ResearchPDF Available

Avoiding the Unthinkable: What Will it Cost to Prevent Tigers Becoming Extinct in the Wild?


Abstract and Figures

A Wildlife Conservation Society report commissioned by the Global Tiger Initiative Suggested citation: J. Walston, K.U. Karanth, and E.J. Stokes. 2010. Avoiding the Unthinkable: What Will it Cost to Prevent Tigers Becoming Extinct in the Wild? Wildlife Conservation Society, New York.
Content may be subject to copyright.
Avoiding the unthinkable: What will it cost to
prevent Tigers becoming extinct in the wild?
Joe Walston1, Ullas Karanth2 and Emma Stokes3
1Joe Walston (corresponding author)
Senior Conservationist
Wildlife Conservation Society (WCS)
2300 Southern Boulevard, Bronx
New York 10460 USA
2 K. Ullas Karanth PhD
Senior Conservation Scientist
Wildlife Conservation Society (WCS)
Director, Centre for Wildlife Studies 26-2, Aga Abbas Ali Road (Apt: 403) Bangalore-560042 INDIA
3 Emma J Stokes PhD
CONSERVATION SCIENTIST / Regional Tiger Monitoring Coordinator
Wildlife Conservation Society (WCS)
House 21, Street 21, Tonle Bassac, Phnom Penh, Cambodia
Suggested citation:
J. Walston, K.U. Karanth, and E.J. Stokes. 2010. Avoiding the Unthinkable: What Will it Cost to
Prevent Tigers Becoming Extinct in the Wild? Wildlife Conservation Society, New York.
Cover Photo: Allan Michaud
Executive Summary
The world’s wild Tiger population is at an historically unprecedented low of about 3,200 animals
with possibly only 1,000 breeding females. Recent declines have affected every range state and
although pockets of conservation success exist, they remain isolated exceptions to the overall
range-wide trend of unremitting losses.
Although over a million square kilometers of Tiger habitat still exists from India to the Russian
Far East, hunting of Tigers and their prey continues to empty Asia’s forests, a problem
exacerbated by habitat destruction and fragmentation. Even where Tigers persist today they are
often represented by remnant populations of a few isolated individuals.
If the unthinkable is to be avoided and the Tiger is not to become extinct in the wild across much
or all of its range, then conservation investments must sharply focus on identifying those sites that
offer the greatest potential for Tiger recovery and prioritise them for immediate, sustained and
intensive protection and monitoring support. While the current international response to the crisis
is growing, many conservation approaches are unfocussed, unnecessarily complex, overly
ambitious and often geographically diffuse. This report defines and identifies a priority sub-set of
areas across the Tiger’s current range, called Source Sites. These sites not only contain the majority
of the world’s remaining wild Tigers but they also have the greatest potential for halting the
decline and initiating a sustained recovery of Tigers. The report provides evidence for Source Sites
as the appropriate spatial scale at which priority interventions should be targeted. The protection
of Source Sites is a pragmatic and achievable goal that will provide quicker and far greater return
on conservation investments than some current approaches. Source sites, by definition, already
have breeding Tigers, are of a spatial scale that is practical to protect, have existing conservation
infrastructure, a legal mandate for protection and, ultimately, have the potential to repopulate
larger landscapes.
The report presents a challenging but straightforward conservation strategy based on proven
examples of sustained Tiger recoveries in landscapes where a Source Site approach has been
taken. It quantifies the potential population increase this strategy could achieve across the Tiger’s
range, and evaluates the costs of implementing this strategy.
Forty-two Source Sites have been identified across the Tiger’s range. Most of these are in India
(18), Sumatra (8) and the Russian Far East (6). Insufficient evidence exists to suggest that
Cambodia, China, Myanmar or Vietnam possess any area that currently qualifies as a Source Site,
though some Potential Source Sites have been identified.
Collectively, Source Sites cover 90,000km2, within which almost 2,200 of the world’s 3,200
remaining wild Tigers currently live. Thus, nearly 70% of all the world’s remaining wild Tigers are
clustered within less than 6% of their current potential range (Tiger Conservation Landscapes) and
less than 0.5% of their historical range.
Just under half of the 2,200 Tigers found in Source Sites are in India. However, only five Source
Sites (12%), all of which are in India, have existing Tiger populations close (>80%) to their
estimated potential carrying capacity. Across the rest of the Tiger’s range outside India, densities in
60% of all other Source Sites are less than half their estimated carrying capacity.
If Source Sites are effectively protected, even at this late stage, their collective Tiger population
would more than double. Given the recent history of unremitting decline in Tiger numbers, this
would be an unprecedented success for the species. Even if there were zero gains in Tiger
populations elsewhere across Tiger landscapes, the projected increases at Source Sites alone would
result in a 170% increase in the world’s wild Tiger population.
Cost estimates are broadly similar across the sites when factors such as cost of living are included.
The cost of effective management, protection and monitoring of Source Sites is estimated to be
US$82 million per year or US$930/km2 (US$9/ha) per year. More than half of this money is
already being committed by range-state governments and, to a far lesser extent, international
donors and NGOs. This leaves a shortfall of US$35 million a year.
While significant investment already exists for Tiger conservation across these Source Sites with
almost US$500/km2 (US$5/ha) per year, on average, currently being spent, much of this money
comes from and is spent in India. When India is excluded from the analysis, the average current
commitment drops to US$365/km2 (US$3.6/ha) per year. India is currently the only country
providing sufficient funds for adequate protection and monitoring of Tiger Source Sites.
However, even where funds are available all is not well. There are currently major differences
between how available resources are being applied in various sites, as well as how efficiently those
resources are being used. In some cases official government budgets for protected areas are not
reaching the ground, while in other cases investments are being used inefficiently or directed into
non-essential activities. Measures of accountability for the effective and sustained protection of
Source Sites are thus paramount. This report provides a set of standards for monitoring
enforcement effectiveness. Additionally, this report suggests that monitoring populations of
Tigers and their prey, for which minimum standards for all Source Sites are provided, is the most
reliable metric of success or failure of conservation efforts.
Source Sites in themselves are by no means the entire solution to the problem of how to save the
Tiger. Only concerted, orchestrated and politically bold commitments by range-state governments,
supported by the international community, and sustained over a number of decades can do that. It
will require a broad range of actions across a variety of sectors that are well documented within
the Global Tiger Initiative (GTI) process. These include reducing global demand for Tiger
products and restricting their illegal trade, improving landscape connectivity, and ensuring
infrastructure development does not impact Tiger landscapes. However, progress on all these
fronts will be futile if there isn't immediate, substantive and sustained progress with Source Sites,
so spreading major resources and effort across these sectors in the absence of the required
investments in Source Sites would be folly.
Source Sites are the highest priorities for urgent and sustained protection and monitoring
interventions. With almost 70% of the world’s wild Tigers confined to less than 6% of their
present potential range, their disproportionate importance cannot be ignored.
ACKNOWLEDGEMENTS ..................................................................................................................1
FOREWORD ........................... ....................................... ........................................ ..............................2
BACKGROUND ...................................................................................................................................3
INTRODUCTION .......................... .................................................... ..................................................5
WILDTIGERSTODAY.............................................................................................................................................................. 5
METHODS ...........................................................................................................................................7
SOURCESITES .......................................................................................................................................................................... 7
ESTIMATEDFINANCIALINPUTSANDNEEDS...................................................................................................................... 8
OTHERMETHODS ................................................................................................................................................................... 9
SOURCESITES ........................................................................................................................................................................10
COSTS ......................................................................................................................................................................................15
DISCUSSION&RECOMMENDATIONS ............................ ............................................ .............. 23
PROTECTIONOFSOURCESITES ..........................................................................................................................................25
SOURCESITESIZE .................................................................................................................................................................26
SETTINGSTANDARDSFORSOURCESITES .........................................................................................................................27
MINIMUMSTANDARDSFORSOURCESITEMONITORING ..................... ........................... 28
LAWENFORCEMENTMONITORING ...................................................................................................................................28
MONITORINGOFTIGERSANDTIGERPREY .......................................................................................................................44
ASOURCESITE? . ............................................ ............................................ .................................... 49
TIGERSUCCESSSTORIES......................... ............................................ ........................................ 52
FINALREMARKS ............................. ................................................ ............................................... 55
REFERENCES .................. ............................................ .............................................. ....................... 56
APPENDIX .......................... ....................................... ..................................... .................................. 65
INDONESIA .........................................................................................................................................................................84
LAOPEOPLE’SDEMOCRATICREPUBLIC ..............................................................................................................91
MALAYSIA........................................................................................................................................................................... 99
MYANMAR ....................................................................................................................................................................... 105
NEPAL................................................................................................................................................................................ 110
RUSSIA............................................................................................................................................................................... 113
THAILAND ....................................................................................................................................................................... 117
VIETNAM.......................................................................................................................................................................... 123
A recent asssement by Karanth et al. (2009) prepared for the GTI-Kathmandu, provides the biological
rationale for the conservation approach presented here.
In preparing this report we draw on the particular experience of the Tigers Forever1 program, a Panthera project,
in collaboration with the Wildlife Conservation Society and government and non-governmental partners to
recover wild Tiger populations at Source Sites across their geographical range.
Additionally, a large number of people contributed to this report, both in writing country reports and
technical sections, and providing comments and criticisms. These include, but are not limited to the
following people in alphabetical order:
Ishtiaq Ahmed, Noviar Andayani, Mike Baltzer, Adam Barlow, Liz Bennett, Nick Brickle, Ravi Chellam,
Sarah Christie, Ruben Clement, Peter Clyne, Nick Cox, Pete Cutter, William Duckworth, Sivanathan
Elagupillay, Tom Evans, Francesc Fàbregas i Soler, Jessica Forrest, Camilla Fritze, Steve Galster, John
Goodrich, Melvin Gumal, Bhim Gurung, Hasnizam Hamzah, Valerie Hickey, Naser Hossain, Luke Hunter,
Yadvendradev Jhala, Arlyne Johnson, Jhamak Katki, Kae Kawanishi, Jenn Kennard, Samba Kumar, Danielle
LaBruna, Nigel Leader-Williams, Song Horng Neo Liang, Matt Linkie, Tommy Lobben, Barney Long, Tony
Lynam, Kathy MacKinnon, Peter Moss, Hannah O’Kelly, Bivash Pandav, Anak Pattanavibool, Kim Pendell,
Edward Pollard, Colin Poole, Alan Rabinowitz, Kent Redford, Tim Redford, Scott Roberton, Bill Robichaud,
John Robinson, Rob Rose, John Seidensticker, P.K. Sen, Bill Shaedla, Loretta Ann Shephard, Dave Smith,
Joe Smith, Men Soriyun, Rob Tizard, Gopal Upadya, Ray Victurine, Chanthavy Vonghamheng, Catty
Walston, Naomi Walston, Hunter Weiler, Beebach H. T. Wibisono.
Considerable content has been sourced from published and grey literature. Many thanks to those authors.
Particular thanks to Rob Rose and Danielle LaBruna of the WCS Conservation Support Program for their
help in undertaking the GIS analyses, preparing the maps and commenting on the report, and to Allan
Michaud for editing and layout. Also, a further thank you to John Goodrich, Samba Kumar, Tony Lynam,
Scott Roberton, Bill Shaedla, Dave Smith and Hunter Weiler, each of who provided extra inputs into the
As part of its commitment to Tiger Panthera tigris conservation and, more specifically, to the Global Tiger
Initiative (GTI), WCS is implementing a Global Environment Facility (GEF) World Bank project with a
number of partners called ‘Tiger Futures: Mainstreaming Conservation in Large Landscapes’. The project’s overall
objective is to ‘mainstream conservation across large landscapes through enhanced protection of Tigers and
their habitats’ with WCS’s role focused on two areas, one of which is to perform an ‘assessment of financial
needs for effective Tiger conservation’.
In looking to define ‘effective Tiger conservation’, consultations were held with a wide range of Tiger
specialists and other conservationists. The following report is the result of these consultations and a
subsequent analysis of what it will cost to implement and sustain effective Tiger conservation at the most
important sub-set of sites for the species across its range.
The presentation of Source Sites across the Tiger’s range is not meant to paint a definitive or static picture.
The report represents ‘known’ Source Sites and it is hoped that these sites be amended, updated and refined
as new evidence is gathered.
The Wildlife Conservation Society saves wildlife and wild places worldwide. We do so through science,
global conservation, education and the management of the world's largest system of urban wildlife parks, led
by the flagship Bronx Zoo. Together these activities change attitudes towards nature and help people imagine
wildlife and humans living in harmony. WCS is committed to this mission because it is essential to the
integrity of life on Earth.
The World Bank is the world’s largest source of development assistance. It works in more than 100
developing economies to fight poverty and to help people help themselves and their environment.
The Global Environment Facility is the largest source of funding for the global environment. It brings 178
member governments together with leading development institutions and others in support of a common
global environmental agenda.
Despite being one the world’s most ecologically adaptable, iconic and revered wildlife species, the Tiger is
close to extinction in the wild. Occupying only 7% of its natural range, and extirpated from some entire
countries, it now numbers about 3,200 animals in the wild. A possibly more significant statistic is that, of the
3,200 animals, little more than a 1,000 will be breeding females (Karanth and Stith 1999, Smith et al. 1998,
Smith 1993, Karanth et al. 2006, Goodrich et al. 2010). Decades of international attention and conservation
support have in some cases succeeded in slowing the speed of decline, but only where range-state nations
have elevated the conservation of wild Tigers to a national priority have there been sustained reversals of this
While some populations have experienced recoveries in the past (e.g. India and Russia in the 1970-90s;
Miquelle et al. 2010) there is currently a collective decline that embraces every range state. Even with greatly
enhanced knowledge of Tiger ecology, significantly improved methods for monitoring Tigers and their prey,
greater range-state conservation capacity, improved understanding of the overall nature of the threats, and
more ambitious conservation strategies, the species has never been in more peril of extinction. Of particular
concern are the recent Tiger population declines in two of the species’ traditional strongholds, India and the
Russian Far East. Even now these two countries together contain more than 50% of remaining wild Tigers,
so the major declines recorded in some areas of India (Jhala et al. 2008) and across much of the Russian range
(Miquelle et al. 2009) have highly significant implications for the conservation of the species.
The reasons for the decline are varied though Tiger poaching has become so intense that entire Tiger
populations have been eliminated from what were once deemed to be secure reserves throughout Asia
(Damania et al. 2008). In some reserves, poaching of prey is a major factor contributing to Tiger declines
(Karanth and Stith, 1999; Karanth et al. 2004), and in some countries, such as Indonesia, habitat destruction
and fragmentation are of equal or higher concern (Linkie et al. 2006).
Although the number of remaining wild Tigers was cited at a recent international meeting of specialists to be
3,200, a proportion of these animals is now isolated in habitat fragments so small and at such low abundance
as to have no realistic hope of recovery. With Tiger numbers at an historically unprecedented low and
continuing their decline, and with scant resources available to stem this decline, it is increasingly urgent to
identify where the remaining defensible strongholds of Tigers are and to act immediately to ensure their
protection. Triage is needed to separate those sites with doomed animals in isolated habitats from those that
have breeding populations embedded in Tiger-permeable habitats that are linked with other known Tiger
populations, thus having the potential to help to re-build meta-populations across conservation landscapes
(Karanth et al. 2009).
As part of this response, a number of governments and non-governmental organizations (NGOs) have
developed conservation strategies at site, national and regional levels. Many of these are ambitious, well
conceived, mutually consistent in their approach, and reflect contemporary knowledge of Tigers and include
recent and high-quality data sets. Of the 13 range states, 11 either have or are developing national action
plans for Tigers.
