The study was conducted in Tshothang Chiwog, Lauri Gewog, South-Eastern Bhutan. The site lies between 27°7′30″ and 27°8′15″ N latitude and between 91°59′0″ and 92°0′00″ E longitude (Figure 1). The topography of the Gewog is rugged with an altitude range of 600 to 3,000 m above mean sea level. Soils in the southern belt of Bhutan covering the study location are mostly fluvisols, which together with cambisols (widespread in the medium-altitude zone) cover ~27% of Bhutan (FAO-UNESCO, 1977). The area experiences a warm humid subtropical climate. Mean maximum temperature recorded at Lauri during the period from November 2014 to September 2017 ranged from 12.1°C (January) to 25.3°C (July) and mean minimum temperature ranged between 11.6°C (January) and 23.6°C (July) (source: The National Center for Hydrology and Meteorology, The Royal Government of Bhutan, Thimphu, Bhutan). Mean annual precipitation during the reporting period was 1,052 mm, characterized by a monomodal distribution pattern with most of the precipitation occurring during the months from June to September. October to March is the dry season. Relative humidity at Lauri ranged from 63% (January) to 84% (June). The forests of the Chiwog broadly fall under the category of broad-leaf mixed coniferous forest.

Tshothang Chiwog consists of three villages, viz., Rashuthang, Makhawoong, and Tshothang with 57 households and a total population of 428. A representative area was selected for the study after a reconnaissance. The criteria used for selection of the study location were: (i) abundances of MPs, and (ii) local farmers' dependence on MP collection to improve their livelihood security.

A vegetation survey was carried out from 30 November 2016 to 12 January 2017 to assess the floristic diversity and structure of the forest, following the quadrat method. A total of 50 quadrats, of size 10 × 10 m, was randomly established to characterize the trees (girth at breast height, GBH ≥ 15 cm). Similarly, a total of 50 quadrats of size 5 × 5 m nested inside the original 10 × 10 m quadrats was used to document the shrubs species (low woody perennials generally <5 m height). All trees and shrubs, and climbers in each quadrat were enumerated. The girth of trees and shrubs were measured respectively at 1.37 m and 2 cm above ground using a measuring tape. For characterizing the herbs, 50 quadrats of size 1 × 1 m nested inside the 5 × 5 m quadrats were used. Specimens were collected from all quadrats and identified by comparing with published reports (Grierson and Long, 1991; Thinley, 2004; Wangchuk and Samten, 2009). All potential MPs of various life forms (trees, shrubs, herbs and ferns, and climbers) were also recorded.

Vegetation data of plant population were quantitatively analyzed for density (no./ha), abundance, frequency, abundance-to-frequency ratio (A/F ratio; Curtis and McIntosh, 1950). The A/F ratio was used to interpret distribution pattern of the species. Simpson diversity index (Simpson, 1949), Shannon-Weiner index, and important value index (IVI; as a sum of relative density, relative frequency and relative basal area for trees and shrubs and as a sum of relative density and relative frequency for herbs and climbers; Muraleedharan et al., 2005) were computed.

The ethnobotanical study was undertaken to gather information on the predominant MPs collected by the local people, characterize the traditional knowledge on therapeutic properties of important MP resources in the locality and to elucidate information on the ecological impacts of MP collection. The socioeconomic survey focused on gathering details such as mode of collection, prospects for cultivation and income generated. Information on ethnobotanical and socioeconomic aspects were elucidated through an open-ended focus group discussion and through semi-structured interviews during the month of December 2016, following standard methods for social surveys. The sample size for the socioeconomic survey was determined using Yamane (1967) formula. Accordingly, a total of 40 respondent farmers (heads of the families) were called in for the focus group discussion and interview. The participants were asked to assemble in the local community hall (tshokhang) on an auspicious day according to the Bhutanese calendar (i.e., 29 December 2016, when the farmers were free from field work). Aspects relating to the spectrum of MPs in their locality (using vernacular names), parts used, habit, and the therapeutic values were discussed. Information on ethnobotanical aspects was instantaneously recorded in a field notebook. In parallel, respondents were also interviewed on socioeconomic status using a semi-structured questionnaire, which covered aspects such as personal characteristics of the respondents (age, gender, education), mode of MP collection (e.g., who collects, major products collected, method, frequency, and season of collection), prospects for cultivation (which MPs are cultivated), annual income level (how much income is generated and where the products are sold), and the current scenario and management of MPs (species rank order of abundance, changing patterns of MP availability in the locality, factors causing depletion of MPs, and presence or absence of guidelines for MP collection/management). Information on ethnobotanical aspects and medicinal properties of the plants were supplemented with published information (State of an Environment, 2007; Wangchuk and Samten, 2009; Wangchuk et al., 2011). The data on socioeconomic variables were analyzed using percentage frequency distribution.

