Yellow pond lily(Nuphar polysepalum Engelm., sometimes classified as Nuphar luteum (L.) Sibth. & Sm. ssp. polysepalum(Engelm.) Beal), was mentioned by 17 consultants. Nineteen cultural groups within the range of this species north of Oregon use pondlily for medicine, while for 2 groups recorded information does not indicate use for medicine (Figure 3). Six Gitksan consultants reported use of the rhizome for TB; four reported its use as a poultice for fractures, three as a poultice for arthritis, and three for poulticing sores and wounds. Two modern consultants, and one archival source report use of the rhizome for a male contraceptive. Use of pondlily rhizome for "sickness", as an appetite stimulant, for liver medicine, and for aches and pains were also mentioned.

Pond lily showed moderate activity against Mycobacterium tuberculosis in vitro, achieving complete inhibition at concentrations of 100 mg 15. It was often mixed with devil's club in antitubercular preparations, which makes efficacy very likely, especially when taken over prolonged periods of time as modern anti-tubercular medicines are. The properties of alkaloids known to occur in the rhizome of Nuphar luteum cannot be neatly correlated with Gitksan ethnomedical uses, although antispasmodic properties of nupharine might aid appetite by suppressing nausea. Other Gitksan uses of yellow pond lily rhizome suggest possible antiinflammatory or analgesic properties, but high levels of steroidal or other known antiinflammatory compounds or analgesics have not been reported for Nuphar luteum. Zhang et al. 20 have suggested that immune suppressant properties of deoxynupharidine, which occurs in N. variegatum as well as the Chinese N. pumilum, may account for the use of Nuphar pumilum in Chinese folk medicine for rheumatoid arthritis, as rheumatoid arthritis is an autoimmune disease.

Indian hellebore (Veratrum viride Ait.) rhizome was mentioned by 13 consultants in recorded interviews, and in various other (unrecorded) conversations, suggesting its widespread collection and use at present Twenty-four other cultures in northwestern North America are reported to have used Veratrum viride for medicine, while one culture (Sahaptin) was reported not to use Veratrum viride rhizome (Fig. 4).

Use of Veratrum viride for purification or as a smudge was mentioned by 9 consultants. Smudges are used as treatment for stroke or mental illness, to counteract evil magic, and to repel ghosts or evil spirits. Use for hunting luck was mentioned by 5 people. Five people made reference the practice of carrying a piece of the dried rhizome for protection from bad luck or evil magic. Three consultants described bathing in an infusion of hellebore rhizome as an external cleanser. Treating skin and hair was mentioned by two people. Two people also described use of Indian hellebore as a "germ killer", which can be put in wash water to cleanse laundry of evil influences, vermin or germs (in the medical sense of bacteria or viruses). Two consultants mentioned use as a snuff to clear the sinuses. (I can verify from observation that it induces dramatic sneezing and running of the nose when taken this way). Uses as a nightmare remedy and as a sleepwalking remedy were also mentioned.

Many of the uses of Veratrum viride are clearly related to cultural notions of spiritual potency of plants, and concepts of cleansing incorporate notions of spiritual as well as physical cleanliness. It is possible that salience, toxicity and/or conspicuous thorniness (as with devil's club, also seen as spiritually powerful and cleansing) contribute to these concepts of potency. The toxicity and potential for loss of life caused by careless use of Veratrum are certainly understood locally and could contribute to its widespread perception as a powerful and dangerous plant. Notions of "dirt" or "contamination" can accommodate to contemporary notions of disease causation by germs, but the local concepts also relate to notions of spiritual as well as physical cleanliness or "dirt". These notions may explain the concept of Indian hellebore as a germ killer; potential insecticidal properties could also be involved in cleaning of laundry.

Most of the known active compounds in Veratrum viride are reported from the roots rather than the rhizomes 21. Gitksan use is primarily of the rhizomes, rather than the attached fleshy roots, though some use is also made of the roots, which may be left attached to rhizome pieces when they are dried. Ethnomedical uses for topical analgesia and skin treatment are confirmed by known chemicals, at least in the roots. The toxicity of Veratrum is recognized by the Gitksan, who do not use it internally. They do not make use of the alkaloids to lower blood pressure; these form the basis of the use of Veratrum viride in modern pharmacology, as a base for synthesis of hypotensive medications 22.

