spring/Summer 2004 Long vowels: (e) as in red (ii) as in peek (aa) as in ‘ah’ JOURNAL OF ETHNOBIOLOGY 91 (00) as between boot and boat Nasal vowel endings: -ens, -aans, -oons, -enz, -iinz, -aanz, and -oonz ACKNOWLEDGMENTS We gratefully acknowledge the contributions of the Lac Seul First Nation Elders who generously provided their botanical knowledge for this study: Shamandy and Bessie Kejick, S thwind, Mary Jane Brisket, Archie Bottle, Alice Kejick, Leo Binguis and Roger Southwind. We especially thank George Kenny for acting as translator and assisting with all aspects of the field work. Some financial support was provided by an NSERC grant to the second author. REFERENCES CITED Atran, Scott. 1985. The nature of folk-bo- tanical life-forms. American Anthropolo- gist 87:298-315. Baldwin, K.A. and R.A. Sims. 1989. Com- mon Forest Plants of Northwestern Ontar- io. ym dering of Natural Resourc- es, Tor Berlin, aa 1992. Ethnobiological Classifi- cation. Princeton University Press, Princeton, New Jersey. Dawson, K.C.A. 1983. Prehistory of Northern Ontario. Thunder Bay Historical Muse- um, Thunder Bay, Ontario. Densmore, Frances. 1928. Uses va ages by the Chippewa Indians. Annual Report 44:275-397. Bite of Atneien Eth- A. and Arthur Cronquist. 1963. Manual of Vascular Plants of the Northeastern United States and Adjacent Canada Nostrand Company, New York. Hunn, Eugene S. 1982. The utilitarian fac- tor in folk besasceiey pap ication. American Anthropologist “e Johnson-Gottesfeld, Leslie M. and Sharon Hargus. 1998. Classification and no- menclature in Witsuwit’en ethnobota- ny: a preliminary examination. Journal of Ethnobiology 18:69-101. Kenny, B. 2000. Ojibway Plant Tax- onomy at Lac Seul First Nation, Ontar- io, Canada. M.S. Thesis (Forestry). pao University, Thunder Bay, On- tari Leighton Anna L. 1985. Wild plant use by e Woods Cree (Nih-thawak) of east- central Saskatchewan. Canadian Ethnol- ogy Service Paper No. 101. Mercury Se- u f Lakes Indian Fish and ee Com- mission, Odanah, Wiscons Ningewance, Patricia 1993. Survival O ry eon Mazinaate Press, Winnipeg, toba. Rowe “ohn S. 1972. Forest Regions of Canada dian Forest Service, Sp; order of i ari e Museum of the City of Milwaukee 4:327- 525 Turner, Nancy J. 1973. Plant Taxonomic Systems and Ethnobotany of Three Contemporary Indian Groups of the Pa- cific Northwest. Ph.D. Dissertation (Bot- any), University of British Columbia, Vancouver. 1987. General plant sp in Thompson and Lillooet, two interior Sa- lish languages of British Columbia Journal : Ethnobiology 7:55-82 “All berries have relations,” sof ati folk plant poring in ompson and Lillooet “erg Salish. edge and Use 7 Plants by the aa tear of British Columbia. Memoir No. 3, Royal British Columbia Museum, Victoria. DOLE Re mite Din Journal of Ethnobiology 24(1): 93-111 Spring/Summer 2004 POPULATION TRENDS AND HABITAT CHARACTERISTICS OF SWEETGRASS, Anthoxanthum nitens: INTEGRATION OF TRADITIONAL AND SCIENTIFIC ECOLOGICAL KNOWLEDGE DANIELA J. SHEBITZ* and ROBIN W. KIMMERER? * College of Forest Resources, University of Washington, Seattle, WA 98195 ® Department of Environmental and Forest Biology, College of Environmental Science and Forestry, State University of New York, Syracuse, NY 13210 ABSTRACT.—Sweetgrass (Anthoxanthum nitens (Weber) Y. Schouten & Veldkamp) is a valued plant among Native peoples of the northeastern United States and eastern Canada, but Haudenosaunee herbalists and basketweavers have reported declines in its population at traditional gathering sites. We integrate traditional ecological knowledge with field and experimental studies to identify and under- stand population trends of sweetgrass. The plant's habitat requirements were also investigated; it was found growing under various environmental conditions. We determined that sweetgrass is declining in sites where it was historically present throughout the northeastern United States. In traditional gathering sites, the lack of controlled burning and unsustainable harvesting may be a factor in its decline, but the greatest threats facing sweetgrass throughout the Northeast are economic development and ecological succession. Key words: sweetgrass, Anthoxanthum nitens, Hierochloe odorata, Haudenosaunee, Iroquois, baskets. RESUMEN.—Fste estudio integra conocimiento ecolégico tradicional con estudios de campo y experimentales para determinar las tendencias poblacionales del sweetgrass (Anthoxanthum nitens (Weber) Y. Schouten & Veldkamp) y las causas de estas tendencias. Los herbolarios y tejedores de cestas Haudenosaunee han sefia- lado declives en las poblaciones de sweetgrass de los sitios de acopio tradicionales. Se estudiaron los requisitos de habitat del sweetgrass; se encontré en condiciones ambientales variadas, por lo que se considera una especie generalista. Este estudio permitié determinar que ademas de declinar en los lugares de acopio, las pob- laciones de sweetgrass también estan disminuyendo en localidades donde estaba hist6ricamente presente en todo el noreste de los Estados Unidos. En los lugares de recogida tradicionales, la ausencia de quemas controladas y la recolecci6n in- sostenible son causas posibles del declive de las poblaciones de sweetgrass, pero las mayores amenazas a las que se enfrenta esta hierba en todo el noreste son el desarrollo econdémico y sucesién ecologia. RESUME.—Le foin d’odeur, Anthoxanthum nitens (Weber) Y. Schouten & Veld- kamp, est une plante fort estimée des Premiéres Nations du nord-est des Etats- Unis et de l’est du Canada. Les vanniers et herboristes de la Ligue des Six-Nations (Haudenosaunee) ont mentionné que les populations de cette plante étaient en déclin aux sites traditionnels de récolte. Nous intégrons le savoir écologique trad- itionnel aux recherches expérimentales et aux travaux de terrain afin de déterminer 94 SHEBITZ and KIMMERER Vol. 24, No. 1 et de comprendre les tendances démographiques des populations du foin d’odeur. Les paramétres définissant les habitats oti se trouve la plante ont été examinés: celle-ci croit sous diverses conditions environnementales. Nous avons pu établir que le foin d’odeur est en déclin dans les sites ow il était historiquement présent a travers le nord-est des Etats-Unis. Quant aux facteurs contribuant au déclin du foin d’odeur parmi les sites traditionnels de récolte, l’absence de briilages dirigés et la cueillette non durable ressortent comme deux facteurs possibles. Toutefois, le développement économique et la succession écologique forment les plus gran- des menaces auxquelles font face les populations de foin d’odeur dans le nord- est des Etats-Unis. INTRODUCTION Anthoxanthum nitens (Weber) Y. Schouten & Veldkamp (=Hierochloe odorata (L.) P. Beauv; CNWG), commonly known as sweetgrass, is a perennial grass native to North America that plays a significant role in the lives of the indigenous people who reside within its range. Although sweetgrass is most frequently used as a ceremonial smudge and incense (English 1982; Kavasch and Barr 1999), its pre- dominant use among the Haudenosaunee (also known as Iroquois), is in basketry (Benedict 1983). Haudenosaunee herbalists and basketweavers interviewed for this study were concerned that sweetgrass populations have diminished and that the plant is now difficult to find in many traditional gathering areas. This observation has also been reported in a publication about the basketmakers of Akwesasne: ‘While sweetgrass grows naturally at Akwesasne and in surrounding areas, it is becom- ing more difficult to locate. . .” (Lauersons 1996:31). This study was conducted in partnership with Haudenosaunee basketweavers, herbalists, and ceremonial lead- ers who are familiar with the ecology and use of sweetgrass. We explore the nature of Haudenosaunee traditional knowledge of sweetgrass, its population trends, and its local and regional distribution. We integrate results from the eth- nographic study with an ecological analysis. Objectives and Hypotheses.—The objectives of this project are twofold. The first objective is to determine if the population of sweetgrass is declining in the north- eastern United States and, if so, to explore possible causes of this decline. The second objective is to understand habitat requirements of sweetgrass. Both objec- tives will be addressed through the integration of Haudenosaunee traditional and scientific ecological knowledge. While the general distribution of sweetgrass is known (Greene 2000; Lynch and Lupfer 1995), its specific habitat requirements are largely unstudied. Pub- lished information concerning the natural habitat of sweetgrass in the northeast- ern United States is limited. Information indicating sweetgrass’s present or his- torical geographical range, the abundance of sweetgrass in those areas, its pop- ulation trends, and indigenous management practices associated with the plant is lacking. Traditional knowledge has the potential to enhance the botanical in- formation that does exist. e hypotheses we tested in this study include: sweetgrass populations are declining throughout the Northeast; development of the landscape poses a sig- Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 95 nificant threat to sweetgrass populations; competition from nonnative plants is a significant threat to sweetgrass populations; sweetgrass abundance is correlated with identifiable environmental variables that characterize its habitat; the Hau- denosaunee maintain traditional knowledge of sweetgrass population trends; and the Haudenosaunee maintain traditional knowledge of its local and regional dis- tribution. Traditional Ecological Knowledge.—Traditional ecological knowledge (TEK) offers a source of biological insight and potential models for conservation biology. It gen- erally encompasses plant geography, plant ecology, and phenology, and often in- cludes information concerning the range and distribution of a species (Kidwell 1973). This knowledge, developed through generations of interactions between native peoples and their lands, can contribute rational and reliable perspectives to the contemporary sciences (Kimmerer 2002; Mauro and Hardison 2000). The knowledge held by Haudenosaunee practitioners concerning the popu- lation trends and habitat requirements of sweetgrass plays a vital role in this project. This knowledge, used in conjunction with a scientific ecological study, contributes to the determination of whether the population of sweetgrass is de- clining, and assists in understanding its habitat requirements. METHODS Ethnographic Methods—The Haudenosaunee consists of six sovereign indigenous nations, whose populations continue to inhabit New York State: the Seneca, Ca- yuga, Onondaga, Oneida, Mohawk and Tuscarora (Grassman 1969; Herrick 1995; Lauersons 1996). Although members of all Haudenosaunee Nations produced bas- kets, it is mainly the Mohawks of the Akwesasne Territory who continue the tradition today (Lauersons 1996). The Akwesasne Territory, or ‘Land Where the Partridge Drums,” is located in the St. Lawrence River Valley near Massena, New York. It is divided by the United States-Canadian border and by the border be- tween the Canadian provinces of Ontario and Quebec (Benedict 1983; Lauersons 1996). It is home to approximately 10,000 Mohawk people, and to the art form of elaborate sweetgrass and black ash basketry (Benedict 1983). The consultants who contributed to this paper are primarily women who are familiar with and use sweetgrass and who are members of the Onondaga and Mohawk Nations, located in central and northern New York, respectively. Some work for this project has been conducted with basketweavers of the Seneca Nation as well, who reside in western New York. Eight formal interviews were conducted with Haudenosaunee consultants who are familiar with the ecology of sweetgrass, most are basketmakers, although herbalists and ceremonial leaders were also interviewed. The interviews took place between February and July, 2001: two elder female herbalists from the On- ondaga Nation, in their late 60s, four female basketmakers from the Mohawk Nation, with ages ranging from 48-70, a male farmer and ceremonial leader in his 70s, and one female basketmaker from the Seneca Nation, aged 46. All of the consultants are fluent in English. During the interviews, participants were guided in discussion through a list 96 SHEBITZ and KIMMERER Vol. 24, No. 1 of topics, but the direction of the interviews followed the participants’ train of thought (Huntington 2000). The interview topics focused on whether the partic- ipants have noticed a change in the distribution of sweetgrass throughout the region, and if so, which factors they thought were responsible for the change. Past and current harvesting practices and land management through controlled burn- ing were also discussed. In addition, we asked them to share information about areas where they currently gather sweetgrass, and/or traditional gathering sites where sweetgrass is no longer found. The formal interviews were tape recorded with written permission from the consultant or handwritten notes were taken if the consultant was uncomfortable with being recorded. Each consultant signed a letter of consent and was compen- sated for his or her time and cooperation. In addition to the eight formal interviews, nine informal conversations were conducted with Haudenosaunee basketmakers. One of the participants of the in- formal interviews was a male basketmaker from the Akwesasne Mohawk Terri- tory who was approximately 70 years old. Eight of the participants were women, five of whom were elders ranging in age from 60-80 years, one female from the Seneca Cattaraugus Reservation in her 50s, and two beginning women basket- makers, aged 20, from the Seneca Allegheny Reservation. These conversations were generally short discussions in which a few questions were asked concerning the basketmakers’ relationship to and use of sweetgrass, as well as her/his meth- od of harvesting and knowledge of past land management practices involving controlled burning. All of the informal conversations were conducted on the Ak- wesasne Mohawk Territory in July, 200 Participant observation was used in visits to the sweetgrass gathering areas in the vicinity of the Akwesasne Territory. The observations included gathering sweetgrass with three generations of women in the Burns family, and their female friends at their grass collection sites in July of 2000 and July of 2001. This process assisted in our identification of sweetgrass and gave us the opportunity to gain an understanding of the sweetgrass habitat characteristics. Ecological Field Methods.—Herbarium records enabled us to ascertain the historic distribution of sweetgrass in the Northeast and to obtain information on its habitat preferences. We consulted collections in four major herbaria in the Northeast: the New York State Museum in Albany, Cornell University, the New York Botanical Garden, and Harvard University. In addition, we visited the H. Lee Ferguson Museum Herbarium to obtain information on sweetgrass sites on Fisher's Island, New York. At each herbarium, sweetgrass specimens collected in the northeastern United States were studied and information regarding date of collection, the col- lection site, associated plants, and environmental conditions of the area were re- corded. This process resulted in over 250 records of sweetgrass throughout the North- east. Of these, 27 sites were described in sufficient detail to find. The sites were located in: New York (14), Massachusetts (4), Connecticut (2), Vermont (3) and New Hampshire (4). These 27 ‘‘sites of record’ were each visited to determine sweetgrass presence and to characterize its habitat. Vegetation at each of the sites of record that was intact (i.e., not lost to de- Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 7 velopment) was studied in order to determine whether sweetgrass was still pres- ent at the site and to assess the relative abundance of sweetgrass and associated species. Vegetation presence and cover were quantified by placement of 30 quad- rats placed in a stratified random design along three 50-m transects. Each quadrat was a circular plot with a diameter of 0.8 m (approximately 0.5 m2). The cover of each plant species within each plot was estimated to the nearest 5%. Species were identified following Gleason and Cronquist (1991). All of the 27 sites of record were visited from mid-July to early September, 2000. By sampling in a relatively limited time frame, there was minimal variation in developmental stages of the vegetation. In addition to vegetation sampling, canopy cover readings were taken and soil samples were analyzed in order to determine if there were significant rela- tionships between sweetgrass abundance and these environmental variables. A Model-A spherical densiometer was used at elbow height to determine the percent canopy cover at three random points at each site. Three soil samples of 7-cm depth were also taken in random points at each site. Each of the soil samples was an- alyzed for texture and pH in a laboratory at the State University of New York College of Environmental Science and Forestry according to standard methods described by Wilde et al. (1972). Each site was photographed and a map of the site was drawn showing prox- imity to water, and the arrangement of the sample plots for future monitoring studies. At the sites where the landscape had been altered through development or succession since sweetgrass was recorded, the vegetation was not formally assessed, but photographs were taken to document the change. In addition to the 27 sites of record, five Haudenosaunee current and past sweetgrass gathering sites were studied. We identified these sites through partic- ipant observation and interviews. The ecological sampling methods used at the sites of record described above were also employed at these gathering sites. Data Analysis.—The data from the 27 sites of record and the five sweetgrass gath- ering sites were included in the analysis. In order to determine which plants occurred most frequently with sweetgrass, the average percent cover for every plant species at each site of record was calculated. The total average plant cover was also calculated. Since we were interested only in those plants that might have statistically strong relationships with sweetgrass abundance, those species that had at least 1% cover over all sites and occurred with sweetgrass in at least three sites were included in the data analysis. A Satterthwaite two-sample t-test was performed for each species using SAS (version 7.0) Statistical Program (SAS In- stitute, Inc. 1990) with the purpose of determining if a relationship exists between these species and sweetgrass abundance. In order to determine if the presence of nonnative species was related to sweetgrass abundance, Satterthwaite two-sample t-tests were performed with the nonnative species collectively, nonnative grasses, and nonnative dicots. Statistical analyses were then performed on the average percent canopy cover, the percent of sand, silt, and clay in the soil, and the soil pH in order to determine if they were related to sweetgrass abundance. Relationships between sweetgrass abundance and the environmental variables were tested using Pearson’s correla- 98 SHEBITZ and KIMMERER Vol. 24, No. 1 oe agri Native American Territories FIGURE 1.—The 32 sites that were visited throughout states, five of w are/were Haudenosaunee sweetgrass gathering s sites and 27 of which are sweetgrass sites of record. tion coefficients and coefficients of determinations (R?) through correlation and regression. These data were analyzed using the SAS (version 7.0) Statistical Pro- gram (SAS Institute, Inc. 1990) with sweetgrass percent cover as the dependent variable. RESULTS Population Status of Sweetgrass Throughout the Northeastern United States.—Sweet- grass was found at 13 of the 27 sites of record, and 4 of the 5 Haudenosaunee gathering sites. The oldest herbarium record where sweetgrass was still present was taken in 1904 from a tidal marsh in Salem, Massachusetts. In 15 of the 32 visited sites, sweetgrass was not found. The dates of the collection of sweetgrass from these 15 sites ranged from 1913 (Percy, New Hampshire) to 1982 (Wheelock, Vermont). These sites and the probable causes for sweetgrass’s absence are illus- trated in Figures 1 and 2 respectively. Population Status of Sweetgrass in Traditional Gathering Sites.—Four of the traditional sweetgrass gathering sites sampled are located within 30 km of the Akwesasne Mohawk Territory, and one is in the vicinity of the Onondaga Nation Territory. In the summer of 2000, sweetgrass was in the four gathering sites located near the Akwesasne Territory: Norfolk, Saint Regis Falls, Dickinson Center, and Ho- gansburg, New York. Two of these sites, Norfolk and Saint Regis Falls, both are considered to be popular sweetgrass harvesting areas and have a high percentage Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 99 250 Sweetgrass herbarium specimens from throughout the Northeast 27 specimen pease in oo detail to “sites of re Stes ies and rnkd 20 with intact vegetation 7 not with intact vegetation (75% of sites of record) (i.e. developed) f N (25% of sites of record) 13 sites with 7 sites with sweetgrass present sweetgrass absent (65% of intact sites) (35% of intact sites) 4 have undergone 3 undetermined cause of succession sweetgrass absence FIGURE 2.—A flow diagram illustrating the number of herbarium specimens studied, the sites of record, the intact and altered sites, the intact sites with sweetgrass present and sweetgrass absent and whether its absence is a result of or undetermined causes. cover of sweetgrass. Dickinson Center, located near the Akwesasne Territory, also has a large amount of sweetgrass. A consultant who harvests sweetgrass from Dickinson Center believes that few people know of and harvest from this site. The remaining site near Akwesasne, in Hogansburg, was once a gathering site, but is no longer visited due to the relatively recent decline of sweetgrass. Sweet- grass was found only in small patches on the Hogansburg site in the summer of 2000. The other gathering site, LaFayette, is near the Onondaga Nation Territory. In this site, a consultant stated that sweetgrass used to line a creek that runs through the area. She stated that she witnessed the population of sweetgrass slowly de- clining over the years until approximately 1998 when none remained. Sweetgrass was not found at this site in the summer of 2000. Habitat Characteristics of Sweetgrass.—The habitats in which sweetgrass was found ranged from wetlands including salt marshes, fens, swamps and marshes to dry roadsides. The variations in values of environmental variables examined for this TABLE 1.