New exposure-based metric approach for evaluating O(3) risk to North American aspen forests.

The United States and Canada currently use exposure-based metrics to protect vegetation from O(3). Using 5 years (1999-2003) of co-measured O(3), meteorology and growth response, we have developed exposure-based regression models that predict Populus tremuloides growth change within the North American ambient air quality context. The models comprised growing season fourth-highest daily maximum 8-h average O(3) concentration, growing degree days, and wind speed. They had high statistical significance, high goodness of fit, include 95% confidence intervals for tree growth change, and are simple to use. Averaged across a wide range of clonal sensitivity, historical 2001-2003 growth change over most of the 26 Mha P. tremuloides distribution was estimated to have ranged from no impact (0%) to strong negative impacts (-31%). With four aspen clones responding negatively (one responded positively) to O(3), the growing season fourth-highest daily maximum 8-h average O(3) concentration performed much better than growing season SUM06, AOT40 or maximum 1h average O(3) concentration metrics as a single indicator of aspen stem cross-sectional area growth.