Paleoecological records have shown shifts in ecosystem composition and structure in conjunction with periods of warming and cooling in the past. However modern observations of physiological processes that cause ecosystem responses to climate changes are mismatched in spatial and temporal scale from long-term paleoecological data. As part of the PalEON Project, we have bridged these two scales and approaches to investigate drivers and mechanisms behind ecosystem change and stability using millennial-scale ecosystem model simulations of forests in the Midwestern and Northeastern United States. We compare stability of climate, modeled ecosystem characteristics, and paleoecological observations in the pre-settlement era (prior to 1850) as well as the modern era that is characterized by anthropogenic-driven climate change (1901-2010). In this multi-model study, we analyzed the responses of a suite of ecosystem characteristics including carbon fluxes, biomass pools, and forest composition to both climate change and interannual climate variability.
Results/Conclusions
Prior to 1850, high interannual variability in meteorology and carbon fluxes such as gross primary productivity (GPP) caused few periods of significant change. In contrast, slow responses to extreme events and internal ecosystem dynamics caused carbon pools and ecosystem characteristics such as forest composition to be highly unstable and at a nearly constant disequilibrium state. Rates and duration of temperature increase after 1900 exceeded those found in our meteorology drivers prior to 1850 at most sites, and this corresponds to unprecedented and consistent increases in GPP across models, space, and time. However, because biomass and composition were perpetually undergoing change in the pre-settlement era, modern shifts in biomass or composition did not necessarily exceed those seen prior to 1850. Although models have been highly criticized for their tendency towards equilibrium and widespread use of equilibrium assumptions at the beginning of model simulations, interannual climate variability can produce a prevalence of disequilibrium states in most models. Lack of ecosystem stability for states like composition or biomass pools prior to European settlement may impact modeled and observed response to modern climate change.