—The location, habitat type, percent sweetgrass cover, percent sand and clay, soil pH, and percent canopy cover of the 13 sites of record and four Haudenosaunee sweetgrass gathering sites where sweetgrass was found in State County Region Habitat % Sweetgrass % Sand % Clay pH % Canopy Connecticut New London Stonington salt marsh x 727 8.7 57 0 Connecticut New London Stonington roadside 0.7 83.5 6.7 5.0 40.8 Massachusetts Berkshire Egremont en 6.8 61.7 8.8 7.6 0 Massachusetts sex Salem salt marsh 14.3 42.6 th. oe 0 New Hampshire heshire Stoddard roadside 18.8 90.0 3.7 55 21.6 New Hampshire Strafford Dover marsh 2.5 71.4 8.5 5.0 49.5 New York ompkins Groton meadow 1.2 52.8 15.7 73 0 New York Rensselaer Taborton roadside 4.7 87.1 4.9 via 42.3 New York sex N. Elba riverbank 17 94.2 1.9 6.1 9.9 New York Jefferson Fargo roadsid 4.3 94.0 2.4 73 18.4 New York St. Lawrence Potsd roadsid 22 87.8 4.4 72 9 New York Suffolk Fisher’s Island brackish marsh 1.0 66.2 4.8 6.5 0 New York Franklin St. Regis Falls roadside 6.0 95.5 2.0 7.2 13.1 New York Franklin Hogansburg meadow 0.3 56.5 30.4 71 0 New York Franklin Norfolk meadow 15.6 70.9 co ia: 6.9 0 New York in Dickinson Ctr. meadow 6.5 85.8 a7 5.7 0 Vermont Caledonia Danville swamp io 67.2 4.3 6.0 56.5 OOT UaaHNWE pue ZLIGHHsS L ‘ON ‘PZ “TOA Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 101 TABLE 2.—Results of correlation and regression analysis conducted on the environmental variables (a = 0.05). The data included in the analysis were collected from the 13 sites of record and four Haudenosaunee sweetgrass gathering sites where sweetgrass was found in 2000. Correlation analysis Regression analysis Environmental Correlation variable coefficient P-value R? P-value % Sand —0.584 0.824 0.177 0.358 % Silt 0.160 0.540 0.177 0.364 % Clay —0.159 0.542 0.177 0 p —0.163 0.532 0.177 0.421 Canopy cover —0.069 0.793 0.177 0.744 study reflect this wide variety of habitats. The pH of the sites ranged from 5.01 to 7.63. There was a wide variation in percent canopy cover over the sweetgrass habitats as well, ranging from 0-56.5%. Sweetgrass was found primarily on sandy soils, however there was a relatively wide variation in the soil texture, from 42.6% sand to 94.2% sand. These environmental data for the 17 sites are presented in Table 1. No significant relationships (a = 0.05) were detected between sweetgrass abundance and the environmental variables of soil pH, soil texture and canopy cover (Table 2). Sweetgrass was found growing among other grasses and shrubs in all of the sites where it was present. It was the dominant species in four of the sites (Ta- borton, Norfolk, Salem, and Stoddard) but was commonly intermixed with other species and was never found growing in pure stands. A total of 141 plant species, and 110 identified genera were found to occur with sweetgrass in the sample quadrats. Sixteen species, including sweetgrass, had a cover of at least 1% of the total area surveyed in the 17 sweetgrass sites. These species and their percent cover over the total area sampled are listed in Table 3. The Satterthwaite t-tests revealed no significant positive relationships (a = 0.05) between sweetgrass and co-occurring species. Significant negative relation- ships were found between abundance of sweetgrass and both wild carrot (Daucus carota) and red clover (Trifolium pratense) (Table 4). Wild carrot and red clover are the only two nonnative dicots listed on Table 4. A significant negative relationship was found between sweetgrass abundance and the presence of the nonnative dicots (a = 0.05). Sweetgrass abundance was not significantly related to either the abundance of the nonnative grasses (Phalaris arundinacea, Bromus inermis, Agropyron repens, and Phleum pratense or all nonnative plants (dicots and grasses) included in the data analysis (Table 5). DISCUSSION Habitat Requirements and Population Status of Sweetgrass Throughout the Northeastern United States.—The presence of sweetgrass in a wide variety of habitats and the absence of significant relationships between sweetgrass and the environmental variables studied suggest that sweetgrass is a generalist and can thrive in diverse 102 SHEBITZ and KIMMERER Vol. 24, No. 1 TABLE 3.—The species present with sweetgrass that had a percent cover of é at = fe = the total sampled area: the 13 sites of record and four I g sites where np tanh was found in 2000. Percent cover Latin binomial Common name (total) Phalaris arundinacea = - reed canarygrass 7.35 Spartina patens Aito: salt-meadow cordgrass 5.83 Anthoxanthum nitens * (Weber) Y. Schou- sweetgrass 5.72 ten & Ve Solidago candensis L.* Canada goldenrod 4.06 Agropyron repens L. Nevski* quackgrass 2.70 Trifolium pratense L.* red clover 2.86 Phleum pratense L.* timothy grass 2.42 Vicia cracca L.* bird vetch 2.01 Poa pratensis L. Kentucky bluegrass eg | Bromus inermis Leyesser* smooth brome grass 1.62 Panicum clandestinum L. deertongue Lae Daucus carota L.* wild carrot 1.44 Panicum virgatum L. switchgrass 1.60 Onoclea sensibilis L.* sensitive fern 1.44 Asclepias syriaca L.* common milkweed 1.10 sete sempervirens L. seaside goldenrod 1.01 s with an asterisk oe were found in three or more of the sampled sites and were included in the data analysis for this study. habitats. In addition to habitat types, the amount of disturbance in areas with sweetgrass also varied. Sweetgrass was found in undeveloped marshes as well as in disturbed areas such as roadsides. There are limitations, however, to the level of disturbance in which sweetgrass can survive. One such limitation is the alter- ation of the landscape through development. We hypothesized that sweetgrass populations were declining throughout the Northeast, in part due to habitat loss through development. Our data indicate that only 75% of the sites of record studied were intact; 25% of the sites had been TABLE 4.—The results of the Satterthwaite t-tests to determine stored pierains be- tween sweetgrass and species present in three or more sites with sweetgrass and constitute 1% or more of the total sampled area (a = 0.05). Absent Present Mean Mean sas Sites (Std. Err) Sites (Std. Err) P-value Agr on repens 12 6.5 (1.8) 5 4.0 (1.1) 0.2678 Asclepias syriaca 14 6.0 (1.6) 3 4.3 (2.0) 0.5205 Bromus inermis 13 6.5 (1.7) 4 3.4 (1.2) 0.1751 Daucus ta 13 6.8 (1.6) 4 2.2 (0.8) 0.0218 Onoclea sensibilis 12 6.0 (1.4) 5 6.0 (3.3) 0.9078 Phalaris arundinacea 14 5.6 (1.4) 3 6.1 (4.8) 0.9259 Phleum pratense 8 6.5 (2.0) 9 5.0 (1.9) 0.6078 Solidago canadensis 13 6.1 (1.7) 4 4.5 (1.2) 0.4627 Trifolium pratense 11 7.9 (1.7) 6 1.7 (0.7) 0.0051 Vicia cracca 13 7A; 5) 4 5.9 (3.5) 0.9649 Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 103 TABLE 5.—Results of Satterthwaite t-tests conducted between sweetgrass abundance and the nonnative dicots, nonnative grasses and nonnative plants that were found in at least three sites with sweetgrass and at least 1% of the total area sampled (a = 0.05). Absent Present Mean Mean Plants Sites (Std. Err) Sites (Std. Err) P-value Nonnative dicots 11 8.0 (1.7) 5 1.6 (0.6) 0.0045 Nonnative grasses 7 7.6 (2.5) 10 4.4 (1.4) 0.2882 Nonnative plants 5 6.5 (2.3) jes 5.4 (1.7) 0.7121 altered due to development. Habitats were lost due to urbanization, the establish- ment and maintenance of recreation areas (beaches, parks), and in one case, the reforestation of agricultural land. The findings from this study, therefore, support the hypothesis that development contributes to the loss of sweetgrass populations. In addition to habitat alteration through development, we predicted that in- vasion by nonnative plants was associated with decline of sweetgrass populations. Our data indicated no significant relationship between sweetgrass abundance and presence of nonnative species. Despite the presence of exotic species at the ma- jority of the sites, sweetgrass was still found at 65% of the intact sites of record. Of the remaining intact sites, most had undergone natural succession to native forest or shrubland and are therefore no longer suitable for sweetgrass. The per- sistence of sweetgrass at most of the sites independent of the presence of non- native plants suggests that sweetgrass is not significantly threatened by nonnative plants. The herbarium records indicated only past presence of sweetgrass, how- ever, not its abundance. It is possible that the nonnative species do influence the sweetgrass abundance to some extent in the areas sampled. It may not be possible, however, to determine the degree to which the surrounding vegetation is affecting the sweetgrass since the herbarium and oral records did not provide records of sweetgrass abundance in the area at the time of its collection. Two nonnative species, wild carrot (a biennial) and red clover (a perennial), were found to have a significant negative relationship with sweetgrass. These species are not considered to be invasive (Gleason and Cronquist 1991). Negative relationships found between sweetgrass abundance and the presence of wild car- rot and red clover may suggest that there is some competition between them. There is an alternative explanation to the negative relationship, however, which is more likely the cause for the negative relationships. The co-occurrence of wild carrot and red clover with sweetgrass is due to the similar habitat pref- erences of the species. Both of the dicots, like sweetgrass, inhabit disturbed areas, such as roadsides, waste places, and fields (Gleason and Cronquist 1991; New- comb 1977; Reed 1971). The negative association detected likely results from en- vironmental preferences within these habitats. Six out of the seven sites in which sweetgrass was found with at least one of these dicots (Groton, Taborton, Ho- gansburg, Fargo, Potsdam, and New London) were within five meters of a road. The majority of the quadrats in which wild carrot and/or red clover were found (69%) was in the transect closest to the road; in contrast, most of the quadrats in 104 SHEBITZ and KIMMERER Vol. 24, No. 1 which sweetgrass was found in these sites were in transects that were not closest to the road (57%). Therefore, the negative association between the presence of wild carrot and red clover and sweetgrass abundance may be due to environ- mental preferences involving the level of disturbance in the transects, rather than to competition. Although negative relationships were found only between sweetgrass and wild carrot and red clover, other nonnative species were found with sweetgrass, some of which are considered to be invasive. These invasive plants include smooth brome grass (Bromus inermis) and quackgrass (Agropyron repens), both perennials that were introduced from Europe and now are commonly found in waste places and roadsides (Gleason and Cronquist 1991; Hitchcock 1935). Sweetgrass’s exten- sive root system and ability to vigorously reproduce vegetatively (Greene 2000) may be responsible for its persistence against these invaders. The most abundant plant which co-occurred with sweetgrass (7.35% cover) was reed canarygrass, (Phalaris arundinacea) a perennial that inhabits marshes, riverbanks, and moist areas (Hitchcock 1935). This species includes native plants as well as commercial genotypes that have European origins. There are no phe- notypic differences between the native and European plants. A difference does exist between the two, however. The European genotype of reed canarygrass has a tendency to grow in monoculture and is often considered to be invasive in many natural wetlands in the United States. It grows vigorously and is able to inhibit and eliminate native species (White et al. 1993). The fact that a negative relation- ship was not found between reed canarygrass and sweetgrass may indicate that the plants found growing with sweetgrass are of the native genotype. Sweetgrass was found in only 48% of the sites of record. This finding indicates that the northeastern sweetgrass population is indeed declining in sites where it was historically present. The population trends that were examined in this study are limited by information that was gathered in the past on sweetgrass habitat. Trends in sweetgrass populations throughout the Northeast were determined by its presence or absence in areas that were previously recorded as sweetgrass hab- itat. It is possible that although sweetgrass was absent from some sites where it was historically present, the species is colonizing other areas. The lack of infor- mation about areas where sweetgrass was absent in the past, however, makes this determination impossible. Ethnographic Findings.—Participatory research is a method of study that provides cross-cultural opportunities for cooperation and communication (Colorado 1988). Participatory observation in this study was important to establish a rapport with the sweetgrass gatherers. Sweetgrass is primarily used by women in basketry, and men do not often gather the grass (Lauersons 1996). Many Haudenosaunee women have shared the harvesting of sweetgrass with family members and friends for countless generations. As Christine Horn, a sweetgrass gatherer in her sixties re- calls: “‘We'd go out, the females in my family. We'd pick berries in June, and sweetgrass in July. It was a way of life at the time, this is what you did.’”! We were taught to recognize sweetgrass by its distinct shiny, light green blades and purple base. To confirm its identification, Theresa Burns told us to crush some of the blade to release the sweet fragrance. One of the women with Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 105 whom harvesting was conducted was colorblind and had no sense of smell, and still recognized sweetgrass by its shine. Theresa Burns instructed us to harvest the sweetgrass by pinching the base of the stem, just above the ground, so as not to disturb the root. Each blade of sweetgrass is picked individually, while taking care not to bend or damage the blade. The act of “cleaning” involves the removal of any brown, dried, or broken blades and keeping the long, bright green grass (Lauersons 1996 and Shebitz, personal observation). Meeting with and interviewing individuals who each have their own use and understanding of sweetgrass contributed various perspectives on the importance of the plant to Haudenosaunee culture and to the environment. When asked to describe the link between sweetgrass and the Mohawk culture, Christine Horn stated: ‘It can’t be separated, it’s just being Indian.’’? All of the individuals who took part in the formal interviews and most of the informal interviewees (five out of six) were concerned that sweetgrass popu- lations are declining in the vicinity of their reservations. When asked about the status of sweetgrass in a formal interview, a Mohawk basketweaver who preferred to remain anonymous stated: “I can’t find it anymore, it’s difficult to find. It’s become evasive. When going to pick it on the Rez, my old favorite spots don’t have sweetgrass anymore.’’? Both the formal and informal interviews revealed the Seneca basketmakers’ beliefs that although sweetgrass was abundant in western New York in past cen- turies, it is now rare, if present at all, in the area. Michele Dean Stock is one of the only Seneca basketmakers remaining. She believes that the absence of sweet- grass and black ash in the area is partly responsible for the fact that traditional Seneca baskets are currently seldom made: To my understanding, there was a time when you can gather sweetgrass on the reservation in certain spots ... there was a time when it was at Allegheny Reservation but it’s been at least 100 years that people haven't been able to find it there.* When asked why they believed that sweetgrass populations were declining, five of the eight participants in the formal interviews stated that they felt that sweetgrass is threatened by nonnative plants. As Onondaga herbalist Otatdodah Homer stated, “I blame the invaders ... Foreign plants from other areas.’’* In particular, four of these participants specifically referred to purple loosestrife, Lythrum salicaria L. Purple loosestrife was found in small quantities at the two harvesting sites, Akwesasne and LaFayette, which were reported by interviewees to be past harvesting areas of sweetgrass. It was not found at any of the other 30 sites visited. Ecological disturbance is one factor that might be responsible for the absence of sweetgrass in LaFayette, which is now a popular park. Also, the interviewee who gathered from this site believes the water in the creek to be polluted. The meadow in Hogansburg (Akwesasne) has been a popular sweetgrass gathering site for the past 50 years. The decline of sweetgrass in the Akwesasne area led me to inquire about the past land management practices in the vicinity of both the Akwesasne and Onondaga territories. Traditional knowledge systems provide insights on the management of re- 106 SHEBITZ and KIMMERER Vol. 24, No. 1 sources and ecosystems (Berkes et al. 2000). One of the goals of the interviews was to gain an understanding of past land management and sweetgrass harvest- ing practices. This topic was covered to determine if a change in these practices might be the cause of the reported decline in sweetgrass populations in the vi- cinity of the Akwesasne Reservation and other harvesting areas. It is possible that sweetgrass is not found on the Akwesasne meadow because of the manner in which it was harvested. Sweetgrass reproduces primarily by its rhizomes (Green 2000; Winslow 2000). To many, sweetgrass is traditionally har- vested by grasping the shoots firmly at the base of the stem and pinching or pulling them until they break loose from the rhizomes and roots, which are an inch or two below the surface (English 1982). Theresa Burns explained that: The way I pick sweetgrass is the same way that my grandmother picks sweetgrass. She never takes the root, so that it can come back next year. As she's picking, she cleans it. I don’t get as much sweetgrass as maybe somebody else does because I like to get it all clean, I don’t like to clean it when I get home ... that’s the way she does it, she cleans as she goes. And she’s very selective as she picks, and I am too.° Not all Haudenosaunee sweetgrass gatherers, however, practice this method of harvesting. Knowledge bases, whether they are western scientific or traditional are both collective and individual in nature. As such, they reflect a diversity of perspectives. All seventeen of the consultants in both the formal and informal interviews reported that some Native gatherers are now taking the roots when they harvest the sweetgrass. Eight stated that they harvest sweetgrass from its root and do not believe that this method affects the sweetgrass population. Thom- as Porter, a Mohawk leader who burns sweetgrass as an incense in ceremonies, stated in a formal interview that “... we take the whole plant, just pull it up, and some root comes off too, but that’s. not a problem, it doesn’t hurt the grass.’’” Onondaga herbalist Jeanne Shenandoah explained the lesson she received from her friend when they went out to pick sweetgrass: She said ““Oh you have to take the roots up when you pick it.” She’d have big bunches of it with the roots. She said if you don’t pull the roots up, it won't stay green. And I thought, you would hope that people would be considerate so as not to take the whole patch, you know? So that it could multiply. I was really shocked when she said ‘Pull the roots.’’® By pulling the entire plant and removing the roots and rhizomes from the ground, that plant's energy storage and primary reproductive means is lost. Whether this action negatively affects the overall sweetgrass population is debat- able. There are documented cases where indigenous harvesting practices that in- volved the digging of subterranean organs of wild plants, such as rhizomes, in fact benefited the overall population of the plant. For example, M. Kat Anderson (1997:149) presents the argument that tillage activities practiced by Native Amer- icans of California “. .. mimicked natural disturbances with which the plants co- evolved, and played an ecological role that is now vacant in many wildlands, where Native Americans can no longer harvest and manage . Five interviewees (three from the Burns family) stated that they were taught Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 107 from their mothers and grandmothers to cut the sweetgrass at the base of the stem, so as not to disturb the root, and that this method was used by their an- cestors. All of these participants are angered when they see people, both Native and non-Native, harvesting sweetgrass from its root and believe that only recently have people begun, in their haste, to carelessly pull the roots of sweetgrass. Otat- dodah Homer stated, “I think people pick it and they didn’t know how to pick it. They would just pull it up from the root. And by pulling it up from the root, there goes the plant! ... Obviously they’re not properly picking.’”’