Assessing changes in distribution of the Endangeredsnow leopard Panthera uncia and its wild prey over 2decades in the Indian Himalaya through interview-based occupancy surveys
A B H I S H E K G H O S H A L , Y A S H V E E R B H A T N A G A R , B I V A S H P A N D A V
KO U S T U B H S H A R M A , C H A R U D U T T M I S H R A
R . R A G H U N A T H and K U L B H U S H A N S I N G H R . S U R Y AWA N S H I
Abstract Understanding species distributions, patterns ofchange and threats can form the basis for assessing the con-servation status of elusive species that are difficult to survey.The snow leopard Panthera uncia is the top predator of theCentral and South Asian mountains. Knowledge of the dis-tribution and status of this elusive felid and its wild prey islimited. Using recall-based key-informant interviews we es-timated site use by snow leopards and their primary wildprey, blue sheep Pseudois nayaur and Asiatic ibex Caprasibirica, across two time periods (past: –; recent:–) in the state of Himachal Pradesh, India. Wealso conducted a threat assessment for the recent period.Probability of site use was similar across the two time peri-ods for snow leopards, blue sheep and ibex, whereas for wildprey (blue sheep and ibex combined) overall there was an% contraction. Although our surveys were conducted inareas within the presumed distribution range of the snowleopard, we found snow leopards were using only % ofthe area (, km). Blue sheep and ibex had distinct dis-tribution ranges. Snow leopards and their wild prey were notrestricted to protected areas, which encompassed only %of their distribution within the study area. Migratory live-stock grazing was pervasive across ibex distribution rangeand was the most widespread and serious conservationthreat. Depredation by free-ranging dogs, and illegal hunt-ing and wildlife trade were the other severe threats. Ourresults underscore the importance of community-based, land-scape-scale conservation approaches and caution against reli-ance on geophysical and opinion-based distributionmaps that
have been used to estimate national and global snow leopardranges.
Supplementary material for this article can be found athttps://doi.org/./S
Introduction
The distribution and habitats of some species have beennegatively affected by human activities and develop-
mental pressures globally (Sanderson et al., ). Largemammals, especially apex carnivores, are often consideredto be umbrella or flagship species for ecosystem conserva-tion (Simberloff, ). Large home ranges and habitatspecificity tend to make large mammals vulnerable torange contraction and extinction (Ceballos et al., ;Michalski & Peres, ). Understanding threats to largemammals and determining their occurrence and changesin distribution are thus important for conservation planningat large spatial scales (MacKenzie et al., ; Karanth et al.,; Taubmann et al., ).
Since the s India has experienced a % increase in itshuman population alongside rapid economic growth, whichhas negatively affected mammal species and their habitats(Madhusudan & Mishra, ; Das et al., ; Singh &Bagchi, ). Studies at national and regional levels have re-ported declines in distribution ranges of some common largemammalian species (Karanth et al., ; Pillay et al., ). Incomparison to the largemammals of the Indian peninsula, thedistribution and conservation status of the high-altitude largemammalian assemblagehave received little scientific attention(Gaston et al., ; Mishra et al., ). Reliable baseline in-formationondistribution range is lacking even for charismaticand threatened species such as the snow leopard Pantherauncia and its primary prey, blue sheep Pseudois nayaur andAsiatic ibex Capra sibirica, in India (Lovari et al., ;Lyngdoh et al., ; Snow Leopard Network, ).
The snow leopard, which is categorized as Endangeredon the IUCN Red List (Jackson et al., ), is the top
ABHISHEK GHOSHAL* (Corresponding author), YASH VEER BHATNAGAR†,KOUSTUBH SHARMA†, CHARUDUTT MISHRA†, R. RAGHUNATH andKULBHUSHANSINGH R. SURYAWANSHI† Nature Conservation Foundation, 3076/5,IV Cross, Gokulam Park, Mysore, Karnataka 570002, IndiaE-mail [email protected]
BIVASH PANDAV Wildlife Institute of India, Dehradun, Uttarakhand, India
*Also at: Snow Leopard Trust, Seattle, Washington, USA; Wildlife Institute ofIndia, Dehradun, Uttarakhand, India; and Saurashtra University, Rajkot,Gujarat, India†Also at: Snow Leopard Trust, Seattle, Washington, USA
Received February . Revision requested April .Accepted June . First published online October .
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
predator and flagship species for conservation of the IndianHimalaya (Snow Leopard Network, ; Bhatnagar et al.,). The Indian Greater and Trans-Himalaya form thesouthern limit of the global snow leopard distributionrange and constitute c. % of the estimated global snow leop-ard range (Snow Leopard Working Secretariat, ).Amongst the five Indian states where snow leopards occur,Himachal Pradesh has c. one-fifth of the potential snow leop-ard habitat (,–, m altitude; Project Snow Leopard,; Snow Leopard Network, ), corresponding toc. , km. Within Himachal Pradesh a large, contiguouspatch of potential snow leopard habitat is constituted by thefour regions of Kinnaur (, km at ,–,maltitude),Lahaul (, km), Spiti (, km) and Pangi (, km).These regions (, km) constitute % of the potentialsnow leopard habitat in Himachal Pradesh. Since the sconsiderable socio-economic changes have occurred in thesefour regions, including improvements in connectivity and theadvent and spread of cash crops (e.g. green peas in Spiti andLahaul, and apples in Kinnaur and Spiti) (Mishra, ; Ranaet al., ; Basannagari &Kala, ). Increased road connect-ivity along international borders has led to a rise in infrastruc-ture projects and unregulated tourism, especially since ,when Kinnaur and Spiti were opened to Indian and inter-national tourists (Snow Leopard Network, ).
