MIKE (Monitoring the Illegal Killing of Elephants)

 

Monitoring the Illegal Killing of Elephants (MIKE)
Preparatory meeting for Southeast Asia

AIT Center, Bangkok, 25th-26th October 1999

MINUTES

Monday 25th October

The meeting was facilitated by Simon Stuart, Co-ordinator of the Species Survival Programme, IUCN and he began the meeting by introducing Peter Rezel.

Welcome by Peter Rezel, IUCN-Asia Office, Bangkok

The participants were welcomed to IUCN's regional office. The participants were wished success in their discussions at this important stage of MIKE's development.

Introduction by Pak Widodo Ramono, Deputy-Chair of the Asian Elephant Specialist Group.

The importance of MIKE for elephant conservation activities in Southeast Asia was emphasised. It was also stressed that within the sub-region MIKE must reflect Southeast Asian conditions and be responsive to needs here.

An introduction to CITES (Convention on International Trade in Endangered Species of Wild Fauna and Flora) by Malan Lindeque of the CITES Secretariat

A general introduction to CITES as well as background documentation was provided by the CITES Secretariat. The importance of adequate national legislation as the basis for the implementation of CITES was emphasised. Many Parties, including those in Southeast Asia need to update and improve their legislation in order to enhance the effectiveness of CITES as a tool for biodiversity conservation and as a framework for international trade. The CITES Secretariat also shares with Asian elephant range States the concern that not enough attention has been given to the conservation of Asian elephant populations within CITES, as the result of the historical focus on the African elephant. Decisions taken by the 10th meeting of the Conference of the Parties to CITES concerning the African elephant have, however, provided an opportunity to direct resources and attention to Asian elephants, largely through the establishment of international monitoring of the illegal killing of elephants as well as trade in elephant products.

This workshop was accordingly arranged to initiate the implementation of monitoring systems approved by the Conference of the Parties and more recently, the Standing Committee of CITES. The objective of the MIKE (and ETIS) system is to assess the impact of CITES policies and decisions concerning elephants and trade in elephant products on elephant populations in the wild, but with the important corollary that this initiative should increase technical capacity at range State level and contribute to national conservation efforts

Presentation on the Elephant Trade Information System (ETIS)

Chen Hin Keong presented an overview of ETIS. He began by outlining what CITES required of ETIS in terms of monitoring trade in elephant products. CITES requested that TRAFFIC manage this system. TRAFFIC has been maintaining a Bad Ivory Database System (BIDS) for some time and it was determined that after further development this should become "the appropriate instrument for monitoring the pattern and measuring the scale "of trade in ivory and other elephant products. ETIS will contain three datasets (Resolution Conf. 10.10 (Rev. CoP12), Annex I): seizure records, law enforcement effort and related supplementary information.

Presentation on Monitoring the Illegal Killing of Elephants (MIKE)

Phil McGowan started by presenting the objectives of the international illegal hunting monitoring system that CITES agreed to in Resolution 10.10. He then provided background on the purpose of a monitoring system and how this is fulfilled. MIKE is site-based and the way in which the sites were selected was indicated. The proposed process that will feed this information into the CITES decision-making process was outlined, together with the suggested flow of information within MIKE: six sub-regions providing data to a central unit. Finally, the various areas of activity that will be funded under MIKE were outlined (e.g. field surveys, capacity building).

At this stage of the meeting, Simon Stuart drew everyone's attention to a flip chart entitled 'Parking Lot'. Any questions, comments, concerns or other points relating to MIKE could be written on this sheet at any time during the meeting and would be discussed.

It was also noted that unfortunately the Forest Resources Conservation Division from Lao PDR was not able to participate.

Agreement of objectives for the meeting (what outputs are required?)

The following list of outputs were proposed and agreed:

 

1.

Understanding of level of government participation in each country.

 

2.

Agreement on pilot phase MIKE sites

 

3.

Structure for managing MIKE in SE Asia

 

4.

Agreement on how sub-regional coordination should be handled

 

5.

Compile an inventory of resources available to MIKE (NGO's, scientific institutions, existing programs, donors etc)

 

6.

Agreement on a strategy for sampling elephant populations in SEAsia

 

7.

A list of research and development needs on methods

 

8.

