Cross-cultural analysis of medicinal plant use practice in four (Gurage, Mareqo, Qebena and Silti) ethnic communities in south central Ethiopia

Background : Ethnic groups throughout the world have developed their own cultures expressed in the form of customs, taboos and traditional healthcare systems. Traditional medicine system is one of the widespread cultures known throughout the world which is very much tied to cultural practices of the community or ethnic group. Medicinal plant treasure found in Gurage and Silti Zones remained poorly characterized and understood. Therefore, this study was conducted in four ethnic groups (viz. Gurage, Qebena, Mareqo and Silti) which have lived in close proximity and contact for many centuries in the two Zones. In the present study unique and shared cultural elements in connection to traditional herbal medicine were examined through investigation of the diversity of medicinal plants. Moreover, attempts have been made to determine similarities among the society in the medicinal plants they have used in general and in medicinal plant species considered culturally most important. Methods : In a study that involved 320 randomly samples informants semi-structured interviews, focus group discussions, participant observation were used and qualitative and quantitative data were collected. Descriptive statistics, rank order priority (ROP), Jaccard similarity coefficient and clustering were used for data analysis. Results : A total of 244 medicinal plant species and a fungal species used to treat human and/or livestock ailments were documented. The number of plants (80 plants, 33 %) with ROP value greater than 50% is considerably smaller than that of plants with ROP < 50% (164, 67 %). Jaccard similarity index and clustering analysis for all cited plants, among the respective studied districts, indicated that grouping generally followed the existing ethnic origin. On the contrary, clustering based on culturally important medicinal plant species (80 plant species, score ROP ≥ 50%) showed the influence of proximity and geographical orientation rather than ethnic relation.


Conclusions:
Culturally most important plants (80 spp.) are widely used and best shared with nearby communities and this could imply current (new) knowledge being practiced in the community. This knowledge must be documented and better utilized in a modern way including modernized use of traditional medicinal plants.

Introduction
Ethnic groups throughout the world have developed their own cultures expressed in the form of customs, taboos, songs, traditional foods and healthcare systems. Traditional medicine selected based on several thousands of years of experience has been major aspect of cultural heritage [1]. Traditional medicine system is one of the widespread parts of culture known throughout the world [2]. Like other kinds of local knowledge, it is very much tied to cultural practices of the community or ethnic group [3,4]. Its role in the healthcare system is enormous and widely recognized. The word 'culture' refers to the characteristics and knowledge of a particular group of people, defined by everything from language, religion, cuisine, social habits, song, story and arts [5]. In the present study, the type of plants used in a group of people in their traditional health care system is considered as part of culture.
Ethnomedicinal knowledge, which develops from the interaction of a given culture with the local biophysical environment or locally available plants, is diverse and sometimes it could be ecosystem and ethnic community specific [6,7]. Factors such as social, ecological, cultural background (incl. religious, linguistic) and ancestral inheritance determine the kind of traditional herbal knowledge developed in a community. Owing to these facts, herbal knowledge varies hugely across different community, geographic setting or ethnic groups [7][8][9]. The influence of cultural background and ancestral inheritance could be seen in the variation of people's perceptions and plant use preference in a community inhabiting same geographical or ecological area, facing similar environmental factors [9,10]. As a result, comparative studies in traditional knowledge and plant use culture between communities or ethnic groups gives an opportunity to investigate how the local flora is understood and used in daily life, health practices and, ultimately for survival under different cultural settings [11].
Various studies conducted in different corner of the world have shown the variation that exist in plant use knowledge and ethnomedicinal healing systems across cultures and agro-ecology [12][13][14][15][16]. Traditional medicinal plants use could also vary among communities within the same ethnic group [17] and geographic area [18]. Thus, medicinal plant use practice can be considered to show the cultural differences that may possibly exist between traditional societies or ethnic groups inhabiting in similar/different geographic locations.
Recently, the knowledge of traditional peoples and accompanied systems is disappearing at increasing rate. Besides medicinal plant treasure found in Gurage and Silti Zones, is not fully explored and hence it remained poorly characterized and understood. In light of these scenarios the present study is conducted to document the traditional medicinal plants known in the societies of the two zones. The study area harbor four ethnic groups (viz. Gurage, Qebena, Mareqo and Silti) which have lived in close proximity and contact for many centuries. In the present study unique and shared cultural elements in connection to traditional herbal medicine were examined through investigation of the diversity of medicinal plants. Moreover, attempts have been made to determine similarities among the society in the medicinal plants they have used in general and in medicinal plant species considered culturally most important.