A total of 157 plant species (38 trees, 19 shrubs, 85 herbs and ferns, and 15 climbers), representing 74 families and 137 genera were identified from the study area (Appendix 1 in Supplementary Material). The most species-rich plant families were: Asteraceae (15 spp.), Poaceae (nine spp.), Solanaceae (seven spp.), Polygonaceae and Urticaceae (five spp. each). Of the total plants encountered in the study area, only 69 species (14 trees, 10 shrubs, 38 herbs and ferns, and seven climbers), belonging to 41 families, and 69 genera were medicinally important (Table 1). The most species-rich families of medicinal plants were: Asteraceae (eight spp.), Apiaceae (four spp.), Polygonaceae, Brassicaceae, Zingiberaceae, and Urticaceae (three species each).

The total density of trees and shrubs was 784 and 1,448 individuals per hectare with a corresponding stand basal area of 24.86 and 5.38 m² ha⁻¹, respectively (Appendix 1 in Supplementary Material). Mountain bamboo (Borinda grossa: 66 trees ha⁻¹), flowering or evergreen ash (Quercus griffithii: 62 trees ha⁻¹), and Himalayan blue pine (Pinus wallichiana: 42 trees ha⁻¹) contributed the three top tree densities and titeypati (Artemisia dubia: 216 individuals ha⁻¹), Indian stinging nettle (Girardinia diversifolia: 208 individuals ha⁻¹) and blackberry (Rubus fruiticosus: 96 individuals ha⁻¹) represented the three top shrub species at the study site. The total per hectare count of herbs/ferns was 1,91,800 and that of climbers 330. Among the herbaceous plants, S. chirata was the most abundant (13400 plants ha⁻¹), followed by saim weed (Chromolaena odorata: 7200 plants ha⁻¹), P. polyphylla (6800 plants ha⁻¹), and elatostema (Elastostema platyphyllum: 6200 plants ha⁻¹). Asteraceae accounted for 23,600 of the total density of 1,91,800 individuals ha⁻¹ (Appendix 1 in Supplementary Material).

Importance Value Index (IVI) expresses the overall importance of a species in a community. Q. griffithii, P. wallichiana, sunda oak (Lithocarpus elegans) and Chinese sumac (Rhus chinensis) exhibited the three to four top IVI for tree species (Appendix 1 in Supplementary Material). Similarly, A. dubia and S. chirata represented the highest IVI for shrubs and herbs, respectively (Appendix 1 in Supplementary Material). R. cordifolia showed the highest IVI among climber species. All plant species showed abundance-to-frequency ratio (A/F) ≥0.05, except for P. polyphylla (Melanthiaceae), which had an A/F-value of 0.04.

The floristic diversity (Simpson's Diversity Index, Shannon-Wiener Diversity Index) and important attributes (Density, Important Value Index, and Equitability) of all categories of plant species are summarized in Table 2. Simpson's diversity indexes were the highest for the herbaceous species (0.97) and the lowest for climbers (0.92) at this study site. Shannon index (H′) values ranged from 3.69 (climbers) to 5.82 (herbs/ferns).

Girth frequency distribution showed a characteristic inverse J-shaped distribution pattern implying a negative exponential relationship between density and size classes, implying adequate regeneration potential for the woody flora (Figure 2). The GBH class of 15 to 29 cm had the highest number of individuals for tree species. Q. griffithii, B. grossa, L. elegans, walnut (Juglans regia) and prickly ash (Zanthoxylum bungeanum) were the predominant species in this size class (data not presented). The GBH distribution pattern of shrubs also showed an inverse J-shaped pattern, except for the GBH class 0 to 5 cm, which was truncated (Figure 3), implying poor regeneration status of the smallest size class, which is a matter of concern.

The ethnobotanical study documented 55 MPs representing 37 families, which included 27 herbaceous plants, 15 trees, eight shrubs and five climbers (Table 3). The most species-rich families were: Asteraceae (five spp.), Lauraceae and Polygonaceae (three spp. each). The data presented in Table 3 also indicate that MPs are important in the local health care delivery system. As much as 67.5% of respondents reportedly used MPs to cure various ailments like cold, cough, headache, body pain, wounds, blood pressure, etc. (data not presented). The farmers in the focus group have indicated as their opinions that S. chirata, A. dubia, C. odorata, and V. nepalense are the commonly used MPs for treating diseases like headache, cough and cold, body pain, and stomach ache. At the national scale, some of these plants are also used for drug discovery, which is known to cure a wide array of diseases. Local people deemed MPs complementary to western medicine.