° Another issue which was brought up in four of the eight formal interviews was the possibility that sweetgrass is being overharvested. The removal of the roots and rhizomes, in conjunction with overharvesting, possibly affects the sweetgrass population of Akwesasne. The Haudenosaunee Environmental Task Force warns that overharvesting particular plant species is a threat that faces the native grasses of Akwesasne. This unsustainable harvesting may eliminate whole generations of new plants as people tend to pick the strongest of plants, leaving the young and frail ones to continue to the next generation. Arquette (2000:57) comments, ‘“Every plant has a leader among their family group. When we target the leader and discard the others, we weaken the entire remaining family group.” Efforts are being made by the Task Force to educate individuals about the im- portance of harvesting sweetgrass sustainably. With the cooperation of the Task Force, Arquette (1999) has written an information pamphlet on preserving and restoring small plants and sweetgrass that instructs gatherers of sweetgrass to pick it sustainably, to not overharvest, and to replant roots from sweetgrass that are picked. In addition to unsustainable harvesting of sweetgrass, the absence of con- trolled burning might be responsible for the decline in sweetgrass populations. Many indigenous societies create small-scale disturbances, such as fire, to “‘nur- ture sources of ecosystem renewal’ (Berkes et al. 2000:1256). Fire is a significant ecological factor in maintaining perennial grasses in grassland ecosystems (An- derson 1996). Fires set by indigenous people were often used to increase yields, recycle nutrients, clear detritus, and promote growth of desired plants in the midst of reduced competition (Anderson 1996). Since some plants used in bas- ketry require burning, the absence of controlled burning, and modern fire sup- pression policies have created difficulties for contemporary weavers (Ortiz 1993). All of the consultants for this study stated that they recall land being burned by their grandparents, mostly for the regeneration of hay. In fact, two of the interviewees remember that the fields from which they used to harvest sweetgrass were burned for hay until approximately 50 years ago. Theresa Burns recalled that: Most of the time what they burned for was hay. So that the hay would come in, they’d always burn it. In the spring, right after the snow went away.... [SJuch a great smell, the burning. I used to walk through [the fields] and just get all full of the grass smoke, it was great. They did that because . .. burning puts all the nutrients back in the soil." The increased abundance of sweetgrass in these areas was probably not the aim of the burning, but a result of it nevertheless. Two individuals who took part 108 SHEBITZ and KIMMERER Vol. 24, No. 1 in informal interviews, both of whom were elders from the Akwesasne territory, stated that burning has been done specifically to encourage sweetgrass growth. The consultants of both informal and formal interviews explained that although some controlled burning is still carried out, the practice has become much less common over the course of the past 50 years. When we inquired why burning was not practiced often, the consultants responded that now people are too con- cerned about burning their neighbor’s homes, there is not a great deal of space left. Thomas Porter explained, “Growing up in Akwesasne, I used to help my family burn our land, and the land around our area... It’s hard to control fire. When wind blows, it could burn the homes and the whole forest.” " At each interview, we expected to hear that the consultant thought that the absence of controlled burning might be responsible for the decline of sweetgrass in traditional gathering areas. This possibility was not brought up in any of the interviews, however, until we explained our theory. The tolerance of sweetgrass to fire (Walsh 1994) was discussed with each consultant. Since fire does not con- sume the underground rhizomes, the grass can recover from burning, while ben- efiting from the increased sunlight and nutrient availability (Lynch and Lupfer 1995). The rhizomes of sweetgrass often sprout after aerial portions are burned and culms arise from among the dead foliage of the preceding year (Walsh 1994). It is possible that the foliage protects basal buds from fire damage in the spring, when the dead foliage is rich in moisture (Walsh 1994). After our perspective was explained, the interviewees agreed that the lack of controlled burning in the vi- cinity of their nation’s territory might be responsible for its current absence in past gathering sites. Through the interviews, the strength of the connection between the Hauden- osaunee people and sweetgrass was made apparent, as was their concern for the fate of sweetgrass. Otatdodah Homer explained there is a fear “... that it’s be- coming extinct... It’s important to our culture and we want to keep it alive, to keep using it... I think that scientists should know that it’s sacred to us native peoples..." CONCLUSION Berkes et al. (2000:1521) stated, ‘Indigenous groups offer alternative knowl- edge and perspectives based on their own locally developed practices of resource use.”” This understanding was central to the research presented in this study. The knowledge possessed by the Haudenosaunee proved to be valuable in identifying population trends and in characterizing sweetgrass ecology and habitat. Detailed knowledge of past and present harvesting techniques and land management prac- tices, such as controlled burning, contributed to understanding of the influences that may be responsible for the difficulties in locating sweetgrass in traditional gathering areas. Most of the threats that face sweetgrass populations throughout the Northeast are no different than the threats that face other midsuccessional species that in- habit moist areas. Habitat destruction brought about through the draining of wet- lands, suppression of natural fires, lack of controlled burning, and ecological suc- cession, has led to the replacement of sweetgrass habitat with altered landscapes. Spring/Summer 2004. | JOURNAL OF ETHNOBIOLOGY 109 These threats are a result of shifts in cultural practices; as the Haudenosaunee have changed their traditional land management practices and urbanization en- croaches upon what remains of the undeveloped landscape. The integration of knowledge bases in this study allowed us to frame and approach the questions concerning ecological requirements and population trends of sweetgrass. Approaching this project from both an ecological and ethnographic perspective enhanced the understanding of sweetgrass for this study, and may prove to be beneficial in future sweetgrass conservation efforts. On the Onondaga and Akwesasne territories, a return to traditional land management practices such as controlled burning and sustainable harvesting practices may be the primary means to ensure that sweetgrass populations persist. The continued presence of sweetgrass in the vicinity of the territories will enable traditions associated with the plant to endure. NOTES ‘Christine Horn, interview, July 12, 2001. 2 See note 1. ° Anonymous interview, February 15, 2001. * Michele Dean Stock, Seneca basketmaker, interview, July 10, 2001. > Otatdodah Homer, Onondaga herbalist, interview, February 15, 2001. ° Theresa Burns, interview, February 15, 2001. 7 Thomas Porter, a Mohawk leader, interview, May 30, 2001. 8 Jeanne Shenandoah, Onondaga herbalist, interview, April 25, 2001. * See note 5. See note 6. " See note 7. 2 See note 5. ACKNOWLEDGMENTS 4 ities and individuals We greatly appreciate the support of the H whose cooperation made this project possible. One of the basketmakers from the Akwes- asne Mohawk Territory, Theresa Burns, was hired for this project as a field consultant. Her assistance was invaluable in locating study sites and introducing us to other basketweavers in the area. Funding for this study was provided by the United States Department of Agriculture Fund for Rural America and the Sussman Foundation of the State University of New York College of Environmental Science and Forestry. 110 SHEBITZ and KIMMERER Vol. 24, No. 1 REFERENCES CITED Anderson, M. Kat. 1996. The ethnobotany of deergrass, Muhlenbergia rignes (Po- aceae): its uses and fire management by Califo: se Indian tribes. Economic Bota- ny 50:41 ion From tillage to table: the in- digenous cultivation of geophytes for food in California. Journal of Ethnobiolo- gy 17:149-169. Arquette, Margaret. 1999. Tsi Shonkwnientho: ervation and All Else, ed. Haudenosaunee Environ- mental Task Force, pp. 52-63. Hauden- osaunee Environmental Task Force, Rooseveltown, New York. Pea Salli. 1983. Mohawk ppasebat ak- s of Akwesasne. In American Indian Basketry, ed. Otis Tufton Mason, ‘Pp. 10- 16. Dover Publications, New Y Berkes, for Johan Colding, ad Carl Fo 000. Rediscovery of traditional eco. os knowledge as a ve m, oo. Ecological Applications 10:1251- ieee P. 1988. Bridging native and western science. Convergence 21(2/ i: 3):49- 68. English, Moira. 1982. Sweet grass—a sa- cred herb. The Herbarist 48:5-9. Gleason, Henry A. and Arthur Cronquist. 1991. Manual of Vascular Plants of North- eastern United States and Adjacent Canada, 2nd lew York Botanical Garden, New York. aoe: T. 1969. The Mohawk Indians and Their Valley. J.S. Lishynsky, New York. Greene, David. 2000. Sweetgrass Management Guide. New York Sea Grant, Buffalo. Herrick, James W. 1995. Iroquois Medical Bot- any. Syracuse University Press, Syra- cuse. Hitchcock, A.S. 1935. Manual of the Grasses of the United States. United States De- partment of Agriculture, Washington, Huntington, Henry P. 2000. Using tradi- tional ecological knowledge in science: methods and applications. Ecological Applications 10:1270-1274. Kavasch, E. Barrie and Karen Barr. 1999. American Indian Healing Arts. Bantam Books, New York. Kidwell, Clara S. 1973. Science and ethno- science. Indian Historian 6(4):43-53. Kimmerer, Robin W. 2002. Weaving tradi- tional ecological knowledge into a bio- logical education: a call to action. Bio- Science 52:432-438. Lauersons, Dctoa 1996. Teionkwahontasen; etgrass is Around a son of oa Dollco Press, Ottaw Lynch, Beth and Bruce ee 1995. Sweet Grass (Wiingashk) Project. Great Lakes Indian Fish and Wildlife Commission. Odanah, Wisconsin. Mauro, FE and P.D. Hardison. 2000. Tradi- tional knowledge of indigenous s and lo- cal communities: international debate and policy initiatives. Ecological Appli- cations 10:1263—1269. Newcomb, Lawrence. 1977. Newcomb’ Wild- sae Little, Brown and Compa- ny, Bo Ortiz, Beth "1993. Contemporary California the environ- Anderson, pp: a: Ballena Press, Menlo Park, Californi Reed, C.F. 1971. Common Weeds of the United ork. SAS Institute, Inc. 1990. Guide, Version 6, Third Edition. SAS Insti- tute, Inc., Cary, North Carolina. CNWG. 2004. Catalogue of New World Grasses (Poaceae), by RJ. Soreng, G. Davidse, P.M. Peterson, Fa ore EJ. oo TS. Filgueir, d O. Mor- rone. [http://mobot. abot Sau W3T/ Search/ nwgc.html] (verified January 21, 2004 Walsh, R.A. 1994. Hierochloe odorata. The Fire Effects Information System [Data base], compiled by William C. Fischer. US. Department of Agriculture, Forest Service, Intermountain Research Sta- tion. Intermountain Fire Sciences Lab- oratory. Magentic tape reels; 9 track; 1600 ee ASCII with Common LISP presen White, DI. E Haber, and C. Keddy. 1 Invasive Plants of Natural Habitats in i Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 111 ada: an Integrated Review of Wetland and __ Wilde, S.A., GK. Voigt, and J.C. Iyer. 1972. Upland Species and Legislation Governing Soil and Plant Analysis for Tree Culture. Their Control. North American Wetlands Oxford and IBH Publishing Co. New Conservation Council (Canada), Cana- Delhi, India dian Museum of Nature, Environment Winslow, Susan. 2000. Propagation proto- Canada, Canadian Green Plan and col for Hierochloe odorata: sweetgrass. Canada, Ottawa. Natioe Plants Journal 1(2):102-103. Journal of Ethnobiology 24(1): 113-138 Spring/Summer 2004 STRUCTURE, PHENOLOGY, FRUIT YIELD, AND FUTURE PROSPECTS OF SOME PROMINENT WILD EDIBLE PLANT SPECIES OF THE SIKKIM HIMALAYA, INDIA MANJU SUNDRIYAL? and R.C. SUNDRIYAL® G.B. Pant Institute of Himalayan Environment & Development North East Unit, Vivek Vihar, Itanagar, Arunachal Pradesh, India @ (manjusundriyal@yahoo.co.in), ® (sundriyalrc@yahoo.com) ABSTRACT.—In the Himalaya a large variety of plants that grow in forest areas are used as food, and a few of them are exploited on large scale. We discuss six wild tree species of the Sikkim Himalaya that yield edible fruit and that are of great economic value to rural people: Baccaurea sapida (Roxb.) Muell.-Arg., Diplok- nema butyracea (Roxb.) Lam., Elaeagnus latifolia L., Eriolobus indica Schn., Machilus edulis King, and Spondias axillaris Roxb. These species, which are an important component of forest diversity, are found in low densities, have poor rates of re- generation, and suffer from overexploitation by the local population. All the spe- cies have market value and significant potential for value addition. The per-tree fruit yield was fair for each species, and plantation stands could generate high economic returns. It is suggested that a large-scale propagation plan is needed for these species so that farmers could cultivate them, which would reduce the pressure on the wild populations of the forests. Key words: wild edible plants, structure, regeneration, phenology, fruit yield, mar- etin, RESUMEN.—En el Himalaya Sikkim se utilizan como alimento una gran variedad de plantas que crecen en las areas boscosas, y pocas de ellas se explotan a gran escala. Examinamos seis especies arboreas silvestres de la regi6n que producen frutas comestibles y que son de gran valor econémico para las gentes rurales: Baccaurea sapida (Roxb.) Muell.-Arg., Diploknema butyracea (Roxb.) Lam., Elaeagnus latifolia L., Eriolobus indica Schn., Machilus edulis King, y Spondias axillaris Roxb. Estas especies, que constituyen un componente importante de la diversidad for- estal, aparecen en densidades bajas, tienen bajos indices de regeneraciOn, y sufren de sobreexplotacién por la gente local. Todas las especies tienen valor en el mer- cado y un potencial de valor afiadido significativo. La produccién por arbol es bastante abundante para cada especie, y una plantacién podria generar altos be- neficios econdmicos. Sugerimos que se necesita un plan de propagaciOn a gran escala para estas especies de forma que los granjeros pudieran cultivarlas, lo que reduciria la presién sobre las poblaciones silvestres de los bosques. RESUME.—Dans le Sikkim himalayen, une grande diversité de plantes, poussant dans les milieux forestiers, sont utilisées comme plantes alimentaires. Plusieurs sont exploitées a grande échelle. Nous discutons de six espéces d’arbres indigenes de cette région qui donnent des fruits comestibles, soit Baccaurea sapida (Roxb.) Muell.-Arg., Diploknema butyracea (Roxb.) Lam., Elaeagnus latifolia L., Eriolobus indica Schn., Machilus edulis King et Spondias axillaris Roxb. Ce sont des espéces dont la valeur économique est importante pour les populations rurales. Ces es- 114 SUNDRIYAL and SUNDRIYAL Vol. 24, No. 1 peéces forment une part importante de la diversité forestiére. Elles se retrouvent a faible densité, possédent un taux de régénération faible et sont surexploitées par la population locale. Toutes les espéces possédent une valeur commerciale et un grand potentiel pouvant augmenter leur valeur. Le rendement des récoltes de fruits par arbre était assez bon pour chacune des espéces. Aussi, les plantations pourraient générer un grand profit économique. Aussi, un plan de propagation est nécessaire afin que les agriculteurs puissent cultiver ces espéces. Cela réduirait la pression exercée sur les populations naturelles que l’on trouve en forét. INTRODUCTION People throughout the tropics depend on their indigenous plants for food security and a multitude of everyday products from medicines to fibers. Large numbers of plants are collected from forests and other wild areas to meet sub- sistence food needs of the people (Arora and Pandey 1996). Such dependence is even greater in the Himalayan region (Samant and Dhar 1997). Wild species pro- ducing edible products could play a prominent role in increasing the income of farm households if cultivated or harvested sustainably (Phillips 1993). Some of these plants are rich in nutrients, and their consumption helps to maintain a balanced diet among the rural population (Sundriyal and Sundriyal 2001a). There- fore more attention should be paid to the natural associations, fruit yield and regeneration of potential wild edible species, as many of them often function as keystone species and provide food not only to people, but to many frugivores as well. In the Sikkim Himalaya, 190 wild plant species used by local communities for food have been screened (Sundriyal 1999; Sundriyal et al. 1998). All of them are collected from wild habitats and to date no effort has been made to cultivate them. We discuss the structure, regeneration, phenology, fruit yield, and market- ing of the six most widely used wild plants of the Sikkim Himalaya that produce edible fruits. Due to their poor regeneration in forest habitats, there is an urgent need for them to be propagated under nursery conditions. The study highlights the condition of these species in the forests of Sikkim and provides related infor- mation that will be useful for their management. MATERIALS AND METHODS The Sikkim Himalaya is the part of Eastern Himalaya, with an elevational range of 300 to 8579 m above msl. Sikkim State has a total area of 7096 km’, which is just 0.02% of the total area of India. It has a population of 505,505, of which 85% lives in villages. The region is rich in cultural and biological diversity. Lepchas, Bhutias, Limbus, and Nepalese are the main ethnic groups; Nepalese are most numerous. Farmers practice double cropping in the valleys and single cropping at higher elevations (Sundriyal et al. 1994a). In Sikkim, 36% of the area is forested, 15% is under cultivation, 10% is pasture and 25% is barren land. The state has subtropical, temperate and subalpine forests as one goes from lower to higher elevations (Rai and Sundriyal 1997; Sundriyal et al. 1994b, Sundriyal and Sharma 1996). All the forests provide diverse products, such as timber, firewood, Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 415 fodder and NTFPs (non-timber forest products). Plants with edible parts are found growing in all the forest stands, are very popular among local people, and a large variety is also brought to markets. The climate of the area is typically monsoonal. Due to its proximity to the Bay of Bengal, it has high rainfall and humidity, which supports the growth of luxuriant vegetation. Average annual rainfall varies from 1200 mm (at 300 m elevation) to 4500 mm (at 2000 m elevation), and over 80% of the rain comes during the monsoon season (June through September). The average maximum temperature varies from 21° to 35°C while the average minimum temperature varies from 12° to 23°C. The winter (December through February) is generally rainless and cold. The minimum temperature during this period varies from 0°C at higher elevations to 9°C in the valleys, while maximum temperature ranges between 13° to 28°C. October to November is the autumn period; spring falls in March and April. Pleasant weather prevails during these two seasons. The legal status of forests may be categorized into Reserve Forest (legally protected areas), Khasmahal Forest (forest area designated for the use of villages for meeting their fuel and fodder needs), Community Forest (owned by a village) and Private Forest (owned by an individual farmer). The Sikkim Forest Depart- ment manages Reserve and Khasmahal forests and village community or indi- vidual farm family manage Community and Private forests. Description of Selected Wild Edible Plants—For the present study species were se- lected with regard to their potential in terms of their importance to local popu- lation; high anthropogenic pressure in the natural habitats; high local demand for the fruits; market value; dearth of information available about their structure, phe- nology, regeneration and growth; and the possibility of incorporating the prop- agation of these species in traditional agroforestry systems. Six wild edible species met these criteria and were selected for detailed study: Baccaurea sapida (Roxb.) Muell.-Arg., Diploknema butyracea (Roxb.) Lam., Elaeagnus latifolia L., Eriolobus indica Schn., Machilus edulis King, and Spondias axillaris Roxb. All are trees except Elaeag- nus latifolia, which is a liana (woody climber). Baccaurea sapida (syn. B. ramiflora Lour., Euphorbiaceae; local name: kusum ‘sapida’) is a small to medium size (up to 10 m height) semi-evergreen tree that grows in the subtropical forests at elevations of 500-1500 m above msl. The yel- lowish fruits are available from May to August. Diploknema butyracea (syn. Bassia butyracea Roxb., Aesandra butyracea (Roxb.) Baehni, Sapotaceae; local name: chiuree ‘Indian butter tree’) is a tree of nearly 15 m height, that grows at elevations of 300-1300 m above msl. The fruits are scented and pulp is juicy and sweet. Elaeagnus latifolia (syn. Elaeagnus conferta Roxb., Elaeagnaceae; local name: mus- lendi ‘bastard oleaster’) is a large evergreen liana that grows at lower elevations in the Sikkim Himalaya. Fruits are collected during March and April and are commonly available in the markets at that time. Eriolobus indica (Rosaceae; local name: mehal ‘Indian crabapple’) is a deciduous tree (9-12 m height) of the lower temperate zone (900-1800 m above msl). Fruits are available July-September and eaten fresh or processed into a form of pickle. 116 SUNDRIYAL and SUNDRIYAL Vol. 24, No. 1 An extract of the fruit is made into chuk, a medicine valued for relieving stomach in. Machilus edulis (syn. Persea fructifera Kost., Lauraceae; local name: pumst) is an evergreen tree of about 20-30 m height that grows in natural forests at elevations above 1700 m. It is considered to be a local variety of avocado (Persea americana). Fruits are commonly found in markets during December to March, and the outer fleshy pulp that comes out attached with the skin is scooped out and eaten. Spondias axillaris (syn. S. acuminata (L.f.) Kurz., Choerospondias axillaris (Roxb.) Burtt. & Hill, Anacardiaceae; local name: lapsi ‘hog-plum’) is a tree that can reach a height of 35 m. It commonly grows in lower hill forests. The ripened fruits are eaten raw. The fruits that are sold at a semiripe stage are pickled, which gives them a shelf life of up to 3-5 years. METHODS Vegetation Structure Analysis and Regeneration—In order to analyze the vegetation structure and regeneration for the six selected species, eight forests were chosen for detailed study based on the preliminary surveys done for areas that are fre- quented by forest dwellers. The vegetation was sampled with 10 x 10-m quadrats. Depending upon the area of each of the forests, between 30 and 40 quadrats were examined (Sundriyal 1999). All tree species and indviduals growing in a quadrat were listed and their cbh (circumference at breast height) was noted. Woody plants with >32.4-cm cbh were counted as trees (Sundriyal et al. 1994b; Sundriyal and Sharma 1996). The frequency, density, abundance, and basal cover (ground area covered by each species) were calculated following the method given by Curtis and Cottam (1956). To measure the dominance of each tree species, the Importance Value Index (IVI) was calculated as a sum of relative frequency, rel- ative density and relative dominance (Mueller-Dombois and Ellenberg 1974). For each stand, the species with the highest IVI was considered most dominant. The A/F (abundance/ frequency) ratio was calculated to interpret the distribution pat- tern of the species; a value <0.025 shows regular distribution, between 0.025-0.05 shows random distribution, and >0.05 shows contagious distribution of the spe- cies in the forest stand (Saxena and Singh 1982). The regeneration of all species was assessed from 20 randomly selected quad- rats of 10 x 10-m size in each forest stand. All plants with <32.4-cm cbh con- sisting of saplings and seedlings of woody species were considered to be regen- erating individuals (Sundriyal et al. 1994b; Sundriyal and Sharma 1996), and den- sity for all species was noted that also included wild edible plants in each stand. Phenological Investigation.