Invariably, these plans reflect the fact that the future of wild Tigers lies in large, well-managed conservation
landscapes. Their strategies relate to achieving this state, though also recognize that this is increasingly
challenging due to shrinking habitat blocks, fragmented populations, burgeoning pressures and limited
capacities and resources. As a consequence, most governments highlight the need to focus special attention
on the last remaining sites where Tigers have a high chance of survival and recovery so that broader
landscape strategies have any chance of succeeding.
Previous approaches suggested that so-called ‘core areas’ were well protected, but that they were often too
small to maintain genetically or demographically viable populations and, as such, efforts should focus largely
on habitat connectivity and building conservation landscapes. While undoubtedly large Tiger landscapes must
be the ultimate goal, basic protection, even of core areas, has lately been proven less successful than was
hoped. Within government strategies, there is an increasing appreciation of this fact and that, far from being
a reductionist approach, concentrating greater resources on ensuring the integrity of these Source Sites is not
only important in itself but is also the touchstone for larger landscape conservation efforts. If nations are not
able to protect these Source Sites, it is highly unlikely that loftier landscape ambitions will be realized.
This report examines this key element of Tiger conservation and assesses its cost. While political
commitment of the range state is undoubtedly the single most influential factor in defining Tiger
conservation success or failure, it is incumbent on the international conservation community to ensure that
lack of funding is not a factor. This report examines where these Source Sites are and how much it will cost
to ensure that Tigers do not disappear due to a global failure to support range states in ensuring protection of
Source Sites. The report goes further to set out recommended standards for law enforcement monitoring and
Tiger and Tiger-prey monitoring at the Source Site level. It further discusses the implication of areas that
currently lack Source Sites and provides recommendations on how the Tiger community can best achieve its
goal of increasing the world’s wild Tiger population.
The report does not attempt to cost out other essential components of a successful recovery of Asia’s Tigers,
such as reducing global demand for Tiger products, reducing their illegal trade, improving landscape
connectivity, or building community engagement in conservation. While these are implicitly recognized as
essential activities in their own right, some are not finance-dependent and others are beyond the scope of this
Wild Tigers today
The world’s population of wild Tigers is at an historically unprecedented low. At a recent gathering of Tiger
specialists, other conservationists and range-state governments in October 2009, the number generally
accepted was 3,200, representing a prodigious and unrelenting decline from the estimated 100,000 that
existed one hundred years ago. A similar decline in available habitat has meant that the species now occupies
less than 7% of its original range (Dinerstein et al. 2007). Ironically, these new population estimates have
never been so accurate.
Most worrying is that declines are happening across the Tiger’s range. Where once India and the Russian Far
East were strongholds for the species, achieving remarkable and admirable increases in Tigers, these are now
both seeing declines on a significant scale (Jhala et al. 2008, Miquelle et al. 2009). Even with these declines,
however, these two countries still harbour more than 50% of the world’s wild Tigers. This is partly due to
their histories of strong, though not always sustained, governmental commitment to protection of Tigers, but
also due to the massive declines in Tigers elsewhere in their range. Over the last 30 years, large areas of forest
habitat have been surveyed for Tigers for the first time, providing at least preliminary data on areas that
previously lacked any. In earlier global estimates for Tigers, these areas received population estimates based
on hunter reports or educated guesses. With few exceptions these estimates now look optimistic, with vast
areas of South East Asia recently found to be void of Tigers, and depleted of prey by hunters. While over 1
million km2 of potential habitat remains, most of these are forests emptied of Tigers.
It is now highly unlikely that any major population of wild Tigers has yet to be discovered. Some gaps remain
in our knowledge, such as sections of the Thai-Myanmar border, yet the sub-optimal nature of these habitats
suggest that low-density populations would be present even in the unlikely event that they were entirely
Now, more than ever, many of the last remaining wild Tigers are limited to protected areas, with some
exceptions. This in itself would not be a disaster if these areas were successfully protecting their resident
Tigers. Indeed, a recent conservative estimate suggested that almost 5,000 Tigers could find sanctuary within
these existing protected areas (Dinerstein 2009). This would represent a 50% increase on the world’s total
population today. Given the unrelenting decline of the Tiger, their prey and their habitats, this kind of
recovery would be a huge achievement that is sometimes lost in the rather more grandiose rhetoric being put
forward about Tiger recoveries.
Collective ambitions though should be for far more than 5,000 wild Tigers, in landscapes where natural
transfers of Tigers occurs between sites, providing further demographic and genetic stability and robustness
against predictable and stochastic pressures. However, as this report will demonstrate, a dilution of focus on
strong protection at Source Sites and an over-emphasis on spreading meagre funds across disparate, less-
tangible activities, has been a major factor in the extirpation of Tigers from so-called ‘conservation
landscapes’. The concurrent lack of intensive monitoring at these sites has also meant that the collapse went
largely unnoticed. Tigers can no longer afford this approach.
It is manifest that all progression towards creating conservation landscapes filled with 20,000 Tigers must
first emanate from Source Sites. Before they can provide this function, however, they themselves need to be
secure, ideally inviolate, with Tigers reproducing well above replacement levels (Karanth et al. 2009). Ensuring
the demographic viability of source populations should therefore be seen as the central objective of
immediate conservation efforts, as the basis for building robust meta-populations across landscapes. To do
this, one must first recognize that what is required to protect and manage an area with Tigers is significantly
different from what is required at other sites. The massive financial incentives for poaching tigers, combined
with the low opportunity costs of hunting, the low risk of detection and capture, and the copious number of
willing buyers for Tiger parts, makes the protection of Tigers, like rhinos, intensive and expensive. Even
today, India stands almost alone in recognizing the level of intensity of protection that is needed, though
even India only achieves this in a few protected areas.
Donors, conservation NGOs and government agencies still employ language that suggests that a standard
well-managed protected area will save the Tiger, making no distinction between the needs of general
protected areas and those that contain Tigers. The premise of this report is that there needs to be a clear
acceptance that not only are Source Sites disproportionately important to the recovery of the Tiger, but also
that the cost of managing them is invariably far higher than assumed. Costs cannot be extrapolated from
existing Protected Area assessments (e.g. Bruner et al. 2004) and need to be calculated based on what is
required, at that site, to protect and monitor Tigers and their prey, not through proxies, remote-sensing,
modelling or extrapolations.
The future of the Tiger rests on Asian range states creating effective Tiger landscapes. The future of these
landscapes ultimately relies on the integrity and sustained protection of these Source Sites. This report
examines where these Source Sites are and what it will realistically cost to protect and monitor them.
Demonstrated success with Source Sites and Tiger conservation
Nagarahole National Park is embedded within the Malenad-Mysore Landscape in the
Western Ghats of Karnataka, India. Under strong government management and with
NGO support, effective protection, successful voluntary relocations of settlements,
and strong scientific monitoring, it has witnessed more than a 400% increase in the
Tiger population over the last 30 years, and has sustained it through challenging times
and increasing pressures. Today it not only contains a globally important resident
population of Tigers, but it is a functioning model of a Source Site, providing surplus
Tigers to disperse across the wider landscapes.
With the adjoining forests of Bandipur, and Bhadra Reserve to the North, Malenad-
Mysore Tiger Landscape now harbours one of the largest populations of Tigers
anywhere in the world, with more than 220 Tigers just within the six Source Sites. With
Tiger densities increasing, more and more animals are moving between these Source
Sites and across the landscape, forming a functioning meta-population.
WCS is now transferring these lessons to seven other sites across the Tiger’s range
through the Tigers Forever program in partnership with Panthera. This is providing case
studies of best practices in recovering Tigers and prey in a variety of landscapes based
on a demonstrably successful conservation approach primarily focused on building
outwards from Source Sites.
Source Sites
A number of sometimes interchangeable terms for identifying important sites for Tigers are currently in use:
Core Areas, Priority Areas, Core Breeding Areas, and Tiger Conservation Landscapes (TCL). For the
purposes of this report we use the term ’Source Sites’ (defined below), though we recognize the value in
retaining alternative terms for different contexts.
Although similarities exist, we have moved away from the existing terms above, because they are ill- or
entirely un-defined, or they have existing connotations that are not shared between countries or published
works. Therefore, to avoid confusion, we have chosen a separate term, defined it clearly, and explained its
relevance to the work.
We have chosen the term Source Sites because there is a clear inference that their value lies beyond just their
own boundaries and they function as a wellspring for the repopulation of Tiger landscapes. It is explicitly
recognized that often Source Sites are themselves too small to harbour long-term, demographically viable
populations robust enough to withstand artificial and stochastic pressures (Woodroffe and Ginsberg 1998,
Carroll and Miquelle 2006). It is an obvious corollary of being a Source Site that it needs to be part of a larger
landscape. As Karanth et al. (2009) suggest, they should be protected, conflict-free zones, where female
Tigers can raise cubs to dispersal stage naturally reproducing well-above replacement levels over their
lifetimes. These sources, therefore, are not likely to be found in buffer zones or corridors and are most often
going to be core Protected Areas.
For Source Sites we follow, with amendments, Karanth et al. (2009) who define them as those areas
embedded within larger ‘Tiger-permeable habitats’ landscapes where Tigers are likely to be reproducing
above replacement levels and therefore have the greatest potential to repopulate the broader landscape.
Simply put, they are those sites that, if they lost their wild Tigers, would prevent or greatly retard any natural
repopulation of the larger landscape.
We define a Source Site as having the following features:
1. Higher densities of Tigers than in the overall landscape within which it is embedded
2. Evidence of current Tiger reproduction
3. The potential to maintain a cluster of >25 breeding females (Karanth and Stith 1999) either alone or
with other connected Source Sites in the same landscape
4. Embedded within a Tiger-permeable landscape with the potential to maintain > 50 breeding females
5. A genuine government commitment to preventing further human in-migration or infrastructure
6. Existing protection capacity or political commitments to establish such capacity in the very near
7. A legal framework in place or being developed for the prevention of poaching or hunting of Tigers
and their prey
These criteria are intentionally designed to include ecological and demographic features, combining physical
and legal attributes that infer something of the practical potential for long-term protection. The number of 25
breeding females comes from published sources (e.g. Karanth and Stith 1999) though is meant to be
indicative. For instance, Jhala et al. (2008) consider 20 breeding females to be an adequate number, while in
the Russian Far East it is considered much higher (Dale Miquelle 2009, pers. comm.). Regardless, even the
lowest number encourages more rigorous evaluation of whether a site has genuine potential to be considered
a priority site for the recovery of Tigers.
The definition is also intentionally flexible enough to accommodate most of those within government action
plans, strategies and prioritization processes. As much as possible, the report follows these governmental
strategies. For example, Malaysia’s National Action Plan for Tigers (DWNP 2008) identifies ‘Priority Areas’
embedded within ‘Tiger Landscapes’ and use criteria that allow for their clean translation into Source Sites
for this report. Where definitions are not consistent with Source Sites, amendments have been made based
on available published or report-based data.
Where the case for any existing area’s inclusion as a Source Site is equivocal, they have been given the
designation of Potential Source Site. This does not infer that the site is not of the highest importance for
Tigers; just that insufficient evidence currently exists. The analysis is deliberately conservative so that the final
costing estimate for preventing the loss of Tigers across their range is a minimum. As more data become
available the promotion of sites to Source Sites from Potential Source Sites should be made and vice versa. It
is important to stress the obvious that Source Sites refer to ‘known’ Source Sites and do not suggest that this
list is in anyway static or definitive.
Maps of the Source Sites are also indicative. The boundaries and names shown and the designations used on
the maps do not imply official endorsement or acceptance. Where there are major uncertainties regarding
boundaries of Source Sites or potential Source Sites, a circle has been used of the same size or of an
approximation. Exact shape-files will be sourced.
Estimated Financial Inputs and Needs
Wherever possible, costs were sourced from those on the ground responsible for management of the Source
Sites, and agencies supporting that management. Government figures were used where available, though
occasionally these figures were not close to the funding levels that the site-based managers were able to
access. Every attempt has been made to access accurate, verifiable data, even if those data sometimes
contradict accepted or published figures.
Costs are limited to core activities of the Source Site and of the agencies directly responsible for protecting
and managing Source Sites, such as the Protected Area agency. At a minimum, these are law enforcement,
law enforcement monitoring, general management, and the monitoring of Tigers and their prey. Sites were
also encouraged to include additional activities they consider essential, though similarly encouraged to restrict
them to those directly linked to Tigers and to keep them conservative. A number of sites have included
community engagement, informant networks, and monitoring of trade routes and restaurants within their
estimates of financial needs. Occasionally costs were challenged and removed where they are deemed
excessive or sufficiently unrelated to Tigers.
The cost of additional or new activities in areas, such as adding essential law enforcement patrols where they
are needed but are currently lacking, have been adjusted using two factors: hunting pressure and logistical
difficulty. As these two factors influence the nature and expense of a response, financial estimates have been
weighted accordingly.
Costs relating to the relocation of communities within Source Sites have been gathered where available
though not included in this analysis. Reasons for this relate to the complexity of the issue and a lack of
sufficiently detailed cost estimates at some sites, rather than just the subject’s political sensitivity. In some
sites the relocation of communities has been an important and successful initiative with positive results for
communities and wildlife, while at others it has been a disaster for both. Voluntary resettlement of families
out of critical Tiger habitats need to be an active part of the conservation debate, and analyses of successful
examples such as those from Bhadra Tiger Reserve in India (Karanth 2007) need to gain wider coverage.
One-off investments such as creation of conservation infrastructure have been gathered and are presented
separately to the recurrent costs. Incremental values – the difference between current and required
investments – have also been calculated. In many cases individual site data have been combined with other
national sites based on the request of those agencies providing the data. Often this has been to ensure that
one site or agency is not highlighted as ‘doing less’ than another, though a number of reasons have been
Costs have been kept to a minimum and the figures should be taken to represent neither the operational
budgets for any site nor the aspirational budgets for their overall management.
Other Methods
The report does not include discussion of sub-species of Tigers. This does not infer a dismissal of the
concept of Tiger sub-species, nor does it endorse their validity, but merely recognizes that it is beyond the
remit of this report.
Brief country profiles have been prepared and included in the Appendix. The country profiles are intended
to provide context to the selection of Source Sites, give an independent overview of the status of Tigers, and
a brief justification for the cost estimates. They are not intended to be conservation strategies or to supersede
existing government action plans. Profiles have been written using various sources and invariably more than
one author.
The report follows Duckworth and Pine (2003) in standardising the use of English names for species and in
considering species names as proper nouns. Thus, the major components of a species name in English are
capitalized. For a more detailed justification for this, see Duckworth and Pine (2003).
The results of the analysis are summarised in Tables overleaf.
The majority of this report focuses on issues of the scale and cost of preventing the extinction of Tigers and
of promoting recovery in the most effective, efficient and realistic manner. However, as money alone is no
solution to the Tiger crisis, we have gone further to propose how and where funds should most effectively be
spent, and how their impact should be measured. As mentioned earlier, this report does not purport to be a
definitive or prescriptive solution to saving the Tiger. It does not consider how to address the global demand
for Tiger products, improve global trade controls, or other aspects of long-term Tiger conservation. Its focus
is very much on the efforts of site-based conservation and how nations and donors can serve the Tiger
through ensuring its continued survival and recovery at the field level.
Source Sites
Forty-two Source Sites have been identified across the Tiger’s range, representing the highest site-based
priorities for immediate, intense and sustained protection and monitoring. Most of these are in India (18),
Sumatra (8) and the Russian Far East (6). Other countries have three (Malaysia and Nepal), two (Thailand) or
one (Bangladesh and Lao PDR). Insufficient evidence exists to suggest that Cambodia, China, Myanmar or
Vietnam have any area that currently qualifies as a Source Site, though Potential Source Sites have been
identified. No evaluation has yet been carried out in Bhutan, but this will be done in the near future.