Medicinal plants also have cultural significance. In fact, the traditional belief systems of the local people (e.g., giving protection from evil spirits) are intricately inter-woven with MP usage in this locality. Species like tree rhododendron (Rhododendron arboretum), oriental thuja (Thuja orientalis), Japanese cedar (Cryptomeria japonica), and A. dubia are used for making smoke (Saang) every morning and evening to ward off evil spirits. Consistent with the tenets of Mahayana sect of Buddhism, it is believed that smoke drives away evils and devils, and serves as a remedy for quenching thirst and hunger of evils and devils. Furthermore, several species of Cupressus spp. and Juniperus spp. are traditionally used for making the mainline or life tree (Sog-shing) in prayer wheels, stupa (Choertens) and statues. Circumambulating and rotating prayer wheels and stupa may accumulate good fortune, purify one's soul, make them healthier and enable them to live longer.

A total of 40 farmers were randomly interviewed (75% male and 25% female) to assess the socioeconomic status of MP collectors in Tshothang Chiwog. The average age of the respondents fell in the range of 23 to 78 years with a mean of 46.2 years. Only five of the total 40 respondents had completed primary education and the remaining 35 did not attend school. Average size of the household was four. In the opinion of the respondents, MP collection is an important income generating activity for the local community. Accordingly, farmers were actively engaged in collecting and selling MPs. Thirty one of our 40 respondents (77.5%) collected MPs. The most important commercial MPs were: S. chirata, P. polyphylla, R. cordifolia, and V. nepalense. MPs were predominantly collected during the winter season (December to February), either by father (38%), mother (5%), children (27%) or both father and children (8%). Farmers also sold the MPs either in the local markets of Woongthi and Jomotshangkha, or to the traders from the Indian State of Arunachal Pradesh. P. polyphylla and S. chirata were the chief MP resources sold to the MP traders of Arunachal Pradesh, which enable the collectors to fetch premium prices. The annual income of such farmers apparently ranged from 30 to 330 USD, and the higher returns improved their socioeconomic status, especially during the last couple of years.

Tshothang Chiwog is characterized by high floristic richness and diversity and abundant MP wealth (Appendix 1 in Supplementary Material; Tables 1–2). In general, the Himalayan forests are characterized by high species richness, diversity and play a significant role in providing many ecosystems services (Chettri et al., 2010). However, anthropogenic factors including climate change are affecting the structure and composition of these forest types and their capacity to provide ecosystem services, including the ability of these forests to provide MPs to meet the society's requirement (Hamilton and Radford, 2007; ICIMOD, 2009). The anthropogenic impacts on the structure and composition of forests are variable from locality to locality, depending on the magnitude of these disturbances. Although some studies have addressed this concern in the north-western Himalayan context (Pant and Samant, 2012; Hanief et al., 2016), such studies have been rarely undertaken in Bhutan (Wangchuk et al., 2016).

The total number of tree and shrub species (woody species richness) presently reported from the broad-leaf mixed coniferous forest of Tshothang Chiwog, Bhutan (Appendix 1 in Supplementary Material; Table 2) can be compared to that of Dochula Pass, Bhutan (123 species ha⁻¹; Wangda and Ohsawa, 2006) and Yakpagang Community Forest in Mongar, Bhutan (150 species ha⁻¹; Buffum et al., 2008). Although 69 MP species occurred at the site (Table 1), only five MPs were commercially exploited (author's observations). The top most species in this respect were: S. chirata, P. polyphylla, R. cordifolia, P. confertiflora, and V. nepalense. The total number of MPs in the present study was greater than that of what Namsa et al. (2011) reported for Sikkim (50 plants species belonging to 29 families used for treating 22 human and four veterinary ailments) and lower than that of what Wangchuk et al. (2011) earlier reported for Bhutan (113 lower elevation medicinal plants). Species composition is often determined and regulated by an array of ecological, environmental, edaphic and anthropogenic factors (Muraleedharan et al., 2005). The occurrence of a large number of plant species in the study area is presumably because of the prevalence of many micro-habitats in the mountainous terrain. The density of trees and shrubs reported from the study site is similar to that of the broad-leaf forests of Dagana Dzongkhag, Bhutan (662 individuals ha⁻¹; Tenzin and Hasenauer, 2016) and the stand basal area similar to that of the logged over plots of Chimithanka within Gidakom Forest Management Unit, Bhutan (29.77 m² ha⁻¹; Covey et al., 2015). The stand basal area, however, was less than that of a humid mixed conifer forest at Dochula Ridge Top, Bhutan (66.4 m² ha⁻¹; Wangda and Ohsawa, 2006).

