Ground-level ozone (O3) is harmful to people and plants (Ainsworth et al., 2012; Fleming et al., 2018). In plants, O3 causes internal oxidative damage following uptake through their stomata, which then slows photosynthesis (Reich and Amundson, 1985; Morgan et al., 2003; Ainsworth et al., 2012), impairs stomatal control (Hoshika et al., 2015), suppresses the land-carbon sink, and indirectly forces climate change (Sitch et al., 2007; Lombardozzi et al., 2012). O3 exposure can also increase plant metabolic costs (Iriti and Faoro, 2009), affect reproduction (Black et al., 2000; Iriti and Faoro, 2009), alter nutrient cycling and biodiversity (Fuhrer et al., 2016), heighten the effects of other environmental stressors (Sandermann et al., 1998; Black et al., 2000; Iriti and Faoro, 2009), and diminish crop yield and quality (Ainsworth, 2017). Although plant species and varieties vary in their sensitivity to O3 (Feng et al., 2018; Harmens et al., 2018; Mills et al., 2018c), nearly all are injured to some degree and O3 is the most damaging air pollutant for most plants (Krupa et al., 2001; Wittig et al., 2009; Ainsworth et al., 2012; Lombardozzi et al., 2012). At present-day levels of O3, injuries are documented in crops, grasses, shrubs, and trees across Europe, North America, and Asia (Chappelka and Samuelson, 1998; Krupa et al., 2001; Baumgarten et al., 2009; Sarkar and Agrawal, 2010; Mills et al., 2011a; Ainsworth et al., 2012; Tang et al., 2013; Feng et al., 2014; Büker et al., 2015; Hoshika et al., 2015). These injuries reduce crop yields and lead to economic losses. It is estimated that O3 has reduced global soybean, wheat, rice, and maize yields about by 5–15%, valued at US $10–25 billion annually (Reich and Amundson, 1985; Van Dingenen et al., 2009; Fishman et al., 2010; Avnery et al., 2011; Tai et al., 2014; Mills et al., 2018c). The magnitude of these impacts and their relevance to food security and carbon storage in the biosphere show the importance of quantifying and understanding trends in O3 and its impacts on vegetation, which is our goal in this work.
Several metrics are used to quantify surface O3 and its impacts. For human health, the maximum daily average over 8 hours (MDA8), a concentration metric, is widely used to predict respiratory injury (EPA, 2011; McDonnell et al., 2012; Turner et al., 2016; Fleming et al., 2018). For assessing plant impacts, some metrics quantify O3 concentration in ambient air while others quantify the flux of O3 into leaf tissue through the stomata. The most widely used concentration metrics are the accumulated O3 exposure over a threshold of 40 ppb (AOT40) and sigmoidal-weighted cumulative exposure (W126) indices, both of which give greater weight to high concentrations (Lefohn and Runeckles, 1987; Hůnová et al., 2003; Avnery et al., 2011; Lefohn et al., 2018; Mills et al., 2018a). Although correlations between these concentration metrics and plant injuries have been reported, the flux of O3 through stomata is a better predictor of plant damage because it reflects the physiological dose to tissues within the leaves (Musselman et al., 2006; Mills et al., 2011a, b; Braun et al., 2014; Büker et al., 2015; CLRTAP, 2017). The phytotoxic O3 dose (POD) metric integrates the stomatal flux over a growing season or other designated time period. The related PODy metric integrates flux that exceeds a threshold (y nmol O3 m–2 s–1) that can be detoxified by the plant (PODy, Mills et al., 2011a, b; CLRTAP, 2017; Mills et al., 2018a). When stomata are closed, high ambient O3 concentrations may not injure plants. Conversely, when stomata are open wide, large fluxes and resulting injuries can occur at low O3 concentrations. For these reasons, flux-based metrics are generally preferred, where they are available, and critical PODy levels have been determined for many plant species (Mills et al., 2011b; CLRTAP, 2017). Indeed, several studies have found that impacts predicted from modeled stomatal uptake differ in size and pattern from impacts predicted from concentration metrics (Mills et al., 2011a; 2018c; Tang et al., 2013). Despite the advantages of POD and related flux metrics over concentration metrics, however, many plant impact studies continue to use concentration metrics because O3 concentration data are much more widely available than stomatal O3 flux data (Fuhrer et al., 1997; Musselman et al., 2006; Van Dingenen et al., 2009; Avnery et al., 2011; Mills et al., 2011a; Braun et al., 2014; Holmes, 2014; Lefohn et al., 2018; Mills et al., 2018a).