A list of factors affecting illegal killing and how they can be measured or how the elephant population is affected.

The proposed objectives of the meeting fell into two broad categories: those relating to policy and management of the pilot phase and those relating to fieldwork. Therefore, it was suggested and agreed that the meeting split into two groups. The first four points would be discussed by a 'policy group' comprising government officials from all five countries present and points 6 and 7 by a 'field methods group'. The latter would also begin to compile the inventory indicated in point 5, and point 8 would be discussed by all later in the meeting.

NB: The outcomes of the two discussion groups are presented separately below. The discussion then

Policy group

Simon Stuart, Malan Lindeque and Liu Yuan met with government officials to address the issues outlined above. The discussion started with the sites as the sites that were taking part in the pilot phase had to be determined before government contributions could be discussed.

Potential pilot phase sites

The site in Indonesia

This could be included in the pilot phase, but concerns were expressed about the size of the site and so Indonesia would come back to CITES with a suggestion for which part should be monitored, or failing that, an alternative site. This would then be discussed with the Secretariat and with the statistician who has provided a globally representative set of sites.

The site in Malaysia

This could be included in the pilot phase, but concerns were expressed about the size of the site. Malaysia would come back to the Secretariat with a suggestion of appropriate boundaries.

The site in Myanmar

This could be included in the pilot phase.

The site in Thailand

This could be included in the pilot phase.

The site in Vietnam

This could be included in the pilot phase.

The site in Lao PDR

Unfortunately, colleagues from Lao PDR were not able to attend the meeting and, therefore, it was not possible to get any opinions as to the likelihood of this site being included in the pilot phase.

In general it was considered that MIKE should, wherever possible allow sites to fit in well with the existing elephant programmes.

Requirement of each site and in each country

What was needed to carry out MIKE activities in each country was then discussed

-

field workers would usually be government staff but with some people from NGO's

-

the most important government contribution would be personnel salaries in the field and in-kind contributions

-

facilities that are required include vehicles, housing and field allowances. There is also a need for porters and guides.

It was pointed out that one of the main goals of MIKE is to build government capacity: a person would often be seconded to the project as the main liaison person in each country (see below).

In the Central African pilot phase it was noted that the governments were not in a position to make any financial contribution from governments.

Government contributions to field surveys

Indonesia

PKA can provide staff, including rangers, biologists, foresters but not their extra salaries that would be required in the field. Training costs would also need to be covered from outside, including training in data analysis. PKA can provide some housing, but the availability of vehicles is very limited. For large surveys, the costs of short-term staff will have to be covered from outside.

Thailand

RFD can provide existing staff for some of the work. Short-term staff that would be needed for large surveys would have to be brought in and funded from outside. RFD can provide vehicles and housing.

Malaysia

PERHILITAN can provide existing staff, vehicles and housing. Funds would be needed for extra staff for surveys and field allowances.

Myanmar

The Forest Department can provide existing staff, including bringing in people from other protected areas for training. It can also provide domesticated elephants for transport within the site and possibly transport to the site. Camping equipment would need to be provided by the project for survey activities and personnel activities.

Vietnam

Need experts for surveys and training and also require funds for forest guiding. Not enough staff to take part and would also need support for housing and vehicles.

Country/site co-ordinators

As each country has only one site in SEAsia, there needs to be only person acting as the liaison point. This is in contrast to other sub-regions where there are sometimes two or more sites per country.

The country/site-co-ordinator would be below director level but would usually be appointed by the director of the relevant government department. There is, however, a need for some flexibility as each situation is different. Where appropriate this position should link with the relevant CITES Scientific Authority.

Across the sub-region it was felt important that country/site co-ordinators should liase with other co-ordinators, including the sub-regional co-ordinator. A Regional Advisory Committee would help this if it was composed of country/site co-ordinators and so should be established.

Sub Regional Co-ordination

How MIKE might be co-ordinated in SEAsia was then addressed. As IUCN had been requested by CITES to undertake the current stage of preparation for MIKE in the sub-region, Simon Stuart left the discussions at this stage. This was to allow a free and open discussion on this issue between the CITES Secretariat and the governments.

