Background
Trinidad and Tobago is located northeast of the Venezuelan coast and has a humid tropical climate. British Columbia (BC) is the western-most province in Canada and has a temperate climate. This paper describes a selection of the ethnoveterinary medicines used for horses in Trinidad and Tobago and in British Columbia. These places are part of a common market in pharmaceuticals and are both involved in the North American horse racing circuit. Since racehorses and jockeys are often in transition from other regions and between Canada (including Woodbine racetrack in Ontario, the Aqueduct racetrack and Belmont Park, both in New York) and the Caribbean, one of the goals of this research was to investigate commonalities in ethnoveterinary medicine between these two regions. Very little research has been conducted on ethnoveterinary medicine used for horses and there are few comparative studies. There are some shared cultural features between Canada and the Caribbean derived from common Amerindian culture, British colonial histories, and substantial and continuous migration from the Caribbean to North America. An estimated 150,000 Trinidadians are currently living in Canada.
The population of Trinidad, just over 1 million people has equal proportions of African-origin and East Indian-origin (39%). Approximately 15% of the population consists of mixed raced persons and the remainder consists of minority groups (>2%) of European-origin, Middle-Eastern-origin and Chinese-origin people. British Columbia has a total population of 4.168 million people. The 1996 census revealed that 50% of the population was of European origin and 27% of Asian origin. The population of Chinese origin is estimated at 253,382. The 2001 Census revealed that the top 10 languages spoken in BC are: English, Chinese (including Cantonese and Mandarin), Punjabi, then five Western European languages, Tagalog and Korean.
There are major differences in vegetation between the two areas. However a few studies have revealed that geographical barriers are lessening in terms of increasingly globalized ethnomedicine. For example, one study conducted south of Trinidad revealed that of 216 introduced plant species used by peoples in northern South America (Brazil, Colombia, Ecuador and Peru), 80% were of European, Mediterranean or Asian origin, 9% were of African origin and 8% were from the New World [1]. Another researcher found that 36% of the taxa used in the Atlantic forests of Bahia, Brazil for which origins could be established came from Africa, Asia and Europe [2]. The plant pharmacopoeia in South America is cultivated, exotic and opportunistic and sourced from home gardens, roadsides and secondary forest rather than indigenous species from the primary forests that were alien to the region's new settlers [2]. Canadians use a wealth of herbs of European origin. However research conducted by the first author in both countries indicates that there are far more herbs of Chinese origin being used in Canada than there are in Trinidad and Tobago [3,4]. We will return to this point later in the paper.
Horse racing has been established in Trinidad since 1828 [5]. There are occasional race days as well on the twin island of Tobago. The only utilised racetrack in Trinidad was moved from the capital city (Port of Spain), east to the refurbished venue at the Santa Rosa Complex (73 hectares) in Arima in 1993. Previously, all races were run on a clockwise turf track. However due to the influence of American-style racing, an anti-clockwise sand track surface circuit was laid. Race distances vary from 350 metres (for 2-year-olds) to 2000 meters. The Santa Rosa Complex hosts 40 race days annually. The Trinidad and Tobago Racing Authority is the body responsible for horse racing. There are several associations associated with horse racing: the Owners, the Stud Farm, the Bookmakers, Grooms and Trainers Associations, the Arima Race Club and the Tobago Race Club. The government Minister in charge of horse racing in 1999 claimed it was not economically viable and relied heavily on government financial support [6]. Creoles (locally born horses) from Trinidad also race in Puerto Rico, Barbados and Jamaica. In the 2004 Agricultural Census conducted by the Ministry of Planning and Development, the Ministry of Agriculture, Lands and Marine Resources and the Central Statistical Office, horses were not counted and therefore data on the horse industry is limited.
The research area in British Columbia consisted of the Lower Mainland, the Thompson/Okanagan region and south Vancouver Island. The racetrack situated in the research area is Hastings Park, in Vancouver, the largest city in the province. The 2001 Statistics Canada Census revealed that there were 53,366 horses and ponies living on 6,820 farms in BC. The horse industry in 2001 was primarily located in the Thompson Okanagan (25%), the Lower Mainland (20%), the Peace River (18%), and lastly 15% in the Cariboo region. A typical horse unit has seven mares on 10–70 acres. The horse racing industry includes between 9,000 and 10,000 horses, generates $198 million annually and creates 4,000 jobs; but horse racing constitutes only 18% of the horse economic sector [7]. Other parts of the sector include recreational and trail riding, competitions, companionship and other kinds of working animals. The total economic activity involving horses in BC contributed 771 million dollars.