Across large parts of the United States and Europe, surface O3 air quality has improved in recent decades, according to many concentration metrics (Cooper et al., 2014; Chang et al., 2017; Lefohn et al., 2017; Fleming et al., 2018; Lefohn et al., 2018; Mills et al., 2018a), after deteriorating for much of the 20th century (Vingarzan, 2004; Shindell et al., 2006; Parrish et al., 2012; Cooper et al., 2014). These recent O3 improvements resulted from policies and technology that reduced emissions of O3 precursors, particularly nitrogen oxides, carbon monoxide, and volatile organic compounds (EPA, 2003; Council of the European Union and European Parliament, 2008; EPA, 2011; EEA, 2016; EPA, 2016). The Tropospheric Ozone Assessment Report (TOAR) concluded that these O3 declines reduced the potential risk of damage to crops and other vegetation in these regions, while recognizing that climate, soil, and plant controls on stomatal conductance also determine the risk of damage (Mills et al., 2018a, hereafter TOAR-Vegetation). While the TOAR-Vegetation report used concentration metrics—principally AOT40 and W126—because long-term O3 flux data are very sparse, the report recommended that stomatal uptake metrics be used in future risk assessments (Lefohn et al., 2018; Mills et al., 2018a).
Although some studies use empirical stomatal models to calculate O3 flux metrics and predict reductions in crop yield (Emberson et al., 2000; Mills et al., 2011a; Büker et al., 2012; Grünhage et al., 2012; Tang et al., 2013; Emberson et al., 2013; Mills et al., 2018b, 2018c), there has been little analysis of decadal or longer trends in POD or whether those trends match the trends in concentration metrics (Colette et al., 2018). POD and concentration metrics can be well correlated, at least on short time scales, under conditions where stomatal variability is limited, such as containers with a single plant species or irrigated and fertilized fields (Cieslik, 2004; Karlsson et al., 2004; Gonzalez-Fernandez et al., 2010; Matyssek et al., 2010; González-Fernández et al., 2014). Under less controlled, natural conditions, however, weather, hydrology, and climate can drive substantial changes in conductance on time scales from minutes to years (Emberson et al., 2000; Büker et al., 2012; Keenan et al., 2013; Clifton et al., 2017). This environmental variability may disrupt the relationship between POD and O3 concentration. The widespread and well-documented reductions in AOT40 and W126 in the United States and Europe may, therefore, misrepresent the benefits for plant health because of the influence of stomata on POD. As a result, there is a need to test whether O3 flux into vegetation (POD) covaries with concentration metrics (AOT40 and W126) on multi-year time scales.
This paper quantifies temporal variability and trends of O3 uptake into vegetation and compares them to variability and trends in O3 concentration over a decade. While past work has documented that POD has low spatial correlation with AOT40 and W126 (Mills et al., 2011a; Ducker et al., 2018), we specifically test temporal trends and relationships. We use a new dataset of O3 fluxes in the United States and Europe (Ducker et al., 2018), which covers the period 2005–2014 when previous studies documented declines in O3 concentration metrics (Chang et al., 2017; Lefohn et al., 2017; Lefohn et al., 2018; Mills et al., 2018a). We will show that POD trends differ significantly from the concentration metrics—specifically AOT40, W126, and mean O3—and we examine the implications of those divergent trends for vegetation health.
We analyze trends in stomatal O3 uptake and O3 concentration in the SynFlux dataset. As described by Ducker et al. (2018), SynFlux calculates stomatal conductance and other components of O3 deposition velocity from measurements at eddy covariance flux towers (Pastorello et al., 2017), with some additional information from remote sensing. The method uses observed water vapor flux, heat flux, momentum flux, leaf area, and standard meteorology variables. Direct measurements of O3 flux are not needed for SynFlux. The stomatal conductance and deposition velocity are then combined with a gridded dataset of O3 mole fractions to estimate stomatal fluxes of O3 into vegetation around the flux tower. The O3 dataset, described by Schnell et al. (2014), is a weighted interpolation of about 4000 of air quality monitoring stations and has horizontal resolution of 1° and temporal resolution of 1 hour. At sites with O3 flux measurements, the gridded O3 dataset reproduces 60–90% of observed daily O3 variability (R2 = 0.6 – 0.9) with mean bias of 5–10 ppb (Ducker et al., 2018). At a broader range of sites, Schnell et al. (2014) estimated gridded O3 errors to be 6–9 ppb (rms). Since O3 errors at a particular site and time affect all concentration and flux metrics simultaneously, the metric vs. metric comparisons shown here are insulated from inaccuracies in the O3 dataset. SynFlux reproduces approximately 90% of the day-to-day variability (R2 = 0.9) in stomatal O3 uptake at flux measurement sites with a mean bias of 20% or less that can mostly be explained by the O3 concentration bias (Ducker et al., 2018).
