Books

Responses of Northern U.S. Forests to Environmental Change
ISBN 0-387-98900-5

Chapter 12: Regional Impacts of Ozone on Forest Productivity

John A. Laurence, Scott V. Ollinger , and Peter B. Woodbury

Mechanistic models simulate changes in forest structure or function by quantifying fundamental mechanisms or processes, whereas statistical models rely on analysis of empirical data. Mechanistic models can therefore be used to extrapolate forest responses beyond conditions that have already occurred, if the correct response mechanisms are understood and modeled correctly.

PnET-II, an ecosystem-scale model used to estimate regional forest production, total ecosystem carbon balances and water yield, and responses to climate change, was modified to include ozone effects on productivity including interactions with other stressors. The approach was to modify the model’s photosynthesis algorithms to include leaf-level uptake-response relationships and allow interaction with factors such as light attenuation, canopy ozone gradients, and water stress. Results for New England and New York showed decreases in predicted annual net primary production (NPP) from 2 to 17 percent as a result of mean ozone levels from 1987-1992, with greatest reductions occurring where both ozone levels and stomatal conductance were greatest. Growth declines were greatest on sites with wetter soils.

Another approach to estimating ozone effects was to couple a model of forest stand development and composition (ZELIG) with a model of tree-level physiological response to stress (TREGRO). TREGRO models the acquisition of carbon, water, and nutrients, and allocates carbon among competing plant parts depending upon resource availability and phenology. ZELIG is a gap-succession model used to simulate succession in mixed stands typical of eastern and northern forests, and has both mechanistic and empirical characteristics. The two models were coupled by passing calculations from TREGRO to ZELIG in order to modify the growth relationships in the stand model. Simulated response to ambient ozone from 1991 using the TREGRO/ZELIG combination for red oak and sugar maple in the Northeast showed a small reduction in red oak basal area growth (2-4 percent over a 100-year simulation) and a compensating small growth increase in sugar maple.

Model predictions include uncertainty due to (1) incomplete regional data, (2) incomplete knowledge of how forests respond to particular stresses, and (3) uncertainty about which processes and parameters should be included in the model. Some of this uncertainty is related to spatial interpolation, aggregation, and scaling errors.

Below: Reduction in net photosynthesis in relation to cumulative ozone uptake (from Reich, 1987) using data from independent ozone fumigation studies conducted on a variety of hardwood seedlings.

Previous: Chapter 11 | Next: Chapter 13