Study Regions

Our nine GLORIA target regions span over 1500 km from the Hengduan Mountains of the Tibetan Autonomous Prefecture (TAP) in northwest Yunnan, China, through Bhutan and eastern Nepal, to western Nepal near Mt. Annapurna (Figure 1). Target regions were selected along this latitudinal, biogeographic, and climatic transect to elucidate the impacts of climatic change among regional patterns of biodiversity, precipitation, temperature, and resource use.

Global Observational Research Initiative in Alpine Environments (GLORIA)

Following GLORIA's multi-summit approach, each target region includes three to four individual mountain summits spanning an elevational gradient between treeline and permanent snowline (~3800–5000 masl) to represent the ecotones: subalpine-lower alpine, lower alpine-upper alpine, upper alpine-subnival, and subnival-nival ( Supplementary Material (etbi-34-03-02_supp.pdf), Table 1). Each summit is divided into eight summit area sections for general plant inventories and uses. Four 1 × 1 m quadrats are placed in the four cardinal directions at five vertical meters from the highest point of the summit; these quadrats provide detailed data on plant species abundance, frequency, and cover. Photographic data were also collected using the standardized methods of GLORIA ( http://www.gloria.ac.at/).

Environmental Data

Environmental data—latitude, longitude, elevation, aspect, slope, and temperature—were taken using GPS units, altimeters, compasses, clinometers, and temperature monitors according to GLORIA protocol. Total annual precipitation and average annual temperature data were compiled from national weather stations closest to each target region ( Supplementary Material (etbi-34-03-02_supp.pdf), Table 2) at comparable elevations. There are no government meteorological stations on mountain tops, so the closest weather station data are from elevations below actual target regions, thus underestimating the actual precipitation received on mountain summits.

Plant Identification

Alpine plants were identified in the field and confirmed with vouchers deposited at the Missouri Botanical Garden (MO) and/or national/local herbaria: the Shangri-La Alpine Botanical Garden (SABG) and Kunming Institute of Botany (KUN), Yunnan, China; the National Biodiversity Centre (THIM), Thimphu, Bhutan; and Tribhuvan University Herbarium (TUCH), Kathmandu, Nepal. Several flora and ethnobotanical references were used for plant identification, distribution, and use information ( Supplementary Material (etbi-34-03-02_supp.pdf)).

Ethnobotany Methods

In addition to conducting standardized GLORIA sampling, this study was carried out in collaboration with traditional doctors (e.g., Amchi trained in herbal medicine), knowledgeable indigenous peoples, and with reference to extensive botanical and ethnobotanical literature (see  Supplementary Material (etbi-34-03-02_supp.pdf)). After obtaining prior informed consent, information on specific plant uses, general habitat uses, and climate change mitigation were recorded. We allowed our indigenous collaborators to interpret “useful” species within their own TEK and cosmologies resulting in a wide range of plants reported to be used for food, medicine, grazing, wood, tools, construction, fiber, dyes, herbs/spices, decoration, poison, and spiritual purposes. The information presented here represents the responses of more than 350 informants interviewed over seven years (2005–2012; see individual studies cited in introduction for sampling and demographic data).

Statistical Analyses

To assess diversity (species number and evenness combined) among and between Himalayan target regions, Shannon-Weiner (H′) indices of diversity were calculated for each individual summit and at each target region (three to four summit units) for species percent cover, frequency, and abundance:

PC-ORD (McCune and Mefford 2006) was used to order species presence/absence, frequency, and percent cover data using the Sorensen index of similarity and Non-metric Multidimensional Scaling (NMS) (McCune and Grace 2002). These methods assist in visualizing the level of similarity among target regions within the dataset. Canonical correspondence analysis (CCA) and a Monte Carlo test of significance (998 runs) tested the significance of environmental variables (aspect, slope, elevation, and precipitation) in explaining variation in species data. Indicator species analysis was performed using PC-ORD following the method of Dufrêne and Legendre (1997) to calculate indicator values, or the level of taxa specificity to a particular location. Significance was determined using a Monte Carlo test of significance (4999 permutations).

Two-way analyses of variance (ANOVA) using R statistical programming (R Core Team 2012) tested the influence of independent variables—ecotone and location—on dependent variables, which include number of taxa (species richness), H′, number of endemic taxa, and percent endemism (transformed by arc sin square root to normalize data). Analyses were run on Himalayan target region data alone as well as on a combined dataset of 17 European GLORIA target regions (Pauli et al. 2012) and our nine Himalayan regions to compare species richness and endemism between two very different alpine areas—Europe and the Himalaya—both experiencing changes in alpine vegetation in response to climatic change. One-way ANOVA in R was also used to assess differences in plant use by country (three Nepal regions, two Bhutan regions, and four Tibetan China regions).