The first issue that was addressed was the criteria that the organisation undertaking sub-regional co-ordination should satisfy. The main requirements were that the organisation:

be well established in the sub-region

have excellent communication skills

should add to sub-regional capacity and ownership

It was felt that during the pilot phase the IUCN office in Bangkok would satisfy all of these in the most desirable manner. It was further agreed that the programme manager (i.e. sub-regional co-ordinator) should be from the sub-region. Subsequently, the potential for co-operation with the Association of Southeast Asian Nations (ASEAN), or even a transfer to that inter-governmental organisation should be assessed.

Field methods group

Sampling strategy

The group started by discussing what was required. During this discussion, part of the written record of the IUCN/TRAFFIC responses to questions at the 41st Standing Committee Meeting (Geneva, February 1999) was read out. These state that "... The monitoring system will help us to understand the process being monitored and hence to make judgements on causal relationships, including the effect of any future decisions to allow legal trade, on the incidence of illegal killing that may be made once the monitoring system is in place. To "prove" that the resumption of legal trade, or any other change in the status of the African elephant within CITES, has led to an increase in mortality from poaching is not something that any monitoring system, could ever achieve regardless of its design. The most that can be achieved is to statistically assess the evidence relating or correlating legal trade with elephant mortality. MIKE is designed to make such assessments."

The other issue that was raised was that of the analytical framework: the specific questions being asked and some of the work done to date would influence the sampling strategy. Therefore, there would seem to be a need for greater clarity in these two areas. In essence, it would be difficult to determine a sampling strategy without this information, mainly because of uncertainty over the requirements for comparable variance estimates between sites.

A third uncertainty was the personnel who would be carrying out the work. The level of training etc required would all depend on the experience of the people carrying out the sampling.

Despite these uncertainties there is a need to make progress towards determining how the pilot phase can proceed in a scientifically defensible manner. Field details are also needed that will allow the pilot phase to be costed.

The over-riding issue in SEAsian forests is that densities of elephants are low. In many, if not all, cases the distribution of elephants within sites is not known and movements are also not clear. Therefore, obtaining enough encounters from which density estimates could be derived is a real challenge for any sampling programme. Possibly at the extreme is Yok Don NP in Dak Lak Province, Vietnam, where an earlier attempt to survey elephants using dung counts had to abandoned because there were insufficient piles of dung found. As the discussion was now turning to site specifics, each site was addressed in turn to consider what methods would be practical in the field. First, however, a list was made of methods that could be considered. This included: village interviews, Barnes dung counts, recce dung counts, direct sighting line transects, camera traps, belt transects (subdivided into set lengths and sign evidence recorded), and aerial surveys (as used by Olivier and Woodford in survey for kouprey in Cambodia in 1994).

The site in Vietnam

This site arguably presents the most severe test of reliable monitoring of elephants within the SEAsian sites. Belt transects were considered the only real possibility here. Numbers of elephants are extremely low. This is a method that has been used in surveys of carnivores. Long transects are walked, which in this case would probably mean across the park in both north-south and in east-west directions. Evidence is recorded in each of the strips. A major concern raised was uncertainty over whether the elephants that use the area are wild or domestic. In some periods, domestic elephants are known to roam in the park and ensuring that footprints or dung that is recorded relates only to wild elephants is seemingly impossible. There is a need to address this uncertainty before monitoring starts.

The site in Lao PDR

This site rivals the site in Vietnam as the most difficult of the six to reliably survey. It is a large area, in which the distribution and movements of elephants is almost entirely unknown, but which is better known than any other site in the country. There are some practical concerns about the site, principally that a part that elephants are known to inhabit is due to become the head pond of a dam in about five year's time. To date access to various parts of the site has been difficult.

The site in Thailand

Collection of field data is possible at this site and the Barnes technique should yield sufficient dung encounters to allow densities of elephants to be estimated. This is also an area where the recce transect method could be evaluated. Further information should be sought on the feasibility of using camera traps at this site from WCS, who have been active here.

The site in Myanmar

Collection of field data is possible at this site and the Barnes technique should yield sufficient dung encounters to allow densities of elephants to be estimated.