We conducted a threat assessment and evaluated changesin the distribution of the snow leopard, blue sheep and ibex inthese regions over a period of decades. We used multi-season occupancy modelling (MacKenzie et al., ) ondetection/non-detection data collected by interviewinglocal people and corrected for imperfect detection acrosstwo time periods (Karanth et al., ; Pillay et al., ;Taubmann et al., ), past (–) and recent (–). The method facilitated analysis of data from past andrecent time periods based on recall of interviewees. Multiplerespondents from a sampling unit were treated as replicatesurveys for the primary sampling seasons, thereby accountingfor imperfect detection. Thus, sampling difficult-to-detectspecies across a large area and multiple time periods is pos-sible using this method. We present distribution maps of thesnow leopard, blue sheep and ibex for the two time periods,using appropriate covariates. We discuss the implications ofour findings in the context of snow leopard conservation pro-grammes at state, national and global levels.
Study area
The study area comprised , km at ,–, maltitude (potential snow leopard habitat, Snow LeopardNetwork, ) covering the contiguous potential snowleopard habitat across the Greater and Trans-HimalayaMountains in Himachal Pradesh (Fig. ). The vegetation inthe Trans-Himalaya is largely dry alpine steppe (Champion& Seth, ), with gentle-rolling uplands interspersed with
steep cliffs and rocky outcrops. The terrain is rugged, withcliffs and sharp ridgelines interspersed with alpine and sub-alpine tracts in the Greater Himalaya. The south-east andnorth-west extremes of our study area are characterized by arelativelymoist climate and tracts of forest.The entire region ischaracterized byPalaearctic conditions, with annual tempera-ture ranging from c. −°C in winter to c. °C in summer.The region is drained by three major rivers, the Sutlej inKinnaurDistrict, the Spiti in Spiti subdivision andpart of nor-thern Kinnaur, and the Chandra-Bhaga or Chenab in Lahauland Pangi subdivisions.
Given the difficult terrain, limited plant productivity andextreme climate, the local agro-pastoralist communitiesinhabit the landscape at low densities (one individualper km in Lahaul and Spiti; per km in Kinnaur; perkm in Pangi). Kinnaur, Lahaul and Pangi are inhabited pri-marily by followers of the Hindu religion, with Buddhists in-habiting only the uppermost villages, whereas Spiti isinhabited primarily by Buddhists. Their traditional depend-ence on natural resources for food (including wildlife), fuel,livestock grazing (sheep, goats, cattle, yaks, yak–cattle hy-brids), fodder, construction material and medicinal plantsis widespread and varies across the landscape (Mishra,). The regional economy has shifted from a subsistenceand barter-based system to being cash-based and market-driven (Mishra, ; Bhatnagar & Singh, ). The adventof the green pea Pisum sativum as a cash crop from on-wards and the opening up of Kinnaur District and Spiti sub-division to tourists in have accelerated these changes.Development and agricultural intensification are providingjob opportunities and attracting non-native labourers to thelandscape. Additionally, migratory herders from Shimla,Kangra and Chamba districts graze livestock in the high-altitude pastures across most of the study area during sum-mer (June–August).
Carnivores in the study area include the snow leopard,the wolf Canis lupus and the brown bear Ursus arctos at. , m altitude, and the common leopard Panthera par-dus andHimalayan black bearUrsus thibetanus at, ,maltitude (Jayapal & Ramesh, ; Mishra et al., ). Amongwild herbivores, the Himalayan tahr Hemitragus jemlahicus,Himalayan serow Capricornis thar, musk deer Moschusleucogaster and Himalayan goral Naemorhedus goral occurin the Greater Himalaya, and blue sheep and Asiatic ibexoccur in both the Greater and Trans-Himalaya.
Methods
Survey design and data collection
We used key-informant interviews with a structured ques-tionnaire to record occurrences of snow leopards andtheir prey for two time periods: – and –
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
(following Pillay et al., and Taubmann et al., ). Thesurvey was carried out during the summer season (June–September) of . The study area was divided into
grid cells (sites) of × km (area km) each, the areabeing similar to the estimated mean home-range size of thesnow leopard (Johansson et al., ).
The questionnaire (Supplementary Material S) con-tained two sections. The first dealt with information relatedto the attributes of the respondents that could affect reliabledetection of the study species (Table ), and the second con-tained questions regarding detections of the study speciesduring the two time periods, including the location, yearand approximate month of each detection.
Upon entering a village we first created a focus group oftwo or three people with thorough knowledge of the site orgrid cell. With their help we prepared resource maps(Suryawanshi et al., ), with approximate locations of set-tlements, agricultural areas, pastures, seasonal use of pastures,number of livestock and areas used for biomass extraction.This information was transferred to printed maps of thearea. During this exercise we did not record species detec-tion/non-detection, as the reports of one person could be in-fluenced by another in such a group. These respondents andothers were later interviewed individually by AG, specificallyto gather data on detection/non-detection of snow leopardsand their prey. The resource mapping facilitated the plottingof detection locations on a map. The respondents includedherders, active and former hunters, cattle herders, medicinalplant collectors, forest department officials, staff and contrac-tual workers, military personnel, tourist guides, porters andphotographers. Key informants were shown images of
study species among images of numerous other carnivoreand mountain ungulate species, such as the Tibetan wolfCanis lupus filchneri, brown bear, common leopard,Himalayan black bear, Himalayan tahr, Himalayan serow,musk deer, Himalayan goral, Kashmir stag or hangulCervus elaphus hanglu, markhor Capra falconeri, kiangEquus kiang, Tibetan gazelle Procapra picticaudata, Tibetanargali Ovis ammon hodgsoni, wild yak Bos mutus andLadakh urial Ovis vignei vignei. Respondents who failed toidentify study species from the images were excluded fromthe analyses, as were those who were unable to provide reli-able information on location of species detection. The iden-tity of the respondents has been kept confidential.