The plant distribution pattern in the study location ranged from random to contagious (highly aggregated distribution). According to Odum (1971), contagious distribution is the commonest, random distribution is found only in a uniform environment, and regular distribution occurs when severe competition occurs between individuals. Hanief et al. (2016) reported similar observations (A/F: 0.01 to 0.08) from the Darhal watershed in Jammu and Kashmir Himalayas, India.

Species diversity contributes to ecosystem richness (Muraleedharan et al., 2005). Herbaceous flora exhibited the highest diversity and species richness followed by trees, shrubs, and climbers, as expected. The Simpson and Shannon-Weiner diversity index ranged from 0.92 to 0.97 and 3.69 to 5.82, respectively (Table 2). High diversity in the study site can be explained by the low to moderate levels of disturbances occurring at the site such as MP collection, grazing etc. This is perhaps consistent with the “intermediate disturbances” hypothesis of Connell (1978), according to which stress limits species at the lower end of the disturbance gradient, whereas increased competition is the limiting factor at the upper end of the gradient. Tenzin and Hasenauer (2016), reported similar diversity (D = 0.88 and H′ = 3.39) for the broad-leaf forests of Dagana, Bhutan. Overall, the floristic composition, diversity, and vegetation structure are comparable to other studies within Bhutan and the north-western Himalayan region (Buffum et al., 2008; Wangchuk, 2013; Wangchuk and Gurung, 2016).

The poor regeneration status of shrub species in the lower size classes (Figure 3) is presumably because of occurrence of constraints like ground fire, browsing, and livestock grazing, which affects the shrubs more than the trees. According to Narayanan and Swarupanandan (1996), fire kills the young seedlings and encourages the invasion of seedlings and saplings. It also causes damages and loss of species richness and diversity (Buffum et al., 2008; Wangchuk, 2013; Wangchuk and Gurung, 2016). Cattle and yaks are major sources of grazing pressure in the study area (author's observations). It was also reported that farmers collect shrubs like Sikkim knotweed (Aconogonum molle) and A. dubia for fodder and green manure purposes.

An array of MP resources are utilized in the study locality and adjoining areas of Bhutan for curing various ailments and the indigenous knowledge system appears to be profound in this respect (55 species; Table 3). The number of MP species presently reported, however, was marginally lower than that from the results of the vegetation study (69 species; Table 1). Such differences are plausible in view of the differences in experience and knowledge of the respondents and the investigators. The concordance between ethnobotanic and vegetation studies for the most species-rich families was also less evident. For example, Asteraceae, Lauraceae, and Polygonaceae formed the dominant plant families (with more than three plants per family) according the ethnobotanical survey (Table 3), while the species-rich families of medicinal plants according to the vegetation analysis were: Asteraceae, Polygonaceae, Apiaceae, Brassicaceae, Zingiberaceae, and Urticaceae (Table 1).

Apart from the medicinal uses, MPs are also used in drug discovery in the indigenous Bhutanese medical systems. Indeed, there are two indigenous medical systems prevailing in the country. These are: (i) the formalized traditional medicine system- gSo-ba rig-pa (pronounced as So-wa Rig-pa) and (ii) the traditional local healing practices or folk medicine (Nang-pai mempa). gSo-ba rig-pa is a codified medical system in the country, which is estimated to use over 200 species of MPs, 35 different types of animal parts and 18 different types of minerals to produce 103 types of drugs (Jamphel, 2011; Choden and Dorji, 2014). The traditional healing practices, however, are transmitted through oral traditions from either father to son or from elderly to young and lacks proper documentation; but uses a wide spectrum of MPs. Regarding the choice of MPs in the indigenous medical systems, factors such as ease of collection, accessibility, and affordability appear to be major determinants (Pannozzo et al., 2012). Results of our ethnobotanical study (e.g., medicinal and other uses of plant resources: Table 3) are comparable to what Namsa et al. (2011) documented previously for the Monpa tribe of Kalaktang, Arunachal Pradesh, India, and Prakash (2011) for the tribals of Uttar Pradesh in India, implying strong similarities in the usage of MP resources by the tribal communities of the Eastern Himalayan belt. Apart from this, MPs in Bhutan have been a cardinal source of food and food supplements, socioeconomic and religious uses too (Wangchuk et al., 2008; Wangchuk and Olsen, 2011).