The site in Malaysia

This is large and very variable site, with terrain differing significantly across the state. With a better definition of the boundary of the site, it should be practical to collect appropriate data. The Barnes method would be appropriate and the recce transects could be evaluated. Camera traps are being used in the Peninsular in a joint PERHILITAN/WCS programme and it should be clarified where they have been used and how successful they were.

The site in Indonesia

This is large and very variable site, with terrain differing significantly across it. With a better definition of the boundary of the site, it should be practical to collect appropriate data. It is worth considering whether altitude could be used to mark the upper limit and whether lowland areas outside the park should be included, essentially creating a lowland landscape. The Barnes method would be appropriate and the recce transects could be evaluated, as could belt transects with a view to their use in the Lao or Vietnamese sites, should they prove successful. Camera traps are being used in this site and it should be clarified where they have been used, sampling protocols etc.

Dung issues

Models

The first comment made was that the steady state assumption, which forms the basis of most conversions from the number of dung piles to elephant densities, is difficult to justify. Crudely, this requires that there is always the same amount of dung around for a given number of elephants (i.e. decay rate = defaecation rate) and that any variation in the amount of dung means that there is a difference in the number of elephants.

The Hiby and Lowell model (the programme DUNGSERV) requies no such assumption and there would be merit in evaluating this during the pilot phase if possible. A second alternative would be to use the recently proposed approach of Barnes and colleagues (J. African Ecology 35: 39-52, 1997) that relates rainfall patterns to the population dynamics of dung piles.

Field calibration

Work that has been undertaken so far suggests that is substantial variation in the dung decay and also in the rate of dung defaecation between sites and that this can also vary at a site in different seasons. Evidence from smaller dung (e.g. that of cattle) suggests that there may also be variation in decay rate across a site. Therefore, both dung decay rate and the defaecation rate should be calibrated at each site, preferably in several places and throughout the pilot phase. This will provide the most appropriate figures for using in the conversion from the number of dung piles to the estimate of elephant densities. Qualitative notes should be made at these places, such as whether pigs or dung beetles suddenly accelerate the decay of dung. This will all help in understanding the decay rate data.

Rainfall data should also be collected at each place where decay rates are being calibrated. This would allow the relationship between rainfall and dung pile dynamics to be established, but this will take longer than the pilot phase. However, if it can be done, it will mean that there will be no reliance on estimates of decay rate in the future.

In conclusion, it was seen that the emphasis had to be on the word 'pilot' in this pilot phase. There are too many unknowns to be able to make realistic prescriptions of the field effort needed. Consequently, year round fieldwork will be needed during the pilot phase to clarify these issues. This does limit our potential to cost the pilot phase of MIKE and thus produce an acceptable proposal. However, the approach that might be adopted was addressed later (see below).

Inventory of related activities

It was realised that there are a variety of elephant activities that are either definite to go ahead or that are planned over the coming twelve months. These could be of considerable benefit to MIKE in terms of either continuing development of particular field activities or conducting trials of other methods (e.g. recce transects) in the sub-region). Therefore, a list was compiled of projects that participants knew about. This is not a complete list and completing this 'inventory' should be a priority during pilot phase implementation so that appropriate links can be established.

Cambodia

Although not included in MIKE at this stage, there is likely to be work that will include elephants in the coming year. These include:

  • FFI - planning elephant surveys.
  • WCS - planning to undertake elephant survey training and to use camera traps.

Because of the potential benefits for MIKE, it was felt that it would be desirable to keep Cambodia informed about the monitoring system. Perhaps it would be possible to include Cambodia in the CITES Mission.

Indonesia

There are a variety of elephant activities underway on Sumatra. Inclusion of Kerinci-Seblat in MIKE means that all four key elephant localities will be subject to some work.

  • WCS - active at Barisan Selatan and are addressing methodological issues.
  • FFI - a relatively large project in Aceh Province (Conservation of Elephant Landscapes in Aceh)
  • WWF - beginning to work in Riau Province as part of the WWF AREAS (Asian Rhinos and Elephants Action Strategies) project

Lao PDR

The World Bank may continue its involvement in activities at this site and if this happens it is likely to include elephant surveys on the Nakai Plateau in more detail than has hitherto been possible.

Thailand

There have been a variety of field activities in Thailand that could inform MIKE protocols in the field. These include:

  • Mahidol University - previous work in Huay Kha Khaeng Wildlife Sanctuary in the Western Forest Complex.