Using meteorological records from national weather stations ( Supplementary Material (etbi-34-03-02_supp.pdf), Table 2), linear regressions of temperature and precipitation change over approximately 60 years (1950–2010) were calculated using Microsoft Excel (2007). For NMS and CCA, 20 year averages of annual precipitation were used to better represent present conditions.

Climate Change in the Eastern Himalaya

Average annual temperatures near target regions have increased by approximately 1.5°C over the last 60 years (1950–2010) confirming global climate change models; total annual precipitation has increased by an average of 362 mm (Figure 4). Global warming and increasing precipitation is clearly evidenced in the eastern Himalaya; there is great cause for concern over Himalayan climate change.

Figure 4.

Himalayan climate change: In support of global climate change models, average annual temperatures (a) near target regions have increased by approximately 1.5°C over the last 60 years (1950–2010); total annual precipitation (b) has increased by an average of 362 mm. (Meteorological station data listed in Supplemental Material, Table 2). Bhutan weather data were available only as single point averages (not graphed).

Alpine Flora

The number of taxa sampled at all target regions totaled 762 with 373 (49%) endemic to the Himalaya ( Supplementary Material (etbi-34-03-02_supp.pdf), Table 3). Target regions have more taxa, more endemics, and more diversity at the lowest (subalpine-lower alpine) ecotone just above treeline (Figure 5); however, percent endemism is not significantly different among ecotones ( Supplementary Material (etbi-34-03-02_supp.pdf), Table 4). The number of taxa, number of Himalayan endemics, and Shannon-Weiner index of diversity (H′) differed significantly among target regions with the greatest number of taxa and endemics found in western Nepal and the greatest diversity found in western Bhutan ( Supplementary Material (etbi-34-03-02_supp.pdf), Table 1). The number of taxa, number of endemic taxa, and percent endemism at Himalayan GLORIA target regions are significantly greater than at European target regions (Figure and Supplemental Material Table 4) demonstrating the extraordinary richness of Himalayan alpine flora.

Figure 5.

Comparisons of Himalayan and European alpine flora—number of taxa and number of endemics. The Himalaya have significantly more taxa, endemic species, and percent endemics (see  Supplementary Material (etbi-34-03-02_supp.pdf), Table 4 for statistics). Ecotones are comparable: (a) subalpine-lower alpine; (b) lower alpine-upper alpine; (c) upper alpine-subnival; and (d) subnival-nival. However, these ecotones are found at strikingly different elevations (Europe: between ~1000–3000 masl; Himalaya: between ~4000–5000 masl).

Ordinations and Indicator Species

Non-metric multidimensional scaling (NMS) produces high correlations for the first three axes (p = 0.001;  Supplementary Material (etbi-34-03-02_supp.pdf), Table 5) and orders summits similarly whether based on species presence, frequencies, or cover (Figure 6a, b, and c, respectively with each individual mountain summit represented by a single point). Species composition is distinguished by precipitation, target region, and country (Figure 6). Although geographic groupings may be overemphasized by NMS due to GLORIA's repetitive sampling design, results show species composition to be more similar to summits within the same target region and country than to summits located within the same ecotone.

Figure 6.

Ordinations of Himalayan Alpine Vegetation: non-metric multidimensional scaling (NMS) ordered summits similarly whether based on (a) species presence-absence, (b) frequencies, or (c) percent cover. Canonical correspondence analysis (CCA) also shows similar trends based on (d) species presence-absence, (e) frequencies, and (f) percent cover. Elevation and precipitation best determined vegetation similarities (p = 0.001), although occasionally (f) slope and aspect (not shown) were also significant. Vegetation is distinguished by target region (from east to west: Da Xue Shan [DXS], Run Zi La [RZL], Ma Ji Wa [MJW], and Mei Li Shui [MLS] in TAP, China; Tampela, Wangchuck Centennial Park [TPL] and Jomolhari, Jigme Dorji Wangchuck National Park [JML] in Bhutan; Kangchenjunga Conservation Area [KCJ], Langtang National Park [LNP], and Manang, Annapurna Conservation Area [MNG] in Nepal), precipitation (precipitation gradient labeled from low to high in a and b), and biogeography (countries: Bhutan, Nepal, or TAP designated by ovals). Ecotones designated by shape: ♦ subalpine-lower alpine (~4000 masl); • lower alpine-upper alpine (~4300 masl); ▪ upper alpine-subnival (~4700 masl); ▴ subnival-nival (~5000 masl) show little similarity in species composition among target regions.

