B041-01
Simulating the Effects of Functional Types of Trees in Individual-Based Simulations on the “Fundamental Response” of Forested Mosaics

Wednesday, 9 December 2020: 16:00
Virtual
Herman Henry Shugart, University of Virginia, Department of Environmental Sciences, Charlottesville, VA, United States, Sassan Saatchi, NASA Jet Propulsion Laboratory, Pasadena, CA, United States, Bin Wang, University of California Irvine, Department of Ecology and Evolutionary Biology, Irvine, CA, United States, Manuel Lerdau, University of Virginia Main Campus, Charlottesville, VA, United States and Huan Zhang, Beijing Forestry University, School of Forestry, Beijing, China
Abstract:
Following a disturbance, The dynamic response in biomass/carbon storage of a forested landscape comprised of tree species with relatively similar growth and death rates resembles the dynamic response of a underdamped second-order spring. This pattern arises from the synchronization of the many ~0.1 ha plots, which sum to the landscapes net carbon storage. After a disturbance, the biomass of the elements of the landscape are all in a state of mass accumulation. Eventually, the semi-synchronized mortality of the large canopy-dominant trees balances growth and the mass accumulation goes to zero and then drops when mortality exceeds growth. The local recovery from these first-wave tree deaths loses synchronization and the landscape settles to a lower quasi-equilibrium on a mosaic landscape. This “Fundamental Response” changes when the synchrony among species is low, as in highly diverse forests. It can be amplified when there forests with a few super-dominant species. In this presentation we apply a relatively diverse individual-based model (IBM) to determine how lumping species into functional groupings (e.g., lumping species as shade tolerant versus shade intolerant; lumping species by genus; lumping all the species into a single functional type). These strategies produce contrasting model performance. Since the reference gap model has 32 species for which the parameters are specified, it is noteworthy that the method of parameter lumping to form functional type also has an effect on model performance. Averaging the parameters into functional groups produces different results than computing weighted averages of species parameters based on species biomass at different ages of forests.