Samburu District is situated in the northern half of the Rift Valley Province in Kenya. It is bordered by five other Districts; Turkana (Northwest), Baringo (Southwest), Marsabit (Northeast), Isiolo (East) and Laikipia (South) respectively. It lies between Latitudes 0°40" north and 2°50" north of the equator and Longitudes 36°'20" east and 38° 10" east of the Prime Meridian (figure 1). It covers approximately 21126.5 square kilometres. It is divided into six divisions, 39 locations and 108 sub locations. It is characterised by high level plateaus, hills and the Rift valley with an altitude up to 2000 m a.s.l.

The study area has a bimodal rainfall distribution from April to May (long rains) and July to September (short rains). The dry season then extends from January to March. The mean annual rainfall is 500 mm. it has mean annual temperatures of 29°C. Samburus are the indigenous dominant ethnic group with Turkana and Maasai having settled in the area. Pastoralism is the major economic activity of the local people. The District has a population of approximately 156125 people17. These rural communities are almost totally dependent on forests and savannah as traditional/herbal medicine for their own health and livestock care.

A reconnaissance survey was made from December 2006 to January 2007 to obtain an impression on vegetation characteristics of the study area. The fieldwork was done in January, February, May and August 2007.

A total of 100 informants in figure 2 were selected purposively 18 based on knowledge, attitudes and practices (KAP) survey with the help of local administrators. They included 14 traditional medicine practitioners (4 females and 10 males), 86 locals (Male: Female = 2:1). Information on knowledge depth of respondents was collected from local elderly people, opinion leaders and the local administrators. Similar responses obtained from the three groups were used to identify knowledgeable traditional healers. The respondents and traditional healers identified were consented to share their knowledge only for the purpose of this study 19.

The methods used for ethnobotanical data collection were semi structured interviews, field observation, preference ranking and direct-matrix ranking according to 1820. These interviews were conducted in vernacular (Samburu) translated by local field assistants, three different field trips were conducted.

Table 2 indicates relevant data collected on: age, sex, and occupation of informants as well as animal and human health indications treated, vernacular plant names, growth form, plant part used, methods of preparation, dosages, route of administration and possible contraindications. Threats to medicinal plants, conservation efforts, beliefs and indigenous knowledge transfer were also documented. These interviews were done in the field in order to avoid the probable confusions with regard to the identity of the medicinal plants 12. The morphological characteristics, habitats and habits of medicinal plants were observed, photographed and recorded during and after the interviews. The key informants for purposes of ranking these species were selected randomly from among all informants 21.

Geographical Positioning Systems readings were also taken at the sites where each medicinal plant was collected (GARMIN, Olathe, USA). Some of the plants were identified in the field by herbalists while most were identified at the Herbarium of University of Nairobi, Botany department using specific taxonomic keys and floras2223.

Ethnobotanical data were entered in to Excel spreadsheet and summarized using descriptive statistics 24.

Wilcoxon's test was used to determine if there was a difference age of respondents and knowledge of medicinal plants used. Chi-square test was used to evaluate the average number of medicinal plant species reported and used by each informant, to determine if there is any significant difference between female and male practitioners with respect to the knowledge and use of medicinal plants. The Spearman rank correlation test was used to determine whether there was a significant correlation between the disease reported and the number of ethnoveterinary medicinal plant species used by each informant for management of the disease. STATA 9.2 IE (Stata Corporation, College Station, Texas, USA) software was used.

There was a highly significant difference between age of respondents and knowledge of medicinal plants (Wilcoxon's test, p < 0.001). The average number of medicinal plants known and used by female and male practitioners was similar (χ² = 8.262, d.f. = 13, p = 0.932). It was observed that informants between 58 and 77 years old mentioned more species than younger informants: 58–67 years old: about 10 per informant; 48–57: about 6; 38–47: about 5; and 28–37: 2 quoted plants, due to larger experience of older individuals. These results also agree with other previous studies25. It was observed that some plants had more than one vernacular name due to use of the Maasai and Turkana dialects in the area.

A total of 28 animal and 26 human ailments were reported by the informants respectively. The frequency of the most cited ailments and the number of medicinal plant species used are also given in figure 3. The most frequently cited animal health problems were; Retained afterbirth (9), Ectoparasites (8), gastrointestinal disorders (5), Theileriosis (4) and helminthosis (3). Human ailments treated cited frequently included; Malaria (18), gastrointestinal disorders (10), helminthosis (5) and rheumatism (4). Respondents had good knowledge and remote diagnosis of the disease and could readily distinguish them on the basis of accepted signs and symptoms. Ailments such as convulsions, hypertension, asthma, yellow fever and infertility were beyond the scope of the present study, it was considered important to record plants that were frequently mentioned for the treatment of such health conditions (see table 1).

Ticks (lntunturi) were the main cattle ectoparasites that the local people controlled using traditional plant extracts. The most frequently used plants for tick control were found to be: Acokanthera schimperi (Lmorijoi), Adenium obesum (Lperantai), Aloe secundiflora (Sukuroi),Psiadia punctulata (Labaai),Nicotiana tabacam (Lkumbao), Euphorbia herechroma (Lpara) represented in table 1.

The respondents classified all intestinal worms under one local name, ntubui and therefore use the same plant extracts for all helminths. The main species used for this were: seeds of Myrsine africana L (Seketet) (45%), Albizia anthelmintica (Lmungutan) (30%) and Warburgia ugandensis (Sokorioi) (18%).