We identified conservation threats to the snow leopardand its wild prey in Kinnaur, Lahaul, Spiti and Pangi basedon discussion with the focus groups and personal observa-tions. The threats were ranked based on area of spread, inten-sity and urgency, following Margoluis & Salafsky (). Weidentified fourteen threats in four broad categories, namely,livestock–wild prey interactions, people–wildlife interactions,developmental activities and other human disturbances. For aparticular threat in a region, a score of –was assigned, with indicating low and severe. Thus, total ranking of a threat(sum of scores for area, intensity and urgency) in a regioncould be –. We considered a score of – to indicatelow threat, – medium threat, and – severe threat.
Data analyses
To define the past time period we chose two landmarkevents: the introduction of the green-pea crop in Lahaul &
FIG. 1 The locations of Kinnaur, Lahauland Spiti Districts and Pangi sub-divisionof Chamba District, in Himachal Pradesh,India. The shaded region denotes ,–, m altitude, the potential snowleopard Panthera uncia habitat inHimachal Pradesh.
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
Spiti in , and the opening up of Kinnaur and Spiti totourists in . We defined the recent time period as the-year period prior to our survey (–). As yearsis a considerably long period to assume population closurewithin sampled units, we assumed that the occupancy esti-mates would reflect the probability of the sites being used bythe study species in the respective time periods. This inter-pretation of the conventional occupancy estimate as theprobability of site use allows for changes within the sam-pling units and relaxation of the closure assumption(MacKenzie et al., ; Taubmann et al., ). This designassumes that the changes in probability of site use withineach sampling unit during each sampling period were ran-dom and described by the specific covariates applied.Probability of colonization (γ) was estimated as the prob-ability of a site that was not used by a species during thepast time period being used during the recent time period,whereas probability of extinction (ε) was estimated as theprobability of a site that was used by a species during thepast time period not being used during the recent time per-iod (MacKenzie et al., ). The probabilities of local col-onization and local extinction can be interpreted asprobability of distribution expansion and contraction, re-spectively, between the two periods.
Data on sightings of snow leopards, blue sheep and ibexgathered during interviews were arranged in a detection/non-detection (/) framework for past and recent time
periods. Each interviewee’s report from a particular sitewas assigned as a replicate survey within the site. Sitecovariates that could influence probability of site use by aspecies were modelled using logistic insertions (Table ).Topographic covariates such as mean altitude above meansea level, and proportions of rugged area in a site (terrainruggedness index) and vegetated area in a site (normalizeddifference vegetation index) were generated using ShuttleRadar Topographic Mission (SRTM) and Landsat data, re-spectively. As snow leopards are known to occur between, m (above the tree line) and , m (approximatelower limit of permafrost and limit of vegetation growth)in the Western Himalaya (Project Snow Leopard, ;Snow Leopard Network, ), we used mean altitude andits square to model probability of site use as a linear andquadratic function of altitude. We expected probability ofsite use to be higher in sites with a greater proportion of ter-rain ruggedness (Snow Leopard Network, ; McCarthy &Mallon, ). A higher proportion of vegetated area was ex-pected to favour the distribution of wild prey. Presence ofmigratory livestock grazing reflected areas that were usedfor grazing only during the peak summer (mid June–midAugust) and were far from villages, and hence not grazedby local livestock. Presence of migratory livestock grazingwas expected to lower the probability of site use by wildprey. The livestock population (i.e. the number of sheep/goats in a site) was expected to negatively influence
TABLE 1 Site and survey covariates used in occupancy analyses to assess changes in distribution of the snow leopard Panthera uncia andwild prey (blue sheep Pseudois nayaur and Asiatic ibex Capra sibirica) between two time periods (– and – in the Greaterand Trans-Himalaya of Himachal Pradesh, India (Fig. ), with description, variable type, and expected relationship with site use and de-tection probability.
DescriptionVariabletype
Expected relationship with site use and detectionprobability
Site covariatesAltitude Mean altitude Continuous Positive correlation with site use by study species
up to c. 5,200 m altitudeRuggedness Proportion of rugged area, using terrain
ruggedness indexContinuous Positive correlation with site use by study species
Normalized differencevegetation index
Proportion of vegetated area Continuous Positive correlation with site use by blue sheep &ibex
Migratory livestockgrazing
Presence or absence of migratory livestockgrazing
Binary Negative correlation with site use by blue sheepand ibex
Livestock population Number of sheep/goats Continuous Negative correlation with site use by blue sheepand ibex
Survey covariatesActive Whether respondent was active or not during
the primary survey periodsBinary Active respondents are more likely to encounter
study speciesTime spent Number of months in which the respondent
was active in the area of knowledgeContinuous Respondents spending more time in area of
knowledge are more likely to detect study speciesFamiliarity Number of years for which the respondent
was familiar with the area of knowledgeContinuous Respondents with higher familiarity are more
likely to detect study speciesAge Age of respondent in years Continuous Mixed effect on encounter of study speciesProfession Categorical variable representing sedentary,
semi-outdoor & outdoor workCategorical Respondents with outdoor work are more likely
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
probability of site use by wild prey. If the confidence intervalof the coefficient of a variable included zero we interpreted itas a statistically non-significant association.
Covariates that might have influenced the probability ofdetection of a species and its reporting to the researcherwere used as survey covariates to model detection probabil-ity (Table ). These included respondent age, profession,duration (in years) of familiarity with the area for whichthey were providing information, time spent in that area an-nually, and a binary variable indicating whether or not a re-spondent visited snow leopard and prey habitat during eachtime period.
We modelled probability of site use and detectionprobability through single-species multi-season models forsnow leopards, blue sheep and ibex, using Presence .(MacKenzie et al., ; Hines, ). We parameterizedour models to estimate empirically historical and recentprobabilities of site use and probability of distribution ex-pansion. Estimates of probabilities of distribution contrac-tion were derived. Models were ranked in ascending orderbased on the Akaike information criterion (AIC, Akaike,; Burnham & Anderson, ).