Collectively the Source Sites cover 90,000km2, about the size of the state of Maine in the U.S. or less than
half the size of Cambodia. Of this, almost 20% or 17,000 km2 is in Russia and only slightly less (16,000 km2)
in India. Within this 90,000km2 it is estimated that 2,200 of the world’s 3,200 remaining wild Tigers currently
live. Thus, nearly 70% of all the world’s wild Tigers are clustered within less than 6% of their current
potential range (Tiger Conservation Landscapes) and less than 0.5% of their historical range1. This
underlines one of the central pillars of this report: Source Sites deserve disproportionate attention; immediate
and sustained efforts to reinforce and intensify protection and monitoring of these sites must be a critical
The importance of India to the Tiger cannot be overstated. Just under half of the 2,200 Tigers found in
Source Sites are in India. However, across Asia, even Source Sites are in real danger of losing the Tiger. Only
five Source Sites (12%), all of which are in India, have existing Tiger populations close (>80%) to their
estimated potential carrying capacity. Across the rest of the Tiger’s range outside India, densities in 60% of all
other Source Sites were less than half their estimated natural densities (see Tables 1 and 2)
Estimates of natural Tiger densities and population dynamics were gathered from published sources (e.g.
Kawanishi et al. 2003, Karanth et al. 2004) or from expert opinions for specific sites. These estimates suggest
that if Source Sites were effectively protected, and Tiger recovery enabled, their collective Tiger population
would more than double. Given the recent history of unremitting decline in Tiger numbers, this would
represent a dramatic and unprecedented success for the species. Even if there were zero gains in Tiger
populations elsewhere, the increase at Source Sites would raise the world’s Tiger population by 175%.
Additionally, the very criteria that define Source Sites suggest that their conservation would be a far more
pragmatic and achievable goal for conservationists than some currently being advanced: Source Sites, by
definition, already have breeding Tigers, they are of a scale that is practical to protect, they have existing
conservation infrastructure, a legal mandate for protection, and the potential to repopulate larger landscapes.
1Current potential Tiger range and historical range are both taken from the Tiger Conservation Landscape analysis
(Sanderson et al. 2010).
The protection of Source Sites should not be considered an end-point to site-based Tiger conservation
efforts, but evidence suggests that where successful, sustained recoveries have occurred (such as in the
Western Ghats in India), the construction of Tiger landscapes began by building outwards from Source Sites
and not the other way around. Even with 30 million people living in the Western Ghats landscape, Tigers are
now moving between Source Sites and breeding successfully. In other Tiger conservation landscapes a lack of
focus on Source Sites and greater efforts on corridors and less immediately urgent activities has meant that
there are far fewer Tigers to utilise these corridors, undermining their intended value.
Map 1. All Source Sites & Potential Source Sites
Map 2. All Source Sites
Map 3. All Potential Source Sites
The cost of effective management, protection and monitoring of Source Sites is estimated to be US$90
million per year (see Tables 3 and 4 for a more detailed breakdown). More than half of this is already being
committed by range-state governments and, to a far lesser extent, international donors and NGOs.
This leaves a shortfall of approximately US$35 million a year.
While the quality of figures varies between sites and countries, the results broadly suggest three major facts:
1) Significant investment already exists for Tiger conservation across these Source Sites with almost
US$500/km2 (US$5/ha) per year, on average, being spent. However, much of this money comes from
and is spent in India. When India is excluded from the analysis, the average current commitment drops
to US$365/km2 (US$3.6/ha) per year. India is currently the only country both able and willing to
provide sufficient funds for adequate protection and monitoring of Tiger Source Sites.
2) There are currently major disparities between how existing resources are being used in various sites.
In some cases official government budgets for protected areas are clearly not reaching the ground,
while in other cases investment is being used inefficiently or directed into non-essential activities. In
some cases the disparity between reported and actual figures was vast. For example, for two major
Protected Areas in important Tiger landscapes, the official government budget for law enforcement
and protection was US$5 million per year, per site. At these same two sites, only eight official patrols
were recorded during the first six months of 2009 (six and two patrols respectively), and both heads of
the Protected Areas claimed that less than 20% of the official budget was made available to them.
3) There is broad and independent agreement from all Source Sites that, regardless of inefficiencies,
considerably more funding is needed to effectively protect and monitor Tigers. It is noticeable that the
cost estimates, which were generally provided independently of each other, were broadly similar across
the sites when factors such as cost of living were included, with some notable exceptions. On average,
the amount required is US$900/km2 (US$9/ha) per year.
As discussed, there is questionable value in comparing the cost per km2 of Tiger Source Sites with estimates
from Protected Areas (PA) that do not contain Tigers, as Tiger Source Sites are an exceptional sub-set of
PAs requiring vastly more intensive and costly protection and monitoring capacities. While this figure is far
higher than many have predicted, it is similar to those required by other species of high commercial value
such as the African rhinoceroses (Leader-Williams and Albon 1988).
Table 2. Source Sites by Country – Present and Potential Populations.
Source Site Name
Current to
Source for Density Estimate
1 Anamalai + Parambikulam 1,304 6 7.5 78 98 20 Estimates
2 Bandhavgarh 448 12 12 54 54 0 Jhala et al. 2008
3 Bhadra 492 3.42 12 17 59 72 Karanth et al. 2004
4 BR Temple 580 6 12 35 70 50 Estimates
5 Corbett 521 22 16 115 83 -38 Jhala et al. 2008
6 Dudhwa+Kishenpur
+Katernia Ghat 750 6 16 45 120 63 Jhala et al. 2008
7 Kalakad Mundanthurai
+Shendurney 903 1.5 4 14 36 63 Jhala et al. 2008
8 Kanha (includes Mukki,
Supkhar,Phen) 1,050 8 12 84 126 33 Jhala et al. 2008
9 Kaziranga 490 16.8 18 82 88 7 Karanth et al. 2004
10 Melghat 788 6.7 10 53 79 33 Karanth et al. 2004
11 Nagarahole + Bandipur +
Mudumulai 1,855 10 12 203 228 11 Karanth et al. 2004
12 Pench (Maharashtra +
Madhya Pradesh) 667 8 12 41 80 49 Karanth et al. 2004, Jhala et al. 2008
13 Periyar 925 2.3 6 21 56 62 Jhala et al. 2008
14 Rajaji 820 2.5 12 21 98 79 Harihar et al. 2008
15 Ranthambore 392 11.46 10 45 39 -15 Karanth et al. 2004
16 Similipal 845 2.2 12 19 101 82 Jhala et al. 2008
17 Sunderbans 2,585 0.84 7 22 181 88 Karanth et al. 2004
18 Tadoba Andhari 625 3.3 12 21 75 73 Karanth et al. 2004
19 Huai Kha Khaeng and
Thung Yai Naresuan 6,424 2.4 8 154 514 70 WCS Thailand 2009/Tigers Forever
20 Kaeng Krachan /Kuiburi 3,884 0.8 5 31 194 84 WWF Thailand 2010
21 Endau Rompin (Johor and
Pahang) 3,558 0.7 2.3 25 82 70 WCS Malaysia in prep. (pop. density =
gross estimate)
22 Taman Negara National
Park 4,343 1.7 2 74 87 15 Kawanishi and Sunquist 2004
23 Belum-Temenggor Forest 3,546 1.8 2.9 64 103 38 Darmaraj and Mohammad 2009 (pop.
density = gross estimate)
Source Site Name
Current to
Source for Density Estimate
Leuser - Ulu Masen
0.6 2.3 61 234 74
24 Ulu Masen (SS1) 4,381
25 South Aceh (SS2) 5,810
Kerinci Seblat
2.2 3.8 140 241 42
26 North Kerinci (SS3) 1,450
27 Central Kerinci (SS4) 2,350
28 South Kerinci (SS5) 2,547
Bukit Tiga Puluh
29 Bukit Tiga Puluh (PSS6) 1,447 2.9 4.6 42 67 37
Bukit Barisan Selatan /
Bukit Balai Rejang
1.8 2.5 50 69 28
30 Bukit Balai Rejang (PSS7) 1,245
31 Bukit Barisan Selatan
(PSS8) 1,506
32 Nam Et-Phou Louey 5,960 0.45 5 27 298 91
33 Sundarbans 4,000 7.50 10.5 300 420 29 Ahmad et al .2009, Barlow et al. 2009, D.
Smith pers comm.
34 Chitwan/Parsa 1,657 5.73 10 95 166 43
35 Bardia 909 1.98 10 18 91 80
36 Shuklaphanta 305 2.62 15 8 46 83
37 Sikhote-
Alinskii/Udegeyskaya 5,056 0.45 0.85 35 54 35 Soutyrina et al. in press; Miquelle
38 Leopardovyi/Kedrovaya
Pad' 2,005 0.35 1.5 7 24 71 Pikunov et al. 2008
39 Ussuriiskii 409 1.20 1.5 5 7 29 Miquelle et al. 2010
40 Lazovskii/Zov Tigra 2,023 0.54 1.25 12 28 57
41 Anyuiskii 4,272 0.20 0.5 8 20 60
42 Botchinskii 3,051 0.10 0.5 4 20 80
Totals: 88,177 2,126 4,435
Table 3. Potential Source Sites by Country.
Country Potential Source Site Name TCL Source Area km2 Source Area (GIS)
Achanakmar ** 552 545
Buxa 37 759 695
Cauvery-MM Hills 67* 1,820 2,001
Dandeli-Anshi 69,70 1,084 864
Nagarjunsagar-Gundla Brahmeswaram ** 1,400 1,184
Kawal ** 893 1,080
Koyna-Radhanagari 71* 1,200 1,605
Kudremukh-Someshwara 66 688 781
Manas 37 500 539
Mukurthi-Nilambur 66* 600 56
Nameri-Pakhui 37 1,205 944
Bori-Satpura 51 1,428 1,478
Valmiki 40 840 401
Wayanad 66 344 368
Thap Lan 19 2,217 2,217
Xe Sap 27 1,525 1,525
Nam Kan 36 1,514 1,514
Dong Ampham 27 1,998 1,998
Nam Kong 27 1,514 1,514
Hukaung Valley WS 37 6,482 6,480
Tanintharyis 19 21,213 21,213
Cambodian Eastern Plains 27 12,990 12,990
Chu Mon Ray 27 tba tba
Dak Rong-Phong Dien ** tba tba
Song Than-Dakprin 27 tba tba
* Source Site partially inside
** Source Site entirely outside
Table 4. Source Site Costings by Country
Source Site Name Source Area
km2 Cost/km2 Total cost
(US$) Current $ inputs
1 Anamalai + Parambikulam 1,304 1,518 1,979,472 1,979,472
2 Bandhavgarh 448 993 444,752 444,752
3 Bhadra 492 1,426 701,395 701,395
4 BR Temple 580 1,518 880,440 880,440
5 Corbett 521 1,426 742,738 742,738
6 Dudhwa+Kishenpur+Katernia Ghat 750 1,426 1,069,200 1,592,395
7 Kalakad Mundanthurai+Shendurney 903 1,518 1,370,754 1,370,754
8 Kanha (includes Mukki, Supkhar,Phen) 1,050 993 1,042,388 1,042,388
9 Kaziranga 490 1,782 873,180 873,180
10 Melghat 788 1,188 936,144 910,008
11 Nagarahole + Bandipur + Mudumulai 1,855 1,426 2,644,488 2,644,488
12 Pench (Maharashtra + Madhya Pradesh) 667 993 662,164 662,164
13 Periyar 925 1,518 1,404,150 1,404,150
14 Rajaji 820 993 814,055 814,055
15 Ranthambore 392 993 389,158 389,158
16 Similipal 845 1,518 1,282,710 759,000
17 Sunderbans 2,585 1,848 4,777,080 4,777,080
18 Tadoba Andhari 625 993 620,469 620,469
19 Huai Kha Khaeng/Thung Yai Naresuan 6,424 964 6,190,000 1,600,000
20 Kaeng Krachan/Kuiburi 3,884 856 3,326,000 1,200,000
21 Endau Rompin (Johor and Pahang) 3,558 1,129 4,018,209 580,000
22 Taman Negara National Park 4,343 706 3,066,120 1,513,000
23 Belum-Temenggor Forest 3,546 706 2,503,476 360,000
Source Site Name Source Area
km2 Cost/km2 Total cost
(US$) Current $ inputs
Leuser - Ulu Masen
24 Ulu Masen (SS1) 4,381 713 3,125,647 see SS 31 for total
25 South Aceh (SS2) 5,810 574 3,335,176
Kerinci Seblat
26 North Kerinci (SS3) 1,450 1,054 1,529,098 see SS 31 for total
27 Central Kerinci (SS4) 2,350 880 2,069,472
28 South Kerinci (SS5) 2,547 771 1,964,875
Bukit Tiga Puluh
29 Bukit Tiga Puluh (PSS6) 1,447 993 1,436,464 see SS 31 for total
Bukit Barisan Selatan / Bukit Balai
Rejang Selatan
30 Bukit Balai Rejang (PSS7) 1,245 1,243 1,546,941
31 Bukit Barisan Selatan (PSS8) 1,506 1,028 1,547,337 14,000,000
32 Nam Et-Phou Loeuy 5,960 295 1,756,287 280,000
33 Sundarbans 4,000 1,264 5,054,000 1,750,000
34 Chitwan/Parsa 1,657 1,782 2,952,774 929,648
35 Bardia 909 1,782 1,619,838 334,500
36 Shuklaphanta 305 1,782 543,510 76,000
37 Sikhote-Alinskii/Udegeyskaya 5,056 754 3,811,058 1,700,224
38 Leopardovyi/Kedrovaya Pad' 2,005 1,730 3,468,461 121,546
39 Ussuriiskii 409 763 311,863 149,151
40 Lazovskii/Zov Tigra 2,023 699 1,414,482 266,535
41 Anyuiskii 4,272 634 2,708,814 197,695
42 Botchinskii 3,051 115 350,966 43,203
Totals: 88,177 82,285,604 47,709,588
The Function and Importance of Source Sites
The results of this analysis demonstrate just how critical the few remaining Source Sites are for the future of
the wild Tiger. If the first step in recovering Tigers is to stop the current decline, and to do this across as
much of their range as is possible, then Source Sites should be priorities for urgent and sustained protection
interventions. With almost 70% of the world’s Tigers within less than 6% of their current potential range,
their disproportionate importance is manifest. Beyond this, however, it is useful to understand the longer-
term value of Source Sites.
In a landscape where Tigers, Tiger prey, and the threats to them both are all homogenously distributed, the
concept of Source Sites is redundant. However, nowhere is this the reality for Tiger landscapes today and
there is no likelihood that this will be the reality in the foreseeable future, if ever. There are two features of
Tiger landscapes that demand the presence of highly protected Source Sites:
1. Tiger Landscapes are almost invariably human-dominated
Tigers living at optimal or ‘natural’ densities across large landscapes is no longer a rational objective.
The presence or influence of humans will always ensure that there will be areas where either Tigers
are absent or at significantly lower densities. This in itself may not necessarily be a barrier for
creating and sustaining functional meta-populations of Tigers, but it will always increase,
proportionately, the value of having sites where these impacts are minimized.
2. Tiger Landscapes are dynamic and future impacts unpredictable
Tiger landscapes will always be prone to both predictable and stochastic events that will cause short-
term declines of Tigers, such as mismanagement of reserves, rapid economic/infrastructure
development, social unrest, disease outbreaks, spiking global prices for forest products, national
political/ideological shifts, transboundary tensions, and the loss of individual conservation
champions. All these factors are possible negative impacts that landscapes will need to be able to
endure and which are beyond the standard capacity of conservation agencies to influence. Indeed,
they are sometimes beyond the capacity of national governments to deal with in the short-term.
Tiger meta-populations must therefore be robust enough to weather these storms. Smaller areas
under high-intensity protection will have greater chances of withstanding these pressures, thereby
greatly enhancing the resilience of the overall meta-population.
