Find a SECReT supervisor
Information for overseas students
View SECReT animation
Download SECReT brochure

Agent-Based Modelling of Wildlife Poaching

22 March 2013


Joanna Hill

Wildlife poaching is a serious and growing problem in many developing countries. Much of the research on spatial patterns of poaching conducted hitherto have used so called “top-down” approaches, which include the statistical analysis of observational data, and the mapping of spatial concentrations of poaching incidents. The mechanisms that generate poaching patterns are then inferred. Computer simulation is an alternative method that employs a “bottom-up” approach, whereby the researcher specifies a theoretical model proposed to explain the phenomenon, and then tests this using a computer simulation to examine the extent to which the model generates the behaviours it seeks to simulate. Agent based modelling (ABM) is one particular method in the toolbox of complexity science, and one that is particularly appropriate here.  Such models comprise two basic components: 1) agents or actors who engage in behaviours described by simple condition action rules; and, 2) the environment they inhabit. The aim of this thesis is to understand space-time patterns of wildlife poaching using ABM.  Semi-structured interviews of communities which live adjacent to Murchison Falls and Queen Elizabeth National Park (QENP) in Uganda will be undertaken in order to understand why and how poachers hunt wildlife. This information will be crucial for developing abstract models which will test optimal foraging and crime pattern theories. Snares surveys will also be undertaken in QENP for the purposes of model validation. These facsimile models will then assess how different ranger strategies influence the rate of the number of poaching incidents detected.