Approaches for modeling spatial dynamics of forest insect disturbance: The integration of GIScience, complex systems theory and swarming intelligence

Forest ecological systems are constantly being changed by natural disturbances such as insect infestations, fires and diseases among others. These events can result in the death of trees over areas of several thousand hectares. In western Canada, including the province of British Columbia, extensive outbreaks of mountain pine beetle (MPB) have been occurring during the last decade, raising concerns about the health of these forests and the ability to deal with these issues. For this reason, the development of forest insect infestation models has become an active research topic for scientists from many different disciplines, and geography is not apart from this issue. The insect disturbance phenomenon is a complex process that is inherently linked to space and time. Interactions between insects such as the MPB and the forest ecosystem display a wide variety of complex system properties. Accordingly, complex landscape patterns of tree mortality emerge from interacting MPB individuals that act at local host tree levels. Complex systems theory modeling approaches such as cellular automata (CA) and agent-based modeling (ABM), allow simulations of spatial interactions, which can describe the ecological context in which insect populations spread. The objective of this research is to develop and implement several spatio-temporal modeling approaches that are based on the integration of complex systems theory, swarming intelligence (SI) and geographic information systems (GIS). In particular, this dissertation introduces novel modeling approaches for generating forest patterns emerging from MPB disturbance. MPB behaviours observed in nature are simulated using SI algorithms that depict their indirect communication, collective behaviour and self-organized aggregation in a forest ecosystem. Thesis findings demonstrate that forest patterns of MPB disturbance can be realistically depicted and simulated when collective aggregation behaviour of MPB, forest structure and spatial dynamics within the system are considered and analyzed simultaneously. Approaches are implemented in the context of MPB disturbance in British Columbia, Canada. This dissertation presents novel contributions to the study of the dynamic changes of forest cover resulting from forest insect infestations by means of complex systems theory, swarming intelligence and GIS. The thesis main contributions are in the fields of GIScience, Landscape Ecology, Environmental Resource Management and Geography.;Keywords: agent-based modelling (ABM); complexity theory; swarming intelligence; forest infestation, geographic information system (GIS); mountain pine beetle (MPB); cellular automata (CA).