Even in areas where Source Sites may not play such an obvious role, such as the Russian Far East, recent
declines in Tigers and their prey have highlighted their value. Too small to maintain long-term Tiger
populations individually, the Russian Source Sites nonetheless have higher densities and reproductive rates
for Tigers and prey species than the overall landscape average (Miquelle et al. 2010). As such, they now play
an important role in both resisting declines that have happened elsewhere (as they are generally better
protected) and in helping the meta-population recover.
In summary, Source Sites provide robustness and stability to otherwise dynamic and unpredictable
landscapes. In a way they are bank accounts holding Tigers. When times are good, investments can be spread,
portfolios diversified, and the size of the bank account is less significant as a result. However, when times are
bad, investments fail, it is the cash in the bank (Tigers at Source Sites), which suddenly gains pre-eminent
importance. Tigers are now going through their greatest, longest and deepest recession. Now is not the time
to take risks with the few Tigers we have left. We need to save, both actually and metaphorically.
Ironically, if we are successful at conserving Source Sites and increasing the world’s Tiger population to
healthy levels, whatever that may be, then the value of Source Sites will lessen. They will not, however,
realistically ever become redundant. Human-dominated landscapes will always be prone to events, both
predictable and stochastic, that will cause declines of Tigers. By having Source Sites, preferably inviolate ones,
we increase the robustness of the meta-population and augment landscape recoveries for the species.
Source Sites will always be most effective when habitat connectivity between them is maintained or created.
Connectivity ensures that Source Sites are more than just the sum of their parts. Conservation efforts cannot
afford to lose connectivity between Source Sites, though too often conservationists confuse the addition of
new protected habitat to areas where Tigers live (e.g. an extension to a Protected Area), and the
establishment of ‘corridors’ for the movement of Tigers between Source Sites. The former does not
necessarily provide connectivity, the latter does. Both are important, but are very different in form and
Some current landscape maps show large tracts of ‘corridors’ that patently are no such thing, even lumping
disconnected Tiger habitat under the term corridors. A corridor, by definition, is a passage between two or
more sites. If these sites do not exist, then neither does the corridor.
A muddling of these concepts has serious impacts on conservation planning and, ultimately, Tigers.
Corridors have very specific functions and make no demands for there to be resident breeding Tigers within
them. As such, the conservation effort required is lower than that required by Source Sites; high density of
prey species is not a prerequisite nor is the intensity of law enforcement effort. However, sufficient
protection should be afforded that allows a proportion of dispersing Tigers to safely move through a
permeable habitat matrix, and this too will require resources and effort. Therefore, the scale of corridors
should be defined by those factors that enable Tiger movement among Source Populations. There can be
little use in assigning >10,000km2 of sometimes isolated forest fragments as corridors, as is the case in at least
one high-profile Tiger landscape. To confuse terms or use broad-brush techniques of mapping only serves to
misdirect limited resources into low priority activities or to spread resources too thinly so that they are
rendered ineffective.
Below is a list of area categories prioritized by level of protection needed. Ideally, all these sites would receive
concurrent effort and this report does not advocate a piece-meal approach or the lumping of all conservation
resources into just the first one or two categories. However, with only 3,200 Tigers left in the wild and
chronically few resources available, prioritisation cannot be ignored or effort diffused. This list is intended to
be indicative only.
1. Source Sites (SS)
2. Protected Areas with Tigers (PATs)
3. Corridors between SSs and/or PATs
4. Additional habitats contiguous with SSs and PATs
5. Additional habitats contiguous with corridors
6. Additional habitats within the landscape that are currently not connected
Protection of Source Sites
All evidence from Asia and Africa suggests that where high commercial value species exist, they are now
targeted by organised, well-armed gangs connected to global trading networks (e.g. EIA 2006; Damania et al.
2008). The last Tigers are now highly vulnerable to organised poaching and only immediate, intensive and
sustained effort at building up Source Sites populations is likely to withstand this pressure.
In compiling this report, many respondents provided supporting evidence for the above and wanted to make
clear that, at their site, there is an increasing level of threat to Tigers from professional hunters, both local
and non-local. In these situations the direct reward from the sale of a poached tiger vastly outweighs any
hunting costs or opportunity costs, which are for the most part small. This, combined with the fact that the
gang leader or dealer rarely gets caught (often recruiting local skilled hunters), makes Tiger poaching an
extremely profitable business. Note that this is a different additional scenario to opportunistic killing of
Tigers by local people, which may be more effectively addressed by increasing hunting costs and, in the case
of human-Tiger conflict, through local resolution strategies. However, the greatest threat comes from
organized poaching networks and only strict protection of Source Sites will have the desired effect of
reversing the decline, and allowing us to plan and fund larger-scale strategies. In the long-term, it is hoped
that a global decline in demand will result in the reduction of these needs, though no on-the-ground
conservation strategy should plan for this.
There is still, surprisingly, a widespread belief that standard investments in Protected Areas, when used
efficiently, are sufficient to protect resident Tigers. Where Protected Areas have not succeeded, it is often
suggested that mismanagement, corruption or lack of community support was at the core of the failure.
While this has certainly been true in some cases, it is rarely acknowledged that sites where Tigers persist must
be considered an exceptional subset of Protected Areas requiring greatly increased resources and capacities
for protection. It was well stated that “Most range states have legislation in place to protect tigers from
poachers. Unfortunately, the lack of resources for enforcement…result in inadequate patrolling, and have
hampered protection efforts” (Dinerstein et al. 2007). Where Tiger populations are successfully breeding,
such as in Karnataka State in India, Tiger conservation is afforded special recognition and sites provided with
proportionately more conservation funding. This needs to be accepted by governments across the Tiger’s
The costs associated with this level of protection across known Source Sites has been presented here, both in
terms of overall cost (c. US$90 million a year) and costs currently not covered (c. US$35 million a year).
These figures are considerably higher than previous estimates for effective Protected Area management
(Bruner et al. 2004), but in-line with estimates of where species of high commercial value are being targeted.
“The protection of species vulnerable to poaching is a costly exercise. The example of the African elephant is
instructive. During the high point of the African poaching crisis of the 1980s, losses were concentrated in
four states with wildlife management budgets ranging from $0.05 to $0.15 per hectare [$5 to $15 per km2] of
protected areas. In contrast, elephant populations stabilized in South Africa and Zimbabwe, where budgets
were $43 and $4.75 per hectare [$4300 to $475 per km2], respectively. South Africa is where the strongest and
most successful wildlife management model has emerged.” (Damania et al. 2008). The evidence from the
current analysis supports this observation. Approximately US$10/hectare per year will be needed for Tiger
Source Sites. Our estimates appear conservative compared with those in South Africa, even when adjusted
for cost of living.
In Protected Areas where tourism is a source of income, valuable staff time and effort is distracted from core
conservation activities. Most resources end-up being clustered around park headquarters, tourist
accommodation and trails, with management working under perverse incentives to increase revenue. These
sites are rarely as effective as centrally funded Tiger Sanctuaries whose main purpose is to protect Tigers and
their prey. There are exceptions to this, notably Chitwan National Park in Nepal and Kaziranga National
Park in India. However, these sites have habitats able to maintain extremely high densities of prey and Tigers,
and that allow relatively easy viewing of wildlife. These are not representative of the majority of areas where
Tigers persist today across Asia, where densities and visibility are lower.
Community-based conservation programs are also rarely effective in preventing organized criminal networks
in tropical forests (Leader-Williams and Milner-Gulland 1993), though often these programs are sensibly not
designed to achieve this. There are many examples: Zimbabwe during the CAMPFIRE project (Taylor 2009),
Zambia’s community-based programs with rhinos (Leader-Williams et al. 1990) and, more recently, Tiger
losses in the Terai Arc (Karki et al. 2009). What is needed is for wildlife authorities to direct increased
manpower into law enforcement patrols and, where possible, detection rates should be further increased by
also improving intelligence networks in the areas surrounding protected areas or Source Sites. Evidence
suggests that in the face of organized poaching, wildlife authorities should focus as a priority on increasing
detection rates (by an order of magnitude) with a second emphasis on severe penalties for those involved in
serious crime (e.g. dealers or gang leaders). Evidence further suggests that imposing stiffer penalties alone is
insufficient (e.g. the complete extirpation of the rhino from Zambia where penalties were harsh but detection
and effort intensity inadequate), particularly where detection rates are low and magistrates fail to uphold the
penalties. Furthermore, organized gangs frequently buy-off officials, and simply consider such penalties as
part of their hunting costs. Conversely, improving detection rates, or the perceived risk of getting caught, is
something often entirely within the wildlife authorities control and which can be implemented immediately.
If the success of Tiger Source Sites is to be judged by increases in Tiger numbers, then the management of
these sites must be empowered to focus on this challenging task. Serious consideration should be given to the
feasibility of establishing a range-wide network of Tiger Sanctuaries or Reserves. Not only would this provide
clarity and purpose for managers and staff, but would allow sustainable financing mechanisms for the
recovery of Tigers to be more effectively focused. However, without a system of accountability, a waning of
effort on the ground may occur unnoticed. Thus, the minimum necessary conditions would be to establish
Tiger Reserves/Sanctuaries, or their equivalent, which are afforded high levels of protection funding and on-
the-ground enforcement personnel, and which employ accepted standards for monitoring and accountability
for both law enforcement and Tigers populations and their prey.
Source Site Size
As stated, Source Sites are not intended to represent areas that, by themselves, maintain Tiger populations
capable of isolated, long-term survival. Their more limited size partly represents where the majority of Tigers
are now, but also where there is relatively high potential for conservation success. Since conservation success
here often means protection from poaching of Tigers and their prey and this is a costly exercise, it is worth
considering the issue of scale.
The collapse of rhinos and, to a lesser extent elephants, in Africa and the subsequent responses provides
useful lessons. With rhinos, it took a global population collapse for conservationists to recognize that
focussed, intensive interventions were necessary. While differing in their ecology and requirements, the
Tiger’s own collapse is not entirely different and yet seems to be having the reverse effect, pushing debates
on unrealistically large landscapes with over-arching ambitions of high-value animals wandering unmolested
through corridors of human-dominated landscapes in some of the most densely populated areas on earth.
While Tigers undeniably require far larger areas than rhinos, and small isolated fragments are insufficient,
selecting vast landscapes is not rational, realistic or constructive. It is also important that prioritisation maps
for Tiger conservation represent objective priorities for Tigers and do not represent institutional priorities.
It is widely accepted that the risk of extinction is reduced in large populations, which are less susceptible to
the effects of inbreeding depression, demographic stochasticity and genetic drift (e.g. Soulé 1986, 1987;
Caughley 1994). Thus, conservationists, especially international NGOs, give priority to large landscapes.
While the theoretical principle is not challenged in this report, its pragmatic application is, and was
demonstrably seen to fail for the Black Rhinoceros Diceros bicornis in Africa. Lacking sufficiently high
resources for protection, one large Black Rhinoceros population after another was lost from large Protected
Areas in East and Central Africa (Leader-Williams and Albon 1988). Meanwhile, Kenya re-established the
Black Rhinoceros in small well-protected sanctuaries. The success of this approach threw further practical
doubt on the theoretical ideal of protecting large populations, and suggested that efforts should be
concentrated on smaller areas (Leader-Williams and Albon 1988, Leader-Williams et al. 1990). While there are
ecological reasons why Tigers need larger areas than the Black Rhino, the principle holds true.
While we should embrace opportunities to conserve large intact areas containing Tigers, this must be
balanced with the practical realities of protecting large areas. More specifically, there is an economy of scale,
or "big is best, small is feasible" (Leader-Williams 1989). We should not confuse our long-term ambitions for
having well-populated Tiger landscapes with the shorter-term need to focus on realistic objectives that can
realise these ambitions.
Setting standards for Source Sites
The effective protection of Tigers and their prey is the responsibility of range states. The passing of strong
laws, establishment of Protected Areas and the creation of government bodies charged with wildlife
conservation has been commonplace throughout these countries and demonstrates a significant level of
political commitment critical to successful Tiger conservation. However, far less common has been the
translation of this political commitment into effective on-the-ground protection of wildlife, especially
commercially high value species such as the Tiger. It is important to distinguish between political
commitment, which can take many forms and be somewhat nebulous, and field-based application of that
commitment, which is a sub-set of the former and which can be measured.
In the next two sections of this report we set-out suggested standards for law enforcement monitoring and
for monitoring of Tigers and their prey. If adopted, they form a powerful tool to enhance protection efforts,
reduce inefficiencies, bolster field staff morale, improve communication and planning, and measure its overall
impact in the number of Tigers and prey. The standards are already proving popular and effective at a
number of sites and are now being formally adopted by countries such as Thailand.
minimum standards for source site monitoring
Law Enforcement Monitoring
The importance of law enforcement and monitoring in the successful management of protected areas has
been widely documented in a broad range of contexts (Leader-Williams, 1993; Bruner et al., 2001; Hilborn et
al., 2006; Byers & Noonburg, 2007; Dobson & Lynes, 2008). For the majority of Tiger Source Sites
management is weak and law enforcement capacity poor (Dinerstein et al., 2007; Gratwicke, 2007). We argue
that a key aspect of securing Source Sites is improving management effectiveness and strengthening law
enforcement capacity. Effective law enforcement requires information on where, how and by whom illegal
activities are undertaken, and the ability to apply this knowledge strategically to reduce poaching and other
illegal use of natural resources - often in the context of limited human and financial means. To achieve this, a
site-based mechanism is required that can capture up-to-date data and convert them into useful information
in a timely fashion, presented in such a way as to be easily understood by protected area and wildlife
Secondly, effective enforcement requires a transparent and accountable monitoring system with which to
evaluate the progress and performance of law enforcement agencies in reducing threats to wildlife. To
achieve this, appropriate indicators need to be selected and standardized protocols for data collection and
analysis need to be adopted.
This chapter presents guidelines for establishing site-based law enforcement monitoring (LEM) programs
that will improve both our understanding of threats to Tigers in Source Sites, and the management capacity
to respond to these threats effectively. The guidelines are aimed at site-managers and practitioners
implementing patrol-based and/or intelligence-based law enforcement approaches in Source Sites. We
provide an overview of key concepts, data collection standards and recommended management tools. We
highlight the practical and technical challenges involved in the design and implementation of LEM programs
and provide recommendations for avoiding common pitfalls. This chapter is not intended as an exhaustive
step-by-step manual, rather, as a set of basic principles for site-based LEM. As a minimum standard we call
for managers to adopt a standardized approach to law enforcement monitoring in all Tiger Source Sites.
Law enforcement monitoring is a tool for improving law enforcement effectiveness; it can provide managers
with the information they need to make strategic decisions but it requires that the appropriate legal and
judicial support structures and resources be in place. This chapter should therefore be considered as part of
an overall investment in and commitment to improving law enforcement effectiveness.
Ranger-based law enforcement monitoring
Ranger-based monitoring is the opportunistic collection of data on illegal activities (and other data types
including wildlife) by rangers on wildlife protection patrols (e.g. Gray & Kalpers, 2005). Patrols are typically
conducted regularly and over large areas, and thus have considerable potential to provide managers with
timely and up-to-date information for short-term decision-making. Furthermore, data collection has the
additional benefits of being cheap, relying on existing patrol routes and personnel and requiring a minimum
of specialized skills or equipment.
Ranger-based monitoring can be used as a tool to monitor trends in illegal activities over time. However,
patrols are not systematic surveys and by their very nature are typically deployed in areas where illegal
activities are high, which introduces considerable bias. Furthermore, ranger-based LEM can bias observations
of illegal activities that are predictable in space and time and easy to detect, such as land clearance, over those
activities that are harder to detect or less predictable, such as poaching. As a result, care needs to be exercised
in interpreting the results from ranger-based LEM, and there exist a number of caveats to the use of ranger-
based monitoring methods for quantitative analysis of trends. Rather, a framework is provided here for
evaluating LEM data that will help site managers to make informed decisions regarding patrol deployment
and allocation of resources.
