The importance of individual variation for dispersal and population persistence in a checkerspot butterfly
Individual variation in movement behavior can have important consequences for spatial population dynamics. Individual variation increases leptokurtosis in dispersal distance, for instance, with more individuals dispersing very short and very long distances relative to a homogeneous population. We investigated the importance of including individual variation in predictions of two key conservation-related measures: dispersal distance and the critical minimum patch size, which is the smallest area within which a population can persist based on loss due to emigration. We quantified individual variation from flight path data for the Baltimore checkerspot butterfly using mixed effects models, and incorporated empirically-estimated movement parameters into two versions of an individual-based model (IBM). Both IBMs include demographic stochasticity, but in the first, all individuals move according to the same set of movement parameters, while in the second, each individual has its own unique set of movement parameters. We evaluated growth rates, the critical minimum patch size, dispersal distances, and rates of diffusion in models with and without individual variation, and compared these to estimates from an analytical diffusion-based model that does not include stochasticity or individual variation.
All movement parameters showed substantial among-individual variation, with the greatest variation in move lengths and time spent resting. Across individuals, move length and the probability of resting were negatively correlated, and the probability of resting and time spent resting were positively correlated, indicating that some butterflies were generally more mobile than others. As expected, individual variation increased the mean and maximum dispersal distances, and leptokurtosis of the distributions. The difference in maximum dispersal distance was substantial (~600 m without versus ~2800 m with individual variation), and is likely of significance for conservation. Individual variation also affected the critical minimum patch size, but the difference (~0.5 Ha without versus ~0.4 Ha with individual variation) is unlikely to be ecologically significant. While growth rates with individual variation were higher in smaller patch sizes, it is notable that growth rates were also lower in larger patch sizes, a pattern that is a logical consequence of increased leptokurtosis. Individual variation can alter the outcome of metapopulation models and predictions of minimum area requirements, and we encourage its incorporation where data are sufficient for doing so.