There have been a variety of previous surveys and related activities in and around Khao Yai NP. These include:

  • Wildlife Fund of Thailand (WFT) - in 1992 identified areaas hat were used by elephants
  • Robert Dobias - 1978 estimated that there were about 250 elephants.
  • Lisa Clime (a Peace Corps volunteer) - 1995
  • RFD - a variety of initiatives that are relevant
  • WCS - plans for camera traps
  • WWF-T - no plans to work there

Vietnam

Because of uncertainty expressed above over the possibility of surveying only wild elephants at the site in Vietnam, this was not considered further, although there are elephant initiatives underway in Cat Tien (WWF) and in Tan Phu (FFI - ?temporarily suspended).

Myanmar

There was no knowledge within this group of activities planned or underway in Myanmar.

Tuesday 26th October

At the start of the Tuesday there was a report back from each group. Simon Stuart and Malan Lindeque reported back from the policy group and Arun Venkataraman reported back from the field methods group.

Agreement on pilot phase sites

In discussion it was agreed that the sites in Thailand and Myanmar were suitable for inclusion in the pilot phase. The sites in Indonesia and Malaysia required clarification of the boundaries and practicalities, and some discussion with the Secretariat. Because of the concerns raised over the site in Vietnam, it was considered that a feasibility assessment was required first and that this should be extended to include Lao PDR which was not able to attend the meeting. Cambodia should also included in this feasibility. [Note added subsequently: there is an alternative site identified for one of these sites]

At this stage, both in discussion and in points written on the 'Parking Lot' flipchart, it was obvious that there was need to discuss some of the analysis issues in more detail. It was felt especially important to try and determine who might undertake the data collection and analysis at the site level, and also who would undertake the analysis at other levels (e.g. sub-regionally and globally). These were discussed in turn. Before that, however, it was requested strongly that training be provided in analysis.

Data collection and analysis issues

Site level

The site in Thailand

RFD would undertake data collection.

There is capacity available for any site level data analysis within RFD and WCS or other NGOs, especially within the Wildlife Research Division of RFD (Ronglarb Sukmasung)

The site in Myanmar

The Forest Department would undertake data collection and analysis, but there is a need for training in analysis and increased capacity.

The site in Malaysia

Data collection could be undertaken by PERHILITAN with input from WWF, MNS and local universities.

PERHILITAN and local universities would undertake analysis.

The site in Indonesia

Data collection would be undertaken by PKA, Indonesian Institute of Science (LIPI), NGOs, both FFI (who are active in the area) and locally based, and local universities. Any monitoring at other sites could use NGOs active at that site.

LIPI and local universities could undertake analysis. International NGOs operating under MOUs can help in training in suitable analytical methods.

Analysis at the sub-regional level

There was an overwhelming feeling that this was required because of SE Asian elephant issues. A brief discussion followed as to where such analysis might take place. Mahidol University was proposed as a sub-regional university with the capacity to tackle this task. Because this would require a little more discussion, no decision was taken on this and further consideration was a high priority of the proposed Technical Advisory Committee (see below).

Global issues

MIKE data will be analysed at a central unit in order to fulfil the requirement of CITES Resolution 10.10. As this is the formal mandate for MIKE, it must be fulfilled to the maximum extent possible.

Technical development

In the discussions on field methods and analysis, emphasis was placed on the 'pilot' nature of the pilot phase. This is fine, but there is a need to make progress on resolving the methodological and analytical uncertainties that MIKE currently faces in SEAsia, if not elsewhere. Therefore, there was a clear feeling that a Technical Advisory Committee should be established to address these issues during the pilot phase. This would both provide guidance to the two pilot phases (SEAsia and CAfrica), and also provide the opportunity for the best technical expertise to address the sampling and analysis issues for MIKE as a whole.

Data ownership and access

This issue was note really raised as a point of discussion, but for information it was pointed out that has already been mentioned as in need of clarification. MIKE is overseen by a Subgroup of the CITES Standing Committee and this Subgroup is developing a policy on data ownership and access. Thailand represents the sub-region on this Subgroup and all concerns, comments etc on this issue should be passed to the CITES Management Authority in Thailand so that it can be brought to the attention of the Subgroup.

