Developing adaptive and flexible individual and agent-based ecological models via pattern-guided evolution

Submission note: A thesis submitted in total fulfilment of the requirements for the degree of Doctor of Philosophy to the School of Engineering and Mathematical Sciences, Faculty of Science, Technology and Engineering, La Trobe University, Bundoora.

This thesis explores a novel approach, denoted the pattern-guided evolution (PGE) method, for developing individual and agent-based ecological models that have the capacity to simulate intergenerational adaptation. Intergenerational adaptation, or evolution, can be simulated by implementing model entities or agents with inheritable parametric diversity and modelling entity reproduction. Implementing individual or agent-based models with parametric diversity may also increase their flexibility for model sharing and component reuse. However, it is also likely to increase model complexity. Existing approaches for developing and calibrating individual and agent-based models, such as the pattern-oriented modelling (POM) approach, have already been found to be computationally expensive in many cases. Typically, the POM approach involves searching for parameters that result in models that match real-world data or patterns. The PGE method provides an alternative approach for models have inheritable parametric diversity, which involves guiding the evolution of model entities or agents towards congruence with expected patterns. The results of the simulations conducted for this research show that the method was effective for developing and calibrating a typical and reasonably complex ecological individual-based model having inheritable parametric diversity, built for demonstration purposes using the Repast Simphony agent-based modelling and simulation toolkit. The research findings also suggest that using the PGE method is generally less computationally expensive than the POM approach to model development and calibration. Overall, the method may thus facilitate the future development of increasingly complex models capable of simulating intergenerational adaptation in changing environments. Such models may have increasing utility given the challenges ecologists and environmental scientists and engineers face in a changing world.