—Study of the phenological stages of a species provides information on its functional aspects: growth of buds, leaf fall and leafing out, anthesis, fruiting and seed dispersal in relation to months, season or years that describes seasonal aspects of ecological phenomenon. Understanding phenology is very important for optimal species management under forest conditions. The phenological investigation for each species was conducted in at least two stands. For each species, 5-8 individual trees were marked at each stand to study the phenological observations on leaf fall, leaf bud formation and flushing, flower bud Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 117 formation and anthesis, fruit formation and maturation, and seed dispersal (Ral- han et al. 1985; Sundriyal 1990). Fruit Yield.—Fruit yields were determined by harvesting and recorded individu- ally by tree. For all the species, individuals of different girth classes and heights were marked at different locations well before fruit bearing. Tree girth at breast height, and height (with the help of a bamboo stick) was measured for all selected individuals. Efforts were made to harvest almost all the fruits of small trees. For big trees it was not possible to harvest all the fruits. Therefore, the uncollected fruits were counted visually and converted to biomass value based on the har- vested fruits. This method accounts for nearly 70-80% of total fruit yield. Per-tree fruit yield was plotted against tree circumference at breast height (cbh) for the harvested individuals. Regression equations were developed to assess the rela- tionship between fruit productivity and tree girth. Marketing of Selected Species—Three prominent markets of Sikkim State, namely Gangtok, Namchi, and Singtam, were surveyed every other week for one year. For each fruit type, data about its period of availability, number of retailers in- volved per week in its trade, quantity sold, and gross income to the retailer from its sale. RESULTS Characteristics of Forest Stands and Status of Wild Edible Plants.—Table 1 provides details of all the sampled forest stands ranging from elevations of 400-2600 m in the Sikkim Himalaya. Of eight stands sampled, three are in the Reserve Forest, one is in Khasmahal forest, three are in Community Forest and one is in the Private Forest. The Reserve Forest areas generally do not suffer any severe an- thropogenic pressure from fuel and timber collection. Field visits and discussion with local people revealed, however, that people do collect fodder and various other NTFPs, which also include wild edible plants. Khasmahal Forests are sup- posed to fulfill villagers’ need for fuel, fodder, and timber, and thus bear maxi- mum pressure. The Community and Private Forests generally have higher pro- portions of the species preferred for their subsistence value. The Reserve Forests (Stands I, II, VIII) maintain high diversity of species with high density and basal cover, which is expected in view of low anthropogenic pressure in these forests. Stand II was in the most inaccessible area. Even so, people collect wild edible plants there. The Khasmahal Forest (Stand VII) showed low density and basal cover due to high anthropogenic pressure. The Community Forests (Stands IV, V, VI) were also in fairly good condition, though the Private forest (Stand III) was highly exploited due to individual family demands. The presence and absence of different species along with their common names Is presented in Table 2. Alto- gether 143 tree species were identified. Stand II had the most species (30), closely followed by Stand VIII (29 species). Data concerning the phytosociology of each stand discussed below are presented in Table 3. Stand I. This forest stand occurs at elevations of 400-700 m, and represents a typical subtropical forest of the Eastern Himalayan region. According to Forest Department records the forest stand was originally dominated by Shorea robusta; TABLE 1.—Characteristics of the forest stands/sites sampled for vegetation structure and regeneration of selected wild edible species. Forest stands/Sites Site characteristics I II Tl IV V VI Vil Vil Location Mamkhola —_Raileykhola_Chhota-Sing- Samdung Central Pan- Yuksam Pangthang eae te tam dam (lower) top Darjeeling Darjeeling South Sikkim East Sikkim East Sikkim West Sikkim East Sikkim South Sikkim stead range 400-700 m_ = 450-900 m_— 860-1050 m_~— 800-1400 m_ 1200-1500 m 1200-1600 m 1500-1900 m 1700-2600 m Forest category Reserve Reserve Private Community Community Community Khasmahal Reserve Total Shobe of tree 23 30 5 16 23 12 15 29 and Total no. of wild ed- 7 7 5 4 4 5 3 9 ible species in stand* Dominant tree spe- T. grandis-S._ C. indica-S. _S. wallichii-_A. nepalensis- E roxburghii—__C. tribuloides— E. acuminata— O. lamellosa— cies robusta robusta Albizia E. indica S. axillaris E. indica S. theifolia C. tribulo- ides Standing trees (ha~') 711 280 93 495 420 350 2 Basal area (m*ha~') — 43.52 48.54 19.45 21.66 27.70 51.64 Cee F27 Biotic pressure** Medium Low-medium High High Selective Selective Very high Low Total density of all 180 150 00 105 35 69 5S 155 wild edible spe- cies (individuals | Basal area of all 9.89 11.80 335 4.11 7.35 19.90 16.15 17.2 wild edible spe- cies (m?ha™! Total stand fegencr NR 3820 1655 1710 2060 2786 3138 4700 ation (ha Regenerating of NR 1560 725 695 730 765 503 230 wild edible plants (ha) * Details of the species is provided in Table 2; NR—data not recorded. ualitative assessment based on field observations; and fodder collected in oe forest stand. “Biotic pressure” is a it is based on the number of trees cut and lopped and the quantity of NTFPs sil IVATMGNNS P¥e TVARIGNNS L ‘ON ‘BZ ‘TOA TABLE 2.—Presence (+) of tree species and wild edible plants at different forest stands in the Sikkim Himalaya. Forest stands Species Local name* I I Il IV Vv VI VI VI Acer akan bons & T. kapase — = = rn a x + + Acer oblongu phirphire ~ + - _ _ = = + Albizia procera pec ) Benth. siris sae = + + = — + = Albizia stipulata Boiv. rato-siris = - - _ + = bs 3 Alnus gid oe utish - _ + _ = x = Amoora wallich lahsune + + ~ — _ _ = fs Artocarpus da Fen. ) badar = + a = ee s z 2 Artocarpus sp. kathal ~ 2 + - ~_ = = Bin Baccaurea sapida (Roxb.) Muell.-Arg. usum + + - _ _ = a = Bambusa nutans (Wall. ex Munro) choya bans _ - < i = os mbusa sp. nibabans _ _ = a - C. is . Bauhinia variegata L i + + = 2 = = = 2 Beilschmiedia aac! sola Nees. tarsing - - = a = = es - Betula alnoides Buch saur — = + = i. = = Ps Betula cylindrostachys Wall saur - _ = = = = = - bax malabaricum DC semal — = ns ae ce S ie coat mitis Clarke chuletro - ~ + + + es a ol psa etusa (L.) Faas ayo + _ = ea = a = Callicarpa arborea R guyilo < Bs Fe = - 2 = Se Casearia tomentosa i. barkule = + - = i an Castanopsis indica A. DC. dhalne katus + ao + = as ws = Be Castanopsis tribuloides A. DC. musure katus - ~ 4 - _ pes + a8 Celtis tetrandra Wall. khari = AB ibs S a m1 Cinnamomum impressinervium Meissn. sisi ~ _ _ = sat 2 Cryptomeria japonica Don. hupi _ = oe . = a ee hy Dendrocalamus sikkimensis Gamble. bhalubans — - - +. = i i Be slams strictus (Roxb.) Nees. -bans _ - - + se = a Diploknema butyracea (Roxb.) Lam. chiuree - + + ae : ae = oa Duabanga grandiflora (Roxb. ex DC) lampatey - + _ a 5 co 6 7007 Jewuing /Surids AOOTOISONHLA JO TYNINOL 611 TABLE 2—(continued) Forest stands Species Local name* I Il IV Vv VI Vil VIL Echinoca wie te Benth. rey = ye = = ee + Pisiaphie “givens muslendi — + ka ts a ce laeocarpus sikki Mast. bhadrasey - ~ _ ~ i a + Emblica officinalis (C Coecth) amla ss om oa ma = ‘a el m chinense Benth. setikath _ - _ _ ee re + <3 ohh nora BI. mahua + + - + ~ _ + riolobus indica mehal, likung - - + = re sa =, Erythrin phaledo ~ +- = = =, = Eurys acuminata DC. jhingni = = a = + . + Evodia Vraxiniflia HK.f. khanakpa = is = é. co i. ‘_ Ficus bengalensis L. bar se = + = =F =, Ficus benjamina L. kabra ~ + + _ Ficus cunia Ham khani _ a + + = a = Ficus elastica Roxb. labar + ~ + =a i. Ficus hirta Vahl. khasre ~ = = ae 9 i Ficus hispida L. khasrey = I. e as = ne re Ficus hookerti Co: nebara ~ - + — os = es Ficus nemoralis Wall du fe} rat =: a as ne Ficus roxburghii Roxb. nebara ~ + + + = = Gmelina arborea L. kh + = - nA al oa Gynocardia odorata R. Br. gante ae = a = ay = os Gynocardia odorata R. Br. gantey + * fe +. m= ikki sis Hk. f. lisey _ — _ ian a ea + Juglans regia L. okhar ote = = + ‘ ‘?. eptrum canum Sm ghurpis + = + - - ~ Litsaea citrata BI siltimur ~ + _ + ~_ = 2 oblonga Wall kalipahenle as eh aa ae ~ "a Litsaea polyantha J kutmero _ - oe + a a a Lyonia ovalifolia (Wall.) angeri on SL & + = + Macaranga pustulata King. malata _ ~ + r" i - OCT IVAIMGNNS Ppue ‘TVATMGNNS T ‘ON “2 “TOA TABLE 2—(continued) Forest stands Species Local name* I Il Il IV Vv Vil Vill Macaranga sp. seto-malata Machilus edulis King. lapche kawlo, - + # or 60 } 2 20+ Te 40 + a 10 + 20 + = of 0 i i i. 1 ais re 1 L J 0 " 30 40 50 60 70 80 90 100 110 120 30 50 70 90 110 130 Stem girth (cm) Tree girth cm) M. eduli 200 - S. axillaris _ we eee Y= -82.327+1.33X + 80 Y=-9,2340.149K » wr 70 + 160 } 60 | 8 140 . 50 | 2 p x3 100 + Tr . 80 f+ 30 L $s u 50 F 20 + 40 10 + 20 nu iL. 0 25 80 130 180 230 280 330 380 430 Tree girth (cm) Tree girth (cm) FIGURE 1.—Plant girth (cm) and fruit yield (kg/plant) of different ae edible yak in 9 Sikkim Himalaya. In regression eq (0.192), D. butyracea (0.282), E. latifolia (0.411), E. indica (0.871), M. edulis (0 262), S pe ies (0.545). uation, Y is fruit yield and X is tr irth. R v Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 133 sii B Qt sold (kg) lf Gross income (Rs) Quantity sold and income B.sapida D.butyacea_ E latifoia E indica Meduis S aillais FIGURE 2.—Marketing of the six selected species and gross income from selling the fruits. tation is calculated based on tree-spread (total canopy area covered by each spe- cies) (Table 6). Spondias axillaris, which has the largest area under each individual tree due to its broad canopy, could accommodate a planting of only about 180 mature tree ha~'. In contrast, Baccaurea sapida has the smallest area covered by each tree, so nearly 450 trees ha~' could easily be accommodated in a pure stand. In principle, B. sapida would give the highest income among all the species if grown in a pure stand. (Note that monocropping of trees that evolved in a diverse forest might not be the best way to cultivate them, but at this point in our research it is important to present alternatives to cutting down forest stands; see Bandeira et al. 2002.) Machilus edulis, S. axillaris, and Diploknema butyracea may be grown in TABLE 6. Aap wat economics of production of pilot stands of each individual wild ed- ible species Expected mature tree density in xpec oe Mean tree plantation annual fru Retail Gro: Species spread (m’) (trees ha~') yield* pie prices/kg income “(Rs.) Baccaurea sapida 20.40 450 32.60 10.00 325,980 Diploknema butyracea 36.60 275 14.33 8.00 114,664 Elaeagnus latifolia 44.50 Pam 13.43 8.00 107,424 Eriolobus indica 25.00 400 11.31 8.00 90,464 Machilus edulis 54.40 185 12.43 12.00 149,140 Spondias axillaris 60.00 170 9.75 10.00 97,512 * Calculated on the basis of 75% trees bearing fruits; 1US$ = Rs. 47. 134 SUNDRIYAL and SUNDRIYAL Vol. 24, No. 1 wasteland and can provide substantial income to farmers; all aspects of fruit processing, from simple cleaning and grading quality to the manufacture of juic- es, jams and jellies can add value and double a farmer’s income. DISCUSSION In the Himalaya a large variety of wild growing plants are used as food and some of them are exploited in large quantities (Samant and Dhar 1997). In the Sikkim Himalaya a total of 190 wild plant species are reported that have food value, of which 49 species are commonly used (Sundriyal 1999; Sundriyal et al. 1998). Wild woody plants that produce edible fruits are very rarely studied. Given the lack of information about their natural associations, densities, regeneration, phenology and fruit yields, they do not receive the attention of foresters and researchers in afforestation (Martin 1995). Six popular wild-growing species that produce edible fruits and have the potential to be exploited commercially were studied in detail to obtain information about their place in the forest structure, phenology, fruit yield, and future prospects in the Sikkim Himalaya. The eight forests we studied all fall in subtropical and temperate zones with different man- agement regimes—Reserve Forest, Community Forests, Khasmahal Forest and Pri- vate Forest. The Reserve Forests are managed by the Forest Department, and generally are devoid of any pressure except for the collection of a few non-timber forest products. The management of Community Forests is dependent on the en- tire village community. After villagers harvest a forest resource, they do not re- turn to the same place until the plants have had time to recover. In Khasmahal Forest, however, multidimensional pressure—timber, fuel, fodder, and NTFP col- lection—result in low stand density. The Private Forests are maintained according to individual family need. Species selection is based partly on economic value, mainly as fuel, fodder, and food; density of such stands is also low. The total stand species diversity and density was highest in the Reserve For- ests, followed by the Community Forests, and lowest in the Khasmahal and Pri- vate Forests. Though the total number of wild edible species was highest in Re- serve Forest stands, these species contributed just a quarter of the total stand density. Contribution of wild edible species to total stand density was >35% in Private Forest and >25% in Community Forest, which suggests the villagers ac- tively maintain preferred species. In contrast, the contribution of wild edible spe- cies to total stand density in the Khasmahal Forest was 15%, which suggests people are less concerned with sustainable use of land they do not own. The six species we studied were in high demand for their fruits. With the exception of Machilus edulis and Spondias axillaris, which are upper canopy species, they are middle-story species in the forest stands. Though the collective contri- bution of all wild edible species to total stand density, basal cover, and IVI was significant, an individual wild edible plant species in each stand contributes much less (<7%), which shows that all these species are associated species in the forest stand and such species are very important for maintaining forest diversity (see Phillips 1993). A few wild edible plants are cut down or otherwise stressed de- spite the economic value of their fruits. For example, Diploknema butyracea is also exploited heavily for fodder as well as for its fruits. Similarly, the wood of B. Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 135 sapida is also considered good fuel, while M. edulis and S. axillaris are exploited for timber. The pressure on the wild populations of these species is exacerbated by their low density and low rates of regeneration. Regeneration of all species in the Reserve Forest was better than in any other forest category. This could be attributed to substantial fuel and fodder removal in Khasmahal and Private forests. However, the percentage contribution of wild edible species to total stand regeneration was much higher in Private and Com- munity Forests (35-65%) than in the Reserve Forest. To some extent, villagers manipulate regeneration by uprooting certain undesirable species and promoting wild plants of food and other economic value in Private and Community forests. The six selected wild edible species showed poor regeneration in comparison to many other tree species in their respective forest stands, however. Local people collect fodder from the forest floor. In the process, they unknowingly collect seed- lings of wild edible species. Leaves are central to the adaptation of a plant for growth and competitive success, so it is important to understand the periodic progression of leaf growth for species management (Negi and Singh 1992). November through March is the period of leaf shedding for most of the tree species and 80% of the genera com- plete their bud burst by April in the Sikkim Himalaya (Sundriyal, unpublished data). Generally most wild edible species shed their leaves during winter months. Flowering of the various studied species ends before the monsoon rains. Diplok- nema butyracea and Eleagnus latifolia took the least time from fruit formation to maturation, Baccaurea sapida and Spondias axillaris took an intermediate time, while Eriolobus indica and Machilus edulis matured most slowly. It can be concluded that generally fruit development and maturation take longer for species with bigger fruit. Fruit maturation and dispersal completed before the rains for D. butyracea and E. latifolia, during the rains for B. sapida, and after the rains and before winter for S. axillaris, E. indica, and M. edulis. The time of fruit maturation is directly linked with the seed germination. For example, fruit that matures after the rains and during the winter months have a longer dormancy and viability period than those mature during the rains in the six selected species (Sundriyal and Sundriyal 2001b). Per-tree fruit yield is good enough to provide attractive opportunities for the villagers to collect them. However, fruit collection is highly erratic, because it is not regulated by the community. Often, large tree branches are lopped, because income depends on the amount of fruit collected as quickly and easily as possible. Such practices may lead to depletion of many species from forest areas in years to come. Where forest cover is dense and human population levels are low, forests maintain a good diversity of species, and wild edible fruit trees can yield good harvests. The wild edible species play an important role in enhancing nutrition, particularly for rural populations who can not afford to buy the fruit sold in the market (Sundriyal and Sundriyal 2001a). Plant collectors visit forest stands to collect fruits of the six species that are the focus of this study. If the desired quantity of fruit is not available, they will collect diverse species of economic value to justify their labor expenditure. In addition to the six studied species, the plant collectors harvest fruits of Castanopsis tribuloides, Juglans regia, Elaeocarpus sikkimensis, Mangifera sylvatica, young shoots of 136 SUNDRIYAL and SUNDRIYAL Vol. 24, No. 1 various bamboo species, tubers of Dioscorea, Diplazium, and Agaricus species in large quantities (Sundriyal 1999). Collection of wild plants plays a major role in the rural economy because farmers sell these plants in the nearby markets. For many poor people, selling wild plants is their only source of income. Generally, most of the wild edible plants are available for a short period, and due to their perishable nature they are sold at low prices (Sundriyal et al. 1998). Fortunately, all species studied in this investigation have high potential for value addition (Sundriyal 1999), as is being demonstrated for some wild plants in the other parts of the Himalaya (Dhyani and Khali 1993; Maikhuri et al. 1994). Cultivation of the six most desirable types would therefore reduce pressure on all forest plants. Interviews with the villagers revealed that they are willing to raise all the selected species in their farms. The seeds of these wild edible species show good germination rates when they are raised in a nursery (Sundriyal and Sundriyal 2001b). Unfortunately they are not yet included in the plantation schemes under- taken by any state government or department. Obtaining a supply of such seed- lings is a real problem if these species are to be grown by farmers. There is a need to ensure a supply of quality seedlings to farmers, which will have signif- icant benefits for the adoption of these species in agroforestry systems. Future Prospects.—Disturbance has become widespread in most of the forests in the Sikkim Himalaya. Therefore, information on species composition, growth, re- generation, other phenological characteristics, and sustainable harvest levels is important if we are to rejuvenate the severely stressed forests and individual trees. This is particularly critical for plants of low commercial value but that are nevertheless very useful to local people. In wild plant communities, sustained growth of all species in the presence of older plants is necessary to maintain the health of the stand (Singh and Singh 1992). A large number of people are sup- ported on wild food resources in remote areas, and population growth will lead to a greater demand for food and other necessities in near future. In some forest stands, the density and regeneration of the species discussed in this report is low and conservation efforts should emphasize these species. As wild food plants are exploited mainly by the local plant collectors, they receive little attention from foresters; therefore, the natural populations are quickly getting depleted. These species are crucial for maintaining diversity of the stands (FAO 1985; Herzog et al. 1998). The wild edible plants may bring sufficient returns if grown in the traditional agricultural system in hilly areas. Each of the six species yields a good quantity of fruits and thus provides high economic returns if properly main- tained. An assessment of the economics of the pilot stands for the six selected species reveals that all can produce good returns, especially Baccaurea sapida. Fur- thermore, a simple value addition could enhance the benefits. The wild edible species are also used for fruit, fodder, fuel, and timber collection. Therefore, cul- tivation of these species as part of an agroforestry program perhaps will also contribute to the conservation of genetic resources. If properly planted and cul- tivated some of these species can replace staple or commercial fruits, thereby contributing handsomely to the economy of the subsistence farmers in the moun- tains. Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 137 ACKNOWLEDGMENTS The authors thank Prof. A.N. Purohit for guidance and Dr. L.M.S. Palni for needful suggestions. The Director, G.B. Pant Institute of Himalayan Environment & Development is thanked for providing facilities. The study was funded by the grant provided by CSIR, Government of India, New Delhi. REFERENCES CITED Allen, S.E. 1989. Chemical Analysis of Ecolog- ical Materials, 2nd ed. Blackwell Scien- tific Publications, London Arora, R.K. and A. Pandey. 1996. Wild Ed- in} Plants of India: Conservation and Use. Indian Council of Agricultural Re- search. National Bureau of Plant Genet- ic Resources, ae Delhi. Bandeira, EPS. de E, J. Lopez oo an be V.M. Toledo. 2002, Tzotzil Maya ethn cology: nae perception and man- agement as a basis for coffee agroforest et Journal of Ethnobiology 22:247— eee I. T. and G. Cottam. 1956. Plant eel ogy Workbook: Laboratory Field Referen ene Burgers Publishing Co., Min- Dhyani, PP and M.P. Khali. 