Canonical correspondence analysis (CCA) constrained the ordination of target regions and species by their relationships to environmental variables and produced highly significant correlations between species composition and both elevation and precipitation (p = 0.001;  Supplementary Material (etbi-34-03-02_supp.pdf), Table 5), although occasionally slope and aspect were also significant (Figure 6d, e, and f).

Indicator species are species that distinguish an individual summit or target region from all other summits and regions. All nine Himalayan target regions and nearly all summits within each region (29 out of 33 total summits) have significant indicator species (p < 0.05;  Supplementary Material (etbi-34-03-02_supp.pdf), Table 6), illustrating the unique vegetative character of each region and summit. Additionally, indicator species distinguishing Himalayan target regions and summits are most often Himalayan endemic species (74% or 51 endemics out of 69 significant indicator species) indicating the distinct vegetation that characterizes the Himalayan flora as well as each region and summit.

Plant Use

Across all Himalayan target regions, half of the alpine flora is reported by our local collaborators as being used, predominantly for medicines and animal grazing (Figure 3;  Supplementary Material (etbi-34-03-02_supp.pdf), Table 7). Although Nepal has the highest percentage (68%) of useful plants, there is no statistical difference among countries in percentage of useful plants as reported by local collaborators (df = 2; F = 0.42; p = 0.66).

Climate Change Mitigation

Interviews with and observations of Himalayan peoples managing their natural resources through TEK uncovered at least five major ways that they mitigate climate change: 1) biodiversity conservation; 2) afforestation; 3) agroforestry; 4) soil amendment; and 5) a carbon negative livelihood. Himalayan peoples are not always aware that their TEK is mitigating climate change yet it does and should be noted and potentially emulated elsewhere.

Himalayan Alpine Vegetation and Climate Change

Climate change models for the eastern Himalaya predict greatly increased temperatures and precipitation (e.g., Solomon et al. 2007). These predictions are supported by weather station data showing temperature and precipitation increases over time (Figure 4). Elevation is also a proxy for temperature (~6°C decrease/1000 m elevation; Fang and Yoda 1988). Our results show the significance of elevation and precipitation in determining Himalayan alpine vegetation, and in turn corroborate predictions that climate change—increasing temperature and precipitation—will greatly affect Himalayan diversity. The predominance of Himalayan endemics among significant indicator species ( Supplementary Material (etbi-34-03-02_supp.pdf), Table 6) reveals the importance of endemic plants in differentiating alpine vegetation. These threatened species may be particularly affected by climate change if, as in Europe (Gottfried et al. 2012; Grabherr et al. 1994; Pauli et al. 2012), they are outcompeted by faster growing, more aggressive lower elevation species extending their range to higher elevations with a changing climate. Most suitable alpine habitat in the Himalaya lies below 5000 masl, above which soil is scarce and scree habitats dominate. The outlook for useful, rare, and endemic Himalayan alpine species, left with nowhere to go, is of great concern for conservation.

Comparative Himalayan and European GLORIA Results

The extraordinary species diversity in the eastern Himalaya (Figure 5;  Supplementary Material (etbi-34-03-02_supp.pdf), Tables 1 and 3) is largely due to extrinsic environmental factors including temperature, precipitation, elevation gradients, and biogeographic mixing of independently diverse temperate, tropical, and high Tibetan plateau floras; these factors are not independently monitored by GLORIA methodologies. Himalayan plant endemism is high (49% of all species found;  Supplementary Material (etbi-34-03-02_supp.pdf), Table 3) and indicates the unique and independently evolved flora in the Himalaya.

The present study greatly extends and agrees with our previous work (Salick et al. 2009)—biogeography, elevation (with covariable temperature), and precipitation determine alpine vegetation. These results are also confirmed elsewhere (Giorgi et al. 1997; Gottfried et al. 2012; Kohler and Maselli 2009; Pauli et al. 2012; Thuiller et al. 2005; and others). Our previous studies using repeat photographs show observable effects of climate change: tree and shrub lines are moving up mountains; glaciers are melting; and alpine meadows are decreasing in size (Moseley 2011; Salick and Moseley 2012; Salick et al. 2005). Local inhabitants relate similar observations (Byg and Salick 2009; Konchar et al. in prep.). Furthermore, this study uniquely adds an abundance of quantitative data that will be used to monitor the effects of Himalayan climate change on alpine vegetation into the future.

The Global Observational Research Initiative in Alpine environments colleagues in Europe (Gottfried et al. 2012; Grabherr et al. 1994; Pauli et al. 2007, 2012; and others: see  www.gloria.ac.at) have re-monitored 66 mountain summits at 17 target regions from northern Scandinavia to Greece and from the Scottish Cairngorms to the Russian Urals. Their results show that vegetation extends to higher elevations than previously, that species numbers have increased (except where precipitation is decreasing), and that rare and endemic species make up a smaller proportion of the flora. After a seven year interval, we have just begun to re-monitor the nine Himalayan GLORIA target regions (33 mountain summits). Since temperatures and precipitation are expected to increase yet more in the Himalaya, we hypothesize that Himalayan alpine vegetation will be more strongly affected by climate changes than European alpine vegetation; this hypothesis awaits further study.