We examine trends in O3 concentration and POD in the summer growing season over the ten-year period 2005–2014. This decade has the greatest number of sites in the SynFlux dataset and longer periods would significantly reduce the number of sites in the analysis. All SynFlux sites with at least eight years of observations in the period are used, which results in 32 qualifying sites: 10 in the United States and 22 in Europe. These sites are listed in Table S1. The sites sample ecosystem types that are widely distributed in the United States and Europe. Of the 32 sites, 21 are forests (10 needleleaf, 7 broadleaf, and 4 mixed), 5 are grassland, 2 are crops, 3 are savanna or shrubland, and 1 is wetland.
At each SynFlux site, we calculate POD0, POD3, AOT40, W126, and daytime mean O3 for each summer growing season, defined as June-September. PODy is the cumulative daytime stomatal flux above a threshold y nmol m–2 s–1 during these months, which is contained in the SynFlux dataset described by Ducker et al. (2018). Recommended thresholds vary by species, so we use POD3, a predictor for damage in several vegetation types (Mills et al., 2011a; Büker et al., 2015; CLRTAP, 2017), to test whether the threshold affects our results. We integrate the stomatal flux over times when the sun is at least four degrees above the horizon. When some stomatal flux data are missing, the integral of available fluxes in each month is scaled up by the fraction of missing data; the scaled monthly integrals are summed to obtain PODy for the summer growing season. Months with fewer than 100 observations are discarded from the analysis because of the greater uncertainty in the monthly integral. When one of the four growing season months is missing, we scale up the remaining months in the same way to get total PODy for the growing season; if two or more months are missing, the PODy is treated as missing for that year. The AOT40 and W126 metrics are calculated from the gridded O3 dataset following previously documented methods (Ducker et al., 2018; Lefohn et al., 2018). Unlike some studies, we do not apply a three-year running mean, so our W126 and AOT40 values describe O3 concentrations in the summer growing season of a single year. Gaps in the O3 concentration data, although rare, are treated similarly to PODy, by scaling up the AOT40 or W126 by the fraction of missing data. All metrics are calculated over the same growing period, June–September, at all 32 sites. Accumulating the O3 metrics over site-specific growing months could alter the metric values or their trends, but is less likely to affect whether PODy and concentration metrics have consistent temporal variability, which is the main focus of this work. In cases where PODy is missing for a particular site and year, the AOT40, W126, and mean O3 are also discarded for consistent analysis of trends and variability. For some analyses, we remove mean spatial differences between sites by computing anomalies, , where x represents a metric value at a particular site and is its 10-year mean at that site. To compare fractional changes among metrics with different units, we also normalize values using . Figure S1 shows time series of all normalized metrics.
We estimate the linear trends in O3 metrics at each site using ordinary least squares regression: x = a + bt, where x represents a metric value, t is time, and a and b are fitted parameters. To test if two metrics have the same trend, we normalize the metrics, as described above, and add interaction effects to the regression model: x′ = a + bt + αC + βCt, where C is a categorical variable for metric type (C = 0 for metric one, C = 1 for metric two) and α and β are fitted parameters expressing the differences in the intercept and slope, respectively, for the two metrics. If β is significantly different from zero, then the metrics have different trends. We will use this approach to compare pairwise the trends in POD0, POD3, AOT40, and W126, thus highlighting the relationship between the flux and concentration metrics. In addition to standard p-values of regression coefficients, we use Fisher’s combined probability test, a meta-analysis method, to assess whether an ensemble of p-values collectively provide evidence of an effect (Fisher, 1934). We also assess the temporal co-variability of metrics in three additional ways. First, we pool the anomaly data from all sites and years (n = 299) and compute the coefficient of determination (R2) for each pair of metrics. Any correlation among the anomalies is strictly from temporal co-variability since mean spatial differences have been removed. Second, we calculate the coefficients of determination site-by-site for each pair of metrics (m = 8–10 years) and then average the resulting R2 values across sites (n = 32). Finally, we correlate the temporal trends, described above, for each pair of metrics (n = 32 sites). Analyses are performed in Python using the statsmodels module for statistical tests (Seabold and Perktold, 2010). We use the graphical format of Cooper et al. (2014) and Mills et al. (2018a) to visualize trend results in Figures 1 and 2.