1993. Fruit yield and economics of jelly and jam production from fruits of some prom- ising Ficus (fig) tree Caan Ecology of Food and Nutrition 30:169-17 FAO. 1985. The Role of Minor ips in Nu- trition and Food Security. The Committee of Agriculture, Eighth Session, March 1985. Food and Agriculture Organiza- tion, Rome. . 1998. Domestication and commer- cialization of non-timber forest prod- ucts in agroforestry systems. Non-wood Forest Products 9 (reprinted), Food and wood forest products: experience and biometric principles. Non- orest Products 13, Food and Agriculture Or- ganization, Rome. Gangwar, A.K. . Ramakrish- nan.1990. Ethnobotanical notes on some tribes of Arunachal Pradesh, northeastern India. Economic Botany 94-105. Herzog, E,, D. Gautier-Beguin, and K. Mull- er. 1998. Uncultivated plants for human nutrition in Cote d'Ivoire. In Domesti- cation and Commercialization of Non-tim- ber Forest Products in Agroforestry Sys- tems. Non-wood Forest Products 9 (re- printed), Food and Agriculture Orga- nization, Rome. Maikhuri, R.K, R.L. Semwal, A. Singh, an M.C. Nautiyal. 1994. Wild fruits as a contribution to sustainable rural devel- opment: a case study from the Garhwal Himalaya. International Journal of Sus- tainable Development & World Ecology 1: Martin GJ. 1995. Ethnobotany: a Methods Manual. Chapman & Hall, London. Mueller-Dombois, D. and H. Ellenberg. 1974. Aims and Methods of Vegetation Ecology. John Wiley, piel York. India. pedestal Journal of Biometeorol- 2 ogy 36 Phillips, a 1993. The potential for har- Rai, S.C. and R.C. Sundriyal. 1997. Tourism and biodiversity conservation: The Sik- kim Himalaya. Ambio 26(4):235-242. Ralhan, P.K., R.K. Khanna, S.P. Singh, and J.S. Singh. 1985. Phenological character- istics of the tree layer of Kumaon Hi- malayan forests. Vegetatio 60:91-101. Manual of Analysis of Graw Hill Publishing Company Limit- ed, New Delhi. Samant, S.S. and U. Dhar. 1997. Diversity, endemism and economic potential of wild edible plants of Indian Himalaya. International Journal of Sustainable Devel- t & World Ecology 4:179-191. Saxena, A.K. and J.S. Singh. 1982. A phy- tosociological analysis of woody spe- 138 SUNDRIYAL and SUNDRIYAL cies in forest communities of a part of Kumaon geen le Vegetatio 50:3-22. Singh, J.S. and S.P. Singh. 1992. Forests of Himalaya. ae Prakashan, Nain ital, India. Sundriyal, M. 1999. Distribution, saison tion and Nutritive hae of Some Wild Edible Plants in the Himalaya. Ph. D. Thesis, H.N.B. Garhwal Univer- a Sundriyal. 1998. Wild plants of food value in Sikkim Himalaya: aspects of diversity and uti lization. In Perspectives for Planning asd Development in North East India, eds. R.C. Sundriyal, U. Shankar, and T.C. Upreti, pp.138-149. Himavikas Occasional Pub- lication No. 11, G.B. Pant Institute of Hi- malayan Environment & Development, Kosi-Katarmal, Almora, India. . 2000. Potential of wild edible pa ne fe Sikkim ag conser- . Journal of Non-timber si Prot 7(3&4):253-262. —— . Wild edible plants of the Sikkim ae Nutritive values “4 selected species. Economic Botany 5. 77-390. Vol. 24, No. 1 . 2001b. Seed germination and re- sponse of stem-cuttings to hormonal treatment in six wild edible fruit spe- cies of Sikkim Himalaya. Indian Forester 127:695-706. Sundriyal, M., R.C. Sundriyal, E. Sharma, and A.N. Purohit. 1998. Wild edibles and other useful plants from the Sik- kim Himalaya, India. Oecologia Montana :43— Sundriyal, R. C. 1990. Phenology of some temperate woody species of the Garhwal arpa International Journal of Ecology and Environmental Sciences 16:107—-117. Spiel 2 R C. and E. Sharma. 1996. An- thropogenic pressure on tree structure and biomass in the temperate forest of Mamlay watershed in Sikkim. Forest Y¥.K. Rai. 1994a. Hill agroforestry sys- tems in south Sikkim, India. Agroforest- ry si alg 26:215-235. Sundriyal, R. C., E. Sharma, L.K. Rai, and S.C. Rai. 1994b. Tree structure, regen- avery and woody biomass itor in aught ee forest of Mamlay w shed in ikkim Himalaya. Vegetatio Journal of Ethnobiology 24(1): 139-161 Spring/Summer 2004 PLURALISTIC MEDICAL SETTINGS AND MEDICINAL PLANT USE IN RURAL COMMUNITIES, MATO GROSSO, BRAZIL MARIA CHRISTINA DE MELLO AMOROZO Departamento de Ecologia, IB, Universidade Estadual Paulista (UNESP), C.P 199 - 13506-900 Rio Claro, Sao Paulo, Brazil (mcma@rc.unesp.br) ABSTRACT.—The use of medicinal plants and other traditional forms of treating illness in rural communities of Mato Grosso State, Brazil, that nowadays also depend on modern health care facilities is assessed. Forty-four households were surveyed about disease events and the use of modern health care facilities, me- dicinal plants, and consultation with practitioners of traditional medicine during the six months prior to research. All manufactured medicines, medicinal plants, and other therapeutic products present in the household at the time of the inter- view were recorded. Ninety-three percent of households reported the use of at least one of the modern medical services available as well as the use of medicinal plants during the previous six months. About 120 plant species were recorded. The associated use of modern and traditional medical services and the importance of medicinal plants in a context of social and economic change are discussed. Key words: rural communities, medicinal plants, medical pluralism, Brazilian sa- vannah. RESUMO.—O objetivo deste trabalho foi avaliar o emprego de plantas medicinais e outras formas tradicionais de tratar doencas, em comunidades rurais de Mato Grosso, Brasil, que atualmente contam também com servigos médicos modernos. Foi feito um levantamento em uma amostra de 44 domicilios, sobre episédios de doenga e recurso as diversas opgdes terapéuticas disponiveis, modernas e tradi- cionais, durante os seis meses anteriores 4 pesquisa; registraram-se todos os medi- camentos industrializados, plantas medicinais e outros produtos de uso terapéu- tico presentes no domicilio no momento da entrevista. 93% dos domicilios rela- taram uso de pelo menos um dos servigos médicos oficiais e de plantas medicinais nos seis meses anteriores. Cerca de 120 espécies de plantas foram registradas. Discute-se 0 uso conjunto de facilidades médicas modernas e recursos terapéu- ticos tradicionais, como plantas, bem como sua importancia num contexto de mu- danca social e econdmica. RESUME.—L objectif de cette recherche était d’évaluer |’utilisation des plantes médicinales ainsi que d’autres approches traditionnelles dans le traitement des maladies parmi les habitants des communautées rurales du Mato Grosso, au Bré- sil. Aujourd’hui, ces communautés ont également recours a des centres de mé- decine moderne. Une enquéte s’est déroulée auprés de 44 foyers. Elle portait sur les maladies survenues dans les foyers, la fréquentation des centres de médecine moderne, l'utilisation de plantes médicinales et la consultation de guérisseurs traditionnels, au cours des six mois précédent |’étude. Lors des interviews, on prenait note de tous les médicaments commercialisés, plantes médicinales et au- 140 AMOROZO Vol. 24, No. 1 7 dans tres produits ayant des propriétés thérapeutiques qui se trouvaient pré les différents domiciles. Pendant ces six mois, 93% des foyers ont utilisé au moins l’un des centres de médecine moderne mis 4 leur disposition tout en faisant appel aux plantes médicinales. Environ 120 espéces de plantes ont été identifiées. De plus, dans cet article, l'utilisation paralléle des centres de médecine moderne et des ressources thérapeutiques traditionnelles, ainsi que l’importance des plantes médicinales sont discutées dans un contexte de changements socio-économiques. INTRODUCTION Therapeutic pluralism is common throughout the world and can be under- stood as the coexistence, within the same society or group, of a number of health care alternatives with diverse origins and treatment foci, representing different systems of medical practice and ideology (Janzen 1971; McGrath 1999; Stoner 1986). Some researchers have found that patients and their relatives, when con- fronted with illness, may make use of the multiple treatment options available, even when there are clearly established limits and functions ascribed to the dif- ferent coexisting medical systems. From the viewpoint of users, then, they would be rather complementary than contradictory (Brunelli 1987; Colson 1971; Hamnett and Connell 1981; McGrath 1999). People take a pragmatic view of treatment and are willing to try whatever may be effective. me scholars stress the importance of the role played by social and political factors in legitimizing different medical systems within the same society and al- lowing them to coexist (Janzen 1971; McGrath 1999). In this view, medical plu- ralism provides patients and their families with an array of disease concepts and treatment alternatives that may be employed not only to obtain resources like prestige, power and material resources, but also to negotiate social relations and define cultural identity (Crandon-Malamud 1991). Likewise, the persistence of folk medicine in cosmopolitan settings where modern health care system is well es- tablished is sometimes explained as a means of the subordinate classes to resist impositions of the dominant medical ideology (Loyola 1991). Both pragmatic and sociopolitical views contribute to explain medical pluralism in different settings. In Brazil, different therapeutic traditions have contributed to the formation of folk medicine. From the sixteenth century on, contact between Iberian and indig- enous peoples of various ethnic groups created a complex combination of elements from European and autochthonous medicines; it is often difficult to identify the origin of specific aspects of the folk practices as indigenous, European, or the result of contact (Holanda 1994). African slaves taken to Brazil to work in agri- culture and mining further contributed to shape folk medicine. A feature shared by these different therapeutic traditions is the use of plants, at least to some extent, to treat illnesses. At present, folk or traditional medicine in rural areas of Brazil still retains many aspects of the medicine practised in colonial times. For example, some prac- tices related to humoral theory, especially those aimed at ‘‘purifying” the blood and at maintaining health through hot and cold balance, are still very common; the therapeutic use of excreta (like urine and feces) is also found among rural people (Amorozo and Gély 1988; Fleming-Moran 1975; Queiroz 1984). The relative Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 141 isolation of these populations has contributed to the maintenance of these ideas and to their continuing reliance on local specialists: curers, midwives, and ben- zedores—who heal by praying on (blessing) the ill person—among others. More- over, these populations exploit their environments very efficiently in search of therapeutic elements; in general they have a solid and long-standing knowledge about procurement and use of medicinal plants. In the past two or three decades, modern facilities have been brought to coun- tryside; in some areas, government medical services are locally available to rural populations. The introduction of modern medicine adds another option to the pluralistic base already established and does not eliminate people's use of tradi- tional medicine. Instead, in many instances, traditional and modern procedures are employed together (Alexiades and Lacaze 1996; Candido 1987; Elisabetsky and Setzer 1985; Wagley 1988). But the increasing influence of national culture certainly leads to changes in local medical settings; some kinds of traditional practitioners may disappear or their roles within community life may change (Queiroz 1980). Under the influence of cosmopolitan lifestyles and easy access to modern medical facilities, people may reduce their use of medicinal plants (Nolan and Robbins 1999), The aim of this work is to assess utilization of traditional forms of treating illness—chiefly the use of medicinal plants—by dwellers in rural communities that nowadays can also depend on available modern health care facilities and to discuss factors that may be affecting this use. STUDY AREA The study site is located in Santo Antonio do Leverger Municipality, Mato Grosso State, Brazil, on the left edge of The Cuiaba River, near the Pantanal and about 30 km by paved road south of the state capital, Cuiaba (Figure 1). The dominant natural vegetation is the cerrado (Brazilian savannah), which is, to a certain extent, altered by human activities. This region, formerly occupied by various indigenous peoples, including the Bororo (Viertler 1990), was settled in the beginnings of the eighteenth century by paulistas—descendants of Portuguese and aboriginal peoples (mainly from the Tupi group). They came from southeastern Brazil after the discovery of gold. African slaves, possibly of Banto origin (Bandeira 1988), were taken to work in the mines and in sugar cane spirits and sugar factories. The area near the Cuiaba River soon became important for supplying food for people working in mining. The occupation of these lands had been feasible only after the subjugation or expulsion of the local indigenous populations. When the gold mines were exhausted, the sugar industry began to finurish. In the nineteenth century, sugar and cane spirit production became the most important economic activity. That industry remained very influential in regional politics until the first half of the twentieth century. Later, its importance declined for number of reasons, among them the construction of technologically more ad- vanced factories in other regions (P6voas 1983). Mato Grosso State remained relatively isolated from cosmopolitan influences. Its economy stagnated until the mid 1970s. Since that time, the settlement of 142 AMOROZO Vol. 24, No. 1 —Q— STATE ROAD —™ COUNTRYSIDE ROADS N re 5 10 a A My == RIVERS km te 1d LAGOONS, LAKES FIGURE 1.—Study area. Source: Fundacéo de Pesquisa Candido Rondon, 1991. farmers from the southern region of Brazil and the establishment of agroindustrial centers in the state brought about changes in land use and social and economic conditions. In the study area, economy is based on subsistence agriculture, artisanal fish- eries, and production of manioc flour for home consumption and sale in the urban market. Modernization—provison of water and electric services, introduction of telecommunication facilities and the enhancement of tourism activities in the Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 143 area—is rapidly changing this picture, and contact with the national society is intensifying. Until a decade or two ago, traditional medical services included various types of specialists: midwives, curers, benzedores, and some practitioners of Afro-Bra- zilian sects, locally known as tenda or congdé. Since modern medical facilities have become more accessible, fewer of these specialists are providing services. For ex- ample, midwives are no longer training their replacements; at the time of this study, there remained only three or four very old women said to have been in demand as midwives in their time. Curers who have a deep knowledge of me- dicinal plants and who know how to diagnose ailments by feeling the patient's pulse are now rare, and none were to be found in the area. In contrast, there are many benzedores. They constitute the most important local representatives of tra- ditional medicine at present. Besides “blessing” the ill person, they also prescribe medicinal plants or allopathic remedies (Amorozo 1999). Also, a great number of people, mainly those over the age of forty, control a vast repertoire of knowledge about medicinal plants. In previous research, about 230 plant species employed for therapeutic use have been recorded (Amorozo 2002). Nowadays, state-run health services include a health care center and a hos- pital in the small town of Santo Antonio and three health care posts in the nearby rural communities. These services attend to the population at no cost and supply, depending on availability, some of the remedies prescribed by the physicians. The health center operates daily, offering services in general medicine, pediatrics, gy- necology, and dentistry. It also performs some simple laboratory analyses. The health posts operate only certain days of the week and offer more limited services. In town, there is also one private physician who attends to the local population for free and two commercial drugstores. METHODS Fieldwork has been done in the area since the early 1990s, focusing on the ethnobotany of medicinal plants (Amorozo 2002) and on disease concepts and treatment (Amorozo 1999). For the present study, two quarters in the town of Santo Antonio and two rural communities (Varginha and Barreirinho), approxi- mately seven kilometers away from the former, were selected; all these sites have easy access to government medical services. In June 1998, a survey was conducted in a sample of 44 households (20 in the rural zone and 24 in town). Households were selected by systematic sampling, using a sampling interval suited to ensure 20% of the total in the selected places (adapted from Bernard 1988). In general, interviews were held with female householders, but in some cases, male head or other members of the family also took part. Questionnaires recorded data about socioeconomic position, disease events, and the use of health facilities (both mod- ern and traditional) by family members in the six months previous to the research. Vernacular names of medicinal plants used in this period were also recorded. In addition, all the medicines present in the household at the moment of the inter- view were inventoried; plants and plant material that the interviewed person considered of medicinal use were recorded—whether growing near the house- hold, gathered fresh, or dried. The majority of species had already been collected 144 AMOROZO Vol. 24, No. 1 % 1 - Local health care center 100 - 2 - Local hospital 3 - Local private physician 80 + 4 - Medicine in the city 5 - Medicinal plants 60 - 6 - Benzedor 7 - Tenda 40 20 4 0 WJ 1 2 3 - 5 6 N Therapeutic resources Bhouseholds HM sampled population FIGURE 2.—Use frequency of therapeutic resources (number of households = 44; sampled population = 204). and identified (Amorozo 2002), so collection of botanical material was made, whenever possible, only for those plants that lacked voucher specimens.' Data about medical consultations in the area were also gathered in the health center files, encompassing 3320 consultations by general practitioners between June 1997 and June 1 An index of therapeutic resources utilization was calculated by age group as follows: TRUI = =" Nri/N, where Nri is the number of users of resource i in the age group, and N is the number of persons in the age group. This index rules out the problem of different numbers of people in the age groups, besides con- densing data about use frequency in a single figure, making comparisons among them easier. For statistical comparisons, Kendall correlation coefficient, chi-square (Siegel 1975) and binomial test (Ayres et al. 2000) were used. RESULTS The sample comprised 204 persons, on average 4.6 per household. Forty-one of the 44 households (93%) reported the use of at least one of the local state-run medical services during the six months previous to the research. Among the avail- able medical services, the local health care center and posts were the most visited by respondents and their families (86% of households and 44% of sampled pop- ulation). Plants were the most used traditional therapeutic resource; 93% of house- holds and 61% of sampled population used medicinal plants to treat at least one case of illness in this period (Figure 2). Plant-based recipes can be prescribed by anyone, but older, knowledgeable people and benzedores are the most sought for such remedies. Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 145 70+ @ Medicinal plants iat Z Local health care center 60-69 RY ///Jes | P3333 ae 1 sosee Ei Local private physician 50-59 eee [1 Local hospital si i Benzedor 2 40-49 Wi SY 3333 C) Medicine in the city © @ Zenda a) 30-39 < 20-29 10-19 09 0 0.5 1 L5 2 Beles 3 Utilization Index FIGURE 3.—Indices of therapeutic resource utilization by age group. Differences in the use frequency of therapeutic resources among age classes are to be expected, since people in the prime of life tend to be healthier. In fact, utilization indices (TRUI) were greater for children under 10 and for adults over 40. In the six months prior to research, 93% of children in this age group and 75% of adults over 40 used medicinal plants; they also went to health care center and posts more frequently than did the intermediate age class. Benzedores were most sought for treating children under 10 and, to a lesser extent, adults over 40 (Figure 3). Differences among age classes (0-9; 10-39; and 40 and older) by type of therapeutic alternative used were of no statistical significance for medical fa- cilities in Cuiaba, the local private physician and the local hospital. They were significant for the local health care center and posts (x? = 10.32, p<0.01), medic- inal plants (x? = 14.24, p<0.001) and benzedores (x* = 34.22, p<0.001).? For the comparisons between age classes, all the results were significant, except for the use of health care centers and posts by people under 10 and people aged 40 and older. People aged 10-39 used medicinal plants and sought benzedores significantly less than either other class (Table 1). Use of therapies of any kind by people over 14 was 63% higher among females than among males. majority of illnesses reported in the interviews received some kind : treatment (Table 2). In general, for ordinary indispositions or minor diseases, self- medication was the first step of treatment; it consisted mainly of medicinal plants (chas caseiros) and allopathic medicines that can easily be obtained eco sae ical prescriptions.’ If symptoms persisted or if there were a sapieisgarn ealt service available, people would seek further treatment, even in cases OF Common 146 AMOROZO Vol. 24, No. 1 TABLE 1.