Our analyses yieldedmultiplemodels with similar AIC va-lues, thus indicating that no one model, by itself, could ex-plain the observed variation in the data adequately. Thus,following a multi-model inference approach (Burnham &Anderson, ), we considered a set of models with a cumu-lative Akaike weight of$. to yield the best approximatingmodel by using model averaging (Burnham & Anderson,; Symonds & Moussalli, ). Models that did not con-verge were dropped from the model set before model aver-aging. We plotted the model-averaged estimates of past andrecent occupancy to develop probabilistic maps of past andrecent distributions of the snow leopard and its wild prey.
Results
We carried out systematic key-informant interviewsacross Kinnaur, Lahaul, Spiti and Pangi. For analyses weused of these interviews, having excluded two respon-dents who provided unreliable information. Respondentsreported detection/non-detection of snow leopards andtheir wild prey from a mean of grids per respondent(range –). The mean number of respondents who reportedsighting snow leopards per grid was . ± SD . (range–), whereas the mean number who reported sightingwild prey was . ± SD . (range –). Detection datafor each species are summarized in Table . We developed competing single-species multi-season occupancy modelsfor the snow leopard, for all wild prey combined, forblue sheep and for ibex (Supplementary Table S).
Naïve estimates of the proportion of sites used by snowleopards in the past and recent time periods were . and
., respectively. Naïve estimates of the proportion of sitesused by wild prey in the past and recent time periods were. and ., respectively. When analysed individually,naïve estimates of the proportion of sites used by bluesheep in the past and recent time periods were . and., and by ibex . and ., respectively.
Detection probability
According to null models the detection probabilities of snowleopards, blue sheep, ibex and wild prey were . ± SE .,. ± SE ., . ± SE . and . ± SE ., respective-ly. When modelled with survey covariates, detection prob-abilities of snow leopards, wild prey, blue sheep and ibexvaried between the past (–) and recent (–)time periods (Table ). Age of respondent showed a weaknegative effect on detections of snow leopards, blue sheepand wild prey. Familiarity with the area of knowledge wasan important factor for ibex detection. Time spent eachyear by a respondent in the area of survey had little effecton detection probability. The profession of the respondenthad a positive effect on detection probability for all species.
Site use, and expansion or contraction of distribution:snow leopard
The model-averaged estimate of probability of site use bysnow leopards was . ± SD . (range . ± SE . to. ± SE .) for the past time period and . ± SD .(range . ± SE . to . ± SE .) for the recent timeperiod (Fig. a,b). Probability of site use by snow leopardsshowed weak positive correlation with the terrain rugged-ness index (βtri = . ± SE .; % CI −.–.) andaltitude (βalt.mean = . ± SE .; % CI −.–.)(Table ); however, the coefficients were not statistically sig-nificant. In the % of the area used by snow leopards, theprobability of site use by snow leopards was relatively high
TABLE 2 Number of detections, and number of respondents report-ing detections, of snow leopards, wild prey (cumulative detectionsof blue sheep and ibex), blue sheep and ibex in the Greater andTrans-Himalaya of Himachal Pradesh, India (Fig. ). Multiple re-spondents reported both blue sheep and ibex from more than onegrid, and therefore the detections of wild prey sum to more thanthe detections of individual species.
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
(. .). A relatively higher probability of site use (. .)was recorded in Spiti and in the majority of Kinnaur.Central and western Lahaul and Pangi had relatively lowprobabilities of site use by snow leopards (Fig. a,b).There was no clear evidence of expansion of the snow leo-pard’s distribution (γ = . ± SE .; % CI −.–.).The estimated probability of contraction of the snow leo-pard’s distribution was . ± SD ., indicating no clearevidence for overall contraction, and ranged between . ±SE . and . ± SE . (Supplementary Fig. S).
Site use, and expansion or contraction of distribution:blue sheep
Themodel-averaged estimate of probability of site use by bluesheep was . ± SD . (range . ± SE . to . ± SE.) for the past time period and . ± SD . (range .± SE . to . ± SE .) for the recent time period(Fig. a,b). Probability of site use by blue sheep showed anegative correlation with migratory livestock grazing(βmigrh =−. ± SE .; % CI −. to −.; Table ).There were no clear relationships between probability ofsite use by blue sheep and altitude (βalt.mean =−. ±
SE .; % CI −.–.) or ruggedness (βtri = . ± SE.; % CI −.–.; Table ). A high probability of siteuse (. .) by blue sheep was estimated in the majority ofSpiti and Kinnaur. South-western Spiti and all of Lahaul andPangi showed low probabilities of site use (, .) byblue sheep (Fig. a,b). We found no clear evidence of prob-ability of expansion of blue sheep distribution (γ = . ± SE.; % CI −.–.). The estimated probability of
contraction of blue sheep distribution was . ± SD .(range . ± SE . to . ± SE .), indicating no clearevidence for overall contraction (Supplementary Fig. S).
Site use, and expansion or contraction of distribution:Asiatic ibex
The estimated probability of site use by ibex was the same forpast and recent time periods (ψ = . ± SD ± .; range: .± SE . to . ± SE .; Fig. c,d). Probability of site useby ibex showed a positive correlation withmigratory livestockgrazing (βmigrh = . ± SE .; % CI .–.; Table ).The relationship between ibex site use and mean altitudewas unclear (βalt.mean =−. ± SE .; % CI −.–.)(Table ). High probability of site use (. .) by ibex wasfound in Lahaul and Pangi, and in south-western parts ofSpiti and Kinnaur. Central, south and eastern Spiti, and nor-thern, eastern and south-eastern Kinnaur showed lowprobabilities of site use (, .) by ibex (Fig. c,d). Therewas no clear evidence for expansion of ibex distributionacross the study area (γ = . ± SE .; % CI −.–.). The estimated probability of contraction of ibex distri-bution was . ± SD . (range . ± SE . to . ±SE .), indicating no clear evidence for overall contraction(Supplementary Fig. S).
