Modelling Vegetation Demography, Disturbance, and Carbon Storage in Forests

Abstract

Earth System Models (ESM) have a diverse range of approaches when representing vegetation. These differences can lead to uncertainty in inter-model projections under climate change. Plant demography has been increasingly used by Dynamic Global Vegetation Models (DGVMs) to compromise between tractability within the ESM framework against the complexity of ecosystem dynamics. The primary goal in this thesis is to further develop a new cohort DGVM: The Robust Ecosystem Demography (RED) model. To do this, we review mathematical and modelling techniques that are used to describe the dynamics of forest demography over time. Demographic Equilibrium Theory (DET) is the notion that the size-structure of forests can be sufficiently describe by the Metabolic Scaling Theory of growth (MST) and mortality at equilibrium. This has been validated at scale using forest inventory data in both temperate and tropical forests. The combination of DET and allometry yields functions that can explain forest properties, such as total carbon or growth density, and why such quantities are skewed towards the largest individuals. Using the assumptions of DET as a basis, we further develop RED to include recruitment and competition. RED partitions plants into mass classes, where MST dependent growth rates and plant mortality are applied. RED requires only two inputs, total plant carbon assimilate and disturbance mortality. We split carbon assimilate by a fraction, , into recruitment and, 1ð€€€ , into growing the plants. Using MST we disaggregate plant growth onto each class. Seedlings are constrained by the area not occupied by competing plants, with successful recruits joining the lowest mass class. RED version 1 was driven by UKESM assimilate rates globally, and can be calibrated to find the required mortality rates to fit remotely sensed observations of plant coverage. In RED version 2 we introduce a seed pool and improve the flux scheme, which improves the realism of regrowth. We also adjust the competition by allowing plant coverages to overlap, thereby introducing more realistic diversity into the model. We simulate secondary succession from land-use abandonment and size-dependent disturbances of drought and mortality. Using the RED competition as a guide, we are able to build upon DET by including a competition and recruitment for the lower boundary condition. Closed- Form Demographic Equilibrium Theory (CF-DET) provides insight into how forests are dependent on both their rate of recruitment and mortality. CF-DET suggests forests may optimise certain variables, such as number density or plant height. We also use CF-DET to infer tree demography from top-down site level and remotely sensed observations.