Spatial and Behavioral Determinants of Butterfly Movement Patterns in Topographically Complex Landscapes

The loss of natural habitats and their fragmentation by human activities impose a great threat to biodiversity worldwide. As animal populations become restricted to small and fragmented patches of natural habitat, their viability becomes increasingly dependent upon the arrival of emigrants from elsewhere. Dispersal – the movement of animals away from their original habitat – is thus fundamental to the viability of populations and metapopulations in fragmented landscapes. Ecologists and conservationists seek to understand the processes that determine dispersal patterns, in order to better maintain connectivity between populations. However, dispersal may take complex routes depending on the specific attributes of the landscape and the response of animals to different features of it. Consequently, our ability to predict movement patterns is often limited by a lack of empirical knowledge on animal decision-making processes during dispersal, as well as by the inability to model the varied responses of animals to landscape heterogeneity. This study addressed both problems. I studied movement patterns from a mechanistic, individual-based approach, and developed generalizations on directional movement patterns in heterogeneous landscapes. I focused on topographical heterogeneity, a major source of landscape heterogeneity which receives little attention in the context of dispersal. As a behavioral case-study I examined hilltopping behavior in butterflies. Hilltopping animals seek topographic summits where they meet and mate. Thereafter, mated females descend from summits in search of host plants. Hilltopping is therefore a dispersal-like behavior which leads animals out of habitat patches and directs their movement through the landscape. I developed a systematic, hierarchical approach to investigating directed movements through topographically heterogeneous landscapes. I started at the individual level, using field observations on the hilltopping butterfly Melitaea trivia to deduce the response mechanisms to topography. I then used the results to construct an Individual-Based Model (IBM) of hilltopping, and analyzed movement patterns in virtual and realistic landscapes of increasing complexity. Finally, I validated the model with another field experiment, which assessed the ability of the model to predict movement patterns in a realistic, complex landscape.