<p>The international illegal wildlife trade (IWT) threatens countless species globally. Many solutions to the IWT have been proposed and implemented, the most common being increased security. However, security on its own has been ineffective at protecting all wildlife. Wildlife commodity devaluation strategies have also been proposed and trialled as a complement or substitute to security. The strategy has primarily been applied to protect rhinoceros populations by, for example, dehorning them. The apparently logical expectation is that reducing the value of rhinos’ horns will discourage hunters and protect rhino. Either hunters will choose not to hunt a population with devalued animals or choose not to kill one when it is found. Unexpectedly, however, theory and anecdotal evidence suggest that devaluation might fail, and may even encourage hunting and the killing of devalued animals. This apparently illogical outcome is the subject of this thesis. Games and choice-based surveying were used to study the behaviour of people hunting a commodity for financial gain. These methods were used to understand why devaluation strategies might fail to protect wildlife and to understand when they might be beneficial. Two games and a choice-based survey posing different hunting scenarios were developed to measure hunter behaviour and test hunters’ responses to risk, value and devaluation. Lucky-dip games were rapid, highly replicated games used to test the impact of variation in devaluation and security strategies across multiple populations. Thirty-three lucky-dip games were conducted, each with ten members of the public participating. They were conducted at community events (e.g. fairs and galas) across Wellington Region in February and March of 2018. Scavenger-hunt games, to elicit more complex behaviours and interactions among participants, were conducted over a longer time and larger area but were, therefore, also less replicated. Four of these games were conducted with between 8-20 members of outdoor recreation clubs. They were conducted at public parks across Wellington Region throughout 2018. And lastly, an online scenario choice-based survey presented members of recreational hunting clubs with hypothetical scenarios where the value and likelihood of a successful hunt varied. The survey was distributed through hunting organisations and received 333 responses. The three research methods presented similar scenarios but used different formats to test my ideas among a diverse population of people. Each method involved participants hunting protected items that varied in value with some items having been devalued. Measuring for the trade-offs that people make between risk and reward when making hunting decisions, including choices about where to hunt, whether to kill and, if they did, whether to harvest a commodity, was of particular interest. Devaluation failed to protect commodities and increase commodity survival. In both games, hunters chose to ‘kill’ devalued commodities. Of the devalued items located by hunters, 74% and 100% were ‘killed’ in the lucky-dip and scavenger-hunt games, respectively. This appears to be because risk increased people’s perceived value of partially devalued commodities. Low-risk lucky-dip games resulted in 44% more devalued items being kept by participants compared to high-risk games. Additionally, devaluation reduced people’s perceptions of risk. Compared to lucky-dip games with just risk, games that included devaluation caused a 10% drop in survival at the highest-risk treatment. Moreover, when devalued commodities were worthless in the lucky-dip games, 27% more were ‘killed’ compared to when devalued commodities were worth 25% of the full-value commodity. Therefore, contrary to expectations, partial devaluation may be more successful than complete devaluation. Greater rates of devaluation were also met with greater variation in commodity survival between games. Coefficients of variation for commodity survival increased from 12% to 41% as devaluation rates increased from zero to 100%. Moreover, respondents to the survey ranked the most devalued population (90%) highest 36% of the time, but also lowest 35% of the time. Thus, peoples’ responses to devaluation vary. Risk was a more effective and consistent regulator of hunting behaviour. Average commodity survival was 88% at maximum security treatments but was only 65% at maximum devaluation treatments. My experimental games and surveys do not support the expectation that reducing the value of rhinos’ horns will discourage hunters and protect rhino. The IWT is a complex socio-economic system, and human behaviour is varied in response to risk and reward. Devaluation introduces a complex interaction between the two, rendering risk and devaluation less effective in some circumstances. Every population and situation is unique, and the effectiveness of devaluation will be context-specific. Finally, the research demonstrates that serious games can be applied to the study of criminal hunting behaviour. Other difficult-to-study human subjects and systems could benefit from greater use of similar methodologies.</p>