Importance of changing CO2, temperature, precipitation, and ozone on carbon and water cycles of an upland-oak forest: incorporating experimental results into model simulations

Abstract

Observed responses of upland-oak vegetation of the eastern deciduous hardwood forest to changing CO₂, temperature, precipitation and tropospheric ozone (O₃) were derived from field studies and interpreted with a stand-level model for an 11-year range of environmental variation upon which scenarios of future environmental change were imposed. Scenarios for the year 2100 included elevated [CO₂] and [O₃] (+385 ppm and +20 ppb, respectively), warming (+4°C), and increased winter precipitation (+20% November–March). Simulations were run with and without adjustments for experimentally observed physiological and biomass adjustments.

Initial simplistic model runs for single-factor changes in CO₂ and temperature predicted substantial increases (+191% or 508 g C m⁻² yr⁻¹) or decreases (−206% or −549 g C m⁻² yr⁻¹), respectively, in mean annual net ecosystem carbon exchange (NEE_a≈266±23 g C m⁻² yr⁻¹ from 1993 to 2003). Conversely, single-factor changes in precipitation or O₃ had comparatively small effects on NEE_a (0% and −35%, respectively). The combined influence of all four environmental changes yielded a 29% reduction in mean annual NEE_a. These results suggested that future CO₂-induced enhancements of gross photosynthesis would be largely offset by temperature-induced increases in respiration, exacerbation of water deficits, and O₃-induced reductions in photosynthesis. However, when experimentally observed physiological adjustments were included in the simulations (e.g. acclimation of leaf respiration to warming), the combined influence of the year 2100 scenario resulted in a 20% increase in NEE_a not a decrease. Consistent with the annual model's predictions, simulations with a forest succession model run for gradually changing conditions from 2000 to 2100 indicated an 11% increase in stand wood biomass in the future compared with current conditions.

These model-based analyses identify critical areas of uncertainty for multivariate predictions of future ecosystem response, and underscore the importance of long term field experiments for the evaluation of acclimation and growth under complex environmental scenarios.