| dc.description.abstract | Recent studies have advanced our knowledge of factors that could affect problem solving performance, and also of the positive effects of problem solving ability on fitness measures (the ‘what’ of problem solving). However, a missing linkage exists between this ‘what’ and the corresponding ‘how’. Such linkage requires the understanding of how these factors contribute to problem solving. Therefore, the central aim of this thesis is to examine this ‘how’. The roles of learning and behavioural flexibility in the context of problem solving are shown across the experiments, primarily with laboratory and free-ranging grey squirrels and to a lesser extent with wild red squirrels.
Under a recurring change, laboratory grey squirrels showed a rapid decrease in the number of errors they made per reversal phase in a serial spatial reversal learning task. Such efficiency is achieved by a gradual tactic change, from sequential to integrative tactics, with increased experience. It also involves support from cognitive mechanisms such as attention and inhibitory control. In a puzzle box task, wild grey squirrels showed that they were better problem solvers than the wild red squirrels. However, red squirrels that solved the puzzle box were more efficient than the grey solvers. Detailed analysis of the results showed that learning and flexibility play independent roles in problem solving. Each process is associated with particular traits that to increase efficiency. For grey squirrels, behavioural selectivity (effective behaviours) and persistence increased with increased experience. Flexibility, however, showed minimal positive effect for them, given that it decreased behavioural selectivity. In contrast, flexibility primarily provided a positive effect for red squirrels’ solving efficiency. These results showed that the two species appear to use both similar and different cognitive processes in solving the task.
The discussion gathers the results and explores how learning and flexibility, along with other behavioural traits, vary in their contributions to problem solving performance. As learning and flexibility are definitely not limited in problem solving, the discussion also addresses how these two processes might be involved a construct of general intelligence (‘g’) in animals, and how they are relevant to wilder ecological aspects. | en_GB |