Traditionally, the economic well-being and social status of the rural community of Lauri Gewog is dependent on MP collection (Pannozzo et al., 2012). Accordingly, farmers were actively engaged in collecting and selling MPs. However, the way in which local people collect MPs are dependent on the commercial value it fetches in the market. Although 55 MPs were documented through participatory discussion (Table 3), only five species are commercially exploited. These are: S. chirata, P. polyphylla, R. cordifolia, V. nepalense, and P. confertiflora. The annual income of farmers from MP collection ranged from USD 30 to 330. Wangchuk and Gurung (2016) reported that farmers of Lingzhi and Langthel of western and central Bhutan had also improved their socioeconomic status by the collection of MPs. Studies in Nepal (Ghimire, 2008), and South India (Muraleedharan et al., 2005) also reported similar trends.

The respondents also opined that MP resources in the Chiwog have been diminishing over the years. According to some estimates, in the year 2000, farmers collected as much as 20 tonnes of MPs, chiefly S. chirata. However, it dropped to less than five tonnes by 2006 (Pannozzo et al., 2012), presumably because of the reduced availability in the wild. The situation has further deteriorated now. S. chirata, I. griffithii, Indian bay leaf (Cinnamomum impressinervium), and P. polyphylla are apparently threatened in this locality, as informed by the respondents. Most farmers (45% respondents), cited the ban on shifting cultivation, illegal poaching, encroachment by others (highlanders), forest fires and grazing as probable reasons for the declining MP wealth. More importantly, it could be due to over-harvesting and unsustainable extraction. Although the declining resources warrant domestication of the prominent MPs, none of the respondents except one had successfully cultivated MPs, particularly S. chirata.

Most notably, the way in which local people collect MPs are dependent on their usages and the commercial value it fetches in the market. Farmers highlighted the challenges of lack of farm roads for carrying the produce to the markets and the distance to market as hindrances in getting reasonable returns. Similar difficulties and challenges were reported by Pannozzo et al. (2012) for marketing MPs in another location of Lauri Gewog. Poor accessibility to markets especially for Tamang in Sandhu Marunche and Kanche CF User Groups in Lamjung district of Nepal also resulted in greater dependence on intermediaries and lower sale prices for the MP resources (Gaire, 2005).

The method of MP extraction also has changed over a period of time. Traditionally, farmers of Tshothang Chiwog selectively collected tubers, fruits, or climbers using simple tools such as hoe and knife/sickle. Xhosa and Sotho people of South Africa, also used similar tools to collect MPs (Liu et al., 2007). And the traditional method of extraction is often regarded as eco-friendly. However, contradictory practices were observed in the study area. For example, whole trees were cut down to collect fruits of P. confertiflora and V. nepalense, primarily to avoid the effort of climbing (author's observation). Similarly, the bark was destructively removed from Cinamomum impressinervium. Rhizomes of P. polyphylla also were dug out extensively. These practices would eventually deplete the MP resources of Tshothang Chiwog. Other workers also reported that MPs in some localities of Bhutan have been shrinking over the years (e.g., Wangchuk and Tobgay, 2015). Rai et al. (2000) and Muraleedharan et al. (2005) also reported that overexploitation is a principal reason for the shrinking non-wood forest product resources in Sikkim Himalayas and Western Ghats of India, respectively.

Furthermore, there are no management plans, rules, and regulations for extraction, storage, and utilization of MPs in the Chiwog. The only rule available (formulated within the villages) and that is orally transmitted among the villagers is with respect to the time of collection (when and for how long to collect MPs) and the number of collectors (viz. how many people from each household are allowed to collect MPs). The respondents were also not aware of any collection guidelines for MPs. Interestingly, all respondents (100%) favored the formulation of management rules, strategies, and plans for MP extraction, implying that the farmers are well aware of negative anthropogenic impacts on MPs, although their practices do not corroborate with that line of thinking. The solution to vexed problems such as over-exploitation of MP resources in Bhutan and associated threats to local and regional biodiversity, may lie in carrying out advocacy and awareness programmes, formulating stringent policies for better MP resource management and domestication of the more valuable MPs (e.g., S. chirata and P. polyphylla).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.