Conifer barks and pitches are still widely collected. Uses of Picea and Pinus, and Abies, were mentioned by 9 consultants each (Table 3). The liquid and solid pitches of spruce, pine, and fir are used for wound dressings and taken internally. All three species, and other related species, are widely used in northwestern North America. Thirteen groups in the map area are known to use Abies lasiocarpa (Hook.) Nutt., while 8 use Abies grandis (Dougl.) Forbes or Abies amabilis (Dougl.) Forbes (Fig. 5). For 11 groups there is no record of use of Abies.

As the spruce used by the Gitksan Picea × lutzii, or Roche spruce, is a hybrid swarm of Picea engelmannii Parry and P. glauca (Moench) Voss with P. sitchensis (Bong.) Carr, I have mapped use of all three of these species in northwestern North America (Fig. 6). Seventeen groups are known to use these species for medicines, while for nine groups there are no reports in the literature of medicinal use.

Altogether, 16 groups are reported to use Pinus contorta Dougl. for medicine, while 21 cultures are not reported to use lodgepole pine (Fig. 7) (but many of the more southern groups do use other species of pine such as ponderosa or western white pine).

Abies is used for a variety of purposes: the bark in mixed decoctions for treatment of tuberculosis, or as a tonic, or for internal medicine or lung disease (Additional file 1). The liquid pitch is regarded as the strongest part of the subalpine fir, and is used for TB treatment, as a purgative or cleanser, or for cuts or boils. Pine and spruce pitch and bark have been given together, as consultants often consider them interchangeable. There is high consensus on use of pitch of pine or spruce for infections, burns and boils (9 of 9). Other uses are in tonic "wood medicine" mixed decoctions and for treatment of tuberculosis.

Conifer pitch has antibiotic properties, especially spruce pitch, as Ritch-Krc et al. 23 showed in their research, where they obtained very strong inhibition of bacteria and Candida with the pitch of Picea glauca from central British Columbia, and moderately strong inhibition of bacteria with pitch from Pinus contorta. Interestingly, their assay results found no significant inhibition of bacteria or Candida by Abies pitch, which is generally rated a better medicine by indigenous consultants. However, these negative results could have been influenced by technical difficulties mixing pitch with the growth medium (Turner, personal communication 1999).

Common juniper, Juniperus communis L., is widely used for medicine around the world. Juniper is called "boughs of the supernatural" or la

x

A detailed discussion of the many active compounds found in juniper and their diverse effects is beyond the scope of this paper. Extracts of J. communis showed moderately strong bacterial inhibition of all species tested 14. Numerous active chemicals are found in "berries" and other parts of the plant, including-umbelliferone, camphene, camphor, myrcene, delta-3-carene ("berries"-), alpha-pinene and limonene, rutin, borneol, citronellol, and umbelliferone. These compounds exhibit diverse activities, including: antiseptic, bactericidal, fungicidal and antiviral properties; analgesic, anesthetic, antihistiminic, antiinflammatory, expectorant and antitussive properties; spasmolytic, spasmogenic and sedative properties; antiedemic, antidiabetic, anticancer and cancer preventative, antiatherogenic, and liver protective actions; and reduction of fever 21. In addition, limonene, alpha-pinene and rutin are also known to be anti-nephritic or nephrotoxic, which corresponds to the known kidney damaging effects which J. communis extracts can produce (Oates pers. comm 1987).

Red elderberry, Sambucus racemosa L., is the last medicine mentioned by 5 or more consultants. It is a strong emetic. Perhaps it was mentioned by 5 consultants because of the dramatic and memorable nature of its reported effects! In Gitksan thought dirt or contamination, at both spiritual and physical levels, are implicated in disease causation. Therefore, cleansing with an emetic preparation from elderberry root bark was seen as an effective treatment for a serious ailment like influenza. The other reported use is as part of a spiritual smudge medication. Thirteen cultural groups in the map area used red elderberry as medicine, while there was no record of use for 13 additional cultural groups (Fig. 9). Use of roots or stem inner bark as an emetic was recorded by 7 of 22 other groups which use red elderberry for medicine 9, sources in 2. Elderberry species contain cyanogenic glycosides in roots, stems and leaves, and an unidentified cathartic compound in the bark and roots (24: 148).