Climate Change Mitigations through Traditional Ecological Knowledge

This study substantiates previous work, quantifying half the Himalayan alpine flora as useful, predominantly as medicinal and grazing plants (our work: Byg and Salick 2009; Ghimire et al. 1999, 2005a, 2005b, 2006; Salick et al. 2005; Salick et al. 2009; Salick et al. in press; Salick and Byg 2007; Salick and Moseley 2012; Salick and Ross 2009a; and others' work: Ives 2004; Vedwan 2006; Vedwan and Rhoades 2001). Here, we add a new perspective on human dimensions of climate change in the eastern Himalaya by concentrating on climate change mitigation by Himalayan peoples through TEK.

Few traditional people in the Himalaya understand the causes of climate change. Although they are keenly aware of climate change itself—especially warming, melting glaciers, and unpredictable rain patterns (Byg and Salick 2009)—many people attribute spiritual causes with prayer and propitiation as solutions (Salick et al. 2012). Nonetheless, Himalayan peoples actively mitigate climate change through implementation of TEK. In particular, we found that factors with mitigating effects include sacred site conservation, afforestation, tree crops, soil carbon sequestration through incorporation of mulch and manure, and carbon negative livelihoods. Traditional ecological knowledge in and of itself does not recognize climate change nor its mitigation, but TEK does provide valuable methods of climate change mitigation. These TEK methods of climate change mitigation are valuable not only to Himalayan peoples but also to the world.

Often, discussions of climate change mitigation are dominated by “Reducing Emissions from Deforestation and forest Degradation'' (REDD; e.g., Griffiths 2008; Köhl et al. 2009). However, climate change is too great a challenge and solutions too few and tenuous for us to narrow our options to these two coarse scale variables. In addition to scientific solutions, TEK from indigenous peoples around the world has much to offer (Nakashima et al. 2012). In the Himalaya alone, we have uncovered five prominent climate change mitigation practices that could be adopted elsewhere.

Sacred Himalayan Sky-Island Archipelago for Cultural Heritage and Conservation of Alpine Flora

If the projections from our data are correct—suggesting that climate change, especially temperature and precipitation, threatens the endemic Himalayan alpine flora—then what are our options for conservation? Developing conservation strategies that take climate change into consideration is challenging. Short of reducing climate emissions, fixing excess CO₂ already in the atmosphere, and rapidly developing non-carbon based energy systems—undeniably the only sustainable solution to climate change—conservationists are recommending strategies such as migration corridors (Malhi et al. 2008) and assisted migration (McLachlan et al. 2007). If climate change continues to cause elevation shifts in plant communities and phenology to the threshold of local extirpations and increased biological invasion, neither of these options seems feasible in the Himalaya. The reproductive constraints of many alpine plant populations inhibit reproduction, dispersal, and colonization (e.g., Law et al. 2010; Poudeyal and Ghimire 2011) and corridors would have to span huge elevation gradients over which alpine plant populations cannot migrate.

We suggest a biocultural conservation strategy that unites cultural practices and biodiversity conservation. Sacred peaks abound in the Himalaya, surround most of our research summits (Figure 2), and form an archipelago of “sky islands” among which alpine vegetation might disperse and migrate. Our previous work in the eastern Himalaya has shown that biodiversity is well conserved in sacred areas (Anderson et al. 2005; Salick and Moseley 2012; Salick et al. 2007). We know that conservation strategies not supported by local populations do not work (e.g., Liu et al. 2001; Müller and Guimbo 2011). Sacred mountains are culturally supported and relevant to local ethnic populations (Salick in prep.; Salick and Moseley 2012). Nonetheless, there are still many challenges. Indigenous concepts of Himalayan sacred mountains are spiritual and do not include conservation (Salick et al. 2007; Salick et al. in prep.); sensitive cross-cultural and cross purpose negotiations would have to be judiciously embarked upon. Policy would have to be internationally integrated across an area where governments do not always have diplomatic relationships. Alpine floral areas are small (and getting smaller as lower elevation species move higher) and scattered across vast distances. Island biogeography holds that as island size is reduced, diversity diminishes and that distance determines the likelihood of dispersal (MacArthur and Wilson 1967; Simberloff 1974). Determining the necessary size and distances between these sky islands as a strategy for conservation would be an exercise in applied island biogeography yet to be tested.