—Binomial test results for comparisons of therapeutic resources use between age classes. Age classt (z values) Therapeutic resource x2 1x23 22 Health center 4.06* 1.54 2.66* — plants 5.43* Paes ag Bes bi Benzedor 6.54* a.14* 3.08" ti = 0-9;2= 10-39; 3 = *p < 0.05; results not Soin in unmarked cell. ailments. For example, consultation with a health professional took place in more an 30% of the reported cases of influenza and cold symptoms. Likewise, data from the health care center files show that influenza symptoms accounted for almost 6% of consultations by general practitioners between 1997 and 1998. Con- sultations regarding respiratory and digestive systems ailments, intestinal infec- tions, and helminthiasis accounted for 37% of consultations in the same period. In the rural places studied, in the absence of the physician, who visits only two or three times a week, trained local health agents provided medicines and pri- mary health care. Medical professionals were sought, locally or in Cuiaba, for serious or acute problems (like heart attack, cerebrovascular accidents, trauma, pneumonia) and undiagnosed illnesses. However, this did not prevent patients from seeking alter- native means of healing. For example, two mothers whose young children had been diagnosed with pneumonia reported that they sought government-provided medical care. When the children returned home, the mothers gave them a locally popular tea prepared with some plants and guinea hen feathers in addition to the medicines prescribed by the medical staff. The combined utilization of modern health services and folk medicine ele- ments, mainly plants, occurred in 45% of reported cases of illness; resort only to modern medicine accounted for 29% of cases, whereas treatment exclusively with plants, and with plants and self-medication with allopathic remedies occurred in 18% of cases, mainly for influenza and cold symptoms (Table 2 With regard to the medicines inventoried by the time of the interview, allo- TABLE 2.—Health facilities sought in cases of illness. Cases (%) Type of health care (n = 121)* Medical professionals + traditional medicine (mainly plants) 44.6 Medical — 28.9 Medicinal plants 9.1 Medicinal plants + allopathic remedies 9.1 Allopathic remedies 3.3 Allopathic remedies + medical professionals 2.5 Others a7 No treatment 0.8 * If more than one person in a household had the same disease at the same time and treated the same way, only one case was counted. Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 147 TABLE 3.—Remedies in the households. Frequency = 44) Mean Range SD Total Allopathic — 93.2 S11 0-29 5.19 225 Medicinal plan 88.6 5.86 0-27 6.01 258 Commercial phytotherapics 18.2 0.27 0-2 0.62 12 Miscellaneou 34.1 1.29 0-10 1.66 34 Plants used the previous S 93.2 4.27 0-13 3.08 188 * Topical antiseptics, medicinal soaps, animal products. pathic remedies and medicinal plants were found in most households; the mean number of allopathic medicines per household was around five, whereas the mean number of medicinal plants was around six (Table 3). Some households had a greater number of patent medicines and/or medicinal plants, eventually provid- ing these facilities to neighbors in need. There was a significant positive correla- tion between number of medicinal plants reported to have been used in the six onths previous to research and the number of medicinal plants present in the fotiahsa (Kendall Tau = 0.40437, p<0.001) and also between this latter and the number of allopathic remedies present in the household (Kendall Tau = 0.261316, p<0.01). Allopathic medicines recorded were mainly analgesics and antipyretics, anthelmintics and remedies for digestive system ailments, antibiotics, vitamins and fortifiers, remedies for urinary tract, hypertension and tranquilizers; antisep- tics and antibiotic ointments for treating wounds were also common. About one fifth of the medicines, mainly anthelmintics, antibiotics, vitamins, and analgesics, was supplied free of charge by official health services. Plant species recorded in the households by the time of the visit numbered 111; about 83 species were mentioned to have been used in the previous six months. In total, about 120 plant species were recorded (Appendix 1). Little more than one-fourth of identified species were native to the cerrado or wet areas in the region (Lorenzi 1991; Pott and Pott 1994); cultivated plants accounted for about half of the total plant species. Herbs (37%) and trees (31%) were the most com- mon. Plants cited and/or found in more than 10% of sampled households (Table 4) were mainly cultivated species; at least half of them were exotic species intro- duced into the area. People aged 40 and older kept on average more than twice as many plants and plant material at home as the group under 40; likewise, mean number of cerrado plants and plant material was greater for this age group. Around 30% of these species were employed to treat illnesses or symptoms of the respiratory system; 20% were used for treatment of gastrointestinal ail- ments and intestinal worms. Urinary tract ailments, hypertension, disorders re- lated to female reproductive functions (problemas de mulher) and wounds were also treated by a variety of plants. Seven species were used to prevent or treat mau- olho (evil eye) and envy. It was common to employ the same plant for different diseases. 148 AMOROZO Vol. 24, No. 1 TABLE 4.—Plant species most frequently present/used in households (%, n = 44). Present in Used last house- six Plant species Utilization hold months Gossypium barbadense L. inflammation, ‘’female 23 Zz roblems” Lippia alba (Mill.) N.E.Br. tranquilizer 18 14 Coleus sp. stomach/liver 18 9 Justicia cf. pectoralis Jacq. influenza /cold 18 Z 1 aker stomach /liver 18 fe Eucalyptus sp. influenza/cold 14 18 Stachytarpheta cuore (L.C. Rich.) Vahl cough/influenza 14 11 Punica granatum throat inflammation 14 > Citrus X ia ie i influenza/cold 11 16 Petiveria alliacea L. mau-ol 11 9 Ruta graveolens L. “female problems” 11 5 Machaerium aculeatum Raddi cough/influenza 9 14 Polygonum hydropiperoides Michx. influenza /cold 7 16 Hes pen citratus Stapf. influenza/cold > 16 Hyptis sp influenza /cold z 14 DISCUSSION Data presented above show the combined employment of folk therapies, mainly plants, and modern medicine in almost half the reported disease cases. It is possible that minor complaints were underreported in the recall survey, which could lead to an underestimate of exclusive use of traditional therapies, normally employed in the first place in such cases, or no treatment at all (Brunelli 1987). However, data from the sampled households and from the health center files have shown that even for simple ailments, like influenza or colds, people turn to an official health system professional whenever they can. The spread of official health care facilities, in the present case, adds more treatment opportunities to a context of health care alternatives constructed after the contribution of diverse influences during the last three centuries. To the prac- tices taken by paulistas, already a blend of European medicine of the sixteenth century with medicinal knowledge of indigenous groups from the southeast, ther- apeutic knowledge of local natives was incorporated. A number of cerrado plants used today as remedies were considered eriibo (that is, magic or medicinal plant) for the Bororo Indians—e.g., Hyptis spp., Macrosiphonia longiflora, Anemopaegma ar- vensis, Protium heptaphyllum, Byttneria melastomifolia, among others (Albisetti and Venturelli 1962; Amorozo 2002; Hartmann 1967). Some plants, like Petiveria alli- acea, used by African slaves in the eighteenth century to poison their masters (Santos Filho 1977) were also included in this pharmacopoeia. ew therapeutic options brought about by growing access to modern medical facilities coincide with changes in traditional subsistence and production conditions as well as increased influence of urban ideology on rural areas. The impact of this scenario on former therapeutic practices and practitioners is dif- ferent in each case: for example, for the study area, midwives are no longer nec- Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 149 essary because access to institutional obstetric services is relatively easy. On the other hand, benzedores still play an important role, not fulfilled by modern med- icine. Culturally recognized ailments that are not considered by modern medicine, like quebrante, mau-olho (evil eye), and arca-caida, are treated exclusively by benzi- mento (blessing) (Amorozo 1999). This uniqueness partly explains the great pop- ularity benzedores enjoy in the area. But, benzedores are also sought when the ther- apeutic focus of an illness episode concentrates upon elements of modern medi- cine and in situations of emotional distress, which suggests psychological support plays an accessory role. For instance, the mother of a teenager who was experi- encing a delicate emotional situation reported she had taken her daughter to a psychologist or psychiatrist in Cuiaba and also to a local benzedeira. Medicinal plants were ordinarily employed to treat ailments like influenza, colds, gastrointestinal disorders and intestinal worms, sometimes in combination with allopathic medicines. Plants aimed at treating these ailments were the most frequently kept at home. The use of plants for these purposes is also common in other parts of Latin America (Bennett and Prance 2000; Trotter 1981) and Brazil (Amorozo and Gély 1988; Hanazaki et al. 1996; Silva-Almeida and Amorozo 1998). It is noteworthy that these ailments accounted for a great proportion of consultations in the health center as well. People aged 40 and older kept much more plants/plant material at home and made use of them most frequently. Plants were also commonly used to treat small children; young mothers commonly seek advice and plants from older relatives or neighbors. In sum, the majority of the sampled population used traditional as well as modern therapeutic facilities during the recalled period, taking advantage of all treatment opportunities available to them. The following description illustrates well this therapeutic syncretism. During earlier fieldwork in Santo Antonio, I collected an account of an illness event, where a woman about 40 years old had been struck by facial paralysis of unknown etiology. In the first few days of her disease until the onset of the paralysis, she complained of an intense headache, blurred vision, and weakness. She first tried some home remedies prepared with plants and a commercial phy- totherapic, all recommended by relatives and friends. Then she consulted a phy- sician in town and the next day, a neurologist in Cuiaba; both prescribed allo- pathic medicines. During the first days of her disease, she also sought a benzedor, who blessed her and prescribed medicines and baths prepared with plants and other ingredients. Her illness lasted about one month; during this time, she con- sulted one more physician, a homeopath in the city, three other benzedores, one physiotherapist, besides taking advice from a handful of relatives and acquain- tances. She used whatever medicine was prescribed to her, both internally and externally, which included several remedies prepared with thirteen plant species, six allopathic and two homeopathic medicines, among others. She and her rela- tives also prayed and made promises to the Roman Catholic saints in order to ensure her cure. Though a rather extreme example, this case can give us some insights about the way people in Santo Antonio deal with illness events. In a very short period of time, the patient turned to almost all the therapeutic options available to her, 150 AMOROZO Vol. 24, No. 1 traditional and modern. There was not time enough to evaluate the treatment efficacy before she changed to another therapy and also no complete shift would take place among treatments. Instead, the whole array of therapeutic resources at hand was exploited virtually simultaneously. Moreover, she made no distinction between different approaches by medical doctors, employing both allopathy and homeopathy. A remarkable feature of this situation is the interest of relatives, friends and neighbors in restoring the well-being of the diseased person. Everyone took pride in suggesting a recipe, based on their own experience, that of relatives or hearsay. That is, the community was involved in the illness episode, a situation character- istic of traditional rural settings. On the other hand, the array of therapeutic choices has increased, since people can either seek folk health agents and/or the new options made available by the broadening of official health services coverage. Feierman (1979), discussing the size and composition of therapy-managing groups of kinsfolk in Africa argues that, in periods of great social mobility, the extendend kinship network tends to cross various lines as related to instruction, social class, rural or urban residence, thus broadening the range of therapeutic preferences. Though one can not talk of ‘therapy-managing group’ proper in the present case, social and economic mobility in the area tends to have a similar effect. Also, in Santo Antonio, exchanges with society at large are more intense today due to increased tourism and the development of other economic activities in the region. Other authors have already stressed the predominance of pluralism in Brazilian rural societies undergoing social and economic changes due to ex- ternal influences (Candido 1987; Wagley 1988). It seems that in the case of Santo Antonio, this pluralism is sometimes ex- pressed as a sort of therapeutic opportunism. To a certain extent, the availability of free official health facilities encourages this behavior. But, medical services of- fered locally do not entirely meet people’s expectations. For example, there are many complaints about the frequent change of physicians in the health center and posts, which, according to users, hinders patients attendance. They also view clin- ical and laboratory examinations performed there as superficial, which makes them feel unsupported by medical personnel. Moreover, here, as in many parts of rural Brazil, provision of medical facilities for poor people is a common means of obtaining political advantage and is often used for electoral purposes. This can be done, for instance, by hiring private physicians to tend the population at no cost in times of election. On the other hand, people continue to turn to traditional ways of treating illnesses that are still available. But, as subsistence activities (like manioc flour production and fishing) become increasingly directed toward production for mat- ket and opportunities for wage labor increase, as is the case in the area, people begin to lose control of their working conditions. This limits their ability to follow traditional practices designed to treat illness and maintain health (Feierman 1979)—for example, to stay at home after intake of ‘hot’ medicines, to avoid hard work in agriculture fields during the hottest hours of the day, or certain foods in some situations. Medical personnel and cosmopolitan culture despise local conceptions of health and disease, labeling the traditional ways of healing as inferior, supersti- Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 151 tious and backward. In fact, traditional therapeutic means are challenged: inten- sification of exchanges with national society has introduced new health problems and situations they can not cope with. For instance, the resident physician re- ported that many patients he takes care of, especially women, presented symp- toms and complaints typical of stress. According to him, many of them were anxious about their children’s future, in view of new problems brought about by the ongoing socioeconomic changes in the area, exposure of youths to illegal drugs, prostitution, and so on. So this opportunistic strategy may be due to a feeling of unease about the effectiveness of the therapeutic options available to people, be they traditional or modern, which mirrors in fact their present sociocultural situation. In this changing sociocultural context, medicinal plants are widely used and benzedores are often consulted. In an illness event, like the one described above, they work to reinforce family and communal ties. Knowledge about plants and other therapies is shared and taught through advice and prescription of folk rec- ipes for the diseased person. It is possible that current socioeconomic changes will cause knowledge and use of medicinal plants to decrease in the studied communities for two reasons. First, younger people are becoming increasingly involved in occupations other than those of their parents. They either engage in local wage labor—agriculture, construction, or services—or migrate to Cuiaba, so at least part of the time they do not participate in the sphere of communal life where traditional knowledge is passed on. Second, land tenure and use in the area is rapidly changing. Cerrado vegetation has many species valued in folk medicine, but cerrado tracts, which were formerly common property, are now privately owned by foreigners. This places severe restrictions on their use by local people; moreover, natural areas are being replaced by cattle ranches and weekend houses. CONCLUSIONS Medicinal plant use in Santo Antonio is still an important and living tradition. Nevertheless, easy access to modern medicine, disruption of traditional know]- e transmission and change in land use, with destruction of natural vegetation, will ultimately lead to an erosion in both plant species availability and knowledge about them. Though people will probably continue to use medicinal plants, these will be more and more restricted to cultivated and exotic species, as is already the case for rural areas in more industrialized regions of the country These trends are difficult to reverse, but there are measures that could miti- gate their outcomes. For example, adoption by local official health services of native medicinal plants of known therapeutic efficacy together with measures to conserve cerrado patches and the cultivation of some of these species could con- tribute to maintain this rich lore and, eventually, achieve a better integration of the various ways of treating illnesses in the area. 152 AMOROZO Vol. 24, No. 1 NOTES ' Voucher specimens are deposited in Herbarium Rioclarense (HRCB). * Tenda and Medicine in the city categories were excluded from statistical tests because of great number of cells with zero. * These generally comprise analgesics, antipyretics, drops against blocked nose, but may also include medicines to be used strictly with medical attendance, due to lack of control of sales and employment of medicines ACKNOWLEDGMENTS I wish to thank people of Santo Antonio, for their collaboration, support and friend- ship; the Fundacao de Amparo 4 Pesquisa do Estado de Sao Paulo (FAPESP), for financial support through grant 96/8127-2; José Silvio Govone for statistical advice; Maria Inés de Mello Amorozo for helping with the English version; an anonymous referee for insightful comments and Naomi F. Miller for suggestions and the careful copy-editing of the manu- script. REFERENCES CITED Albisetti, C. and A. J. Venturelli. 1962. En- ciclopédia Bororo—vocabulario e etnografia. Publ. n. 1. Museu Regional D. Bosco/ Conselho Nacional de Pesquisas, Cam- po Grande, Vol. 1 Alexiades, Miguel N. ‘and Didier Lacaze D. 1996. FENAMAD’s Program in tradi- tional medicine: an integrated approach to health care in the peruvian Amazon. In Medicinal Resources of the Tropical For- est: Biodiversity and Its Importance to Hu- man Health, eds. Michael J. Balick, Elaine Elisabetsky, and Sarah A. Laird, pp. 341-365. Columbia University Press, New Yor Amorozo, Maria Christina de Mello. 1999, Medicina tradicional em Santo Antonio o papel dos benze- dores e suas habilidades. Revista Satide e Ambiente 2:48-66. . 2002. Uso e diversidade de plantas medicinais em Santo Antonio do Le- erger, MT, Brasil. Acta Botanica Brasilica 16(0): 189-203. Amorozo, Maria Christina de Mello and Gély. 1988. Uso de plantas medi- cinais por caboclos do Baixo Amazo- nas, Barcarena, Para, Brasil. Boletim do Museu Paraense Emilio Goeldi, Sér. Bot. 4(1):47-131. Ayres, Manuel, Manuel Ayres, Jr., Daniel Lima Ayres, and Alex Santos dos San- tos. 2000. BioEstat 2.0 Aplicagées estatis- ticas nas dreas da ciéncias biol6gicas e médi- cas. Sociedade Civil Mamiraud/CNPq, Belém. Bandeira, Maria de Lourdes. 1988. Territério negro em espaco branco. Estudo an tropol6- gico de Vila Bela. Editora Brasiliense/ CNPg, Sao Paulo. Bennett, Bradley C. and Ghillean T. Prance. 000. Introduced plants in the indige- nous Se of northern South rica. Economic Botany 54:90-102. see H. Russell. 1988. Research Methods in aay Anthropology. Sage Publica- iO. c., Newbury Park. Brunelli Gilio. 1987. Des esprits aux mi- es—santé et societé en transfor- iat chez les Zoré de l’Amazonie brésilienne. MSc. Thesis, (Anthropolo- gie), Faculté des Arts et des Sciences, Université de Montré Candido, Antonio. 1987. Os parceiros do Rio een Livraria Duas Cidades, Sao Pau- pe Anthony C. 1971. The differential use of medical resources in developing countries. Journal of Health ed Social Be- havior 12 (September):226-23 Crandon-Malamud, Libbet. faci "From the Spring/Summer 2004 Fat of Our Souls. Social Change, Political Process, and Medical Pluralism in Bolivia. University of California Press, Berkeley. Elisabetsky, Elaine and Rachel Setzer. 1985. Amazonia. In Studies in Third World So- cieties, ed. Eugene P. Parker, pp. 242- 278. Williamsburg. Feierman, Steven. 1979. III. History of plu- ralistic medical systems. Change in Af- rican therapeutic cape Social Science & Medicine 13B:277-2 Fleming-Moran, Millicent. 1975. The Folk View of Natural Causation and Disease in Brazil and Its Relation to Traditional mike Seana MSc. Thesis, Univer- sity of Flori Hamnett, Michael P. and John Connell. 1981. Diagnosis and cure: the resort to traditional and modern medical practi- tioners in North Solomons, Papua New Guinea. Social Science & Medicine 15B: 489-498 Hanazaki, Natalia, Hermogenes de Freitas Leitao-Filho, and Alpina Begossi. 1996. Uso de recursos na mata atlantica: o caso do Pontal do Almada (Ubatuba, Brasil). Interciencia 21 (6):268-276. Hartmann, Tekla. 1967. A nomenclatura bo- tanica dos Bororo (Materiais para um en- saio etno-botanico). Instituto de Estudos Brasileiros, 6, Sdo Paulo. Holanda, Sérgio Buarque. 1994. Caminhos e fronteiras, 3rd ed. Companhia das Le- tras, Sado Paulo. The comparative study of medical “haoeas as changing social _— Social Science & Medicine Medicinas tradicionais e medicina ocidental na Amazonia, ed. Dominique Buchillet, pp. 125-133. Edigdes CEJUP, Belém. McGrath, B.B. 1999. Swimming from island to island: healing practice in Tonga. Medical Anthropology Quarterly 13(A4): 483- Nolan, Justin and Michael C. Robbins. 1999. JOURNAL OF ETHNOBIOLOGY 153 Cultural conservation of medicinal lant use in the Ozarks. Human Orga- Pgs: 58(1):67-72 t, Arnildo and Vali J. Pott. 1994. Plantas oro P Pantanal. Centro de Pesquisa Agro- pecudria do Pantanal/Servico de Pro- ducao de Informacao, Brasilia. Pévoas, Lenine. 1983. O ciclo do agticar e a politica em Mato Grosso. Instituto Hist6- rico e Geografico de Mato Grosso, Cu- iab. iaba. Queiroz, Marcos de Souza. 1980. Feitico, mau-olhado e susto: seus tratamentos e prevengoes na aldeia de Icapara. Reli- giao e Sociedade 5:132-159. Queiroz, Marcos de Souza. 