Data collection
Data collection in Trinidad took place in 2000, with further work conducted in 2003. Data collection in BC was carried out in 2003. The respondents were ethnically and demographically varied. A selection of both sets of ethnoveterinary remedies is evaluated in the discussion section of the paper using a non-experimental validation method. The Trinidad component of this study was derived from a larger research project on ethnoveterinary medicines used in Trinidad and Tobago [4]. This previous study revealed that the main outcome or synergy in folk medicine is that all the knowledge is available to all ethnic groups in a kind of 'melting pot' and that there are no rigid barriers preventing the spread of knowledge between the various ethnic groups. In order to gain access to the study population the authors worked through previously known individuals and from previously existing social networks in building a snowball sample and hence a network of interviewees [8]. The first contact relating to this study was a race-horse owner (#8 top earner for the period 1994 to 2000); she drove the first author to the initial visit to the racetrack and to the broodmare farm where her horses were kept. She also introduced the author to several of the trainers. When respondents in the horse racing industry were contacted subsequently it was discovered that they already knew about the research from the initial contact.
Interviews in Trinidad took place from July to September 2000 (CL) and in 2003 (KG). The interviews conducted in Trinidad in 2003 reassessed the initial data (a form of triangulation). The research was facilitated by community-based contacts and occupationally based contacts obtained from newspapers. This networking approach was necessary because there is no sampling frame of persons involved in ethnoveterinary medicine in Trinidad. It produced the desired purposive sample of key respondents.
Four visits were made to the sole racetrack; one of these was on a race day. One visit each was made to three of the six brood mare farms in Trinidad, located in North, East and Central Trinidad. At the racetrack, ten trainers and two assistant trainers were interviewed and one retired trainer was interviewed by phone (this sample is one-tenth of all trainers in Trinidad). The sample frame for choosing the trainers was obtained from the sports pages of the three daily newspapers and from the statistics kept at the University of the West Indies library. All of the interviews were unstructured and open-ended. One of the trainers was also a practising veterinarian. Seven of the ten trainers are recorded in the statistics kept on the "top 25" winners (1994–2000) (#3, #4, #6, #7, #9, #14, #18).
Of all of the trainers interviewed two used no ethnoveterinary medicines, 25% were active users while others reported past use in the 1970s or what they had observed others using. Four grooms were interviewed; they were current users of ethnoveterinary medicines. Six owners/breeders or their representatives were interviewed in 2000, two of them by phone. Four were ranked among the "top 25" in winnings (1994 – 1998) (#1, #7, #8, #12); only one used ethnoveterinary medicines. Three of the six veterinarians consistently working with horses were interviewed, two reported their knowledge of ethnoveterinary medicines, one was also a trainer as indicated above, the other a former jockey.
In 2003, four trainers were interviewed (one by phone). One was selected to confirm the previous data; two were interviewed in 2000, but independently selected in 2003; one was new. Additionally a groom, a stable lad, an assistant trainer, a jockey and a recently graduated veterinarian were interviewed.
Ethnoveterinary data for British Columbia was collected over a six-month period in 2003. All available literature about livestock farmers and the secondary literature on ethnomedicinal plants, folk medicine and related fields in British Columbia was reviewed.
A purposive sample of livestock farmers was necessary to target key informants with the knowledge sought. The sample size was 60. The sample was obtained from membership lists of organic farmers, horse breeders and trainers, horse stables, other specialists in alternative medicine and holistic veterinarians.
Interviewees comprised one naturopath, four horse breeders/trainers, two herbalists, one farmer and one headmistress with horses at her school (for girls). All of the respondents used herbal medicines for horses. Two visits were made to each farm or respondent, and to the Hastings racecourse in Vancouver. All of the interviews at the initial stage were open-ended and unstructured. A draft outline of the respondents' ethnoveterinary remedies was delivered and discussed at the second visit in order to confirm the information provided at the first interview. Medicinal plant voucher specimens were collected where possible and were identified and deposited in the University of Victoria herbarium (V).
The plant-based remedies were evaluated for safety and efficacy with a non-experimental method, prior to including them in the draft outline. Published sources such as journal articles and books and databases on pharmacology and ethnomedicine available on the Internet were searched to identify the plants' chemical compounds and clinically tested physiological effects. This data was incorporated with data on the reported folk uses, and their preparation and administration in North America and Europe. For each species or genus the ethnomedicinal uses in other countries are given; followed by a summary of chemical constituents, in addition to active compounds if known. This type of ethnopharmacological review and evaluation is based on previous work and the use of these methods in a previous research study has been published [4,9-11]. The non-experimental validation of the plants is presented in the discussion section of the paper.

Validation workshop
Ten participants with experience in traditional human and ethnoveterinary medicine took part in a participatory five-day-long workshop at the University of Victoria (BC), in October, 2003. In the workshop the facilitator asked participants very specific questions in a supportive environment about the medicinal plants used. Each animal/livestock species was covered in a morning or afternoon session [4,11]. At the horse session the four participants (two horse trainers and two herbalists), introduced themselves and their work and were instructed on the participatory workshop method. The participants discussed the previously produced horse section of the data. There were two editorial assistants/facilitators in attendance. After the discussions, the horse section was edited. In addition, two herbalists in Port Alberni were visited by the ethnoveterinary consultant and the researcher (CL) and the edited horse data was discussed with them. One trainer with horses at the Hastings racecourse visited the researcher after the workshop and discussed the workshop-edited horse data with the researcher and the ethnoveterinary consultant.

Non-experimental validation of ethnoveterinary remedies
The researcher and the ethnoveterinary consultant completed the non-experimental validation of the remedies in advance of the workshop. A low-cost, non-experimental method was used to evaluate the potential efficacy of the ethnoveterinary remedies [9-11]. This method consisted of:
• obtaining an accurate botanical identification of the herbal remedies reported;
• searching the pharmaceutical/pharmacological literature for the plant's identified chemical constituents in order to determine the known physiological effects of either the crude plant drug, related species, or isolated chemical compounds that the plant is known to contain. This information was then used to assess whether the plant use is based on empirically verifiable principles.
Supporting ethnobotanical data and pharmacological information was matched with the recorded folk use of the plant species [12-18], to determine degrees of confidence about its effectiveness. Four levels of confidence were established:
1. Minimal level: If no information supports the use it indicates that the plant may be inactive.
2. Low level: A plant (or closely related species of the same genus), which is used in distinct areas in the treatment of similar illnesses (humans or preferably animals), attains the lowest level of validity, if no further phytochemical or pharmacological information validates the popular use. Use in other areas increases the likelihood that the plant is efficacious.
3. Mid level: If in addition to the ethnobotanical data, available phytochemical or pharmacological information is consistent with the use, this indicates a higher level of confidence that the plant may exert a physiological action on the patient.
4. High level: If both ethnobotanical and pharmacological data are consistent with the folk use of the plant, its use is classed in the highest level of validity and is considered efficacious.