Figures 1 and 2 show that summer daytime mean O3 concentrations decreased at a large majority of sites over 2005–2014. That is 14 of the 22 European SynFlux sites and 7 of the 10 sites in the United States, although only 3 sites had trends with the customary p < 0.05 level. Past studies have found that the highest quantiles of O3 distribution have fallen faster than the median and lower quantiles (Cooper et al., 2014; Lefohn et al., 2018). As a result, W126 and AOT40, which emphasize high concentrations, have stronger declining trends than daytime mean O3. For W126, 28 of the 32 sites have negative trends and 7 of these have p < 0.05. For AOT40, 25 of the 32 sites have decreasing trends and 3 have p < 0.05. No sites had positive trends with p < 0.05 for either AOT40 or W126.
Our analysis methods differ from some past trend studies in that we use gridded O3 data rather than original station measurements. Nevertheless, our concentration trend patterns are very similar to those reported by others (Cooper et al., 2014; Lefohn et al., 2017; Mills et al., 2018a), despite each of those studies using different averaging methods (daytime or 24 hour) and examining different ranges of years. They generally show that summer mean O3 in Europe has decreased or held steady since the 1990s, while concentrations in the United States decreased in the eastern United States and were steady or rising across most of the western and central United States. Like us, those earlier studies also found more consistent downward trends in AOT40 and W126 than mean O3 in all of these regions except the western and central United States (Fleming et al., 2018; Mills et al., 2018a). The TOAR-Vegetation analysis suggests that statistically significant declines of W126 and AOT40 are more widespread than we report (45–50% of TOAR-Vegetation sites in the United States and Europe had declines; Mills et al., 2018a), but this apparent difference is explained by the larger fraction of TOAR-Vegetation sites in areas recovering from severe historical O3 pollution, like the eastern United States and California, and the longer period of the TOAR-Vegetation analysis (1995–2014). Overall, the comparison shows that our trend results for concentration metrics using SynFlux and gridded O3 fields are consistent with TOAR-Vegetation results using station O3 observations.
Trends in POD0 are distinctly different from trends in mean O3, AOT40, and W126, as seen in Figures 1, 2, and S1. Unlike all of the concentration metrics, POD0 increased at more than half (18 of 32) of the sites, although 4 had POD0 declines with p < 0.05. The sign or direction of the POD0 (and POD3) trends also disagree with the concentration trends about as often as they agree. Of the 28 sites with decreasing W126, 16 have increasing POD0. Of the 25 sites with decreasing AOT40, 14 have increasing POD0. Similar patterns appear in the multi-site mean trends. For AOT40, W126, and mean O3 the multi-site mean trends are downward (p = 0.001, p ≤ 0.0001, and p = 0.06, respectively) while the mean POD0 trend is upward (p = 0.9).
The discrepancies between POD0 and concentration trends occur in nearly all vegetation types examined. The POD0 trends have opposite sign to W126 and AOT trends at roughly half of the forest sites (5 of 10 needleleaf, 4 of 7 broadleaf, 3 of 4 mixed forest), grassland sites (2 of 5 sites), and shrubland sites (1 of 2). At the one wetland site all metrics have the same trend sign, which is consistent with stomatal flux correlating with O3 concentration in a moisture-rich environment that promotes stomatal opening (CLRTAP, 2017). Conversely, at both crop sites, POD0 and concentration trends have opposite sign. While the crop sites used irrigation, which relieves water stress, they also rotated crops in some years, which would increase the interannual variability in stomatal conductance and POD0 while having less effect on O3 concentrations, thereby partly decoupling O3 concentration and uptake. Trend disagreements also occur across most of the examined geographical region and climate types, with no clear pattern. While the small number of sites for some vegetation types (particularly non-forests), regions, and climates make it difficult to determine if O3 concentration and flux preferentially decouple in certain environments, it is clear that discrepancies between PODy and concentration trends are widespread.
