The scale of the community is the most complex organizational level through which to understand and predict forest dynamics. Communities hold the diversity that determines functional response to climate, the spatial dynamics through which competition and density dependence play out. Modelling communities requires balancing complexity and mechanism. There exist, however, a great deal of demographic data on tree communities through permanent sample plot networks. We have developed a modeling platform that uses growth and survival data to inform broad scales of inference in forest ecology using the community level as a base. To do this we consider statistical models of demography through evolutionary theory and biological information. Specifically, we build population models that reflect changes in vital rates over ontogeny, the balance between resource acquisition and allocation, and correlations between demographic rates across life-history. To do this we fit vital rate (growth, survival and reproduction) models to data, extract clusters in PCA space from model parameters, and identify key demographic ‘modes’ in the forest. The modes are broadly defined by life form (e.g., shrub, understory tree, or canopy tree), sensitivity to early mortality, responsiveness to resources (implicit in growth rates), and size at senescence.
Results/Conclusions
The demographic modes we identify align with some classic life history strategies, such as position along the ‘slow-fast’ continuum. We also find potential sink species, where resource demands lead to both low survival and low growth. Important to carbon fluxes, we find that several canopy-class demographic modes show distinct senescence strategies. These strategies are not phylogenetically conserved, implying that demographic modes might arise within lineages that diverge within communities. This is an important approach to better understanding how species might be better aggregated into modeling platforms that require large-scale resolution of forest function, such as Dynamic Global Vegetation Models. We propose investigating demographic modes as a way to define plant functional types.