Trade and diffusion of knowledge occur despite significant linguistic differences. In some cases the name may diffuse across language boundaries, reinforcing the concept that the knowledge of use may also have diffused. An example of this is the name for yellow pond lily

g

Recent documented examples of knowledge diffusion across cultural boundaries include anecdotal acknowledgement of learning of a particular use of plants by members of another cultural group: one consultant used soapberries as arthritis medicine; he said he learned this "from the Coast people". Another example includes use of the lichen lungwort (Lobaria pulmonaria); use of this medicine was reputedly taught to the my colleague's great uncle by a healer from Port Simpson, a Coast Tsimshian community. The use of devil's club for diabetes treatment by an elder from Gitsigyukwhla was reportedly learned from someone from Hartley Bay, another Coast Tsimshian community. In the southern Yukon, Catherine McClellan documented that use of Indian hellebore by Tagish people was learned from contact with people on the Coast [Tlingit] (26:226).

Finally, the trade of medicinal plants may diffuse knowledge even beyond the range of occurrence of the plant. The Kaska in the Southern Yukon apparently obtain devil's club inner bark by trade, and have incorporated its use in their medicinal plant repertoire. The Tagish also have to travel over the mountains or trade for devil's club 26. Going in the other direction, various Gitksan people like "caribou leaves" (Artemisia tilesii Ledeb. ssp. elatior (Torr. & Gray) Hult.), the use of which they have recently learned from Tahltan and Kaska people in the Telegraph Creek/Dease Lake areas of northern British Columbia. Some people either travel by car to the Stikine River drainage to harvest a supply or to trade for it.

The modern use of sweetgrass (Hierochloe odorata (L.) Beauv.) and sage (Artemisia spp.) for smudging is another such example; the trade network for obtaining sweetgrass is extensive, and modern use is learned from various Cree and other healers from Alberta. Modern sage use seems to come both from the southern interior of British Columbia and from Alberta and the Plains and is probably spread both by the pow-wow circuit and by sweats and healing ceremonies.

In light of evidence of extensive trade, both in the past and at the present time, I therefore cannot assume that knowledge and use of medicinal plants has been arrived at independently in cultures of British Columbia and adjacent Alaska and Washington. Use of "intercultural consensus" is another way to indicate the increased likelihood of empirical efficacy of plants widely used by a number of different cultural groups.

Careful consideration needs to be given to the conditions under which widespread trade and contact between groups across language boundaries suggests that intercultural consensus in medicinal plant uses should be examined. One of the most obvious issues for a map-based method is, how do you decide what the boundaries of the map area should be? In the case of the Northwest Coast and adjacent interior areas of British Columbia, the southern Yukon, and Washington State, geography and biogeography help to define the area of interest. I think this kind of analysis will have to proceed on a specific region-by-region basis.

Aside from concepts like spiritual cleansing, luck bringing and protection, which cannot be evaluated in biomedical terms, there are several aspects of Gitksan concepts of healing which do affect choices of plant therapies and how they are used. For example, Gitksan people believe that illness must be treated holistically. This means that food and self-restraint or control will be part of the recovery process, and that taking medicines will be incorporated into a healthy lifestyle, as in the prolonged use of tonics. Gitksan in general expect to take medicines over a relatively long period of time (ranging from weeks to months, depending on the nature of the illness), in order for them to be effective. Furthermore, I was often told that one should take a medicine "whenever you feel thirsty", in effect, to replace intake of pure water with intake of medicine, to immerse oneself in therapy.

Another aspect of Gitksan treatment is that dirt or contamination, at both spiritual and physical levels, are implicated in disease causation. Therefore, cleansing (purgative or emetic herbs) may be employed to treat a variety of ailments; the clearest example is the treatment of influenza or inability to eat with a powerful emetic, in order to leave the digestive tract clean and ready to receive healing nourishment. From a biobehavioral perspective one would have to rate use of elderberry root as an emetic to be an effective treatment for influenza and serious sickness; the emetic effects are confirmed by other groups, and by clinical reports of purgative chemicals in the bark and roots of elderberry species. The congruence of the Gitksan and biomedical understandings would not be as great; physicians do not usually consider emesis or purgation to be necessary for recovery from viral illnesses.