Site use, and expansion or contraction of distribution:wild prey
The model-averaged estimate of probability of site useby wild prey (blue sheep and ibex combined) was
TABLE 3 Untransformed estimates of coefficients (β) from the top models of snow leopard, wild prey, blue sheep and ibex model sets.
Snow leopard Blue sheep Ibex Wild preyModel* β ± SE β ± SE β ± SE β ± SE
*Ψ, Probability of site use; γ, probability of distribution expansion; p, probability of detection for the past time period; p, probability of detection for therecent time period; psi, past probability of site use; psi, recent probability of site use. Site covariates: altitude, mean altitude of a site; altitude_squared,square of mean altitude; Ruggedness, proportion of rugged area in a site, derived from terrain ruggedness index; Migratory livestock grazing, presence ()/absence () of migratory livestock grazing in a site; Age, age of respondent in years; Familiarity, no. of years ormonths for which a respondent is familiar withthe area of knowledge; Time spent, time spent in months or days by a respondent in the area of knowledge per year; Prof, sedentary profession (e.g. officeworker); Prof, sedentary and outdoor profession (e.g. forest guard).
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
. ± SD . (range . ± SE . to . ± SE .)for the past time period and . ± SD . (range . ± SE. to . ± SE .) for the recent time period.Probability of site use by wild prey varied as a quadraticfunction of altitude (βalt.mean =−. ± SE ., % CI−. to −.; βalt.sq =−. ± SE .; % CI −. to−.), and increased with ruggedness (βtri = . ± SE .;% CI .–.). There was no clear relationship betweenprobability of site use by wild prey and proportion of
vegetated area (normalized difference vegetation index)(βndvi =−. ± SE .; % CI −.–.) or presence/absence of migratory livestock grazing (βmigrh =−. ± SE.; %CI−.–.; Table ). The probability of site useby wild prey was high (. .) in most of the study area.Central Lahaul, parts of northern Spiti and parts ofKinnaur (especially along the Sutlej River valley) showedrelatively lower probabilities of site use by wild prey(, .). There was no clear evidence for expansion of the
FIG. 2 (a) Past and (b) recent probabilityof site use (occupancy) by snow leopardsacross the Greater and Trans-Himalayamountains of Kinnaur, Lahaul, Spiti andPangi, Himachal Pradesh, India.
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
distribution of wild prey (γ = . ± SE .; % CI −.–.). The estimated probability of contraction of the distri-bution of wild prey was . ± SD . (SupplementaryFig. S), indicating evidence for a marginal overall contrac-tion of distribution (range . ± SE . to . ± SE .).The probability of contraction of the distribution of wildprey was relatively high (ε. .) in % of the sites.These areas were in western Lahaul, northern and easternSpiti, south-western Spiti and Kinnaur (SupplementaryFig. S).
Conservation threats
Threat assessment (Table ; Supplementary Fig. S) indi-cated the major threats across the study area were reducedprey as a result of competition with both local andmigratorylivestock, reduced prey as a result of subsistence hunting bylocal communities, illegal hunting and wildlife trade by bothlocal and immigrant labourer and migratory communities,
and depredation of wildlife and livestock by free-rangingdogs.
In Kinnaur the severe threats identified were reducedprey as a result of competition with local livestock and sub-sistence hunting by the local community, illegal hunting andwildlife trade by immigrant labourer and migratory com-munities, and depredation by dogs. Local livestock grazingwas practised in and around villages all across Kinnaur,whereas migratory livestock grazing was prevalent insouth-eastern and south-western Kinnaur (SupplementaryFig. S). Southern and south-western Kinnaur also experi-enced hunting and wildlife trade by local and migratoryherder communities and immigrant labourers. The depre-dation of wildlife and livestock by free-ranging dogs wasfound to be widespread.
In Spiti reduced prey as a result of competition with localand migratory livestock, and depredation by dogs, were se-vere threats. Local livestock grazing was widespread exceptin a few villages in eastern Spiti on the left bank of the SpitiRiver. Migratory livestock grazing was prevalent in north-
FIG. 3 Past and recent probability of site use (occupancy) by blue sheep (a & b) and Asiatic ibex (c & d) across the Greater andTrans-Himalaya mountains of Kinnaur, Lahaul, Spiti and Pangi, Himachal Pradesh, India.
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
TABLE 4 Threat ranking ( = low threat; = severe threat) based on area (how widespread is the threat; = least widespread; =most widespread), intensity (how severe is the threat; = lowseverity; = high severity) and urgency (how immediate/urgent is the threat; = least urgent; =most urgent), following Margoluis & Salafsky (), for Kinnaur, Lahaul, Spiti and Pangi,Himachal Pradesh, India (Fig. ). Light grey shading indicates a low level of threat, medium grey indicates a medium level of threat, and dark grey indicates a severe level of threat.
Threat
Kinnaur Spiti Lahaul Pangi (Chamba)
Area Intensity UrgencyTotalranking Area Intensity Urgency
Totalranking Area Intensity Urgency
Totalranking Area Intensity Urgency
Totalranking
Livestock–wild prey interactionsPrey reduction as a resultof competition with locallivestock
14 13 5 32 13 13 4 30 7 3 4 14 3 4 5 12
Prey reduction as a resultof competition with mi-gratory livestock
7 14 6 27 12 14 9 35 14 14 10 38 14 14 10 38
Prey reduction as a resultof disease transmitted bylivestock
1 2 3 6 5 6 3 14 2 2 2 6 5 3 4 9
People–wildlife interactionsRetaliatory killing ofsnow leopards & wildprey by local community
2 8 7 17 6 5 12 23 3 8 3 14 4 8 9 21
Prey reduction as a resultof subsistence huntingby local community
13 9 8 30 3 2 10 15 8 9 9 26 9 11 12 32
Illegal hunting & wildlifetrade by localcommunity
6 10 13 29 4 4 11 19 10 10 13 33 10 12 13 35
Illegal hunting & wildlifetrade by immigrantlabourer & migratorycommunities
12 11 14 37 7 8 13 28 12 13 14 39 12 13 14 39
Developmental activitiesImpacts of roads onsnow leopard habitat
4 7 10 21 8 9 8 25 4 7 6 17 7 6 7 20
Impacts of hydro-electricprojects on snow leopardhabitat
western, western and south-western Spiti (SupplementaryFig. S). Depredation of wildlife and livestock by free-ranging dogs is an emerging conservation threat in Spiti.