Law enforcement effort
An important feature of ranger-based law enforcement monitoring is the relationship between law
enforcement effort and the illegal activity encountered on patrols. The relationship between patrol effort and
observations of illegal activity is typically expressed as Catch Per Unit Effort (CPUE), analogous to measures
used in fisheries management to assess the status of fish stocks, and used as an index of relative abundance
for a particular illegal activity or threat indicator, for example, number of snares per km patrolled, or number
of poachers arrested per patrol day (Leader-Williams, 1990).
Quantification of law-enforcement effort is thus required to adjust for varying effort in measuring incidence
of illegal activity over time. Depending on the type of patrolling, law enforcement effort can be expressed in
a variety of ways, from extremely simple measures to measures corrected for unit time, unit area, size of
patrol group and other relevant variables. In general, measures should be kept as simple as possible, with
distance patrolled (for surveillance patrols), number of patrol days, and patrol coverage (or area patrolled)
being three of the most critical measures of effort.
Selecting indicators for illegal activities (the ‘Catch’)
Absolute levels of illegal activity, such as poaching, can rarely be measured due largely to their illicit nature.
Proxy indicators of illegal activities therefore need to be selected and these should be measurable and
sensitive to changes in the level of threat. For example, hunting camps might be a suitable indicator for
poaching if hunters typically travel long distances and spend several days on hunting trips, but would fail to
account for short or day-long hunting trips where camps were not constructed, and therefore underestimate
level of hunting pressure.
As a general rule of thumb, indicators should be defined according to site management objectives. In reality,
these will range from quite general information on human activities, to specific indicators on poaching of key
species. Furthermore, this will very likely vary considerably between sites. To permit a level of
standardization of threat monitoring across different sites (for example either within a national network of
protected areas, or range-wide for a particular species) whilst recognizing flexibility in local conditions at the
site level, a two-tiered approach to defining indicators is recommended:
1. Broad-scale human impacts
Under the Tigers Forever program with Panthera, standardized indicators on human activities were developed
using the unified classification of direct threats developed by the World Conservation Union and
Conservation Measures Partnership (IUCN –CMP; (Salafsky et al., 2008)). This system enables standardized
threat measures within and between sites, and is applicable across a wide range of different types of
monitoring (for example recording observation of human activities by biological monitoring teams), not just
LEM. The system is also scalable and highly adaptable to different local contexts: for example it has been
successfully applied in a number of sites across the Tiger’s range, from the Russian Far East to Malaysia.
2. Site-specific indicators for particular threats to key species
Under the Tigers Forever program with Panthera, a series of specific indicators were also developed for the
poaching of Tigers and their prey that are considered to best reflect the levels of poaching pressure. Principal
indicators include one or more of the following:
Number of illegally killed Tigers
Number of snares or traps confiscated (triggered by Tigers and/or prey)
Number of illegal/un-registered firearms
Number of Tiger/prey poaching or trade violations (seizures, arrests, prosecutions)
These indicators will vary in importance and in specific details between sites. For example, the relative
importance of snaring and firearms to poach Tigers and their prey varies across sites.
Assumptions of Catch Per Unit Effort (CPUE)
The interpretation of CPUE indices for monitoring levels of illegal activity relies on the following
1. Patrol records are reliable accounts
2. Relationship between law enforcement effort and catch is constant
3. CPUE is proportional to true abundance of threat
Guidelines for avoiding some common pitfalls that frequently violate these assumptions are provided below.
Patrol records are reliable accounts
Accurate recording of observations by rangers requires regular
training and supervision, which in turn can address skill levels and motivation of patrol teams – both of
which can negatively impact data quality. Data collection protocols also need to be developed in such a
way as to remove ambiguity and avoid overloading rangers in the field. Law enforcement monitoring
protocols also typically require that observations of illegal activities are recorded only once, thus a
system needs to be put in place to ensure that signs of illegal activities (e.g. snares, camps etc.) are either
removed, destroyed or marked in some way to avoid duplicating records.
Relationship between law enforcement effort and catch is constant
To assume that the
relationship between law enforcement effort and catch is constant implies that all occurrences of a
particular illegal activity indicator (for example, snares or poachers) have an equal chance of being
detected by patrols. In reality there are many examples of why this is often not the case.
For example, patrol deployment typically varies across space and time, or in other words, across zones
and seasons. Thus, at a minimum, interpretation of CPUE indicators for monitoring trends needs to
control for both spatial and temporal variation in patrol effort. Care should be taken to avoid
extrapolating results over areas without any patrol effort at all, or where patrol effort is low or highly
variable – for example, zones that are visited rarely – as these zones provide very little useful
quantitative data for monitoring trends, although may still provide useful anecdotal information to
management on the presence of a particular threat. It is recommended to monitor trends in CPUE
indicators in those areas that are patrolled regularly and to divide these areas into sectors that have a
relatively constant and even coverage of patrol effort from month to month.
Secondly, different types of patrol (e.g. surveillance, intelligence) and means of transport (e.g. foot,
vehicle) can influence the probability of detecting illegal activities if present. For example, foot patrols
are much more likely to locate snares hidden in the forest, than patrols in a vehicle. Vehicle patrols on
the other hand may be more likely to apprehend transport of illegally logged wood or other trafficked
products. Also, patrols that are based on specific intelligence are more likely to result in a ‘catch’ than
routine surveillance patrols. This information, in and of itself, is of interest to managers looking to
maximize return from their investments. Data collection procedures should therefore document patrol
type and means of transport and it is recommended to distinguish between different patrol types and
transport in the analyses and interpretation of CPUE indicators.
CPUE is proportional to true abundance of threat
Perhaps the most important assumption of using
CPUE indicators to monitor threat levels is that the CPUE indicator is directly proportional to the
actual or true level of threat. In other words, if we measure an increase or decrease in the number of
poachers caught, we assume that this represents an actual increase or decrease respectively in the level of
poaching pressure. This is a particularly important assumption for law enforcement monitoring and
allocation of patrol effort as it can potentially lead to a situation where our data give the impression that
a threat is decreasing, when in reality it is stable, or worse, actually increasing.
Two possible violations of the assumption that
CPUE is proportional to true level of threat are
illustrated in Figure 1. Of these, hyper-depletion is
probably the most serious with respect to law
enforcement monitoring and conservation, as it
implies that the threat level is lower than it actually is.
It is almost impossible to tell from the law
enforcement data alone whether this assumption is
being violated. Some idea can be gained however by
assessing whether any of the likely causes are present
in a given context (see Table 5).
Figure 1. Possible relationships between CPUE and abundance
(taken from Milner-Gulland & Rowcliffe, 2007)
In certain instances violation of assumptions can be partly avoided by adopting clear data collection protocols
and appropriate analytical methods. Regardless of the context, it is generally recommended that independent
measures of illegal activities be obtained periodically to verify and calibrate the results from ranger-based
LEM (See Improving the Analysis and Interpretation of Law Enforcement Monitoring Data).
Table 5. A summary of the key reasons why the assumption of proportionality between CPUE and true threat
level may be violated (adapted from (Milner-Gulland & Rowcliffe, 2007))
Type of violation Type of cause Specific example
(CPUE overestimates
actual level of threat)
Inappropriate analysis Aggregating data over a wide area (or timescale) to
include zones (or time periods) with low enforcement
effort, and high levels of threat
Law enforcement strategy Switch in focus of strategy to target a particular activity
at the expense of other illegal activities
Law enforcement strategy Lack of motivation to detect and/or record signs of
illegal activity
Law enforcement strategy When incentives/bonuses are provided for
enforcement staff, satiation may occur once a target has
been reached and teams switch focus to other illegal
Inappropriate analysis Aggregating data over a wide area (or timescale) to
include zones (or time periods) where law enforcement
effort is high (and levels of threat low)
underestimates actual
level of threat)
Poacher (or other violator)
Poacher evades capture by law enforcement agents,
either by avoiding areas that are used predictably by law
enforcement teams or changing technique to one that is
less detectable
MIST: an information management tool for ranger-based LEM
What is MIST?
MIST is an integrated spatial Management Information SysTem (MIST), developed in 1997 through a
collaborative project between Ecological Software Solutions (ESS), GTZ and the Uganda Wildlife Authority
(UWA) for implementation across Uganda’s protected area network. MIST was custom-built to meet the law
enforcement needs of protected area managers by collating standardized data on measures of law
enforcement effort, observations of illegal activities and patrol actions, and converting these into useful
information for management planning1. Because it was designed using a bottom-up approach, it focuses on
the key information and output needs of managers and addresses the technical and practical challenges of
data transfer and data management with limited on-site resources and capacity.
MIST is currently maintained and distributed free of charge by ESS for non-commercial use
( It is implemented in Delphi with ESRI MapObjects to obtain the
GIS functionality and is available as a standalone package comprising a client/server application program and
associated data collection procedures. Both the data collection procedures and the software application were
developed in such a way that they can be tailor-made by the user to reflect differences in issues, objectives, and
threats at local level and in different protected areas or even land-use categories throughout a country. One
of the greatest strengths of MIST is the capacity to provide a platform on which to apply a standardized
approach to the collection, management, evaluation and communication of ranger-based law enforcement
monitoring data, through a user-friendly interface that bypasses the need for complex data-basing skills and
GIS software packages.
MIST is currently employed by protected area and wildlife agencies in sites across Africa and Asia2. As well as
improving management effectiveness, the approach has also succeeded in fostering multi-agency
collaboration in law enforcement efforts and in harnessing a general interest by government and other
agencies in adopting a standardized and transparent approach to the monitoring and evaluation of law
enforcement efforts.
Before starting MIST data collection at a site there are a number of preparatory and planning steps that
should be considered, and which are outlined in Box 1. An initial investment of time and effort into setting
up the system at a site will help to ensure the LEM program is sustainable, efficient and fully supported by
the relevant stakeholders. For managers wishing to implement MIST in multiple protected areas or sites, it is
recommended to first select one or two pilot sites with which to test the system and process, before refining
the process and replicating it on a larger scale.
1 MIST was originally designed to deal with a broad range of different data types of relevance to protected area
management. The MIST GIS component has a total of seven different application tools to aid management planning:
wildlife counts from aerial surveys, tracking of park visitation fees, community-based monitoring of NTFP resource-use,
research observations, anecdotal observations, air patrols and ground patrols. Of these, the ground patrols application is
by far the most developed and well-tested and is the focus of this chapter.
2 MIST is currently being implemented in a diverse range of contexts, from protected areas to wildlife management
programs in logging concessions, from terrestrial to marine ecosystems and in forested and savanna habitats.
Personnel and other resource needs
Defining the personnel and capacity needed at the site-level is a key step to accomplish early in the process.
One of the benefits of ranger-based LEM is that it is typically not necessary to recruit additional personnel
for data collection as existing law enforcement rangers will fulfil this role. It is therefore assumed that
sufficient law enforcement staff are already in place and have been trained in basic enforcement and field
skills. Additional training of patrol teams in MIST data collection procedures however will be required, and it
is recommended to include this as a specific module in standard law enforcement training and refresher
There are two additional roles that need to be filled for the implementation of MIST at a site. These are the
MIST User and, in certain situations, the MIST Database Manager. These are not necessarily full-time
roles but do have differing requirements in terms of location, skills, and responsibilities. In general it is
recommended to train at least two staff members on MIST procedures to ensure sustainability. Basic terms
of reference for these two positions are as follows:
Mist User -
The MIST user is the person responsible for MIST data entry at the site and monthly reporting.
This is an important role and appropriate training in MIST data entry and reporting is required. The MIST
user needs to be able to use a GPS and computer, speak and write basic English1 and ideally be based at the
site. They do not need to have any specialized database or GIS training, but they must understand basic
computer functions (Windows, email, etc). They also need to regularly interact with the patrol teams to
ensure that data collection forms are filled in correctly, and are required to submit monthly MIST reports (or
information as requested) to the site manager.
MIST Database Manager and/or Coordinator -
The database manager is responsible for managing and
maintaining the MIST database. This role is particularly important if MIST is to be rolled out in more than
one protected area and a central coordination database is to be established. This individual would have a
more advanced level of training in computers, and would typically be based in the national or provincial
capital where electricity supply is more reliable. This individual would also be responsible for advanced
features of MIST for customize reporting templates and editing the data structure. The database manager
would communicate regularly with site-based MIST Users, conduct regular quality-control checks on the
MIST data and assist in training/refresher courses for MIST Users and rangers in data-collection protocols.
1 English is currently the default language of the MIST software. Whilst it is possible to customize data collection forms
and reports into local languages, the software application, commands and dialogue windows are all in English. Whilst
translation of the software may be possible to outsource for languages based on the Latin alphabet (e.g. French,
Spanish), languages using other alphabets or character types are not supported by the current programming platform.
Box 1: Summary of steps for setting up MIST-based LEM at a site:
Define resources, personnel and training needs
Define LEM objectives and monitoring indicators
Create MIST data structure for patrol observations
Define patrol staff, stations, type of law enforcement activities and other
patrol effort attributes
Develop MIST data sheets and data collection protocols for patrol teams
Determine reporting needs and map outputs
Spatially delineate management sectors for LEM reporting and compile
other key GIS data layers
Equipment and other resources
MIST is a spatial management information system and requires spatial data to be regularly collected by
rangers. GPS units therefore need to be available to each patrol team. For patrol distances to be measured
with a reasonable degree of accuracy in MIST it is recommended as a minimum for rangers to take position
waypoints (not tracklogs) every 30 minutes when on patrol. GPS battery requirements should therefore be
factored into budgets accordingly.
The MIST User needs to have access to a single computer on-site, on which is installed the MIST database
and which is used for data entry, GPS download and preparation of MIST reports. No other software
packages are required. A constant power supply for the computer is not necessary at the site. Data can be
block-entered once a month or entered on a continual basis depending upon the set-up.
Internet connection at the site is not required to operate MIST. A site should have some mechanism of
sharing MIST reports with the site manager and senior patrol staff on a regular basis. This can be done either
by printing hard copies of MIST reports or through projecting MIST reports and maps on a screen during
monthly patrol meetings. A site should also have some mechanism for backing up MIST data and, in the
absence of e-mail, for providing (and receiving) MIST updates from the MIST Database Coordinator. Updates
can be managed by copying relevant files on a USB flash.
Data collection procedures
Data requirements for ranger-based law enforcement monitoring need to be focused on providing the
necessary information for management without overwhelming rangers with complicated data collection
protocols at the expense of the task at hand - law enforcement.
MIST works on the following principles for data collection:
- Standardized: data needs to be recorded in a consistent and unambiguous way
- Simple: the data collection system must be easy to use with a minimal amount of formal education
and fully localized into regional languages and cultures
- Fast: time spent recording data by rangers must be kept to a minimum
- Flexible: it must be possible to adapt data collection to meet the needs of different users and in
different contexts of natural resource management, reflecting differences in objectives and threats in
different protected areas or wildlife management zones
- Specific: only data which can be processed into information useful for management decision-making
should be collected by rangers
Standard data inputs for MIST include the following:
GPS waypoints (coordinates, dates and time of observations and patrol routes)
Information about the patrol (e.g. patrol dates, names and numbers of rangers, type of patrol and
means of transport)
Patrol observations (e.g. number and type of illegal activities)
MIST uses a standardized nomenclature for patrol observations, which are arranged at four hierarchical
levels, as illustrated in Figure 2. Observations are pre-defined by the user and in the MIST database these
observations appear as look-up lists to facilitate data entry.
