One way of trying to understand how plants grow is to make models of plant growth. We can model the structure or external form of the plant, we can model the internal physiology or function of the plant, or we can model a combination of the two. We can also include a representation of the plant’s environment. In this thesis, I discuss how modelling can be used to help investigate, explain and simulate the interaction between plant form and function.
The thesis looks at the modelling of a number of different aspects of plant growth. I consider how fundamental aspects of plant functioning, such as the acquisition, storage and allocation of resources, can be simulated using a well-established intermediate-level mathematical approach known as canonical modelling. I look at how the development of the structural form of the plant can be modelled using L-systems, and integrated with these canonical functional models. I investigate how the effects of interaction between plants, between parts of plants, and between plants and their environment can be included in these models. Finally, I consider some ways in which the flow of information within the plant
can be represented using computational simulations, such as rule-based, compartment and particle models.
Computational modelling can be a valuable tool in the process of understanding and appreciating the complex patterns of plant growth. However, models are most useful when they act as a simplifying abstraction
of the complexity of reality, and so the question of the appropriate level of model abstraction forms one of the main themes of the thesis. Other important issues addressed include the different types of computational modelling approaches that can be used, the various purposes of modelling, the role that modelling plays in the process of scientific investigation, and the question of what modelling actually is. This thesis shows that these issues are all closely related, and that, for a given modelling problem, the relationship between these issues will have a large influence on which computational modelling approach is most appropriate, and on how this approach can be used.