In early 2015, we heard the first shocking reports of wildlife deaths at waterholes in two Cambodian protected areas in Preah Vihear province. Storks, civets, and even Critically Endangered vultures were found dead, and a group of livestock herders who had stopped off to refresh themselves in the cool water subsequently suffered stomach pains and vomiting. Nearby, pink pellets lay strewn on the soil. The necropsy and lab tests confirmed what was already feared: the water was contaminated with the pesticide Carbofuran, which appeared to have been deliberately placed in at least five different sites.

A dead plaintive cuckoo found at a poisoned waterhole in 2018. Photo: WCS Cambodia

This was a clear threat to the safety and well-being of local communities, as well as to the many threatened species dependent on waterholes. But what was happening here, and why? Colleagues at the Ministry of Environment and at the Wildlife Conservation Society’s Cambodia programme had several theories. Perhaps conflict between farmers was leading to retaliatory poisoning of livestock. Pesticides may have been washed off nearby rice fields. Local farmers could have been using poison to catch wildlife to consume at home. Each of these possibilities would require different forms of intervention targeted at different groups—we needed to learn more.

The research team ready to begin fieldwork in July 2017. From Left to Right: Yim Vichet, Leng Chantheavy & Emiel de Lange. Photo: Emiel de Lange

In July 2017, I began my research into local wildlife poisoning, with the support of two students, Yim Vichet and Leng Chantheavy. We started with an exploratory study in villages outside the two protected areas affected, where we tried to gain a basic understanding of wildlife poisoning. We used this experience to design and refine a questionnaire survey that was structured based on the Theory of Planned Behaviour. Sensitive behaviours, such as poisoning, are difficult to study, particularly when trying to collect quantitative data such as measuring the prevalence of a behaviour across a landscape. To address this, we used the unmatched count technique, a specialized method developed by health researchers to measure the prevalence of drug use. Using this technique allows the respondent to remain anonymous but enables the researcher to make inferences about prevalence of a behaviour.

Vichet and Emiel interviewing a local resident. Photo: Leng Chantheavy

We also wanted to understand wildlife poisoning more holistically: the motivations behind it, people’s perceptions of and attitudes towards it, and the conditions that allow it. To do this, we used qualitative methods, such as interviewing informants and hosting group discussions in local communities. We travelled on motorbikes to visit 10 villages across the landscape. This was a great, but challenging experience. Over the course of 6 weeks, we spoke with a total of 462 people using our survey questionnaire, had in-depth conversations with another 57 key informants, and held 20 focus group discussions.

Video: Travelling in the wet season through the northern plains landscape in Preah Vihear province Cambodia is challenging. To reach some villages, a boat was required. Credit: Emiel de Lange

Our findings highlighted several concerns, and are already being used to design local interventions. We found that poisoning is a form of hunting used by local residents to gather wild meat for consumption at home. It is predominantly practised by young men, and we had reports of this occurring from eight out of the 10 villages we visited. Unfortunately, the unmatched count technique was not successful in providing a plausible estimate of prevalence, but informant estimates suggest prevalence ranges from just a few households up to 40% of households in a village. We heard many stories of cattle and threatened wildlife being killed, including the giant ibis, Cambodia’s national bird. Because of the risk to livestock and human health, as well as the impact  on water sources and fish, poisoning is widely perceived as an inappropriate way to hunt. In some villages, chiefs and other officials have tried to prevent the poisoning by confronting poisoners or holding community meetings.

A summary of evidence for nine statements about wildlife poisoning across ten villages. Green triangles pointing upwards indicate that the hypothesis is supported, and purple triangles pointing downwards indicate it is not. The size of the shape indicates the strength of this evidence. Kulen Promtep Wildlife Sanctuary (KPWS), Chheb Wildlife Sanctuary (CWS)

In 2020, poisoning has become a greater threat than ever before (, and well-informed interventions are urgently needed. Although specialized methods to measure and assess sensitive behaviours need further refinement, detailed mixed-methods research grounded in behavioural theory, as shown in our study, are a vital step towards understanding and acting on emerging conservation challenges.

The Northern Plains of Cambodia in Preah Vihear province. Photo: Emiel de Lange

The article Using mixed methods to understand sensitive wildlife poisoning behaviours in northern Cambodia is available in Oryx—The International Journal of Conservation.

Emiel de Lange is currently completing his PhD at the School of Geosciences, University of Edinburgh. His research draws on the behavioural and social sciences to understand and design more effective conservation interventions. In particular, together with the Wildlife Conservation Society Cambodia, he is looking at the role of communication and community social networks in interventions that aim to reduce wildlife poisoning in Cambodia.