Defining observations in MIST is flexible and can be fully customized to a particular site or context, whilst
still ensuring a minimum level of standardization. For example, in Figure 2 observation categories are
standardized at the Observation Group, Observation and Observation Type level (corresponding to IUCN-
CMP threat definitions), but are site-specific at the Observation Remarks level. Note also that the amount of
detail at the Observation Remark level will depend upon the needs of a particular site. The look-up list
options currently used by Tigers Forever sites are shown in Annexe I. An example of Observation Remarks
developed for a particular Tigers Forever site is given for poaching in the Nam Et Phou Louey National
Protected Area, Lao PDR in Annexe II.
Figure 2. Hierarchical structure of observations in MIST
MIST outputs
MIST outputs take the form of reports, maps, tables and charts. Standard output formats come pre-installed
with the software application. In addition, data-specific outputs using a standard template can be created on-
demand and fully-customized output templates can be created and added by the user.
MIST reports and maps are the primary means of direct and regular feedback to site managers. As such, they
should be designed in such a way as to be user-friendly and easily understood and tailored to specific local
needs and cultures. Examples of typical outputs include:
Indicators of illegal activities (expressed as CPUE)
Distribution maps of illegal activities for monitoring and planning
Patrol and ranger performance indicators (including number of patrol days and distance patrolled)
Patrol coverage maps
Standardized reports to meet institutional requirements
MIST information flow
In order to provide site managers with prompt up-to-date information it is vital that MIST is fully integrated
into the management planning cycle and that regular and direct feedback in the form of MIST reports and
outputs are provided. MIST information flow at the site level, including roles and responsibilities at each step
of the management cycle is illustrated in Figure 3. Regular feedback to patrol teams on performance and
outputs can, in turn, contribute considerably to a better team spirit and motivation.
In order to provide a coordinated flow of information from the site up to national-level management for
planning and resource allocation, sharing of MIST data is performed by a process of data replication,
whereby a central database, housed for example in the relevant wildlife agency HQ, receives regular update
files (via a USB flash drive or email) from protected area or site-based databases. Site-users have access only
to their individual site database, which is operated on a stand-alone PC. Conversely, users at the wildlife
agency HQ can access information on multiple protected areas from a single central database. User privileges
at all levels can be controlled accordingly (see Figure 4).
In order to maintain a standardized approach to data collection and reporting across all protected areas or
sites, database management and editing is typically performed at the central – or national – level. Edits to the
database structure (such as changes to the observation structure or reporting templates) are then
disseminated to individual sites via the replication process. The result is a one-way flow of data from the sites
to the central database and a one-way flow of database and software updates from the central database to the
Figure 3. MIST implementation process at the site level
Figure 4. MIST: Data and information flow and user access at site and national level
Collection Managers
Area Agency HQ
Protected Area
Data entry and import
Database replication (via email or USB flash drive)
Information for management planning (e.g. soft/hard copies of reports, maps and tables)
Access to information at HQ level (via LAN network)
Access to information at PA level
Intelligence-based law enforcement monitoring
For wildlife crimes that are rare, unpredictable or highly covert operations, patrol-based law enforcement
monitoring should be bolstered by local intelligence-based monitoring to improve detection – and
ultimately deterrence - rates. This is particularly the case for wildlife crime that is both organized and
lucrative, and where the potential reward for poachers and traders outweighs the costs of getting caught
by patrol teams, for example Tiger poaching. Intelligence-based law LEM also has the potential to quickly
identify emerging trends in illegal activities, and, if implemented effectively, can both complement and
strategically enhance patrol-based law enforcement approaches.
Local intelligence-based LEM relies primarily on establishing effective informant networks in areas
immediately surrounding Source Sites. Informants can take a variety of forms from agency staff to
community-members, and can vary in the type of arrangement from salaried staff, rewards for pro-bono
information, to anonymous reports via crime hotlines. Information can be actively sought or passively
received by the relevant law enforcement agency. There is no standard prescription for setting up an
effective informant-network, but in most cases these are developed at the discretion of one or a handful
of key individuals on-site. Effective informant networks take time to develop both in their quantity and
A mechanism is then needed for converting informant reports into verified information that can be acted
upon by the relevant enforcement agencies in a timely manner. This system should be able to integrate
intelligence from different informants in order to build up a more complete and accurate picture of the
crime and its perpetrators, as well as identify key reliable informants and gaps in the intelligence-network.
The system also needs to be able to evaluate the efficiency of law enforcement agencies in responding to
information, and pursuing crime reports through to a successful conclusion, for example arrest and
prosecution. Finally, any system needs to be able to identify and monitor trends in the type and nature of
illegal activities.
For the purposes of evaluating trends in illegal activities, data collected through informant networks
present a number of challenges. What we want to know is the
actual level of a covert activity
example Tiger poaching), whereas the data points we have available are
informant reports of this
. If reports are few, is this because law enforcement is effective and there is no poaching, or
because the informant network, intelligence and level of reporting is poor? Similarly, if the number of
reports is high, is this because enforcement is poor or because the informant network and level of
reporting is well-developed?
There are therefore at least two important variables that we need to understand: reporting effort and law
enforcement effectiveness. This is a similar concept to our catch/effort indicators for ranger-based LEM.
The difference with intelligence-based monitoring is that reporting effort and law enforcement
effectiveness are not directly measurable (they are latent variables); we therefore need to develop
appropriate proxy indicators, that are measurable, and which reflect these two processes.
Figure 5. Conceptual basis of intelligence-based law enforcement monitoring, showing latent variables
(law enforcement effectiveness and reporting effort) and examples of proxy indicators for monitoring listed
beneath them
Identifying proxy variables
Reporting effort reflects the efficiency of the informant network in reporting illegal activities. This is
often a complex system involving reports from multiple sources using multiple methods of data recording
obtained by the law enforcement agencies by multiple means. Measures for quantifying the reporting
effort will depend upon the size and coverage of the informant network, who the informants are, if the
information was actively or passively received and whether or not the information was paid for or
rewarded in some way. At a minimum these parameters should be recorded and monitored. Ideally, the
relationship between these parameters and the access to information about illegal activities should be
evaluated. There will likely be additional site-specific parameters depending on the set-up at a particular
The effectiveness of law enforcement operations in reducing illegal activities depends both on the
efficiency with which law enforcement agencies can respond to informant reports (
the ratio between
total reports received and total reports that are acted upon
), and the rate at which a law enforcement
response results in a successful outcome (
the ratio between total law enforcement responses and the
total successful response
s). The definition of ‘success’ will depend upon the site and the particular
illegal activity, but for an activity such as Tiger poaching, a successful law enforcement response might be
an arrest followed through to the appropriate sentence by law. Again, these are two possible proxy
indicators for law enforcement effectiveness that are likely to influence the true level of illegal activity and
are important to monitor. There will likely be other additional site-specific parameters depending on the
local context. One example is the role of media exposure as a deterrent, particularly in successful law
enforcement outcomes in response to illegal activity.
Information management tools for intelligence based LEM
A number of commercially available and valuable tools exist in order to assist law enforcement authorities
in conducting intelligence-led investigative approaches to deterring and solving crime. These range from
sophisticated applications for developing intelligence networks to performance-monitoring tools for
managing and tracking criminal cases (see for a relatively thorough
review of currently available software for crime analysis). Although none of these were developed with
wildlife law enforcement as their primary focus, many of these tools are of considerable use and interest
to wildlife law enforcement agents – particularly in tackling organized criminal networks involved in
cross-border wildlife trade, and some are currently being used to varying degrees.
There is no single commercially available management tool (such as MIST) that addresses all of the
standards raised here in managing, monitoring and evaluating site-based informant-based law
enforcement approaches. Custom-developed in-house systems exist within particular agencies and
organizations, but there is little coordination or adherence to any minimum standards for monitoring
crime and law enforcement effectiveness. Moreover there exist no management tools of this type which
are scalable to a broad range of local contexts and cultures. There is therefore a need for a standardized
approach to site-level informant-based law enforcement monitoring, and a suitable tool with which to
address this1.
There is not likely to be a single management tool that can address all site-based law enforcement
monitoring needs (intelligence and patrol-based law enforcement approaches) and different sites will have
differing requirements and resources available to them in deciding which tool to use. We advocate instead
the adoption of standardized and complementary monitoring approaches, aimed at improving our
understanding of illegal activities and thus improving the effectiveness of law enforcement strategies in
addressing them.
Improving the analysis and interpretation of law enforcement monitoring data
Statistical approaches for the quantitative analysis of LEM data
There are currently no standard ‘off-the-shelf’ statistical approaches or models for the quantitative
analysis of law enforcement data. However, a number of standard modelling approaches exist that can be
applied to law enforcement data in order to incorporate sampling error and data uncertainties and
improve our inference of patterns and trends. Modelling approaches should not be seen as a panacea to
the inherent challenges of law enforcement monitoring and are only as good as the data on which they are
based. Their utility will nevertheless be greatly enhanced by data that is collected according to the
minimum standards presented in this chapter.
Regression models can be used to determine the relationship between indicators of illegal activities and
key predictors such as measures of law enforcement effort, including type of patrol, distance patrolled
and number of patrol staff (Jachmann, 2008) in order to extrapolate levels of illegal activity over time. A
similar practice is used in constructing standardized indices of CPUE in commercial fisheries, which are
subject to similar sources of bias as patrol-based data (see (Maunder & Punt, 2004) for a review of recent
Occupancy-based models (Mackenzie et al., 2002) also have potential for investigating the spatial patterns
and processes of illegal activity indicators over time, and examining the relationship between occupancy
and key co-variables. Occupancy-based models have the additional advantage of estimating the
probability of detecting illegal activities, if present, and to examine the relationship between detectability
and co-variables such as patrol type or law enforcement strategy. These models are particularly well-suited
to patrol-based data, given their relative robustness to missing values and unequal sampling effort over
space and time.
1 In response to this, the WCS, in consultation with Ecological Software Solutions have designed a fully open-source and site-
based wildlife crime information management system. This system is currently being piloted in Sumatra, Indonesia, but will have
broad applicability to intelligence-based law enforcement approaches at the site and national level. For information about the
Wildlife Crime Database and to track progress in its development, see
Models of illegal behaviour based on enforcement data have also been combined with population models,
to test management assumptions and examine the effectiveness of different law enforcement
interventions in reducing illegal off take and impacting target species (Milner-Gulland & Leader-Williams,
1992; Hilborn et al., 2006; Byers & Noonburg, 2007).
Finally, simulation models are particularly well-suited to examining trends in illegal activities from
enforcement data as they can explicitly address limited data, uncertainty in available data and incorporate
additional information from a wide range of different sources (Burton, 1999; Pitcher et al., 2002).
Independent assessments of illegal activities
One of the main challenges for law enforcement monitoring programs is that data collected by law
enforcement teams is likely to influence the very variables we are interested in measuring. Given this, and
the many inherent assumptions of enforcement-based monitoring, we recommend that periodic checks
and balances are undertaken in order to verify and calibrate the results of LEM and to confirm the course
of management interventions taken. Of particular relevance are independent (i.e. non-enforcement related)
assessments of illegal activities. Examples include wildlife survey teams recording indirect signs of illegal
activities on systematic transects or plots, or community-based questionnaires focusing on direct
reporting of illegal behaviour (see (Gavin et al., 2009) for a recent review of the costs and benefits of
different methods for recording illegal behaviour). These data can be combined with law enforcement
monitoring data through a process of triangulation to obtain a more holistic picture of the status of
threats at a particular site. Moreover, in a conceptual framework, these data can be used to test
assumptions governing driving factors of particular threats and the expected outcomes of law
enforcement activities. For example, if poaching of ungulates in a protected area was assumed to be
driven by demand from local wild meat restaurants and markets, then data from systematic market
surveys on wild meat availability could be combined with patrol-based law enforcement data on poaching
infractions to test our assumptions of how law enforcement interventions are reducing threats (de
Merode et al., 2007).
Coordinating biological and law enforcement monitoring programs
There are a number of site-specific factors that can influence the propensity of different threats and their
drivers. For example, wildlife abundance and distribution will likely influence both the spatial distribution
of and intensity of poaching – particularly of rare and highly sought-after species such as Tigers, elephants
and rhinos (Leader-Williams, 1990; Jachmann, 2008). Information on wildlife abundance and distribution
can, in turn, inform the strategic deployment of law enforcement teams to protect those species.
Moreover, regular and reliable data on wildlife populations from statistically-rigorous survey methods can
function as an important and periodic barometer for assessing law enforcement effectiveness. It is
strongly recommended that a mechanism for regularly integrating information from both law
enforcement teams and wildlife survey teams is established at a site, and used to feedback regularly to
management planning and deployment of law enforcement operations.
Monitoring of Tigers and Tiger prey
Why Monitor Tiger Populations at Source Sites?
The goal of Tiger conservation efforts is to address the current crisis and recover Tiger numbers. Therefore,
such efforts must measure their own effectiveness. While general conservation projects can afford ‘soft’
metrics (e.g. money spent, meetings held, awareness raised, employment generated, human welfare gains
made, etc.) for measuring successes, these surrogates have proven inadequate for the challenges posed by
the Tiger crisis. Tiger population recovery (or lack of it) can be reliably assessed only by direct measurement
of Tiger densities (number of Tigers/100 km2) at the targeted recovery site (Karanth et al. 2009). While in
the past the lack of adequate monitoring methods may have compromised the value of making this attempt,
today we have no such excuses.
As with other commercially valuable species such as rhinoceroses, regular and statistically reliable estimates
of Tiger density are critical if managers are to detect rapid changes in populations that occur, for example,
from poaching. Measuring changes in relative Tiger densities or spatial distributions (habitat occupancy)
across wider landscapes is of secondary importance as changes in distribution occur more slowly.
Additionally, measuring potential Tiger carrying capacities through assessments of prey base (ungulate
densities) is also important, particularly for sites where present status of Tigers is uncertain.
Monitoring Tiger Source Sites
Wild Tigers are potentially distributed across Asia over 1.1 million km2 (Dinerstein 2006). However, they
survive in remnant, scattered populations threatened by poaching for trade, conflict-related killings and prey
depletion driven by local hunters. At present, the sizes of individual Tiger populations or the extent of
habitat/area occupied by Tigers within larger landscapes are not well known. The need to objectively assess
populations to guide species recovery is thus critical. However, Tigers are secretive, with wide-ranging daily
movements (~5-30 km/day), and occur at relatively low densities (<1-20 Tigers/100 km2), even in prime
habitats. All these factors pose challenges that defeat counting methods that rely on direct visual detections.
Investigators must therefore use non-visual methods to assess Tiger status.
Tiger populations typically need to be assessed at two spatial scales: individual Tiger populations at specific
Source Sites (e.g. reserves), and measurement of Tiger distribution or habitat occupancy over wider
landscapes and regions (Karanth and Nichols 2002, 2010). In this chapter we focus on monitoring Tiger
populations at Source Sites, and the reader is referred to relevant literature (Mackenzie et al. 2006, Karanth
and Nichols 2002, 2010) with regard to landscape or regional level monitoring of Tiger distributions and
habitat occupancy (Mackenzie et al. 2006, Karanth and Nichols 2002, 2010).
Tiger populations can be monitored directly to estimate population size or indirectly by counting of
ungulate prey animals to answer a slightly different question: How many Tigers can an area potentially
support? The former approach is more relevant to fine-grained monitoring of known source populations of
Tigers, whereas the latter approach is more appropriate for areas where Tigers may be difficult to monitor
or the potential of the area to be a source population itself needs to be assessed. These two basic
approaches are covered in separate sections below.
The Conceptual Framework
There is now a clear scientific recognition that conventional statistical tools based on formal hypothesis
testing (e.g., analysis of variance, power analysis) are not always appropriate (Anderson and Burnham 2000),
particularly for observational data gathered in non-experimental contexts (Williams et al. 2002, MacKenzie et
al. 2006, Royle and Dorazio 2008). Such is often the case with field surveys of Tigers. Therefore, we
recommend methods that explicitly model both the system being observed (Tiger populations) and the
survey process itself, followed by confrontation of alternative plausible models with survey data.