Study area
Gurage and Silti zones are situated in Southern Nations, Nationalities and Peoples Region (SNNPR) and located in south central Ethiopia. Gurage Zone is located at 7°40'0''-8°30'0''N and 37°50'0''-38°40'0'' E with altitudinal range stretching between 1000 -3600 m a.s.l. and covers an area of 5893.5 km 2 [19]. On the other hand, Silti Zone is located at 7°40'0''-8°10'0'' N and 37°50'0''-38°40'0'' E with altitudinal range stretching between 1640 -3277 m a.s.l. and cover an area of 2537.5 km 2 . The geographical locations of the two zones and the studied districts are shown in Figure 1. between Gurage and Silti zones led to extensive intermarriage between the four ethnic groups (Gurage, Silti, Mareqo, Qebena) and maintain inter-ethnic contacts [20]. This mutual influence has been shown in the respective ways of living which most likely resulted from this cultural mix [20]. Extensive cultivation of ENSET (Ensete ventricosum), traditional housing, artifacts, and mode of production are the best-known shared cultural practices of the zones and at large in SNNPR [21]. ENSET is the main food crop together with Hordeum vulgare (barley), pulses, potatoes, and cabbage. The major cash crops are Catha edulis (CHAT), Coffea arabica (BUNA), Eragrostis tef (TEFF) and Guizotia abyssinica (NOUG). Animal husbandry is also part of subsistence farmer's way of life. Amharic language is widely spoken by the communities and sometimes it is used as lingua franca in the zones. Based on the recent classification of potential vegetation types as described in Friis et al. [22], the study area is dominantly characterized by the dry evergreen Afromontane forest and grassland complex (the undifferentiated Afromontane forest subtype) ( Figure 2).

Informant selection and Data collection
A total of 320 informants (40 from each study district) were selected randomly following Gomez-Beloz, [23]. Semi-structured interviews during walk-in-the-woods or by using plant 'props' (freshly collected plant material or photographs), focus group discussions, participant observation and market survey were applied for data collection [24][25][26].
Ethnobotanical information including list of all medicinal plants and parts used; kind of humans/livestock ailments treated and medicinal plants that are sold in the local market were documented. Interview was conducted in Amharic and local languages. Written permission to conduct the research was obtained from the respective zone and district administrative officials. Prior informed consent was obtained from each informant before every interview.

Plant collection and Identification
All plant specimens were collected, dried, identified and deposited in National Herbarium (ETH) of Addis Ababa University. Identification was made by using Flora of Ethiopia and Eritrea [27][28][29][30][31][32][33][34], in comparison with authenticated specimens from the herbarium and later confirmed by senior taxonomists of the herbarium.

Data analysis
Corrected fidelity level (rank order priority) was used to identify the most culturally important medicinal plants from each study districts [35]. Rank order priority (ROP) was calculated using the formula: ROP = FL C RPL. Fidelity Level (FL) is used to quantify the percentage of informants confirming the use of a plant species for the same major purpose. It is computed as: FL(%) = Ip/Iu C100 Where Ip refers to the number of informants who indicated a specific medicinal plant species is used to treat the same major ailment, and Iu is the total number of informants who mentioned the plant for treating any ailment. A high FL value (near 100%) for a plant indicates that all of the use-reports mentioned the plant for a specific treatment, regardless of the number of times mentioned. Whereas, a low FL value is obtained for plants that are used for many different purposes, and/or known by few informants. Jaccard similarity coefficient was used to compare medicinal plants report similarity between the districts. In order to analyse the variability in medicinal plant species among districts, two presence or absence matrices were created. First matrix considered all medicinal plant species cited by informants and the second matrix considered plant species that score minimum rank order priority of 50% (ROP ≥ 50%). The similarity between any pair of study site in terms of medicinal plant species mentioned was calculated using the Jaccard similarity coefficient:

See Formula 1 in the Supplemental Files.
where JI is the Jaccard similarity index, 'a' is the number of species shared by or common to any compared pair of study sites, and 'b' and 'c' are the number of medicinal plant species reported solely in one study district (b for one study site and c for the other). JI values range between 0 and 1, whereby a value of 1 indicates complete similarity. Then the similarity coefficient for each pair of study site was used to obtain a dendrogram using unweighted pair-group method analysis (UPGMA; links a new item to the arithmetic average of a group) [38]. Cluster analysis is generally used to group study sites into categories based on their dissimilarities or partition heterogeneous elements into relatively homogenous groups [39].