In Lahaul reduced prey as a result of competition withmigratory livestock, illegal hunting and wildlife trade bythe local community, and immigrant labourer and migra-tory communities, and depredation by dogs, were the severethreats. Migratory livestock grazing was pervasive at rela-tively high densities (Supplementary Fig. S). Areas grazedby migratory livestock were also prone to illegal huntingby migratory herders and depredation of wildlife by dogs(migratory herders are accompanied by guard dogs).Hunting was prevalent among the local community of west-ern Lahaul. Immigrant labour camps were abundant alongmajor roadways in Lahaul, and our respondents consistentlyreported labourers to be involved in illegal hunting andwildlife trade.
In Pangi the severe threats were reduced prey as a resultof competition with migratory livestock and subsistencehunting by the local community, illegal hunting and wildlifetrade by the local community and immigrant labourer andmigratory communities, and depredation by dogs. Similarto Lahaul, migratory livestock grazing was widespread inPangi and these areas experienced illegal hunting and dep-redation of wildlife by free-ranging dogs. Hunting for sub-sistence and for wildlife trade, by both local people andimmigrant labourers, was prevalent.
Discussion
We set out to examine changes in distribution of the snowleopard and its wild prey, and their current conservation sta-tus across the Greater and Trans-Himalaya Mountains ofHimachal Pradesh between two time periods, past(–) and recent (–). We used a recall-basedkey-informant survey to collect information on detection/non-detection of snow leopards and their wild prey overthe two periods. These data were analysed using a single-species, multi-season site-occupancy framework accountingfor imperfect detection. We recommend caution in inter-preting the effects of variables on site use as causation. Forexample, normalized difference vegetation index data wereused to represent the proportion of vegetated area in a site;however, these data need not represent palatable forage forwild prey, and may explain the absence of a significant posi-tive relationship between proportion of vegetated area andsite use by wild prey.
Our results suggest that snow leopard distribution re-mained largely unchanged between the two time periods,as we did not find any evidence of overall expansion or con-traction of snow leopard distribution in the study area,whereas there was a marginal contraction in the distributionof wild prey (blue sheep and ibex combined). Snow leopardT
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
distribution essentially overlapped with wild prey distribu-tion, with blue sheep and ibex showing mutually exclusivedistributions with marginal overlap.
Within our study area we recorded high probabilities ofsnow leopard site use across large patches in Spiti andKinnaur (Fig. ). These were areas where migratory livestockgrazing either did not occur or occurred at low intensity, andthe intensity of other threats, such as hunting and natural re-source extraction, was also low (Table ; SupplementaryFig. S). Among areas that showed a relatively high probabilityof contraction of snow leopard distribution (SupplementaryFig. S), % were in south-western parts of Kinnaur andSpiti that experiencedmigratory livestock grazing and huntingby local and migratory people as well as by immigrantlabourers. Migratory grazing was prevalent in c. % ofthe snow leopard’s distribution within our study area(Supplementary Fig. S), and was relatively intense (c. live-stock per km) in c. % of the area, especially in Lahaul andPangi (Supplementary Fig. S). Intense migratory livestockgrazing, along with hunting by migratory herders and immi-grant labourers in Lahaul continue to be serious threats to theotherwise contiguous snow leopard habitat, from Pangi to thewest, Spiti to the east andKinnaur to the south-east. Our studyarea is also contiguous with the snow leopard habitats ofHemis in Ladakh (Jammu and Kashmir) to the north.
We found that within our study area % of the area trad-itionally considered to be potential snow leopard habitat(Project Snow Leopard, ; Snow Leopard Network,) had a negligible probability of use by snow leopards.This is despite the fact that areas above , m altitude(mainly rocky areas with snow and ice) and below, m (forested zone) were excluded from our survey.Large-scale projections of snow leopard distributions andpopulations have relied on extrapolation of informationbased on opinions and geophysical modelling (e.g. SnowLeopard Working Secretariat, ; Snow Leopard Network,; McCarthy & Mallon, ). Our results caution againstsuch extrapolations, as they are likely to substantially overesti-mate snow leopard occurrence.
Blue sheep and ibex distributions in our study area werelargely mutually exclusive, overlapping marginally along theSutlej and Spiti Rivers (Fig. ). Blue sheep occurred in theeastern and south-eastern parts of the study area, whichare less rugged, with rolling mountains (Fig. a,b). Ibex oc-curred in the western and south-western parts of the studyarea, which are more rugged, with steeper mountains(Fig. c,d). Thus, in Himachal Pradesh, blue sheep andibex provide an important prey base for snow leopards, inmostly mutually exclusive areas (Suryawanshi et al., ).We found that the overall distribution of wild prey appearedto have declined by %between the two study periods. Areaswhere contraction occurred (Supplementary Fig. S), simi-lar to areas with contraction of snow leopard distribution,were in south-western Kinnaur, and central and western
Lahaul, which experienced intense migratory livestockgrazing, widespread presence of free-ranging dogs, andillegal hunting and wildlife trade by local and migratoryherder communities and immigrant labourers (Table ;Supplementary Fig. S).
