A rule-based model for the functional analysis of vegetation change in Australasian grasslands

Grasslands encompass a broad array of vegetation and climatic zones.We describe the first developments towards a rule-based functional model for predicting vegetation structure in Australian and New Zealand pastures and rangelands. The approach aims to predict the combined effects of climate and disturbance by humans and grazing livestock, and to provide a level of resolution needed for predicting changes in pastures and rangelands. We enlisted expert knowledge to develop: (l)a minimum set of critical traits; (2) rules relating site variables to favoured-plant attributes; (3) rules relating attributes to plant functional traits, and (4) rules relating plant functional types to likely plant communities. We tested the resulting model by deriving some simple predictions of plant communities of some existing pasture and rangeland sites in Australia and New Zealand, with differing climatic and human disturbance inputs. The results indicate that this first model is able to predict plant communities with varying success rates, and with the best results in cases where there are extreme climates or high management inputs. Key sensitivities in the model where further research is required include: (1) the urgent need for more explicit understanding of the key plant functional attributes favoured bydiffering climates and disturbance regimes, (2) the functional relationships between these plant functional attributes and recognisable plant functional types in vegetation, and (3) the assembly rules for the coexistence of these different plant functional types in major plant communities. The same understanding is required for subsequent process-based modelling development. [ABSTRACT FROM AUTHOR]