We provide a simplified overview below on key conceptual issues involved in monitoring Tigers or their
prey species (Williams et al. 2002, Karanth and Nichols 2002). Animal populations can be reliably estimated
from ‘sample counts’ obtained in the field. This context can be represented by the following simple general
estimator that relates the field counts (C) of Tigers or prey animal species to the ‘real numbers’ of these
animals (N) in the population of interest:
ˆ= p
Where N
ˆ = Abundance estimate
C = Count statistic
ˆ = Estimated proportionality constant (detection probability) relating to the count
statistic and abundance.
However much we may dislike detection probability p, this is a ‘nuisance parameter’ that afflicts all types of
animal surveys which will not go away if simply ignored (as several current methods do). Our preference,
therefore, is for Tiger and prey counting approaches that explicitly model and estimate this detection
In the case of Tigers, we recommend an approach based on ‘capture-recapture sampling’ and in the case of
prey species we recommend ‘distance sampling’ methods, to estimate detection probability (see Williams et
al 2002; Amstrup et al. 2004, Buckland et al. 2001, Royle and Dorazio 2008, for detailed conceptual
treatments). We note that these two intensive methods are practical to employ only at the scale of, at
maximum, a few thousand square kilometres or so. Thus, they are very relevant to monitoring of Tiger
populations at most Source Sites, which typically exist at this spatial scale. At the wider spatial scales of
landscapes/regions, these methods are impractical, although their results can be integrated profitably with
data from habitat occupancy sampling feasible at such wider spatial scales.
Sometimes, application of methods capable of direct estimation of detection probability even within Source
Sites may be constrained by local factors of ecology or logistics (e.g. theft of camera traps, lack of
accessibility due to difficult terrain, flooding etc.). Under such circumstances Tiger or prey signs (tracks,
dung) have to be counted and the encounter rate data analyzed to yield some sort of an ‘index’ that is
expected (hopefully) to reflect true abundance of animals. In such cases, the term p in the above equation
can be thought of not as detection probability, but simply as coefficient that relates number of signs
counted (C) to real Tiger numbers (N), if not exactly, at least in some sort of a monotonic manner.
However, there is often poor correspondence between rigorous estimates of Tiger density derived from
intensive methods and indexes generated from simple field surveys of Tiger signs (see Table-1), which
renders the use of indices questionable. If use of indices becomes inevitable because of logistical
constraints, we recommend indices that at least try to incorporate detection probability over indices which
simply ignore it. Under such circumstances, we recommend another sampling approach known as habitat
occupancy modelling (MacKenzie et al. 2006), which explicitly estimates detection probabilities under a
specific modelling approach developed by Royle and Nichols (2003). This may yield a more robust index of
animal abundance.
In the sections that follow, we provide a brief overview of the field and analytical methods that are useful
for estimating and monitoring populations of Tigers and ungulate prey species in Source Sites. The
concepts, field methods and analytical options related to these approaches are not explained in detail here
and the reader is referred to relevant literature. However, we emphasize here that the methods we
recommend have received substantial theoretical development, validation through simulations, and
application in the field on a variety of species in different ecological settings - and most importantly - have
been refined through the process of scientific peer-review and publication. Equally important, their utility
for monitoring Tigers and prey species has been empirically tested for over two decades by WCS
researchers in India and at most of the Tigers Forever sites. Although alternative methods are sometimes
offered for monitoring Tiger and prey populations, we note these lack a comparable track record
established through the process of peer-reviewed science.
Assessing Tiger Abundance and Population Dynamics
In capture-recapture sampling of Tiger populations, multiple ‘samples’, each consisting of several
identifiable individual Tigers, are drawn from a Tiger population of unknown size (N = abundance) over a
short survey duration when the population is assumed to remain unchanged. From the frequencies with
which distinct individual Tigers are captured in these detection histories, the ‘detection probability’ (p) is
modelled and estimated. Thereafter, the unknown Tiger abundance (and density) can be computed.
Furthermore, such capture-recapture surveys conducted sequentially over multiple years permit estimation
of even more critical demographic parameters such as survival, recruitment and movement, providing a full
understanding of Tiger population dynamics (Amstrup et al. 2006, Karanth et al. 2004, 2006). Considering
natural variations, as well as pressures of direct hunting and prey depletion (Karanth and Stith 1999) which
now impinge on most Tiger populations, the importance of such data for ‘real-time’ tracking of fate of
these populations becomes obvious.
Central to the effective application of capture-recapture methods is the samples of individual Tigers from a
population. Currently, identifiable ‘captures’ of individual Tigers are possible using camera-trap photos
(Karanth and Nichols 2002) or DNA extracted from Tiger scats or hair collected in the field (Mondol et al.
2009). Statistical discrimination of individual Tiger identifications from track-shape (Riardon 1998, Sharma
et al. 2005) and scent recognition by trained dogs (Kerley et al. 2007) have also been tried out, but under
small scale controlled experiments. The most widely practiced method of subjective Tiger track
discrimination under India’s ‘pugmark census’, has now been abandoned as not useful after a quarter
century of field application (Jhala et al. 2008, Karanth et al. 2003).
We emphasize that sound survey designs and analyses are not irrelevant theoretical abstractions. They must
be shaped by the ecology of the Tiger population and the logistical context of surveys. Specific insights on
ecological variables such as potential Tiger density and home range size, as well as application of local field
craft, are essential components of field surveys. We focus below on some key practical considerations, but
refer readers to the literature on details of equipment or statistical methods (Karanth and Nichols 2002,
Source Sites of Tigers must be surveyed to cover an area large enough to capture at least 10 or more
individual Tigers, in multiple sample periods, within a short survey duration of 45-60 days to meet the
demographic closure assumption. In the case of camera trap surveys, a trapping intensity of about 500
trap-nights/100 km2 must be aimed for and distance between camera traps set to ensure all animals in the
sampled area have some probability of being captured. Meeting these criteria involves deployment of
dozens rather than a handful of cameras, as most typical surveys do. Even in the case of sampling Tiger
populations for capture-recapture analysis using fecal DNA, the above guidelines are relevant. Furthermore,
care should be taken to collect and store scats and conduct subsequent laboratory procedures according to
careful protocols (Mondol et al 2009). We refer the reader to extensive literature available on proper conduct
of capture-recapture surveys of Tigers and analyses of resulting data (Karanth and Nichols 2002, 2010;
Karanth et al. 2004, 2006; Dattatri and Karanth 2008; Tigers).
Assessing Potential Tiger Numbers by Estimating Prey Densities
There may be ‘potential Source Sites’ within Tiger range where immediate application of capture-recapture
surveys using DNA or photographic captures may not be practical, e.g., where Tiger densities are extremely
low. However, since the Tiger:prey ratio of 1:500 is established reasonably well (Karanth et al. 2004),
potential carrying capacity of a site to hold Tigers can be assessed by estimating density of large ungulates
which are the principal prey species of Tigers. However, in many cases, prey will be below their carrying
capacity due to overharvest, so this may underestimate potential carrying capacity of prey.
Where actual densities of prey species and terrain permit line transect surveys (a class of distance sampling
methods, see Buckland et al. 2004 for a review), these can be applied. If the basic assumptions of 100%
detectability on the transect line, absence of systematic, undetectable, evasive animal movement, and,
accurate counting and distance measurements of animal groups are all satisfied, line transect method works
very well (Buckland et al. 2004, Karanth and Nichols 2002). Line transect method has yielded excellent
results since 1986 in WCS projects in India and since 2006 in Thailand. However, if ungulate densities are
too low, or vegetation too thick to permit placement of transects or easy passage of personnel, this method
may not be practical or will require a disproportionate amount of field effort. In some cases where the
aforementioned techniques are not practical, investigators may be compelled to rely on counts of signs of
prey species, such as dung and tracks, to get an assessment of relative prey densities. There are two
approaches possible in this context. One involves first conducting standard line transect surveys
incorporating detection probabilities to estimate densities of animal dung piles rather than of animals
themselves. Thereafter, animal densities are derived from dung densities by applying correction factors in
the form of species-specific daily defecation rates and estimates of dung decay rates. After extensive trials in
the 1990’s, WCS researchers in India have found this “dung count corrected by decay and defecation rates”
approach to be ecologically unreliable and logistically impractical for typical ungulate prey species in the
tropical forest regions of southern Asia. Therefore, we do not recommend this method for assessing Tiger
carrying capacity at potential Source Sites.
The second approach is underdevelopment. For sites where line transect methods are no practical, we are
developing an alternative ungulate density monitoring approach based on the idea of occupancy modelling
using counts of track and dung. These occupancy methods involve field surveys of habitat patches (grid
cells) and data from simple detection or non detection of signs. The analytical approach is rooted in an
innovative model developed by Royle and Nichols (2003). This method yields a potentially robust ‘index of
ungulate density’ which explicitly accounts for imperfect detection of ungulate signs. Initial calibrations
against line-transect-derived ungulate densities in southern India show good correspondence, thus hold
promise for eventually deriving density estimates from these occupancy surveys. More extensive field trials
have been recently completed in Malaysia and Myanmar and results are currently under analysis.
Finally, a method has more recently been proposed for estimating prey density from camera trapping rates
instead of individual recognition, which is currently not possible for most Tiger prey species (Rowcliffe et
al., 2008). This method has not yet been field tested on Tiger prey but, based on field trials on ungulates in
Tanzania, shows potential as a cost-effective alternative to line-transects in remote areas provided model
assumptions governing prey movement can be met (Rovero et al., 2009).
Because ungulate densities fluctuate somewhat naturally and prey species are also prone to elimination by
local hunters, it is important to keep track of their densities periodically. Ideally, if resources permit, such
prey monitoring should be undertaken on an annual basis for each Tiger Source Site. If resources are a
major constraint, prey monitoring should be conducted at least once in 2-3 year time frame. We refer
readers to the literature on distance-sampling and occupancy sampling approaches for the proper conduct
of field surveys to estimate prey densities (Buckland et al. 2001, 2004, Royle and Nichols 2003, MacKenzie et
al. 2006, Dattatri and Karanth 2008, Tigers).
Tiger Monitoring Conclusions
Based on over two decades of designing and implementing monitoring schemes for Tigers and prey species
in a variety of ecological and social-capability contexts in throughout Asia, our conclusions with regard to
monitoring of Tiger Source Sites are summarized below:
1. Tiger-prey monitoring has to be based on a sound statistical basis and firmly rooted in modelling
approaches that explicitly deal with the critical issue of estimating probabilities of detecting animals or their
signs during the field surveys. Methods that ignore this key nuisance parameter, which in fact obscures
parameters of interest to us, such as Tiger or prey density and vital rates that drive changes in these, are in
our opinion obsolete.
2. Instead of trying to reinvent statistical wheels, monitoring programs should strive to benefit from
available peer reviewed literature, software and instructional manuals, which exist to support application of
these methodologies to monitoring of Tiger and prey populations at Source Sites.
Based on the above considerations we recommend following specific monitoring methods, with a proven
track record (or under advanced stage of development):
3.1. Annual estimation of Tiger population size and density using closed model capture-recapture analyses
based on photographic (camera trap) or non-invasive DNA sampling.
3.2. Estimation of rates of apparent survival, recruitment, temporary emigration and transience across
multiple years using open model capture-recapture analyses of above mentioned genetic and photo-capture
data by means of annual sampling to link closed model analyses across multiple years.
3.3. Estimation of densities of ungulate prey species using line transect (distance) sampling methods.
Thereafter, use these prey density estimates to assess potential carrying capacity of the area for Tigers.
Where line transect methods are not possible, use indices based on encounter rates with animal signs
derived from field surveys conducted in an occupancy modelling and estimation framework.
It is ironic that during past three decades, even as Tiger range and habitats have shrunk, and Tiger
populations plummeted, monitoring efforts have been somewhat stuck in a quagmire of obsolete statistical
ideas.. Although Tiger monitoring programs are eagerly rushing to embrace new and improved material
tools in the form camera traps, GIS software, GPS units and DNA kits, they have not been equally eager to
explore the rapid advances made in the field of allied statistical methods. These new methods substantially
overcome key constraints of older survey methodologies still being used. Conservationists must now rapidly
adopt these newer and freely available intellectual monitoring tools, as readily as they are acquiring the
material tools in the market at substantial cost.
What does it mean if a Tiger Conservation Landscape
does not contain a Source Site?
This report has highlighted the fact that many areas that have recently supported Tigers no longer have an
obvious Source Site, though this should in no way be considered definitive. Some Potential Source Sites
might qualify as Source Sites once further surveys are undertaken, while it is also possible that areas not
identified as either within the report could also contain breeding Tiger populations capable of recovery.
However, an entire TCL without Tigers has obvious and profound implications for Tiger conservation. It
suggests that natural recovery of Tigers is highly unlikely. As more and more landscapes are found to be
devoid of Tigers, the possibility of reintroduction of Tigers into these landscapes by translocation of wild
animals might be considered. Reintroduction from captive sources is a vastly more complicated process,
which should not be considered so long as there are appropriate wild sources of Tigers (IUCN 1998,
Breitenmoser et al. 2001, Hunter and Rabinowitz 2009, Jule et al. 2008). Regardless of the source,
reintroduction should be considered the very last option (Johnsingh and Madhusudan 2009, Christie 2010).
Unfortunately, in many landscapes, obvious alternatives have run out.
However, before reintroduction can even be considered, conditions must be created in which a source
population can thrive and Tigers can persist in a landscape surrounding that source. We propose the
following conditions must be met before reintroducing Tigers to an area:
Factors leading to the extirpation of Tigers have been greatly reduced or completely
In most cases, where reintroduction might be considered, the cause of extirpation will
have been overharvesting of Tigers and/or prey. In this instance, a measurable and sustained
reduction of hunting threats is necessary, and most easily demonstrated through the recovery of
prey populations. While the complete removal of all threats is unlikely to be possible anywhere, and
some losses of a founder population must be expected, all threats must be limited to extremely low
levels. More rarely will reintroduction be possible in areas where Tigers have been lost due to
habitat destruction. While Tigers often thrive in partially-logged or manipulated habitats (Linkie et
al. 2008, Lynam et al. 2007, Miquelle et al. 1999a, Darmaraj and Mohamad 2009), there is little
likelihood of successful Tiger reintroduction unless the area has management that is dedicated to
prey recovery, has been made conflict-free, and is largely devoid of people.
There is a defined, delimited and inviolate core area large enough to hold at least 25
breeding females (c. 75 Tigers).
The need for inviolate areas is essential given the fragility of a
founder population and the high number of potential challenges facing translocated Tigers in
deciding to remain within the area and avoiding conflict. The severe impacts of human-caused
mortality of Tiger populations are well documented (Kenny et al. 1995, Goodrich et al. 2009, 2010).
At the very least, the area must be free of poaching of both Tigers and prey, free of livestock, and
have no people living inside its boundaries (Karanth et al. 2003, Karanth and Gopal 2005, Miquelle
et al. 2010). Further, all resource extraction activities should be eliminated.
There is a well-defined landscape that is Tiger permeable and has the long-term potential
to sustain 50 breeding females (c. 150 Tigers).
The ultimate target should be to have resident
Tigers across these landscapes, thus the definition of their boundaries should realistically reflect this
potential. That is, habitat quality and prey density should be sufficient to support Tigers throughout
the landscape and connectivity between patches must be sufficient to permit dispersal. The target
numbers are indicative and there is some debate as to whether this number should be higher. For
the purposes of this report we have been conservative, but recognize that this number may need to
be increased under certain conditions.
Prey densities are sufficient to support 25 breeding females in the core area and, longer-
term, an additional 25 breeding females in the surrounding landscape.