1984. Hot and cold classification in traditional Iguape medicine. Ethnology 23(1):63-72. Santos Filho, urgo. 1977. Histéria geral da medicina brasileira. Vol. 1. HUCITEC/ rade 2 Universidade de Sao Paulo, Sao nies Sidney 1975. Estatistica nao-paramé- a (para as ciéncias do comportamento). Editora McGraw-Hill do Brasil, Sao o Amorozo. 1998. Medicina olen no distrito de Ferraz, unicipio de Rio Claro, estado de Sao ac Brazilian bal of Ecology 2:36- Paulo. susie tiecoraniag Maria de Fatima and Maria Christina de Mell dies Bradley P. 1986. evening edica ee from the United States-Mexico border. Journal of Ethnopharmacology 4 Viertler, Renate Brigitte. 1990. A duras pe- st6rico ~ relagdes entre indios bey e * civiliza no Mato Grosso. “Sada Sao Paulo ip Fanos em 16). Geo, “ Charles. 1988. Uma comunidade zénica: estudo do homem nos Trépicos, es : Clotilde da Silva Costa. Itati ea “a de Sao Paulo, Sao Paulo. from Amazon Town: a Study ofh Man in the Tropics). APPENDIX 1.—Plant species present in the peer “es the last six months. VN = voucher number; it refers either to herbarium (HRCB) Mic fe or collector number (CA = M.C. Amorozo; agg = rva n.i. = not identified; il. = identified, but not collected; LF = life form; tre = tree; srb = schrub; sbt = shrublet; hrb = i = epiphyte; vin = vine; hpar = hemiparasite; # = purchased; c = cultivated; s = spontaneous; * = most common species; ioe = pedis eu FAMILY / Species Vernacular name VN LF Habitat / occurrence C/S ACANTHACEAE Justicia cf. pectoralis Jacq.* anador, anador-de-planta, = CA277 hrb homegarden c aspirina ALISMATACEAE Echinodorus sp. chapéu-de-couro CA310 hrb swamp s ALLIACEAE Allium sativum L. alho il. hrb cultivated field in river c bank AMARANTHACEAE Alternanthera brasiliana (L.) Kuntze terramicina 30790 hrb homegarden c ANACARDIACEAE a hae pgm L. caju il. tre homegarden c Mangifera indica L. manga, mangueira il. tre homegarden c APIACEAE Pimpinella sp. erva-doce — hrb # — APOCYNACEAE Catharanthus roseus (L.) Don. bom-dia 30791 hrb garden c Hancornia speciosa Gomez mangabeira-mansa 30805 tre cerrado s ARAC Dieffenbachia sp. comigo-ninguém-pode il. hrb garden, homegarden c ARECACEAE Acrocomia aculeata (Jacq.) Lodd. ex Mart. boucaiuveira il. tre cerrado s OZOUOWYV I ‘ON ‘$Z ‘TOA APPENDIX 1—(continued) FAMILY /Species Vernacular name VN LF Habitat / occurrence c/S ASTERACEAE Acanthospermum hispidum DC. cabeca-de-garrotinho 30707 /30709 hrb homegarden, weedy s Artemisia absinthium L losna, nov6nica-macho, CA432/CA313 hrb homegarden c noz-vémica Artemisia verlotorum Lamotte artemije 30664 hrb homegarden c Bidens pilosa picdo-pre 30838 /30839 hrb homegarden s Dendrathema grandifolia (Ram.) Tzv. camomila-branca, camomi- 30677 hrb homegarden c la-amarela ni. camomil hrb homegarden c Vernonia brasiliana Druce assa-p 30705 srb fallow s Vernonia condensata Baker* boldo, bordo, sara-tudo, 30830 /30840 tre homegarden c cura-tudo BIGNONIACEAE Crescentia cujete L. CA409 tre homegarden c See ni gee a Mart. pé-de-anta te srb cerrado, homegarden s/c Tabebui a (Manso) B paratudo 30755 tre cerrado, homegarden s as Bixa orellana L. urucum il. tre homegarden c BORAGINACEAE Cordia insignis Cham. calcao-de-velho 30747 srb cerrado s CACTACEAE Pereskia cf. grandifolia Haworth ora-pronobis CA417 tre homegarden c os hanced eiaaeg Bauhinia sp. pé-de-boi, unha-de-boi 30851 tre cerrado s i 38 desvauxii ee Killip sene 30766 sbt cerrado S Hymenaea courbaril L. v. jatoba-mirim, jatoba CA289 tre cerrado s stilbocarpa Senna occidentalis v) Link fedegoso 30849 sbt fallow s Tamarindus indica tamarindo, tamarino 30763 tre homegarden s AOOTOISONHLA JO TYNANOL £00z Jourung /Sutids ccT APPENDIX 1—(continued) FAMILY /Species Vernacular name VN LF Habitat / occurrence C/S CAPRIFOLIACEAE Sambucus australis C. et S. sabugueiro 30760 srb homegarden c CARICACEAE Carica papaya L. mamdao, maméo-macho il. tre cultivated field, home- c garden CARYOCARACEAE Caryocar brasiliense Camb. pequizeiro, picueiro 30666 tre cerrado s CECROPIACEAE Cecropia pachystachya Trec embatiba 30681 tre homegarden s CHENOPODIACEAE Chenopodium ambrosioides L. erva-de-santa-maria, santa- il. hrb homegarden s maria CONVOLVULACEAE Ipomoea batatas (L.) Lam. batata-doce il. hrb cultivated field c COSTACEAE Costus arabicus L. cana-de-macaco, caninha- CA44 hrb homegarden s/c de-macaco CUCURBITACEAE iffa sp. buchinha, buchinha-paulis- il. vin # ta Momordica charantia L. sao-caetano, melao-de-sao- 30820 vin homegarden, fallow s caetano DILLENIACEAE Curatella americana L. lixeira 30808 tre cerrado s EUPHORBIACEAE maesyce caecorum (Boiss.) Croizat sete-sangrias 30803 hrb cerrado s 9GT OZOYOWV L ‘ON ‘FZ ‘TOA APPENDIX 1—(continued) FAMILY /Species Vernacular name VN LF Habitat / occurrence C/S Jatropha gossypiifolia L. pinhao-roxo 30797 srb homegarden c ni. quebra-pedra hrb cultivated field s Phyllanthus orbiculatus L.C. Rich quebra-pedra 30811 hrb cerrado s Phyllanthus stipulatus (Raf.) Webster quebra-pe CA405 hrb homegarden s Ricinus communis L. mamona, 6leo-de-ricino 30798 /30800 srb homegarden S Seem m dasycarpum aa Yakovl. genciana, quina-genciana CA294 srb cerrado s Cais cajan (L.) Mills feijao-andu CA355 srb homegarden c Dipteryx alata Vog. cumbaru 30813 tre cerrado S cf. Eriosema campestre Benth. bacimo 30852 hrb cerrado s Machaerium aculeatum Raddi* espinheira, espinheiro CA269 tre swamp S FLACOURTIACEAE Casearia sylvestris Sw. cha-de-frade CA5 srb/tre cerrado s LAMIACEAE oleus sp.* boldo, bordo il. srb homegarden c Cunila microcephala Benth. poejo, apoejo 30687 hrb homegarden c Hyptis crenata Pohl. eR sa Piste hortelé-do-campo, hortela- — 30842/30674 sbt cerrado s -nepine St.Hil. ex Be da-vargem Hyptis suaveolens P tapera-velha 30688 hrb fallow, ruderal Ss Leonotis nepetacfolia | (L.) R. Brown cordao-de-sdo-francisco 30689 hrb homegarden, weedy s Mentha arvensis L. var. piperascens Malinv. — vick 30671 /30675 hrb omegarde c Mentha sp. hortelazinho, hortela hrb homegarden c Ocimum gratissimum L. alfavaca 30822 /30823 sbt homegarden ¢ ni. hortela-gordo CA422 hrb homegarden c LAURACEAE Persea americana Mill. abacate il. tre homegarden c LILIACEAE Sansevieria sp. espada-de-sao-jorge ib hrb garden, homegarden c LOGANIACEAE Strychnos pseudoquina St. Hil. quina CA251 tre cerrado s #007 Jeutuns /Burids AOOTOISONHLA JO TVNUNOL ZS1 APPENDIX 1—(continued) FAMILY /Species Vernacular name VN LF Habitat / occurrence c/s LORANTHACEAE Psittacanthus calyculatus (DC.) C. Don enxerto-de-passarinho 30826 hpar homegarden, wetlands Psittacanthus sp. enxerto-de-passarinho CA448 homegarden, wetlands Ss LYTHRACEAE Lafoensia cf. replicata Pohl. mangabeira-braba MC48 tre cerrado s MALPIGHIACEAE Camarea ericoides St.Hil. arnica 30668 hrb cerrado s Heteropterys pannosa Griseb. n6-de-cachorro, raiz-de- MC3 sbt cerrado s santo-antonio Malpighia glabra Linn. acerola il. srb homegarden c MALVACEAE Abutilon sp. 1 marva, mave 30738 /30739 hrb homegarden s/c Abutilon sp. 2 mave 30737 hrb homegarden s/c Si saplaii barbadense L.* algodao, algodao-de-casa_ =: 30735 srb homegarden c sem nome (no name) il. hrb homegarden s SS aricie Tibouchina cf. clavata (Pers.) Wurd. cibalena CA445 srb homegarden c MIMOSACEAE Stryphnodendron adstringens (Mart.) barbatimao CA264 tre cerrado s Coville MONIMIACEAE Siparuna guianensis Aubl negra-mina 30740 tre cerrado s CEAE Brosimum gaudichaudii Tréc. —* CA441 tre cerrado s Dorstenia asaroides Gardner caiapia CA7 hrb cultivated field, cerrado s MUSACEAE Musa X paradisiaca L. bananinha, bananeira il. hrb c cultivated field, home- garden 8ST OZOYOWV I ON ‘¥Z TOA APPENDIX 1—(continued) FAMILY /Species Vernacular name VN LF Habitat / occurrence C/S MYRTACEAE E tus sp.* eucalipto CA308 tre street c Eugenia unifor L. pitanga CA377 tre homegarden c Psidium guajava L. goiaba, goiabeira, goiaba- _i.l. tre oe cultivated c e Syzygium cumini (L.) Skeels jambo CA431 tre homegarden c OXALIDACEAE Averrhoa carambola L. carambola 30700 tre homegarden c ORACEAE Passiflora edulis Sims maracuja il. vin homegarden c PHYTOLACCACEAE Petiveria alliacea L.* guiné 30685 hrb homegarden ¢ PIPERACEAE Piper tuberculatum Jacq. jaborandi, jaguarandi 30669 / 30670 srb cerrado, homegarden S ACEAE Coix lacryma-jobi L. conta-de-nosso-senhor, 30711 hrb homegarden c erva-de-santa-maria Cymbopogon citratus Stapf.* capim-cidrei CA300 hrb homegarden c Saccharum officinarum cana-de-acuica il. hrb cultivated field c Vetiveria zizanioides (L.) Nash capim-santo 30710 hrb homegarden c POLYGONACEAE Polygonum hydropiperoides Michx.* erva-de-bicho 30686 hrb swamp s PUNICACEAE Punica granatum L.* roma il. tre homegarden c RUBIACEAE Rudgea viburnoides (Cham.) Benth. douradinha, erva-mula 30742 /30782 srb cerrado s FOOT Jews /Butds XDOTOISONHLA JO TVWNANOL 6ST APPENDIX 1—(continued) FAMILY /Species Vernacular name VN LF Habitat / occurrence C/S RUTACEAE Citrus sp. 1 lima-de-umbigo CA341 tre homegarden c Citrus sp. 2 lima-das-pecas CA343 tre homegarden c Citrus X ‘aurantiifolia (Christm.) limao-galego, limo, li- CA291/CA339/ tre homegarden c Swingle/Citrus x limon (L.) Osbeck mao-taiti Citrus auran te um L.* laranja CA340/CA342 tre homegarden c Ruta graveo ruda it. hrb homegarden c Se niLAeACiA Scoparia dulcis L. vassourinha 30727 /30728 hrb homegarden S SIMAROUBACEAE Simaba trichilioides St.Hil. calunga 30722 sbt cerrado s SOLANACEAE Solanum cf. comptum Morton joa 30719 sbt cerrado, swam: s ni. beladona CA433 srb homegarden c STERCULIACEAE Guazuma sp. chico-magro 30726 /30720 tre cerrado, homegarden s cagmatae Lantana cam cambara 30691 sbt fallow s Lippia alba Mill) N.E.Br.* erva-cidreira, cidreira-de- | CA407/CA408 sbt homegarden c rama Stachytarpheta cayenensis (L.C. gerbao, gervao MC2 sbt homegarden s Rich.) Vahl.* VOCHYSIACEAE Vochysia divergens Pohl. cambara 30846 tre cerrado, near swamp s ZINGIBERACEAE Alpinia zerumbet (Pers.) Burtt & Smith colonia 30693 hrb homegarden c 09T OZOUONV I ON ‘FZ TOA APPENDIX 1—(continued) FAMILY /Species Vernacular name VN LF Habitat/occurrence C/S INDETERMINATE iL babosa — hrb c a cravo — tre c 3, cancerosa — — c 4. douradinho — — 5 espada-de-nossa-senhora _— hrb c 6. jacar — tre S ce jequiti — tre s 8. quina-do-morro — tre? — FOOT JeuTUNs /ButIds KOOTOISONH.LA JO TVNUNOL T9L POE 3 Journal of Ethnobiology 24(1): 163-172 Spring/Summer 2004 BOOK REVIEWS DARRON A. COLLINS Book Review Editor Food in the Ancient World from A to Z. Andrew Dalby. 2003. Routledge, London and New York. Pp. xvi + 408. $65.00 (hardcover). ISBN 0-415-23259-7. Andrew Dalby has rapidly carved a name for himself as the expert on the ethnobiology of ancient Greece and Rome, and one of the leading experts on food during that period. Here he assembles a monumental amount of scholarship into a dictionary that will be a basic reference for a long time to come. This is the sort of book that one expects to see being produced by a whole team of senior scholars; to have such a work produced by one relatively young scholar is nothing short of incredible. Compiling such a dictionary for the Maya peoples, for example, would be hard enough; think of the task when one has to deal with thousands of original sources (ranging from vast books to scattered papyri and ostraca) and tens of thousands of secondary sources. Yet Dalby wears his knowledge lightly. He has made the entries notably readable, with wry and humorous asides in many cases. Sometimes these are his own; more often they are quotes or tags from the classic comic writers. The entries cover foodstuffs, including those rare and exotic; ancient food writers and medical authorities, again including even the most obscure; and plac- es noted for their food and wine. Dalby is very conscious of the enormous im- portance of that latter commodity in ancient civilization and has long and thor- ough entries on all aspects of wine culture, from manufacturing techniques to famous vintages. Research in ancient-world ethnobiology has made great progress lately, thanks to archaeology and to scholarship such as Alan Davidson's work on Med- iterranean seafood, but Dalby still had to sort through and evaluate a vast amount of nonsense, including early guesses at identification of species. He also has to tread lightly around some still-fiery controversies, such as the one over spices in the ancient world. J. I. Miller’s Spice Trade of the Roman Empire (1969) assumed that cinnamon, cardamom, pepper, and the like were flowing into Rome in vast amounts. This position was sharply attacked by Patricia Crone in Meccan Trade and the Rise of Islam (1987); Crone held that words like kinnamon and kardamomon referred originally to local Mediterranean plants, being transferred only very late to Asian commodities, and that the spice trade never amounted to much. Dalby follows recent archaeological findings in taking a cautious middle ground, some- what closer to Miller. Readers may wish he had said more about the issue, but his handling of it is delicate and sensible. ; oe “Even Homer nodded,” as the ancients said, and there are mistakes in this book. Most are trivial errors in nomenclature. Significantly, Gallus gallus is labeled as “Gallus gallinaceus” on page 83. On page 169, the hazel hen is called a “sand- 164 BOOK REVIEWS Vol. 24, No. 1 grouse,” which it is not. More serious is the entry on the banana (p. 44): “fruit of a palm domesticated in New Guinea... .’’ The banana is, of course, not a palm, and only a rather obscure species (Musa fehi CLG Bertero ex E. (DE) Vieillard) was domesticated in New Guinea. The common banana (Musa x paradisiaca) is an artificially created hybrid probably ‘‘stemming” from Malaysia. No doubt an ex- pert in the classics would find more errors than I have, but at the very least this is a notably reliable work. Archaeobiologists will find this book particularly useful. The archaeological information is reasonably up-to-date, though not always. (Panicum millet is said to have been domesticated in the Caucasus, p. 218; most recent evidence supports China as the source, but the question is still very open.) Archaeologists will want to supplement Dalby’s book with site reports. Ethnobiologists frequently need references on the ancient world, if only be- cause students and the public are often aware of, and very interested in, ancient Greece and Rome. This is clearly the reference of choice, and is a very worthwhile book to add to one’s library. E. N. Anderson Department of Anthropology University of California Riverside, CA 92521-0418 REFERENCES CITED Crone, Patricia. 1987. Meccan Trade and the Rise of Islam. Princeton University Press, Prince- on, New Jersey. Davidson, Alan. 2002. Mediterranean Seafood: a Comprehensive Guide with Recipes, 3rd ed. Ten Speed Press, Berkeley. Miller, J. I. 1969. The Spice Trade of the Roman Empire. Oxford University Press, Oxford. —— s Palates of the Past: An Ethnoarchaeological Study of Crop Cullti- vation and Plant Usage in India. Seetha Narahari Reddy. 2003. Ethnoar- Aaantoered Series 5. International Monographs in Prehistory, Ann Arbor, gan. Pp. xi + 175; photographs, glossary. $65.00 (hardcover), ISBN 1-879621-37-1. $47.50 (paper), ISBN 1-879621-36-3. It is a pleasure to read a well-written book that so masterfully welds together archaeology and ethnography about human and animal plant use. The focus of this volume is millets, Harappan sites in the state of Gujarat in western India, and the complementary interactions between pastoralism and agriculture. Al- though ethnoarchaeological modeling of crop processing for archaeological ap- plication premiered with the outstanding studies on wheat and barley by Hillman and Jones in the 1980s, nowhere before has anyone looked at millets. Millets are relatively small-seeded annuals with growth habits that range from strong- stalked, compact-headed inflorescences (Type A crops) to multiple weaker stalks with looser panicles (Type B crops). Domesticated millets originated in southern and eastern Asia as well as Africa, and wild millets are found just about world- Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 165 wide. I believe Reddy’s observations on millets have general application for ar- chaeological assemblages that include small-seeded annual grasses. Reddy blends together three lines of evidence: her ethnographic study of crop cultivation and processing in India, the paleoethnobotanical record from two Har- appan sites, and carbon isotope studies of dirt and cattle bones from these two sites (millets are C4 plants, whereas most of the associated weeds in the area are C3). Her goal is to model the interplay of crop processing with end product use (e.g., human food, animal fodder, or both) to aid in the interpretation of archae- ological evidence. Reddy’s ethnographic observations focused on intensive summer monsoon cultivation of Type A crops (Pennisetum typhoides (Burm. f.) Stapf & C.E. Hubb. and Sorghum bicolor (L.) Moench) in Gujarat in western India, and extensive winter riverbank opportunistic cultivation of the Type B crop Panicum miliare Lam. in Andhra Pradesh in southeastern India. Whereas Type A, thick-stalked and com- pact-headed crops (P typhoides, S. bicolor, and Eleusine coracana Gaertn.) are har- vested one to three stalks at a time, a strategy that selects against inclusion of weeds, weeds are frequently included in the group harvest of Type B thin-stalked crops (e.g., P miliare and Setaria spp.). Other factors that contribute to the inclusion or exclusion of weeds at harvest include the habit of the weed (prostrate vs. erect) and, of course, whether the weed ripens at the same time as the crop. Reddy presents an interesting twist by pointing out that even Type B crops can be har- vested weed-free when, as in her case study in Andhra Pradesh, the weeds are prostrate and below the level at which the crop stalks were cut for harvest. Where- as the previous models by Hillman (1984) and Jones (1987) of when and how weed seeds may be processed out of a crop focused on the three weed charac- teristics of winnowability (ratio of seed surface area to weight), sievability (seed size), and seed headedness, Reddy found the interplay of seed size (small vs. big), headedness (a continuum from free to headed), and weight/aerodynamics (a combination of seed weight, shape, and aerodynamic appendages) to be more germane to the crop-processing strategies. By concentrating on the properties of the weed seeds themselves, separate from any specific crops or “stages’’ of crop processing, Reddy provides a model that is transferable for use with other crop types. Her handling of multiple combinations of crop/weed characteristics or har- vest/processing/use choices is skillful, leading the reader through a labyrinth of possibilities rather than presenting an oversimplification of human behavior that marks too many archaeological studies. Not only were crops represented by their end use (e.g., as human food, animal fodder, or both), but also as byproducts that could contribute to animal fodder. Products and byproducts may be used green, fresh, or dry; they may be stored or not at various points along the processing continuum; and, likewise, are suit- able or not for trade or exchange. Choices made about the location of each of the processing steps affects the likelihood of exposure to fire (resulting in preserva- tion) or the likelihood of even finding that location for archaeological study. In Chapter 4, Reddy supplies illustrated chart models for what products, byprod- ucts, and archaeological findings may be expected at each step along the way of processing either a Type A (thick-stalked, single-headed) or Type B (thin-stalked, loose panicle) crop. Incidentally, she points out that Chenopodium album L. is pro- 166 BOOK REVIEWS Vol. 24, No. 1 cessed as a Type B crop in India: Chenopodium is another widespread genus fre- quently consumed by humans or supplied to animals as fodder. To round out her understanding of implications for archaeological interpretation, Reddy conducted experimental charring of millets to investigate the likelihood that seed stalks could be preserved. What is used for animal fodder may depend on the season or ripeness (green vs. dry), and certainly reflects preferences by species (e.g., whether it enhances milk production or not). Archaeologically, plants used for animal fodder may be represented both in dung (used as fuel or as plaster) and in the isotopic ratios in animal bone. Reddy briefly reports on her pilot study of dung and hearth samples at two archaeological sites in Gujarat. Additionally, she analyzed a total of nine sediment samples from the two sites and 22 cattle bones for an indication of C3 vs. C4 plants. She found that the general background vegetation at both sites yielded C3 soil, as expected. The analysis of cattle bones, however, was mostly unsuccessful and the results inconclusive. In Chapter 7 Reddy models animal feeding for domesticated herbivores such as cattle, pigs, sheep, and goats. Reddy then applies her models to the two sites—Babar Kot, a very late Mature Harappan site dating to 2200-2050 B.C., and Oriyo Timbo, a Late Harappan site dating to 1900-1800 B.C. She makes a strong argument that Babar Kot was a substantial settlement practicing year-round sedentary agriculture focused on summer and winter crops of millets, legumes, and oilseeds and on animal hus- bandry. Millets were grown for human consumption. Although millet byproducts likely were used for fodder, she could not prove it. Oriyo Timbo, on the other hand, represents a seasonal settlement used by seminomadic or semisedentary pastoralists who may or may not have been growing their own millet. However, they certainly were bringing highly processed millet to that location for human consumption. Both sites reflect the complementarity of agriculture and pastoral- ism in the Harappan sphere of influence. I highly recommend this book—to archaeologists working in Africa or Asia; to ethnographers working with agricultural or pastoral societies; to paleoethno- botanists who work with small-seeded annual crops or weeds; to social scientists who study agriculture and pastoralism; and to students of ethnoarchaeology. It is a well-written, outstanding example of how ethnographic studies may enhance archaeological interpretation. The text lays out a very complex web of interrela- tionships and weaves them together to present one of the most realistic recon- structions I have seen of the complex daily decisions that had to be made in the past. I warn readers that the models with drawings can only be appreciated by reading the text. The volume is nearly free of printing/editing errors, and I was puzzled by only one statement by the author, classifying Brassica (mustard) as a legume (pp. 113, 122, and 128). My original interest in this volume was sparked because it reported on Oriyo Timbo, a site where I worked in 1981-1982 (Reddy reports on the 1989-1990 season). During 1982-1983 excavations at nearby Rojdi, I observed seasonal cul- tivation of a river bottom when drought reduced the river to scattered puddles and pools. What struck me most at the time was that what had been a large, seemingly permanent river the year before was now a naked river bed. What Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 167 strikes me now is that Reddy has provided a model for how even such ephemeral cultivation practices may be reflected in the archaeological record. Gail E. Wagner University of South Carolina Department of Anthropology Columbia, SC 29208 REFERENCES CITED Hillman, G. 1984. Interpretation of Archaeological Plant Remains: the Application of Eth- nographic Models from Turkey. In Plants and Ancient Man: Studies in Paleoethnobotany, eds. W. van Zeist and W. A. Casparie, pp. 1-41. A. A. Balkema, Rotterdam. Jones, G. 1987. A Statistical Approach to the Archaeological Interpretation of Crop Pro- cessing. Journal of Archaeological Science 12:311-323. Nature, Culture, and Big Old Trees: Live Oaks and Ceibas in the Landscapes of Louisiana and Guatemala. Kit Anderson. 