Incorporating cognition into models of animal movement and predator – prey interaction

by Chloe Ingrid Bracis

Institution: University of Washington
Degree: PhD
Year: 2015
Keywords: foraging; memory; model; movement; predation; predator-prey; Applied mathematics
Record ID: 2060596
Full text PDF: http://hdl.handle.net/1773/27579


Incorporating cognition, i.e., learning and memory, into models of animal movement is increasingly important as models seek to answer more complex questions where individuals' prior experiences shape their choices. Two example are foraging behavior and predator avoidance. While models of predator – prey dynamics exist, the impact of cognition on movement and predator – prey interactions is largely unexplored. This dissertation presents a flexible, continuous-space, and continuous-time model incorporating an animal using memory to navigate a landscape of heterogeneous resources. The forager balances attraction to food with avoidance of predators in making movement decisions. Two streams make up the resource memory: a repulsive stream that drives the forager away from recently visited areas and an attractive stream that draws the forager back to high quality areas. The predator memory is solely repulsive. The model is used to examine questions related to the advantage of added cognitive complexity for animals in the context of foraging and balancing the food – safety trade-off with predators. First, foraging without predators is considered and several movement processes are compared: a simple correlated random walk; kinesis, a correlated random walk that switches between searching and feeding behaviors; and memory-informed movement. The model is used to examine for which landscapes the added cognitive complexity of maintaining memory is advantageous and to analyze the behavioral differences between using and not using memory. In general, a landscape where there is a larger payoff for finding a resource patch, whether in size, value, or difficulty in locating, favors memory. While memory-informed search can be difficult to differentiate from other sensory-driven search behavior, disproportionate spatial use of higher value areas, higher consumption rates, and consumption variability all point to memory influencing the movement direction. Next, predators are introduced that vary in their temporal predictability and in their correlation with the prey's resources. Memory outperforms simpler movement processes most for patchy landscapes and more predictable predators, which may be more easily avoided once learned. In these cases, memory aids foragers in managing the food – safety trade-off, as particular parameterizations of predator memory reduce predator encounters while maintaining consumption. Non-consumptive effects are highest in landscapes of concentrated, patchy resources and especially when predators are highly correlated with the forager's resources. These non-consumptive effects are also seen with the shift away from the best quality habitat compared to foraging in a predator-free environment. Finally, learning is examined in more detail with naive foragers introduced to new landscapes as well as predators introduced partway through the simulation. Most non-extreme learning rates provide the forager with sufficient information. In general, foragers that are low to moderately exploratory in new habitats are…