Thus far, I have documented a neat fit between ethnomedical use of plants and potential or known phytochemicals with physiological effects. In the course of the literature search to confirm Gitksan use of specific plants by finding other groups using various plants and to locate information on phytochemicals and activities of plants used by the Gitksan, I could not help but notice multiple records of use and active phytochemicals or assays of plants found in Northwest British Columbia which are not used by the Gitksan. In total, some 70 plants present in the Gitksan territories which are used medicinally by other groups are not known to have been used by the Gitksan. Some of these have a number of active phytochemicals and even clinical studies supporting their efficacy. Additional file 2 summarizes known properties of phytochemicals in the following two plants which are easily available in Gitksan territory, but not used for medicinal purposes by Gitksan people as far as I have been able to discover.

Kinnikinnick (Arctostaphylos uva-ursi (L.) Spreng.) is widely used in phytotherapy and herbology for treatment of urinary tract infections 27; it also contains substantial amounts of arbutin 21. The Gitksan, however, only collected the fruits for winter storage, but made no medicinal use of the plant.

Wild mint (Mentha arvensis L.) is used in many healing traditions around the world and contains many active phytochemicals, including menthol, in relatively large amounts (Additional file 2). It is highly valued by Kaska elders, for example, from whom I learned its uses after only a short period of fieldwork with a few elders. However, Gitksan elders to whom I described or showed the plant did not recognize the plant, and did not use it.

Although rather problematic for a naive empiricist model of medicinal plant use, the plants not used for medicine provide information about cultural selectivity. Many factors might be responsible for their non-use, ranging from the purely stochastic, to symbolic or contextual. One possible approach to understanding some of this variation would be through use of methods used for elucidating diet breadth in optimal foraging analyses. Some people have suggested that an erosion of ethnomedical and ethnobotanical knowledge might be responsible for these seeming gaps in the use of available and active potential medicinal plants. While this may explain individual instances, I find it unsatisfying as an explanation for such a large number of unused plants. Finally, a seminal paper by Glenn Shepard on Amazonian medicinal plant use suggests that cultural factors that create strongly contrasting ethnomedical models of the nature of illness and modes of healing may cause adjacent groups within a broadly similar ecological context to make quite different choices of medicinal plants to use in treating illness, and may dictate quite dissimilar modes of application or use 28.

My analysis was limited by the quality of the data. Both the completeness of my own data, and the extent of recorded medicinal plant knowledge of other groups within the map area make quantitative analysis of intercultural consensus impossible. The medicinal plant knowledge of some cultures is relatively well described, while for other groups, little reliable documentation of plant use is available. Another difficulty is presented by judging what kinds of uses are similar, given a vast range of descriptions of ailments and uses, driven both by cross-cultural translation of symptomology and disease causation, and historical changes in our understandings within western culture. A recent concern by indigenous peoples regarding the appropriation of their medicinal plant knowledge means that some groups prefer to keep their healing knowledge private, limiting the possibilities of comparison.

The non-uniformity of healing plant knowledge within societies between specialist healers and lay people, and between men and women, compounds the difficulties of obtaining a relatively complete sample of healing plants used for any one culture. I have attempted to overcome this in my own data by consulting with a number of different men and women from several villages. However, none of my consultants were publicly acknowledged herbal healing specialists, and many people expressed the concern that much healing knowledge had already passed on with the Elders. Recorded data from other cultures may be based on many fewer consultants, and botanical identifications in older work may be questionable.

These caveats also apply more generally, both to the "negative evidence" of plants not apparently used by the Gitksan, and to plants used by the Gitksan for which no confirming records of similar use by other cultures can be found. Sometimes even when only one culture appears to have healing plant knowledge, the plant in question may still be known to contain phytochemicals that make it likely to be therapeutic for the conditions it is used for from the perspective of a physiologically based analysis. This may be the case for the use of red osier dogwood (Cornus stolonifera Michx.) to treat fractures, a use not reported elsewhere in the literature.