Constructing Tiger
landscapes will require building outwards from areas with high densities of Tiger prey. In many
cases, areas will have depressed prey densities and incomplete prey communities, thus the focus of
early efforts must be on re-building biomass of large prey, such as large cervids, bovids and suids,
which are one of the key determinants of carrying capacity for Tigers and Tiger density (Karanth
and Nichols 1998, Karanth et al. 2004). This is best done through first selecting core areas that can
be made inviolate, where all human activity is strictly controlled and wildlife populations can
recover unencumbered. The wider landscape should also be capable of maintaining an additional
minimum of 25 breeding females, though these are likely to be at lower densities, thus requiring
larger areas.
Intensive law enforcement programs will be required to ensure the recovery of these areas well in advance
of any reintroduction, with recovery of prey serving as a useful indicator of whether law enforcement
efforts are effective and sufficient to prevent poaching of reintroduced Tigers. However, law enforcement
may need to be increased following the release of Tigers because of the greater incentive to poach them.
Details on protocols for law enforcement and law enforcement monitoring are provided above. Recovery of
extirpated Tiger populations will require exceptional government commitment with recovery strategies
integrated into national and regional development plans. This should be a precursor to the following actions
and investments, and are also necessary to achieve the conditions listed above.
Clarification of authority.
It is essential that governmental authority over the management of the
area is unequivocal and straightforward. Although landscape conservation will usually involve
multiple agencies, it is important that ultimate responsibility is not diluted by governance
complexities. We suggest that a dedicated Tiger authority be convened with a clear terms of
reference, staffed by qualified specialists rather than political appointees, suitably funded, and given
sufficient authority to be both effective and accountable.
Landscape planning.
Concurrent with the above activities, efforts should begin by establishing
the conditions at a landscape-level for the eventual reintroduction of Tigers (e.g. Ranganathan et al.
2008). At the heart of this should be a conflict prevention and mitigation strategy for issues such as
livestock management and direct human-Tiger conflict. These efforts should be detailed in a
landscape strategy, setting out the specific conditions that need to be achieved for the landscape so
that it can maintain a functional meta-population of Tigers. Ambitious but achievable goals with
time-specific, explicit objectives and measurable indicators of success should be established.
Initiation of rigorous prey monitoring programs.
A two-tier program of Tiger prey monitoring
will be required from the beginning, both to feed into adaptive management planning and to
inform decisions about when, how and where the Tiger translocations should occur. The first level,
focused at the Source Site, should be intensive and follow the guidelines in the preceding sections
of this report. The second should be implemented at the landscape level and follow occupancy
protocols (e.g. Mackenzie et al., 2006)
Translocation strategy
. Concurrent with all of the above, specialists with experience in Tiger
biology, genetics, wildlife health, wildlife reintroduction, carnivore-human conflicts, and law
enforcement should be convened to assist the relevant governments with the development of a
translocation and reintroduction strategy. These specialists should include practitioners with
experience of large carnivore restoration in other regions of the world, field veterinarians, Tiger
ecologists, law enforcement officers, monitoring experts, and logisticians.
While consultations should be conducted before and throughout the process, it
will be especially necessary to hold a number of meetings with key stakeholders in the landscape,
from central government to local authorities, farmers and communities, and relevant agencies. If
sufficient support, or at least acceptance, is not established for Tiger reintroduction at all levels, it is
unlikely to succeed. Thus, significant political commitment and resources will need to be invested
before, during and after any reintroduction.
Recovery strategy.
A post-release strategy should be developed that includes detailed contingency
plans for all likely post-release scenarios. It is important that training in these scenarios is provided
to project staff well in advance of the relocation.
Monitoring Tiger recovery.
Prior to reintroduction, a plan must be developed for monitoring
recovery of the Tiger population. All released animals should be equipped with telemetry collars
(Goodrich and Miquelle 2010) and the plan should include monitoring of reproduction and new
animals born into the population, e.g. via camera-trapping or non-invasive DNA sampling
(Karanth and Nichols 1998, Karanth et al. 2002, Mondol et al. 2009).
These guidelines are meant to be broadly indicative and not prescriptive. We strongly recommend, however,
that an international working group be convened to develop detailed guidelines for the possible
translocation of wild Tigers.
Tiger Success Stories
Although it is instructive to understand where and why Tiger conservation efforts have failed, it is possibly
more constructive to examine where they have succeeded. Too often, however, high-profile Tiger ‘success’
stories have focused on the promotion of perceived improvements for general conservation rather than
saying anything about Tigers themselves. This is due in part to a lack of investment in monitoring at these
sites, which would have provided this critical information but, for whatever reason, the most valuable
lessons must surely be drawn from landscapes where there have been demonstrable and sustained
recoveries of Tigers.
Russian Far East
Following intensive commercial hunting of Tigers in Russia in the early 20th Century, the Tiger population
was close to extinction and down to 20-30 individuals in the 1940s (Kaplonov 1948, Henry et al. 2009). In
1947, Russia became the first country to ban trade in and hunting of Tigers, which led to a dramatic reversal
in the population decline (Miquelle et al. 2010). When the collapse of the Soviet Union occurred in 1992,
political and economic chaos, as well as access to international black markets resulting from re-opened
borders, resulted in soaring poaching rates on Tigers. At this key junction, the creation of an anti-poaching
team, Inspection Tiger, was critical in bringing the situation under control. At the previous census in 2005,
the Tiger population had increased and stabilized to between 428 and 502 individuals, largely as a result of
effective law enforcement, a feat made all the more remarkable by the fact that the majority of these
individuals are part of one contiguous population (Miquelle et al. 2010). The Siberian Tiger Monitoring
Program, set up by WCS in collaboration with Russian scientists, NGOs, and government organizations,
was established as an annual early-warning system for identifying potential declines in Tiger and prey
populations between each full population census conducted roughly every 10 years. Recently, the
monitoring program has demonstrated its utility by indentifying a short-term but significant decline in both
prey and Tiger numbers since 2005. Most conservationists feel, however, this trend can be reversed with
revised legislature and a renewed commitment to anti-poaching efforts. If this monitoring program were
not in place, it is doubtful whether any response would have been put in place at this time.
Central Western Ghats, India
The Malenad-Mysore Landscape in the Western Ghats of Karnataka, India presents a multitude of success
stories in recovering wild Tigers and their prey. The Nagarahole National Park, Karnataka State, under
strong management stewardship ensuring effective protection, has witnessed more than a 400% increase in
the Tiger population over the last 30 years, and has sustained that through challenging times and increasing
pressures. Most importantly, sustained protection and monitoring efforts in Nagarahole have, in recent
years, succeeded in securing its status as a key Source Site in the landscape, maintaining its key function for
putting Tigers out into the wider landscape. Together with the adjoining forests of Bandipur National Park,
it is one of the two largest source populations of Tigers anywhere in the world. The landscape now has
more than 220 Tigers just within its Source Sites and with dispersal pressures increasing, more and more
animals are moving between Source Sites, forming a varied and functioning meta-population.
Building on the success of Nagarahole, WCS is now transferring these lessons to other Source Sites within
the Malenad-Mysore Landscape and other parts of the Western Ghats. The Bhadra Tiger Reserve, declared
a Project Tiger Reserve in 1998, witnessed the culmination of a major village re-settlement scheme in 2002,
with approximately 4,000 people re-settled to a new location outside of the Reserve. Under careful
surveillance and monitoring by WCS, prey densities inside the Reserve were seen to double between 2005
and 2008, and the declining Tiger population, has now stabilised. As prey populations begin to recover,
Tiger populations are expected to follow.
Increase in Tiger populations in the Nagarahole National Park
The notable elements of success in this landscape have been a State government that considers Tiger
conservation a priority, the creation of inviolate core areas through voluntary, thoughtful, long-term and
transparent relocation programs, and the application of strict protection for both Tigers and their prey. It
has also pioneered the Source Site conservation model, whereby a strong focus on building-up core areas
and not allowing conservation efforts to be diluted or distracted, has now resulted in a functioning meta-
population in one of the most densely populated conservation landscapes on earth.
The Russian Far East and the Western Ghats share few geographic, demographic, cultural or ecological
features. In fact, the Tiger is one of the few notable commonalities. So what is it about these sites that have
made them arguably the only two landscapes to have effected and sustained long-term Tiger recoveries?
The first is the most obvious: genuine political commitment. Although each has had their problems, both
Project Tiger in India and Inspection Tiger in Russia were instrumental in the recoveries, representing the
manifestation of the commitment from their governments. India and Russia developed and, almost
uniquely, implemented Tiger-specific policies and provided political backing for tough decisions and strong
The second feature is that they created a physical task force for these commitments, rather than just adding
the entire burden on existing staff within, say, the protected area network. Furthermore, they provided
funds sufficient for effective implementation. At least in India’s case this was far in excess of the
international communities combined contribution for Tiger conservation. Many countries have made verbal
commitments to Tigers, most have strong legislation in place, but only India and Russia have been both
able and willing to initiate, fund, and maintain these commitments over a significant period of time.
The third is relative stability and wealth. It should be noted that some countries such as Nepal, Cambodia,
Laos and Bangladesh have each made significant strides towards foundations for Tiger conservation,
though each have been hampered either by internal strife or lack of sufficient available funds.
While India and Russia have been the long-term examples of success, it is encouraging to see other
countries such as Thailand now making similar progress.
Huai Kha Khaeng Wildlife Sanctuary, Thailand
The Huai Kha Khaeng Wildlife Sanctuary represents a model Source Site for Tiger conservation in South-
East Asia. In 2006, WCS, in collaboration with the Thai government, launched a Law Enforcement
Monitoring program in Huai Kha Khaeng, and in 2008, under the Tigers Forever collaborative program with
Panthera, enforcement operations were intensified through the launch of the ‘Smart Patrolling’ system, using
MIST to monitor law enforcement efforts and improve management effectiveness accordingly. In line with
the Smart Patrolling strategy, the government has allocated some 170 patrol staff to law enforcement efforts
in Huai Kha Khaeng. The primary current threat to Tiger recovery in Huai Kha Khaeng is poaching of
Tiger prey. Between 2006 and 2009, LEM data indicated that a sustained law enforcement effort had
succeeded in reducing and stabilizing incidences of prey poaching. These data were also able to focus law
enforcement teams in areas of high poaching threat. In 2004, annual monitoring of Tigers and their prey
was also initiated by WCS in collaboration with government partners. Annual monitoring data indicate that
between 2007 and 2009, Tiger densities in Huai Kha Khaeng increased from 1.74 (SE 0.24) inds/100km2 to
2.39 (SE 0.29) inds/100km2. Replication of this model throughout the Western Forest Complex of
Thailand has enormous potential for recovering – and securing - Tiger populations in this landscape and in
mainland South-East Asia.
Nam-Et Phou Louey National Protected Area, Lao PDR
The Nam Et Phou Louey National Protected Area (NPA) is a key site for Tiger recovery under the Tigers
Forever collaborative program with Panthera, and characteristic of many Source Sites and potential Source
Sites in South East Asia, where sufficient habitat remains, but where Tiger and prey populations have been
extremely depleted and efforts are currently focused on reversing the overall trend in population decline. In
2007, under the Tigers Forever program, the Lao government, with the full support of the WCS Lao Program,
considerably scaled up enforcement efforts in the Nam Et-Phou Louey NPA with the implementation of a
MIST-based LEM program. Patrol staffing, effort and coverage increased four-fold between 2007 and 2009,
with a focus on reducing poaching of Tigers and prey in the core zone, and on clamping down on the
possession of illegal firearms. Baseline surveys of prey populations conducted in 2008 indicated that the
current prey population can support approximately 20 adult Tigers (Karanth et al. 2004). In 2009, minimum
population estimates from DNA-analysis of Tiger scat, indicated at least nine individual Tigers are currently
using the core zone.
Source Sites in themselves are by no means the entire solution to the problem of how to save the Tiger.
Only concerted, orchestrated and politically bold commitments by range-state governments, sustained over
a number of decades can do that. It will require a broad range of actions across a variety of sectors that are
well documented within the Global Tiger Initiative (GTI) process. However, progress on these fronts
before Source Sites are secured is the equivalent of constructing stories on a building with fragile
There is a very real danger that in all the encouragingly ambitious discussions about how tall the building
should be (or how many Tigers do we aim for?), that we forget just how fundamental the foundations are.
Today, we believe that some grandiose plans for constructing Tiger landscapes might not reflect the
frangible nature of these foundations, what it will take to reinforce them, or how best we should go about
doing this.
If there were already more than enough funds to secure these sites and fund the other essential activities,
then there would be no issue. However, this report has shown that far great resources are going to be
needed than are currently available, and very quickly, if Tigers are not to be lost from more countries across
their range. This report has aimed to identify where we should build (Source Sites), how we should do it
(bottom up), and how much more is needed to do it (US35 million a year). We believe that this price is a
bargain we can’t afford to miss.
This report is also intended to link investments with results. Beyond sourcing a price, we have presented a
set of guiding protocols for how to monitor and guide this investment, and how best we can show our
conservation investors - whether they be governments, banks or beer companies - how we turn a profit by
increasing wild Tiger numbers.
While the international community spent US$34 million on Tiger protection from 1998 to 2003 (Christie
2006), most of this came from western nations, western-funded multi-laterals, and western companies. As
Asia increasingly becomes the driving force behind the world’s economy it is humbling to see how
insignificant $35 million per year really is in some contexts:
1. The Asian Development Bank (ADB) spent US$274 million on technical advisors, US$1.5 billion
on private sector assistance, US$10.5 billion in loans and $811.4 million in grants. This was in 2008
2. This year Tiger Beer is running a series of promotional events to coincide with the Chinese Year of
the Tiger. The company alone made well over US$100 million profit last year, selling beer.
3. One of Malaysia’s banks made a profit of US$210 million during the same period Singapore
Airlines made nearly US$200 million profit. That period was the first three months of 2008.
Asia may no longer have the large intact landscapes that once allowed the Tiger to walk from the Pacific
Ocean to the Caspian Sea, but it certainly now has the capacity, wealth and knowledge to take control of the
wild Tiger’s destiny.
Ahmad IU, Greenwood CJ, Barlow ACD, Islam MA, Hossain ANM, Khan MMH, Smith, JLD. 2009.
Bangladesh tiger action plan 2009-2017. Dhaka (Bangladesh): Bangladesh Forest Department,
Ministry of Environment and Forests, Government of the People’s Republic of Bangladesh.
Amstrup, SC, McDonald TL, Manly BFJ, editors. 2006. Handbook of capture-recapture analysis. Princeton
(NJ): Princeton University Press.
Anderson DR, Burnham KP, Thompson WL. 2000. Null hypothesis testing: problems, prevalence, and an
alternative. Journal of Wildlife Management 64: 912-923.
Anon. 2007. Vietnam Forestry Development strategy 2006-2020. Promulgated with Decision No.
18/2007/QD-TTg by the Prime Minister. Hanoi, Vietnam.
Balmford A, Whitten, T. 2003. Who should pay for tropical conservation, and how could the costs be met?
Oryx 37: 238-250.
Baltzer MC, Nguyen TD, Shore RG, editors. 2001. Towards a Vision for Biodiversity Conservation in the
Forests of the Lower Mekong Ecoregion Complex. Hanoi and Washington D.C.: WWF
Indochina/WWF US.
Barlow ACD, Ahmed MIU, Rahman MM, Howlader A, Smith AC, Smith JLD. 2008. Linking monitoring
and intervention for improved management of Tigers in the Sundarbans of Bangladesh. Biological
Conservation 141: 2031-2040.
Bell D, Roberton S, Hunter PR. 2004. Animal origins of SARS coronavirus: possible links with the
international trade in small carnivores. Philosophical Transactions of the Royal Society B. 359, 1107-
Breitenmoser U, Breitenmoser-Würsten C, Carbyn LN, Funk SM. 2001. Assessment of carnivore