Across Pangi, Lahaul and northern and south-westernSpiti, the ibex is the only large wild prey of the snow leopard.These are also the areas where migratory livestock grazing,depredation by free-ranging dogs, and illegal hunting andwildlife trade by local and migratory herder communitiesand immigrant labourers are pervasive and intense(Table ; Supplementary Fig. S). Migratory livestock graz-ing is a long-standing, widespread and continually changingpractice (Saberwal, ; Axelby, ) in parts of Kinnaurand Spiti and all of Lahaul and Pangi, covering almost theentire estimated ibex distribution. The use bymigratory her-ders of these relatively steep areas, which are also thehabitats of the ibex, resulted in a positive correlation be-tween probability of site use by ibex and the presence ofmigratory livestock grazing. There is evidence of interfer-ence competition between migratory livestock and ibex(Bagchi et al., ). Migratory livestock displace ibexfrom pastures through forage removal and direct disturb-ance. Understanding impacts of migratory livestock grazingon forage availability and ibex densities (exploitative com-petition) at a finer spatial scale would be useful.
Blue sheep occur extensively and at relatively high dens-ities in Spiti, especially along the eastern bank of the SpitiRiver and parts of the western bank (Suryawanshi et al.,), where the terrain is relatively gentle and ibex arelargely absent. These are also areas used only for local live-stock grazing rather than migratory livestock grazing (USL,), hence the negative correlation between site use by bluesheep and migratory livestock grazing.
Snow leopards and wild prey occurred in large parts ofour study area and were not restricted to the seven protectedareas that cover % (, km) of the area. HimachalPradesh thus mirrors a global pattern (Johansson et al.,) in which large parts (c. , km) of snow leopardand wild prey distribution lie outside protected areas. Thismakes landscape-scale participatory conservation ap-proaches, such as India’s Project Snow Leopard () im-plemented in the Upper Spiti Landscape of HimachalPradesh (USL, ), and the Global Snow Leopard andEcosystem Protection Program (Snow Leopard WorkingSecretariat, ) currently being implemented, more ap-propriate in the context of snow leopard and wild ungulateconservation in the Indian Himalaya.
Acknowledgements
We thank the Himachal Pradesh Forest Department(Wildlife Wing) for financial and logistical support; James
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
E. Hines for valuable inputs at various stages of the workand during preparation of the manuscript; our survey re-spondents for their kind cooperation; and LobzangGialson for invaluable assistance during field work. TheWhitley Fund for Nature has been a long-term supporterof our research and conservation work, and we are thankfulfor support from the Whitley–Fondation Segré PartnershipFund.
Author contributions
AG, KS and YVB designed the study. AG conducted thefield work and wrote the article. AG, KS, KRS, YVB andRR analysed the data. YVB, BP, CM, KS and KRS reviewedthe article.
References
AKAIKE, H. () Information theory as an extension of themaximumlikelihood principle. In Second International Symposium onInformation Theory (eds B.N. Petrov & F. Csaki), pp. –.Akaemiai Kiado, Budapest, Hungary.
AXELBY, R. () ‘It takes two hands to clap’: howGaddi shepherds inthe Indian Himalayas negotiate access to grazing. Journal ofAgrarian Change, , –.
BAGCHI, S., MISHRA, C. & BHATNAGAR, Y.V. () Conflictsbetween traditional pastoralism and conservation of Himalayan ibex(Capra sibirica) in the Trans-Himalayan mountains. AnimalConservation, , –.
BASANNAGARI, B. & KALA, C.P. () Climate change and applefarming in Indian Himalayas: a study of local perceptions andresponses. PLoS ONE, (), e.
BHATNAGAR, Y.V., MATHUR, V.B., SATHYAKUMAR, S., GHOSHAL, A.,SHARMA, R.K., BIJOOR, A. et al. () South Asia: India. In SnowLeopards. Biodiversity of the World: Conservation from Genes toLandscapes (eds T. McCarthy, D. Mallon & P.J. Nyhus), pp.–. Academic Press, London, UK.
BHATNAGAR, Y.V. & SINGH, N.J. () Nomadism in the IndianChangthang: changes and implications on society and biodiversity.InGlobal Change, Biodiversity and Livelihoods in Cold Desert Regionof Asia, pp. –. Bishen Singh & Mahendra Pal SinghPublication, Dehradun, India.
BURNHAM, K.P. & ANDERSON, D.R. () Model Selection andMultimodel Inference. nd edition. Springer, New York, USA.
CEBALLOS, G., EHRLICH, P.R., SOBERÓN, J., SALAZAR, I. & FAY, J.P.() Global mammal conservation: what must we manage?Science, , –.
CHAMPION, F.W. & SETH, S.K. () A Revised Survey of the ForestTypes of India. Government of India Press, Nasik, India.
DAS, A., KRISHNASWAMY, J., BAWA, K.S., KIRAN, M.C., SRINIVAS, V.,KUMAR, N.S. & KARANTH, K.U. () Prioritization ofconservation areas in the Western Ghats, India. BiologicalConservation, , –.
GASTON, A.J., GARSON, P.J. & HUNTER, JR, M.L. () The status andconservation of forest wildlife in Himachal Pradesh, WesternHimalayas. Biological Conservation, , –.
HINES, J.E. () PRESENCE . Software to estimate patchoccupancy and related parameters. United States GeologicalSurvey, Patuxent Wildlife Research Center, Laurel, USA.
Http://www.mbr-pwrc.usgs.gov/software/presence.html [accessed August ].
JACKSON, R., MALLON, D., MCCARTHY, T., CHUNDAWAY, R.A. &HABIB, B. () Panthera uncia. In The IUCN Red List ofThreatened Species : e.TA. Http://dx.doi.org/./IUCN.UK..RLTS.TA.en [accessed July].
JAYAPAL, R. & RAMESH, K. () Management Plan for Rupi-BhabaWildlife Sanctuary, Himachal Pradesh [April –March ].Wildlife Institute of India, Dehradun, India.
JOHANSSON, Ö., RAUSET, G.R., SAMELIUS, G., MCCARTHY, T.,ANDRÉN, H., TUMURSUKH, L. & MISHRA, C. () Land sharing isessential for snow leopard conservation. Biological Conservation,, –.
