Abstract: China holds the greatest proportion of the snow leopard’s (Panthera uncia) global range and is central to their conservation. The country is also undergoing unprecedented economic growth, which increases both the threats to the snow leopard and the opportunities for its conservation. In this paper we aim to review published literature (from 1950 to 2014) in English and Mandarin on snow leopard ecology and conservation in China in order to identify thematic and geographic research gaps and propose research priorities. We first retrieved all publish items that considered snow leopards in China (n = 106). We extracted from these papers 274 reports of snow leopard presence in China. We then reviewed a subset of papers (n = 33) of this literature, which specifically focused on snow leopard ecology and conservation within China. We introduced a thematic framework that allows a structured and comprehensive assessment of findings. This framework recognizes 4 critical and interrelated topics underpinning snow leopard ecology and conservation: habitat (distribution and protected area coverage); prey (distribution and abundance, predator–prey relationships); human interactions (hunting and trade, livestock interactions and conflicts); and the underlying policy context. Significant gains in knowledge as well as research gaps and priorities are discussed with reference to our framework. The modest quantity and limited scope of published research on the snow leopard in China calls for a continued and intensified effort to inform and support national conservation policies.

Abstract: Declining prey populations are widely recognized as a primary threat to snow leopard (Panthera
uncia) populations throughout their range. Effective snow leopard conservation will depend upon reliable
knowledge of food habits. Unfortunately, past food-habit studies may be biased by inclusion of nontarget
species in fecal analysis, potentially misinforming managers about snow leopard prey requirements.
Differentiation between snow leopard and sympatric carnivore scat is now cost-effective and reliable using
genetics. We used fecal mitochondrial DNA sequencing to identify scat depositors and assessment bias in
snow leopard food-habit studies. We compared presumed, via field identification, and genetically confirmed
snow leopard scats collected during 2005 and 2012 from 4 sites in Central Asia, using standard forensic
microscopy to identify prey species. Field identification success varied across study sites, ranging from 21% to
64% genetically confirmed snow leopard scats. Our results confirm the importance of large ungulate prey for
snow leopards. Studies that fail to account for potentially commonplace misidentification of snow leopard
scat may mistakenly include a large percentage of scats originating from other carnivores and report
inaccurate dietary assessments. Relying on field identification of scats led to overestimation of percent
occurrence, biomass, and number of small mammals consumed, but underestimated values of these measures for large ungulates in snow leopard diet. This clarification suggests that the conservation value of secondary prey, such as marmots (Marmota spp.) and other small mammals, may be overstated in the literature; stable snow leopard populations are perhaps more reliant upon large ungulate prey than previously understood.

Abstract: Snow Leopard poaching and trafficking – referred to herein as Snow Leopard crime – is revisited 13 years after TRAFFIC’s first report on the subject, Fading Footprints: The Killing and Trade of Snow Leopards (Theile, 2003). This report builds on a preliminary analysis published in May 2016 (Maheshwari and von Meibom, 2016). It addresses a major information gap concerning the linkage between retaliatory killing for livestock depredation and poaching for trade, and the scale at which both are taking place. The focus is on 12 Snow Leopard range countries: Afghanistan, Bhutan, China, India, Kazakhstan, the Kyrgyz Republic, Mongolia, Nepal, Pakistan, Russia, Tajikistan and Uzbekistan. There is little evidence of illegal trade in Snow Leopards outside these countries.
Two sets of data were developed in the research for this report. The first is a Snow Leopard crime database containing records of seizures (legal actions taken by government authorities) and observations (reports of Snow Leopard killing, capture or trade, including market surveys). The database contains records dating back to 1989 (which are discussed in Annex 1), but the analysis focuses on the period since the release of Fading Footprints, the first TRAFFIC report: 2003-June 2016. Seizures are a function of law enforcement effort, effectiveness and publicization, as well as the magnitude of illegal trade, and so observations are an important component of the analysis, particularly for countries where few seizures are made or reported. However, detailed observations are not regularly published, and may be are biased toward countries where there is more effort, so a simple multiple choice survey was designed for Snow Leopard experts. Completed by 42 of them in 2016, and covering all 12 range countries, the survey asked experts for their total number of known cases, case outcomes, and reasons for killing Snow Leopards.
Based on the average number of cases known to experts over the average of nine years spent working in their geographic areas of knowledge, 221-450 Snow Leopards were estimated to have been poached annually since 2008. With the average rate of poaching detection estimated by experts at less than 38%, these numbers could be substantially higher. Of these, 55% are killed in retaliation for livestock depredation, 21% killed for trade and 18% taken by non-targeted methods such as snares. Although retaliatory killing is estimated to account for roughly half of Snow Leopard poaching (55%), experts estimate that there is a 50-50 chance (48%) that a poaching attempt will take place after a depredation incident. On average, experts estimate that 60% of retaliatory and non-targeted poaching incidents result in an attempt to sell; accounting for differences in this estimate between countries, a total of 108-219 Snow Leopards potentially enter into illegal trade. Over 90% of annual Snow Leopard poaching is estimated to occur in five range countries: China (103-236), Mongolia (34-53), Pakistan (23-53), India (21-45) and Tajikistan (20-25).