2003. University of Texas Press, Austin. Pp. 183 + photographs. $19.95 (paper). ISBN 0-292-70212-4. Anyone interested in big trees who has visited Tikal in the Guatemalan Petén will immediately recognize the towering ceiba (Ceiba pentandra (L.) Gaertn.) on the book’s front cover. Loving big trees and having conducted ethnobotanical research in Guatemala as a graduate student at Tulane (in New Orleans), I was initially drawn to the book by those associations. I became curious almost im- mediately: why these species and not others; what kind of conclusions could the author possibly draw; will the book be theoretically useful or “just” a good read? The book is divided into five chapters and neither the first nor the second chapter did much to satisfy my curiosity. Chapter One, Introduction: Human Tree Relationships, served its purpose in detailing where the text would go, summa- rized nicely how big trees have shaped human imagination, and outlined how trees might shed some light on the trajectory of culture. Chapter Two, Dances with Trees: Notes from the Field, is an amalgam of stories from the author's fieldwork in the two locales. The black and white photographs are many and quite excellent, as are many of the stories. However, I finished reading the chapter feeling that the descriptions were somewhat shallow and anecdotal, especially on the side of the Guatemalan ceiba. The chapter's concluding section, Patterns and Questions, did not answer the question that kept coming to me: “Where is this going to go?” Chapter Three, Natural History: The Secret Lives of Ceibas and Live Oaks, is very well done. The botany, ecology, and natural history was a real pleasure to read—very accurate, detailed, and nicely written. The comparative maps and as- sociated discussions documenting the ‘‘natural’’ versus the “cultural distribu- tions of the two trees were very revealing and insightful. As in previous chapters, the photographs were excellent, telling, and perfectly parallel with the text. Hu- man behaviors that have affected the distribution of the trees were elucidated and the meaning of the text and the reason for its authorship started to emerge. 168 BOOK REVIEWS Vol. 24, No. 1 Chapter Four, Cultural History: How Trees Develop Character, was also ex- cellent. Again, the information on the Louisiana live oaks (Quercus virginiana P. Mill.) was more detailed. The ceiba received approximately 16 pages of text and the live oaks about 26. Nevertheless, the discussion on the cultural associations of the live oaks was wonderful in terms of detail, breadth, and writing. And it was in this chapter and especially in the chapter’s closing section where the two trees’ similar ‘’role” in the cultural context was highlighted: ‘Within the cultural landscape, live oaks and ceibas occupy some remarkably similar niches.” That in itself may not be very remarkable, but the degree and extent to which this is true is quite extraordinary. The discussion on the control that land tenure, town struc- ture, and architectural style exerts on the life histories of these trees is very en- gaging. But the final chapter, Coda: Charismatic Megaflora and the Making of Landscapes, would be the one that would decide whether or not the text would come to any striking or at least significant conclusions. On page 154 of the final chapter the author admits, ‘Sweeping generalizations concerning cultural attitudes toward these big trees are pointless, I found.” Nev- ertheless, by her examination of these fantastic trees, Anderson uncovers some interesting, generalizing points: trees are not passive but active participants in the creation of landscapes; trees acquire symbolic meaning over time; individual humans, through tree planting and care, can play an enormous role in the trans- formation of landscapes; and, certain trees achieve favored status in particular ways. These insights are valuable. They also help answer some of my concerns relating to the book’s purpose. Simply stated, the two trees share many of these characteristics listed above in common—they play a similar “role” in relation to humans. But these characteristics were discovered post facto and do not clarify why such comparative investigations were initiated. Also, the chapter’s penulti- mate section, ‘‘Nature and Culture,” does not treat this widely discussed dichot- omy with enough detail. The wealth of information on historical ecology, which could shed significant light on the question, generally speaking and where these two trees are concerned, is not addressed. Anyone who loves trees and loves to think about the meaning of trees should absolutely consult Anderson’s text. The book is informative, very well written, replete with many superb photos, and reaches some interesting conclusions. As for my first question: even though I still do not fully understand the author's decision to write about ceiba and live oak, and the book left me wanting more detail and analysis, its many charms compensate for these criticisms. Darron Collins Director of Ethnobotanical Field Studies The Amazon Conservation Team Arlington, VA 22203 What Place for Hunter-Gatherers in Millennium Three? Thomas N. Headland and Doris E. Blood (eds.) 2002. SIL International and International Museum of Cultures Publications in Ethnography 39. Dallas, Texas. Pp. 129. $19.00 (paper) ISBN 1-55671-132-8. Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 169 This edited volume considers the plight and survival prospects of the Agta hunter-gatherers of the Philippines. It also asks whether any hunter-gatherers will remain intact as autonomous ethnic groups through the twenty-first century. The volume contains seven contributed chapters, photographic documentation of the Agta during the past thirty years, and a complete bibliography of all publicly avail- able published works on the Agta. The contributed chapters discuss the current situation of the San groups of Southern Africa, the forager-farmer symbiotic groups of the Congo Basin, the Negritos of Southeast Asia, and a detailed examination of deforestation in the Philippines. In general what we learn is that all these groups are facing bleak prospects for continued survival as independent ethnic entities, and that environmental destruction is a major factor in their demise. The first chapter by Hitchcock describes the battle of San peoples for land against a tidal wave of ranching expansion, the struggle for resource use rights in the face of restrictive park regulations and tourism priorities, and a battle for dig- nity and self-determination in the face of bald racism by both Blacks and Whites who consider the San to be a “natural” servant class. The second chapter by Bailey provides demographic documentation of a threatened population of Efe foragers in the Central African rainforests. The Efe show extremely low fertility due to seasonal food shortages and sexually transmitted diseases. They are also culturally op- pressed by more politically and economically powerful neighbors and, because of high rates of hypergyny (Efe women marrying outside men), the group is in danger of both genetic and cultural extinction. High rates of deforestation and land inva- sion are the greatest and most immediate threats to the Efe people. The remainder of the volume examines the case of the Agta people in the Philippines. The Agta have lost most of their land—and access to natural resourc- es necessary for their survival—and have become victims of economic exploita- tion, modern infectious diseases, and a loss of hope and self-esteem. Such prob- lems lead to alcoholism, depression, and hypergynous mating between Agta women and higher status outsiders, destroying the will to live and the drive to succeed for many Agta men. The situation of the foraging societies described in this volume mirrors the threats faced by many small-scale tribal societies and minority ethic groups around the world. oot The detailed examination of the Agta case is illustrative of a larger historic process of conquest and assimilation or elimination of the “Negrito” peoples in Southeast Asia over the past few millennia including alarming yearly reports about the near extermination of the natives of the Andaman Islands. This sad portrayal of the plight of the Philippine Negritos has been foreshadowed by books such as Eder’s On the Road to Tribal Extinction (1992) and Early and Headland’s Population Dynamics of a Philippine Rainforest People (1998). : Racism and ruthless silent conquest are the two processes underlying this historic displacement, and all indications suggest that both African groups dis- cussed in the text are experiencing many of the same problems as the Agta. In- deed, I have seen the same processes among tribal societies in the Amazon forests where I have worked for nearly thirty years. Darcy Ribeiro documented forty years ago that nearly half of all native tribes in Brazil that were known at the beginning of the twentieth century were extinct by 1950. In some cases such as the Agta, Amazonian tribes, the Penan of Borneo, and the Central African forest 170 BOOK REVIEWS Vol. 24, No. 1 peoples, timber and mining interests along with rampant deforestation and rapid colonization have been to blame. But a larger worldview suggests that defores- tation alone is not the cause of tribal extermination. Indeed there is no valuable timber to be had in the Kalahari desert, or in the Paraguayan Chaco (where hunter-gatherer groups have experienced the same ethnic destruction as reported for the Agta), nor in Tierra del Fuego, where the Ona were exterminated within fifty years of their first peaceful contact with Europeans. Instead, pure unadul- terated conquest of territory has led to the recent demise of hunter-gather pop- ulations just as it led to the near extermination of California natives in the nine- teenth century. The battle for tribal survival is a human rights battle. Groups that should be allies must be persuaded that natives share with them mutual interests. Because habitat destruction is often a root cause of tribal destruction, conservation orga- nizations sometimes lend valuable support to native peoples. But almost as often conservationists have been behind the movement to forcibly remove native peo- ples from their traditional territories in order to set up ‘‘people-free’”’ parks and reserves that support a cadre of outside biologists and tourism enterprises but leave native peoples on the fringes of their ancestral lands, as beggars and cheap labor for the “foreign’’ enterprises. The U.S. led the way in this process eliminating all native resource use rights in National Parks and public lands as they were created over the past two centuries. A similar disenfranchisement of native peo- ples has been standard fare in Africa, Asia, and South America. Arrangements that allow native peoples resource use rights, such as in the Okopi Wildlife Refuge described in Bailey’s chapter, must emerge as the standard arrangement between natives and conservationists. Conservationists would do well to read this book, which highlights some the difficult decisions to be faced when considering wheth- er to prioritize ‘‘nature’’ conservation over ‘‘culture’’ conservation. There is no empty natural habitat in the world (unless the native group that once lived there has been recently exterminated) and there hasn’t been for many thousands of years. Because most native groups cannot and do not desire to rapidly make the transition to a “developed” economy, they are excellent guard- ians of natural habitat. They need natural resources to survive, and thus have a stake in the preservation of natural habitat. If international conservation organi- zations would recognize this fact and fully incorporate it into their long-term agenda, both hunter-gatherers and the wild resources that sustain them would face a much-improved prospect for survival in the third millennium. Kim Hill Department of Anthropology University of New Mexico Albuquerque, NM 87131 REFERENCES CITED Eder, James F. 1992. On the Road to Tribal Extinction. University of California Press, Berkeley. Early, John and Thomas Headland. 1998. a ation Dynamics of a Philippine Rainforest People. University of Florida Press, Gainesvi Spring/Summer 2004 JOURNAL OF ETHNOBIOLOGY 171 Plant Resins: Chemistry, Evolution, Ecology, and Ethnobotany. Jean H. Langen- heim. 2003. Timber Press, Portland, Oregon. Pp. 586 + illus. $49.95 (hardcov- er). ISBN 0-881925-748. Jean H. Langenheim’s text seems certain to become the most widely read and cited book about plant resins for many years to come. Dr. Langenheim is professor emeritus of biology and a research professor at the University of California, Santa Cruz and has been one of the most eminent researchers in the field of plant resins for over 40 years. This work finally replaces Vegetable Gums and Resins (1949) by Frank N. Howes as the most up-to-date and comprehensive treatment of this subject. This ambitious book presents an integrated view of plant resins that includes their formation, composition, defensive functions for the plants, and their impor- tance to the many insects and mammals (including people) that use them. The first of the book’s three parts covers plant resin production with chapters em- phasizing the definition and chemistry of different resin types, an overview of the evolution of resin-producing plants, and a description of the structures plants use to secrete and store resin. Part II covers the geological history and ecology of resins. Its chapters describe the knowledge and mysteries of amber (fossil resin) and the interactions between plants, herbivores, and resins. The third section of Plant Resins covers resin ethnobotany, with chapters ex- plaining the past importance and future use of resins. One overview chapter within Part III provides fascinating accounts of the importance of amber and resins to people from the Stone Age to the present. The chapter illuminates the numerous times throughout history that procuring and trading resins have had profound impacts on cultures and economies across the globe, especially in the case of amber and various incenses. Later, more specific chapters detail the origins and uses of three major classes of resins—oleoresins, fragrant and medicinal bal- sams, and varnish and lacquer resins. The best chapters make connections be- tween the history of resin use with the biological aspects of the resin source and formation. Some examples, however, do not have the same level of well-round- edness, but do provide the basic facts about all of the common and many of the obscure resins found throughout the world. a — One of the book’s simplest and most important contributions is in providing clear definitions of what plant resins and their subcategories are and, just as important, what they are not. These definitions give readers a solid foundation for understanding the many facets of this book and in helping sort through other works that may use ambiguous or inaccurate meanings about resins and other plant exudates. For example, Langenheim points out (p. 27) that the trade term “essential oil,” which refers to volatile terpenes in some plants, is misleading out of the industry context because these compounds are “neither essential to plant metabolism nor are they true oils; essential refers to their essence or fragrance, and oil to their feel.” The clear descriptions of the basic classes of terpenoid and phenolic resins are particularly helpful for nonchemists, because they help explain how differences in resin composition influence the degree to which a resin re- 172 BOOK REVIEWS Vol. 24, No. 1 mains fluid or hardens after re el to air, a key property that affects its eco- logical function and/or human u ny ethnobiologist who has putas See plant resin use by people will gain a deeper appreciation for resins by learning about them through the eyes and efforts of diverse investigators cited in this book. Plant Resins provides a vivid example of the way ethnobiology can advance as an interdisciplinary field by synthesizing information and insights from chemists, ecologists, and anthropol- ogists. The author writes about each subject area with clarity, confidence, and precision. The glossary will help readers with no scientific background to easily understand most of the information and concepts in the book. Dr. Langenheim draws heavily on her own extensive work and brilliantly connects the salient points from hundreds of other researchers in the field. When she believes other researchers’ conclusions are not supported by solid evidence, however, she does not accept their views without reservation. On a slightly critical note, each paragraph contains numerous references, al- though every phrase does not have a corresponding citation. This format makes the text more readable than scientific journals, but sometimes leaves the reader wanting to know the specific source of a particular point. The meticulous illus- trations including maps, plant and plant parts, chemical structures, and flow charts complement the text well. The color and black and white photographs bring to life the processes relating to resin formation and harvest. Dr. Langenheim’s prowess as a scholar has allowed her to bring together an impressive array of the past and current explorations about plant resins into one book. Her length and breadth of experience with the diverse aspects of this topic have also enabled her to critically analyze the information from a holistic stand- point and recommend topics that should be investigated in the future. Some of the enticing puzzles that researchers will face include elaborating resin formation processes in plants, unraveling the complex physiological and ecological resin interactions with insects, and investigating new ways that people can utilize plant resins. Plant Resins will become the standard reference for this subject because it brings together the many dimensions of plant resin research and will inspire the next generation of researchers to probe new resin mysteries. Campbell Plowden Forest Resources Cooperative Extension enn State University University Park, PA 16802 REFERENCE CITED Howes, Frank N. 1949. Vegetable Gums and Resins. Chronica Botanica, Waltham, Massachu- setts. ATO OTICE TO AUTHORS 3 1753 00317 9899 The Journal of Ethnobiology’s current “Guidelines for frathors: sa gauge in volume 22, number 1. It is also posted on our website, at w /guidelines/. Careful scrutiny of recent issues : the Journal should provide models of fouls ad style for any manuscript you may wish to subm Those submitting manuscripts for consideration for publication in the deep shoutd send three hard copies. Note that initially illustrations should be sent as high-qu ot or print-outs. Send diskettes and original art or photographs only after consultation with the Editor. Manuscripts submitted in inappropriate style and format will be return Manuscripts should be sent to: NAOMI F. MILLER Editor, Journal of Ethnobiology University of Pennsylvania Museum—MASCA 33rd and Spruce Streets Philadelphia, PA 19104 (nmillerO@sas.upenn.edu) BOOK REVIEWS We welcome suggestions on books to review or actual reviews from readers of the Journal. If you submit a book review for consideration, please send two hard copies and one on diskette. Send suggestions, comments, or reviews to: DARRON COLLINS Amazon Conservation Team 4211 N. Fairfax Dr. Arlington, VA 22203 (dcollins@amazonteam.org) WEBSITE ethnobiology.org For information about the Society of SS na activities, - Journal, and membership /sub- scription forms, consult our website. F suggestions about our website, please contact: DOUGLAS TRAINOR (feedback@ethnobiology.org) SUBSCRIPTIONS Other correspondence to the Society and subscription inquiries for the Journal of Ethnobiology should be addressed to the Secretary /Treasurer, C. Margaret Scarry, Departmen nt of Anthropology, Alumni Bldg., CB# 3115, University of North Carolina at Chapel Hill, Chapel Hill, NC 27599-3115 (scarry@email.unc.edu) Annual subscription rates (subject to change; consult our website): For institutions in the U.S., Canada, and Mexico—$80; vies a cera for individu- als in the U.S. and Canada—$35; for individuals in Mex n Central and South America—$35; for individuals in Africa, Asia, huni pe 1 Europe—$45; foc students in the U.S., Canada and Mexico—$25; for students elsewhere—$35. Checks made out to the Society of Ethnobiology should be payable in U.S. dollars; we cannot accept other currencies. Defective or lost copies of the Journal wit be replaced if a written request is received within one year of issue. For information on back issues contac t Paul E. Minnis Department of Anthropology, University of Oklahoma, Norman, OK 73019; (405) 325-2519 (minnis@ou.edu). CONTENTS ETHNOBIOTICA AN EMPIRICAL ASSESSMENT OF EPAZOTE (Chenopodium ambrosioides L.) AS AFLAVORING AGENT IN COOKED BEANS Michael H. Logan, Kimberly D. Gwinn, Tina Richey, Beth Maney, and Charles T. Faulkner Plantago spp. AND Bidens spp.: A CASE STUDY OF CHANGE IN HAWAIIAN HERBAL MEDICINE Christian Palmer NOMENCLATURE OF BREADFRUIT CULTIVARS IN SAMOA: SALIENCY, AMBIGUITY, AND MONOMIALITY Diane Ragone, Gaugau Tavana, Joan M. Stevens, Patricia Ann Stewart, Rebekka Stone, Paul Matthew Cox, and Paul Alan Cox THE CATEGORY OF ‘ANIMAL’ IN EASTERN INDONESIA Gregory Forth OJIBWAY PLANT TAXONOMY AT LAC SEUL FIRST NATION, ONTARIO, CANADA Mary B. Kenny and William H. Parker POPULATION TRENDS AND soon CHARACTERISTICS OF SWEETGRASS, Anthoxanthum niten TEGRATION OF TRADITIONAL AND SCIENTIFIC celica KNOWLEDGE Daniela J. Shebitz and Robin W. Kimmerer STRUCTURE, PHENOLOGY, FRUIT YIELD, AND FUTURE PROSPECTS OF SOME PROMINENT WILD EDIBLE PLANT SPECIES OF THE SIKKIM HIMALAYA, INDIA Manju Sundriyal and R.C. Sundriyal PLURALISTIC MEDICAL SETTINGS AND MEDICINAL PLANT USE IN RURAL C OMMUNITIES, MATO GROSSO, BRAZIL Maria Christina de Mello Amorozo BOOK REVIEWS 51 113 139 163