Use of tobacco cultivars as biological indicators of ambient ozone pollution: an analysis of exposure-response relationships.

A summary is presented of the numerical analysis of the results from a study of exposure in an open-top chamber conducted at two locations in the north-east USA, utilizing tobacco (Nicotiana tabacum L.) cv. Bel-W3 (sensitive) and cv. Bel-B (tolerant) as differential indicators of ambient O(3) pollution. At both study sites, Bel-W3 was significantly more sensitive than Bel-B. There were differences in the weekly O(3) exposure dynamics and the consequent foliar injury scores on Bel-W3 between the two study sites. During the individual weekly exposures, the bottommost fully expanded leaf (leaf no. 1) on Bel-W3 was more sensitive than the second fully expanded leaf (leaf no. 2). There were no statistically significant differences in the injury scores on Bel-W3 leaf no. 1 between the non-filtered air open-top chamber and the chamberless, ambient field plot treatments at both study sites. In Mallow's critical-point best regression, among the many O(3) descriptors tested, number of hours during each week with O(3) concentrations > 40 ppb (N40) and > 60 ppb (N60) or the corresponding sums of their concentrations (SUM40 and SUM60) proved to be the best predictors of foliar injury on Bel-W3, leaf no. 1. The regression used N40 and N60 or SUM40 and SUM60 together and did not identify each variable alone or by itself as being important. Independent of this, all the R(2) values could account for only about 30-32% of the variability of the foliar injury responses, although these values were statistically significant. Covariance time-series analysis between weekly O(3) concentrations > 40 ppb and the corresponding foliar injury scores on Bel-W3 leaf no. 1 showed that foliar injury was in best spectral coherence with the O(3) exposure (> 40 ppb) during periods of moderate periodicity (variance). The overall results suggest that the visible foliar responses of tobacco Bel-W3 can be used as a qualitative, but not necessarily as a quantitative indicator of relative ambient O(3) pollution on a generalized temporal or spatial scale.