Factors affecting the illegal killing of elephants and elephant populations

MIKE requires not only the monitoring of elephants, but also of factors that might influence their illegal killing. In SEAsia, elephant populations are sometimes also affected removal of live elephants and other activities that do not involve the killing of elephants. Therefore, it was felt that to gain a true picture of factors operating of elephant populations, these other activities might be considered.

A list of factors was extracted from the original MIKE proposal and presented. This was:

1.

Levels of human activity (development activities, forestry, roads etc)

2.

Law enforcement effort (anti-poaching, staffing levels, effectiveness of legal system etc)

3.

Levels of other criminal activities

4.

Civil strife (for SE Asia the general availability of arms in some areas might be included here, although there is no civil strife)

5.

Spread of poaching

6.

Proximity to international boundaries.

7.

Extent of community involvement and socio-economic conditions. Attitude surveys might address this issue

 

 

The following should also be considered for their impact on elephant population numbers in SEAsia

8.

Human-elephant conflicts

9.

Habitat alteration and fragmentation

10.

Changes in habitat used by elephants

11.

Translocation and re-introduction of elephants

12.

Presence of domesticated elephants

13.

A long-term research (or other) presence in the area

Practicalities of fieldwork in the pilot phase and how to cost it

Despite all of the uncertainties in working out how the pilot phase should proceed in the field, there is a need to produce a realistic and transparent costing for this phase. Therefore, a small group met to try and produce a practical and defensible way forward.

Background

When she surveyed the elephant population in Hua Kha Khaeng, Khun Mattana Srikrajang used two biologists and four rangers to carry out 78 transects, each of two km. This took three months. It was felt that one transect team might comprise 4-6 people depending upon the density of the vegetation through which the transects are carried out. Therefore, to ensure that this is adequately budgeted for, it should be assumed that each team would consist of six people, which is also the number proposed for CAfrica.

The approach for the pilot

As stated repeatedly throughout the meeting, it is the unknown variation in encounter rate and therefore the variance associated with elephant population density estimates that is the critical issue throughout SEAsia. We need to try and reduce this uncertainty. Therefore, it was proposed that the pilot phase be conducted in two stages. First, 100 km of transect will be walked and dung counted. This will give dung encounter rates at each site. From this first stage, it will be possible to develop an estimate of variance, which can then be used to propose adequate sampling intensity after appropriate analysis. This proposed sampling intensity would then be trialed during the second stage, later in the year.

The first stage (including analysis which must be costed in) should be carried out as quickly as possible to allow for more time to be devoted to the second stage, as it will probably involve more work. This means that whilst it is possible to cost the first stage of fieldwork, it will not be possible to cost the second stage in any detail now. This approach does, however, mean that the problem of determining sampling intensity can be tackled in the most sensible way.

In Indonesia, 2-3 km can be covered in a day by one transect team. This means that 100 km would be covered in 50 team days. As we wish to complete the first phase as quickly as possible, two teams in the field simultaneously (i.e. 12 people) would mean that the work could be completed in 25 team days. Add in resting periods etc and the fieldwork should be finished in 1.5 months. This should give sufficient basis for costing this part of the pilot phase proposal.

Layout of transects

Sampling effort will be stratified to reflect the variation in the site: i.e. transects will be arranged to represent the major factors that affect elephant distributions. In Khao Yai it was stated that it is important to understand that human use appears to be very important in influencing the distribution of elephants. It should be possible to take a map of Khao Yai and draw on areas of high and low human use, and habitat types. Based on the distribution of these two influences, the position of transects could be stratified. Similar data on Kerinci-Seblat are held in Geographic Information System (GIS) and this would help in this stratifying process for this site.

Summary and close

The two groups then reconvened and reported back. As all objectives for the meeting that had been agreed after the presentations on the first day had been met and as there remained no other outstanding issues Simon Stuart drew the meeting to a close. He thanked the IUCN office in Bangkok for making all of the arrangements and for welcoming the participants at the start of the meeting. The AIT Center had carried out the arrangements very well. The participants were thanked for travelling to the meeting, many at their own expense and for their patience in grappling with so many unknowns. Malan Lindeque finished by thanking Simon for facilitating the meeting.