Given the uncertainties about population numbers, as well as the low rate of poaching detection, it is difficult to assess the impact of this offtake on the viability of the species. Snow Leopard range is used as a proxy for Snow Leopard population numbers; most national Snow Leopard population estimates are derived from extrapolating study site densities across likely range. Although China had by far the highest number of seizures and observations (309 Snow Leopards from 2003-2016) and the highest annual poaching estimate, its share of Snow Leopard crime was not disproportionate to its large share (at least 60%) of Snow Leopard range. Countries flagged for having disproportionate shares of crime levels relative to share of range included Afghanistan and Russia (seizures and observations), and Nepal and Pakistan (annual poaching estimates). China and Russia were most frequently identified as destinations for animals poached in other countries.
The expert survey indicates that the scale of Snow Leopard crime is more serious than apparent from the annual average number of Snow Leopard seized (18) or observed (34) from 2003-2016. This could be in part due to the challenges of law enforcement in the Snow Leopard’s remote montane habitat. Indeed, the survey found that an average of 23% of known cases were investigated by authorities, and only 14% prosecuted.
The minimum number of Snow Leopards in the seizures and observations database fell by 43% from the first half of the analysis period (2003-2009) compared to the second (2010-June 2016) (from 451 to 259). However, the decline was in the number of Snow Leopards observed in trade and in market surveys, which fell by 80% (from 280 to 54), with the largest decline taking place in China. There were more market surveys in the first half of the analysis period (13) than the second (5), but they
TRAFFIC report: An Ounce Of Prevention: Snow Leopard Crime Revisited xi
were repeated in the same places (Kabul, Afghanistan and cities in western China), and far fewer skins were seen (for example, 60 skins in the Chinese city of Linxia in 2007, compared to one in 2011). The numbers of Snow Leopards in other observations were roughly equivalent for the two periods (108 in the first and 100 in the second), but the numbers in trade observations fell by 46% (from 52 to 28). Otherwise, the number of Snow Leopards seized rose by 16% (from 115 to 133), and the number of individual seizure cases rose by 77% (from 44 to 78). The number of poached Snow Leopards seized doubled (from 31 to 60), and the observed number of poached Snow Leopards also increased by 14% (from 56 to 64). The number of Snow Leopards in trade seizures was the same in both periods (55), and the number smuggled roughly equivalent (29 seized in the first period, and 24 seized and observed in the second).
There are three possible interpretations of this situation of rising numbers of Snow Leopards poached (as measured by seizures and observations), steady numbers in smuggling and trade seizures, and steeply declining numbers in trade observations and market surveys. It could be that the limitations of available data and the authors’ inability to collect all of it has resulted in an incorrect picture. It is apparent that illegal trade has become more clandestine and difficult to detect in most countries, so that secondly, it could be increasing, as indicated by the apparent rise in poaching numbers. However, the number of Snow Leopards seized in large cases (more than 3 Snow Leopards per case), indicative of organized trafficking activity, declined from 60 in the 2003-2009 to 23 in 2010-2016. This points to a third possibility: that trade (and perhaps demand) is declining, possibly due to increased enforcement, but local people continue to opportunistically sell Snow Leopards they poached primarily to protect livestock.
With skins being the main Snow Leopard product type in trade (78%), the primary motive for buyers appears to be for display, with some observations of skins hanging on walls in homes and restaurants, as well as stuffed taxidermy specimens. Priced in the thousands of US dollars, skins have been described as a “symbol of wealth and power.” However, there probably exists very little in the way of a definable consumer segment deliberately seeking out such items. They are most likely purchased opportunistically – “impulse buys” – and most consumers probably only buy one in their lifetime. Once in a home, the illegal possession has very low probability of detection, and moreover law enforcement authorities may be reluctant to investigate in such situations. The purchase itself also has a low probability of detection, as indicated by the sharp decline in observed numbers of Snow Leopard skins being offered for sale. While growing personal wealth in Asia has been highlighted as a primary driver of illegal wildlife trade, poverty is also recognized as a driver, and the Snow Leopard trade may be more driven by rural people in Snow Leopard habitat attempting to make money and make up for livestock losses to predators than by wealthy people placing orders for luxury household decorations. Unlike the demand-driven Tiger trade (Annex 2), to which it otherwise bears many similarities, the market for Snow Leopards may be more a function of supply, and actions should focus on the communities living near Snow Leopards to reduce incentives to poach and sell. This notion is reflected in the aphorism behind the title of this report: an ounce of prevention equals a pound of cure. Preventing livestock losses, offsetting the costs of losses and improving community support for Snow Leopard conservation are the most important approaches to tackling the problem of Snow Leopard trafficking.
