E16. Spatially explicit population modelling

Population dynamics have been modelled in many areas of ecology using a variety of methods, including both analytical and process-based techniques. Consideration of population dynamics began with analytical and experimental approaches, such as Lotka-Volterra predator prey models (1926), Leslie's stage-structures (1945); and Birch's estimation of the intrinsic rate of population growth (1948). More recent approaches include individual-based, cohort-based and metapopulation models. However, both application and theory are fraught with nuances: a simple discrete logistic growth curve can produce both multiple stable states and complex chaotic behaviour, and the Lotka-Volterra a hysteresis cycle, depending on a change in value of a single parameter (as shown by May; 1972, 1975). Geographical space adds another level of complexity across which a population dynamic will express itself, with landscape heterogeneity promoting either stability or instability in population dynamics, depending on the system. Studying spatially explicit populations thus provides further computational and inferential difficulties for the modeller.

This session aims to broach the broad spectrum of approaches to spatially explicit population modelling, from the latest theory through to computation, inference and calibration. Real world applications of spatial-temporal models to population surveillance and management are presented from a wide range of ecological studies. Throughout the session we will look to examine points of synthesis between studies and explore the relative merits of a diverse range of approaches.

A proposed structure for this session is one of four mini-sessions of 3-4 speakers each, with the topics:

  1. Theory and Models
  2. Computation and Complexity
  3. Inference and Calibration
  4. Applications

Each of the first three mini-sessions will be chaired by an expert, who will pool together at the end of the last session in a panel discussion of spatially explicit population models.