2020 ESA Annual Meeting (August 3 - 6)

OOS 44 Abstract - The potential for vegetation demographic models to simulate compositional turnover across environmental gradients and with climate change

Lara M. Kueppers1,2, Polly Buotte3, Thomas L. Powell4, Rosie A. Fisher5, Charles D. Koven4, Jacquelyn Shuman6 and Chonggang Xu7, (1)Climate and Ecosystem Sciences Division, Lawrence Berkeley National Laboratory, Berkeley, CA, (2)Energy and Resources Group, University of California, Berkeley, Berkeley, CA, (3)University of California Berkeley, Berkeley, CA, (4)Earth and Environmental Sciences Area, Lawrence Berkeley National Laboratory, Berkeley, CA, (5)Centre Européen de Recherche et de Formation Avancée en Calcul Scientifique, Toulouse, France, (6)National Center for Atmospheric Research, CO, (7)Earth and Environmental Sciences Division, Los Alamos National Laboratory, Los Alamos, NM
Background/Question/Methods

Projecting the future distribution and composition of forest ecosystems is critical to projecting changes in water, carbon and energy fluxes, as well as habitat for other species. Past efforts to predict future tree species and forest distributions using bioclimatic envelope approaches have either ignored or struggled to capture ecologically realistic transient dynamics due to dispersal and demography. At the same time, the composition of any given forest could be robustly predicted via transient demography and competitive interactions among species but lacked spatially extensive application. Neither approach has fully realized the interactions among CO2 enrichment, warming, and droughts as interactive factors driving change. Finally, the vast diversity of species and functional traits that influence tree and forest responses challenges our ability to work across large regions regardless of model framework. Vegetation demographic models (VDMs) link physiological responses to climate and CO2 with demographic rates and competitive interactions across large spatial domains, and collapse diversity into discrete ecological strategies defined by functional traits. Using the VDMs ED2 and FATES, we ask, How does specification of ecological strategies influence functional composition of forests and its sensitivity to environmental variation and change over time?

Results/Conclusions

Sensitivity simulations in mixed conifer forests in California and in tropical forests in Panama reveal that coexisting ecological strategies depend on a small but important set of functional traits that influence establishment, growth and mortality. The persistence and abundance of traits and strategies are themselves regulated by environmental variability and disturbances, such as fire. In California mixed conifer forest, this critical subset of traits influences growth and mortality of new recruits, as well as growth of large trees, which together determine coexistence of the dominant and co-dominant species types (ponderosa pine and incense cedar). In Panamanian tropical forest, traits that differentiate “safe” from “efficient” hydraulic strategies can be roughly orthogonal to traits that differentiate shade “intolerant” and “tolerant” strategies, with coexistence regulated by temporal variation in precipitation that permits storage of rarer strategies in mature tree cohorts. Experiments for tropical forest suggest that if diverse strategies are maintained in the forest, modest shifts in hydroclimate may not lead to loss of important ecological strategies. Together, early results with VDMs suggest that simulated forest composition is highly sensitive to specification of ecological strategies as these interact with environmental and disturbance factors, suggesting some promise in capturing transient change – or buffering of change – in forest composition driven by climate change.