Landscape Fire Succession Modelling
The model I developed during my PhD research utilises the Rule-Based Community-Level Modelling (RBCLM) system developed by McIntosh and others for vegetation modelling with qualitative knowledge in circumstances where quantitative data for model parameterisation is sparse. The two key attributes of vegetation change addressed by the RBCLM approach are:

1. direction of transition between discrete land cover (vegetation) classes
2. rate of transition between these discrete land cover classes.

By considering vegetation change at a broad vegetation type level in these terms allows qualitative understanding of vegetation dynamics to be translated into a formal, spatial model at the landscape scale. The dominant environmental constraints on these broad land cover classes are water availability and light availability. The vegetation state-and-transition model procedure is represented by a flow chart.

Screenshot from the Landscape Fire Succession Model

Seven broad land cover classes are considered including pine, ‘transition forest’, deciduous, oak and shrubland land-covers.Each land cover has distinct life history traits and reproductive strategies. The pine and oak classes are considered as directly analogous to ‘seeder’ and ‘resprouter’ plant vegetation types (respectively). Seed dispersal is also explictly represented.

A grid-based fire dynamics model, considering environmental variables such as topography and climate, is integrated with the vegetation dynamics model. Subsequent to disturbance the model represents succession between land-cover types as following one of two potential regeneration pathways.

A paper published with colleagues in Environmental Modelling and Software (Millington et al. 2009) describes the model and sensitivity analyses. In the future this model will allow us to examine many aspects of landscape change including impacts of potential climate change, potential impacts of the introduction of exotic plant species, and the importance of soil moisture gradients for vegetation dynamics.