Recommendations focus on addressing the leading cause of Snow Leopard poaching (retaliatory killing/Human-Wildlife Conflict) as well as measures to stem illegal trade, and are primarily targeted at the 12 Snow Leopard range countries. They are aligned with existing recommendations and planned actions, including CITES recommendations, draft Decisions and consultant’s reports around implementation of Resolution Conf. 12.5 (CITES 2015, 2016; Nowell and Pervushina, 2014); the Global Snow Leopard and Ecosystem Protection Program (GSLEP, 2013, 2015, n.d.); the SLN’s Snow Leopard Survival Strategy (SLN, 2014); and WWF’s Snow Leopard Species Action Plan (WWF, 2015 and Sharma, 2016). There was also an informal discussion about recommendations to address poaching and illegal trade at the Second China Snow Leopard Forum, held in Urumqi, Xinjiang province 24-26 August 2016 (B. Weckworth, Panthera, pers. comm.).
Recommendations are grouped according to four primary actors in Snow Leopard conservation: 1) governments of Snow Leopard range countries; 2) communities living in Snow Leopard range; 3) conservation organizations and Snow Leopard experts; and 4) donor governments and agencies.

Abstract: Rapidwarming in High Asia is threatening its unique ecosystemand endemic species, especially the endangered
snow leopard (Panthera uncia). Snow leopards inhabit the alpine zone between snow line and tree line, which
contracts and expands greatly during glacier-interglacial cycles. Here we assess impacts of climate change on
global snow leopard habitat from the last glacial maximum (LGM; 21 kyr ago) to the late 21st century. Based
on occurrence records of snow leopards collected across all snow leopard range countries from 1983 to 2015,
we built a snow leopard habitat model using the maximum entropy algorithm (MaxEnt 3.3.3k). Then we
projected this model into LGM, mid-Holocene and 2070. Analysis of snow leopard habitat map from LGM to
2070 indicates that three large patches of stable habitat have persisted from the LGM to present in the Altai,
Qilian, and Tian Shan-Pamir-Hindu Kush-Karakoram mountain ranges, and are projected to persist through the
late 21st century. These climatically suitable areas account for about 35% of the snow leopard's current extent,
are large enough to support viable populations, and should function as refugia for snow leopards to survive
through both cold and warm periods. Existence of these refugia is largely due to the unique mountain environment
in High Asia, which maintains a relatively constant arid or semi-arid climate. However, habitat loss leading
to fragmentation in the Himalaya and Hengduan Mountains, as well as increasing human activities, will present
conservation challenges for snow leopards and other sympatric species.

Abstract: Applied ecology lies at the intersection of human societies and natural systems. Consequently, applied ecologists are constantly challenged as to how best to use ecological knowledge to influence the management of ecosystems (Habel et al. 2013). As Hulme (2011) has pointed out, to do so effectively we must leave our ivory towers and engage with stakeholders. This engagement is especially important and challenging in areas of the world where poverty, weak institutions and poor governance structures conspire to limit the ability of local communities to contribute to biodiversity conservation. These communities often bear disproportionate costs in the form of curtailed access to natural resources, ecosystem services, and developmental
programmes, and also suffer wildlife-caused damage, including injuries or loss of human life, and economic
and psychological impacts (Madhusudan & Mishra 2003). It is well-recognized that conservation efforts in large parts of the world historically have been perceived to be discriminatory by local people (Mishra 2016). The need for engagement with local communities is therefore embedded in the 2020 Aichi biodiversity targets and is widely thought to be critical to the success of conservation efforts. However, although the need for engagement is clear, as ecologists and practitioners we often have little formal training in how we should engage with local communities and how we can recognize the pitfalls and opportunities provided by developing genuine partnerships. The practical challenges of achieving effective engagement are considerable (Agrawal & Gibson 1999; Waylen et al. 2010, 2013), and such forays are fraught with difficulties and ethical considerations (Chan et al. 2007). When they are done badly, conservation interventions
can damage relationships and trust, and lead to serious injustice to local people and setbacks for ecological
outcomes (Duffy 2010). Much has been written on knowledge exchange and participatory research approaches (e.g. Reed et al. 2014 and references therein). This Practitioner’s Perspective
seeks to focus on the next logical step: the elements that practitioners and researchers need to consider when
engaging with communities to effect conservation. Engagement around the management of protected areas
has been discussed and formalized (e.g. Dudley 2008). Considerable literature has also emerged, particularly
from Africa, on the use and co-management of natural resources, commonly referred to as community-based natural resource management or CBNRM (e.g. Fabricius 2004; Roe, Nelson & Sandbrook 2009; Child & Barnes
2010). There have been attempts to draw general principles for CBNRM (e.g. Thakadu 2005; Gruber 2010). In
the related field of community-based conservation, however, while there have been efforts to draw lessons (e.g. Berkes 2004), little exists in terms of frameworks or guidelines for effectively working with local communities to effect biodiversity conservation in multi-use landscapes
(Mishra 2016). The eight principles for community-based conservation outlined here (Fig. 1) build on ideas developed in fields as diverse as applied ecology, conservation and natural
resource management, community health, social psychology, rural development, negotiation theory, and ethics
(see Mishra 2016). They have been developed, challenged and tested through 20 years of community experience andour own research on the endangered snow leopard Panthera uncia and its mountain ecosystems, in South and Central Asia. We suspect that with contextual adaptations, their relevance for applied ecologists and practitioners may be universal.