Fifty four (54) plant species of ethnopharmacological importance were gathered and documented throughout the study period (table 1). These medicinal plants were distributed among 50 genera and 33 families. Analysis of the growth forms of these medicinal plants revealed that, shrubs constituted the largest number or proportion with 31 species (56%), followed by trees 15 (28 %), herbs 5 (9%) and lianas 4 (7%) respectively shown in figure 4. Ethnobotanical knowledge was passed on by word of mouth. Knowledge of ailments such epilepsy, hypertension, venereal diseases, impotence, was generally restricted to the elders and traditional medicine practitioners represented in figure 5.

Leaves were the most frequently used plant parts constituting 4 %, followed by roots (3 0%), stems (10%), fruit/seeds (8%) and whole plant (4%) in figure 6. The majority of informants (42%) mentioned Myrsine africana L. (Seketet) as medicinal for the treatment of various animal and human ailments.Seketet was thus the most popular remedy in the study area, followed by Carissa edulis F. (Lamuriai) (5%), Salvadora persica L. (Sekotei) (30%), Albizia anthelmintica Brongn. (Lmungutan) (27%) and Clerodendrum myricoides Hochst. (Lmakutikuti) (22%).

The number of species frequently used in each family was cited as; Apocynaceae (6), Mimosaceae (5), Euphorbiaceae and Rhamnaceae (4) and Asteraceae (2) other families were represented by at most one species shown in figure 7.

The preparation of the medicines employed several methods; hot decoction (48%) followed by cold decoction (19.4%) and homogenization by pounding or powdering (6.5%) respectively in (table 1). The majority of these preparations were drawn from mixtures of different plant species for the treatment of a single ailment. Oral administration (8%) was the predominant route of administration followed by dermal and nasal administrations (20%).

Many medicinal plants in the study area were mainly threatened by anthropogenic and natural factors. The majority of medicinal plants declined due to deforestation for construction, tools, firewood, fodder, agricultural expansion and ceremonial purposes. Drought, overgrazing, bush fires had reportedly affected a significant number of medicinal plant species.

About 47% of the informants had sufficient awareness in conserving some medicinal plant species that were relatively scarce in their surroundings. In situ protection of plants (fencing plants in their natural habitat, refraining from excessive cutting, debarking and uprooting and protection from fire) and ex situ conservation by cultivation of some plants as live fence and in nurseries were undertaken by the locals. Moreover, some of them were keen to inform responsible bodies or authorities of any illegal logging, deforestation and bush fires.

Majority of local healers preferred to collect medicinal plants solely to preserve their secrecy sometimes accompanied by the chosen family member(s). The ethnobotanical knowledge is transferred to that trustworthy family member by word of mouth rather than through a well organized written script 23. Some of the ethnopractitioners were reluctant to pass on their plant use knowledge even to their families leading to the fragmentation and loss of the indigenous knowledge system and eventually medicinal plants 12.

This study revealed that traditional medical healers and pastoralists in Samburu District had sound knowledge of traditional medicine, from whom about 54 indications (animal and human) and a total of 56 medicinal plants of importance were recorded.

The continued reliance of Samburus' on traditional medicines is due to economic circumstances, which place modern health facilities, services and pharmaceuticals out of the reach of the majority of the population. It is also attributable to the widespread belief in the effectiveness of many traditional therapies 27.

The current investigation indicates that leaves are the most collected plant parts for medicinal purposes and this situation could be a severe threat to some rare and slowly reproducing medicinal plants. The practice of exploiting perennial plant parts, such as roots of relatively slow growing woody species, can result in a decline in both, the size and distributions of populations of the exploited species, and ultimately result in the local extinction of these populations 28.

The use of plants is evenly distributed for management of both medical and veterinary conditions this finding was contrary to earlier findings in the same geographical zone which reported insignificant veterinary use 1.

The majority of plant preparations were drawn from mixtures of different plant species for the treatment of a single ailment and similar results had reported elsewhere 26. This was contrary to the findings of other researchers in other countries where most of the remedies were prepared from a single species 29. This could also be ascribed to the differences in the socio cultural landscapes, indigenous knowledge on synergetic effect of different medicinal plants and vegetation types in the current study area 29.

The most frequently used methods of preparation were hot decoctions, cold decoctions, powdering and grinding respectively. The prepared medicines were mainly administered through oral (98%), dermal (1.5%), and nasal (0.5%), routes concurring with the previous findings in Ethiopia 30.

The measurements used to determine the dosages were not standardized and depended on the age and physical appearance of the patient, sociocultural explanation of the illness, diagnosis and experience of individual herbalist 27.

The naming of diseases by local people when compared to conventional systems, at times did not distinguish between diseases and symptoms of diseases. This is because local disease nomenclature is based on symptoms of diseases and not according to aetiological information 3132.

While conducting this study, some informants raised some concern on false promises about getting the feedback. They agreed that scientific methods are better in revealing harmful effects of herbs. In the drug development research and bioprospection, biological activity based on ethnomedical uses seems as a better approach compared to randomly selected plants 33

We are currently undertaking field trials, in-vitro and in-vivo tests of these plants for antihelmintic, antiparasitic, antiplasmodial, antibacterial and cytotoxic activities to confirm the therapeutic properties claimed by informants.