KARANTH, K.K., NICHOLS, J.D., HINES, J.E., KARANTH, K.U. &CHRISTENSEN, N.L. () Patterns and determinants of mammalspecies occurrence in India. Journal of Applied Ecology, ,–.
LOVARI, S., VENTIMIGLIA, M. & MINDER, I. () Food habits of twoleopard species, competition, climate change and upper treeline: away to the decrease of an endangered species? Ethology, Ecology &Evolution, , –.
LYNGDOH, S., SHROTRIYA, S., GOYAL, S.P., CLEMENTS, H., HAYWARD,M.W. & HABIB, B. () Prey preferences of the snow leopard(Panthera uncia): regional diet specificity holds global significancefor conservation. PLoS ONE, (), e.
MACKENZIE, D.I., NICHOLS, J.D., HINES, J.E., KNUTSON, M.G., &FRANKLIN, A.B. () Estimating site occupancy, colonization, andlocal extinction when a species is detected imperfectly. Ecology, ,–.
MACKENZIE, D.I., NICHOLS, J.D., ROYLE, J.A., POLLOCK, K.H.,BAILEY, L.L. & HINES, J.E. () Occupancy Estimation andModeling: Inferring Patterns and Dynamics of Species Occurrence.Academic Press, San Diego, USA.
MADHUSUDAN, M.D. & MISHRA, C. () Why big, fierce animalsare threatened: conserving large mammals in densely populatedlandscapes. In Battles Over Nature: Science and the Politics ofConservation (eds V. Saberwal & M. Rangarajan), pp. –.Permanent Black, New Delhi, India.
MARGOLUIS, R. & SALAFSKY, N. () Is Our Project Succeeding? AGuide to Threat Reduction Assessment for Conservation. BiodiversitySupport Program, Washington, DC, USA.
MCCARTHY, T. & MALLON, D. (eds) () Snow Leopards.Biodiversity of the World: Conservation from Genes to Landscapes.Academic Press, London, UK.
MICHALSKI, F. & PERES, C.A. () Anthropogenic determinants ofprimate and carnivore local extinctions in a fragmented forestlandscape of southern Amazonia. Biological Conservation, ,–.
MISHRA, C. () Socioeconomic transition and wildlifeconservation in the Indian Trans-Himalaya. Journal of the BombayNatural History Society, , –.
MISHRA, C. ()High altitude survival: conflicts between pastoralismand wildlife in the Trans-Himalaya. PhD thesis. WageningenUniversity, The Netherlands.
MISHRA, C., BAGCHI, S., NAMGAIL, T. & BHATNAGAR, Y.V. ()Multiple use of trans-Himalayan rangelands: reconciling humanlivelihoods with wildlife conservation. In Wild Rangelands:Conserving Wildlife While Maintaining Livestock in Semi-AridEcosystems, st edition (eds J.T. du Toit, R. Kock & J.C. Deutsch), pp.–. Blackwell Publishing, Oxford, UK.
MISHRA, C., MADHUSUDAN, M.D. & DATTA, A. () Mammals ofthe high altitudes of western Arunachal Pradesh, eastern Himalaya:an assessment of threats and conservation needs. Oryx, , –.
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
PILLAY, R., JOHNSINGH, A.J.T., RAGHUNATH, R. &MADHUSUDAN, M.D. () Patterns of spatiotemporal change in large mammaldistribution and abundance in the southern Western Ghats, India.Biological Conservation, , –.
PROJECT SNOW LEOPARD () Project Snow Leopard. Ministryof Environment & Forests, Government of India, New Delhi,India.
RANA, R.S., BHAGAT, R.M., KALIA, V. & LAL, H. () The impact ofclimate change on a shift of the apple belt in Himachal Pradesh.In Handbook of Climate Change and India (ed. N.K. Dubash), pp.–. Earthscan, Abingdon, UK.
SNOW LEOPARD WORKING SECRETARIAT () Global Snow Leopardand Ecosystem Protection Program. Bishkek, Kyrgyz Republic.
SURYAWANSHI, K.R., BHATNAGAR, Y.V. & MISHRA, C. ()Standardizing the double-observer survey method for estimatingmountain ungulate prey of the endangered snow leopard.Oecologia,, –.
SURYAWANSHI, K.R., BHATNAGAR, Y.V., REDPATH, S. & MISHRA, C.() People, predators and perceptions: patterns of livestockdepredation by snow leopards and wolves. Journal of AppliedEcology, , –.
SYMONDS, M.R.E. & MOUSSALLI , A. () A brief guide to modelselection, multimodel inference and model averaging in behaviouralecology using Akaike’s information criterion. Behavioral Ecologyand Sociobiology, , –.
TAUBMANN, J., SHARMA, K., UULU, K.Z., HINES, J.E. & MISHRA, C.() Status assessment of the Endangered snow leopard Pantherauncia and other largemammals in the KyrgyzAlay, using communityknowledge corrected for imperfect detection. Oryx, , –.
USL (UPPER SPITI LANDSCAPE) () Management Plan for UpperSpiti Landscape Including the Kibber Wildlife Sanctuary. HimachalPradesh Forest Department, Shimla and Nature ConservationFoundation, Mysore, India.
Biographical sketches
ABH I SHEK GHOSHAL is interested in applied ecological research andcommunity-based conservation. YASH VEER BHATNAGAR is interestedin conservation and management of snow leopard habitat. BIVASH
PANDAV is interested in landscape ecology and human–wildlifeinteractions. KOUSTUBH SHARMA is interested in species distributionmodelling and population ecology. CHARUDUTT MI SHRA isinterested in theoretical and applied research and community-basedconservation. R. RAGHUNATH is interested in geographical informationsystems and remote sensing. KULBHUSHANS INGH R. SURYAWANSH I
is interested in population ecology, prey–predator relationships andcommunity-based conservation.
https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001107Downloaded from https://www.cambridge.org/core. IP address: 54.39.106.173, on 28 May 2020 at 02:18:16, subject to the Cambridge Core terms of use, available at
