The common brushtail possum (Trichosurus vulpecula) has become a highly invasive pest in New Zealand since its first introduction in the mid-1800s. A range of control methods are used to manage this species, but most fail to catch or kill the last few individuals in the target population. Addressing why these last individuals evade control is vital to successful long-term pest management. One potential cause of avoidance is social learning; that is, the social transmission of information from one possum to the next, whereby an individual may learn to avoid control devices without the need for risky direct interaction.

Social learning has never been studied before in this species, and only rarely in metatherian/marsupial mammals. The focus of this thesis is, therefore, to develop our understanding of the social learning capabilities of possums and consider the implications for future management strategies. I first focus on whether possums can socially learn, using demonstrator and observer possums. I found that demonstrator possums could learn to open domes on a puzzle device progressively faster over consecutive nights of exposure, showing that possums could learn to open the puzzles in the appropriate order. I then tested if this information could be socially transmitted. Possums that had observed demonstrator possums solve the puzzle completed the puzzle more quickly and targeted the correct domes more often than the possums that had to learn from their own interaction with the toy. This provides the first documented evidence of social learning in possums.

I then build on this finding by exploring the real-world problem of possums evading aversive stimuli such as baiting stations and traps. I first tested whether possums learn to avoid an aversive stimulus. However, avoidance was not learnt by the demonstrators as their attraction to the food reward overcame the aversive stimuli. Consequently, social transmission to the observers could not be examined, but they did confirm the patterns of social learning seen in the first experiment. The limitations of these findings and implications for possum pest management studies are discussed.

Finally, I explore broader possum learning patterns using collated data from the previous two experiments. Here, retention of learning and differences in learning between the sexes are explored. By using a small subset of possums observed in both experiments, retention of learning was evident after a 3-month break between trials. Differences in learning between the sexes were also found, with females being faster than males to approach the puzzle. However, no differences between the sexes were found regarding the time between approaching the puzzle and deliberately opening a dome. These findings may have important implications for conservation management regarding possums surviving control operations and the spread of social learning among possums, so the need for further research is noted.

Social learning may prove to be an underestimated aspect of the perseverance of this pest species in its non-native habitat. With the understanding gained from this research, more efficient possum eradication strategies could be implemented, particularly as the learning capabilities of marsupials continues to be explored.