Similarity among the community based on all cited medicinal plants
Computed dissimilarity coefficient using all cited medicinal plants was above 0.5 (Table   1). In relative term, Mareqo and Muhir-Aklil showed the smallest similarity (JI = 0.26) and highest similarity (JI = 0.47) was obtained between Silti and Wulbareg districts.
High cophenetic correlation coefficient (0.92) was obtained using the UPGMA (average) clustering method, unlike the single, complete and ward methods that had cophenetic correlation coefficient of 0.90, 0.88, and 0.73, respectively. High cophenetic correlation coefficient indicates that the resulting dendrogram is a good fit of the reality. The dendrogram formed, using UPGMA (average) clustering method clustered the study sites broadly into two groups that are fairly close (Figure 3). In the first cluster A1, grouped   (Table 4). Seventy-nine percent of the medicinal plant species have local names in one or more local languages of the studied districts that are also sometimes used similarly or with a little differed intonation among the communities. In a few cases one local name is used for many species that have similar medicinal use. For example, local name 'Kureshe' was used for Crinum abyssinicum, Sauromatum venosum, and Tacca leontopetaloides which are used to cure livestock ailments (anthrax and blackleg).

Relative healing potential of medicinal plants: Rank order priority (ROP)
High ROP value indicates good healing potential of a plant and the tendency of informants relying on specific medicinal plant species for treating the respective diseases reported.
In the present study, the results showed that the number of plants (80 plant species, 33%) that score ROP value greater than 50% is considerably smaller than that of plants with ROP < 50% (164 plant species, 67%), despite the fact that these plants are with more frequent uses. This may be probably due to the diminishing popularity of many of the herbal medicine used among the study groups as stated in Friedman et al. [35]. It is also probable that the peoples are becoming selective and only use plants that are accepted in the culture as being more effective. Plants with computed ROP value at least 50% are referred here as culturally important medicinal plant species. Of 80 culturally important plant species, 37 species attained ROP values equals to 100% and might be taken as highly regarded and widely used in the community. These plants are widely known in the community and believed to be more effective [40,41] (Silti, Qebena). The high FL for these species indicated the outstanding choice of informants for treating specific illness [41]. High FL also indicates the similarity of use reports for a given species whereas low FL are obtained for plant species that are used for many different purposes [40]. In these use categories, high number of medicinal plants is used in the treatment of abdominal pain. This could suggest the high prevalence of the disease in the study districts. This might also indicate that peoples usually use more plants (communicate the knowledge) in order to cure frequently occurring ailments.
In deed this support the mere fact that a problem leads to a solution.

Cross-cultural view of medicinal plants use similarity among the study groups
Computed dissimilarity coefficient using all cited medicinal plants was above 0.5, reflecting the fact that co-presence of plant species cited were less than 50%. By using this value, the clustering displayed the similarities between the studies sites as mostly

Nomenclature of medicinal plants -Ethnotaxonomy
Local languages are essential for transferring locally specific traditional knowledge that is vital for conserving the local environment and respective uses [52]. Local plant names occasionally had information for understanding their use or other property of the plants.
This was reflected in the present study as some plants (4 plant species) were named after the ailment which the plants are used against and few more names (19 plant species) indicate the morphology (growth form, truck color and leaf shape), being poisonous, or test and odor. Naming system using the disease treated and colors was similarly reported in an ethnobotanical study of Kafficho, Ethiopia by Awas and Demissew, [6].

Ethics approval and consent to participate
Written permission to conduct the research was obtained from the respective zone and district administrative officials. Prior informed consent was obtained from each informant before every interview.

Availability of data and materials
All data generated or analysed during this study are included in this published article and its supplementary information files are attached as Additional file 1, 2 & 3.   Jaccard similarity indices (0-1), 1 = similar; 0 = dissimilar. Note that the matrix is symmetrical about the diagonal Table 3.