Some differences in trends should be expected because of statistical fitting errors in each slope estimate stemming from errors and uncertainty in the measurements that underlie each metric. Nevertheless, if stomatal conductance and deposition velocity were steady, the normalized O3 concentration and normalized POD should have similar trends. Using a regression model of normalized data with interaction effects (Section 2), we find that the fractional trends in POD are indeed different (p < 0.05 level) from the fractional trends in other metrics at 7 sites, regardless of which concentration metric is chosen. Many more sites have trend differences with marginal or low significance (0.05 < p < 0.2) (9 for AOT40 and mean O3, 11 for W126). The statistical strength of these results may be underestimated due to the random errors in SynFlux (section 2), which inflate the POD0 variance and diminish the significance (p value) of differences from concentration metrics. Mean biases do not affect the POD0 trends, so they should not affect the relationship to concentration metrics. In an aggregate meta-analysis (Fisher’s combined probability test), the skew towards zero of all 32 p-values gives strong evidence that the normalized POD0 trends systematically differ from the concentration trends (p ≪ 0.001), for all normalized concentration metrics. The divergent trends for POD0 and concentration metrics are likely explained through stomatal conductance, which is driven by many factors such as weather, hydrology, and climate. Rising stomatal conductance increases O3 uptake into plants, while decreasing ambient O3 concentration through dry deposition (Solberg et al., 2008; Emberson et al., 2013; Kavassalis and Murphy, 2017). Regardless of the causes, however, the divergent trends indicate that the common AOT40 and W126 metrics have limited utility for tracking changes in O3 impacts on vegetation.
While the O3 trends in our 10-year study period are consistent with past studies that investigated other periods, as shown above, the trends for any given site and metric can vary depending on which years are analyzed. Individual years with extreme or missing data can sometimes discernably affect trend estimates, which may contribute to a few apparently large differences between some neighboring sites (e.g. POD0 trends in Italy). Calculating trends over a longer time period could reduce the influence of individual years, but at the expense of having fewer sites in this analysis. However, the occurrence of an extreme value in one metric but not another (e.g. W126 vs. PODy) at a single site is still a meaningful indicator that those metrics have different temporal variability. In addition, the regression model that tests differences in trends between metrics accounts for the greater uncertainty in the individual trends that result from extreme values, yet we still find statistically significant different trends across metrics.
Another measure of the usefulness of concentration metrics is their temporal correlation with POD, shown Figures 3 and 4 and Table S2. Mills et al. (2011a) recognized that the spatial pattern of POD can be quite different from the O3 concentration pattern, particularly when considering sites with contrasting climate and vegetation. Ducker et al. (2018) further showed that their spatial correlations are very low (R2 ≤ 0.05 for POD0 vs. AOT40, W126, or mean O3). We quantify the temporal correlation between PODy and other metrics in three ways (Table S2). The first approach, seen in Figure 3, which pools data from all sites, reveals no meaningful correlation between POD0 and any of the concentration metrics (R2 < 0.01) while all pairs of concentration metrics are strongly correlated (R2 = 0.7 – 0.9). Results are unchanged when using POD3 in place of POD0 (R2 < 0.01, Figure S2). The second, site-by-site approach also shows that the concentration metrics are closely correlated with each other (R2 = 0.72 – 0.84 averaged across sites; Table S2) but not with POD0 (R2 = 0.13 – 0.14). Both of these approaches mean that none of the concentration metrics can predict well the interannual variability of POD0. Finally, the linear trends shown in Figure 4, which are fitted to data at each individual site, indicate that POD0 trends are essentially unrelated to trends in any concentration metric (R2 ≤ 0.05; Figure 4; Table S2), while all of the concentration metric trends have considerable common variability (R2 = 0.5 – 0.9). The trend correlations are again similarly weak when using POD3 (R2 ≤ 0.1; Figure S3). This means that neither AOT40 trend nor W126 trend has any skill in predicting the PODy trend.
While it is already widely recognized that variations in stomatal conductance complicate the relationship between O3 concentrations and stomatal uptake (Musselman et al., 2006), these results go a step further. Our results show that the conductance changes under common environmental conditions are sufficiently large and important that W126 and AOT40 trends are poor predictors of PODy trends. AOT40 and W126 might still be useful for assessing ozone extremes for other applications, however. Thus, the widespread decreases of W126 and AOT40 in large parts of the United States and Europe, while favorable, are not robust indicators for improved plant uptake or health. In fact, we have shown that there has been no widespread improvement in PODy at sites in these regions.
By many metrics, O3 air quality has improved in large parts of the United States and Europe over the last two decades in response to policies and technological improvements that reduced emissions of O3 precursors. Past work and our results show that there are downward trends in mean O3, AOT40, and W126 metrics at a majority of sites that we studied in the eastern United States and Europe. These metrics are widely used to assess the impacts of O3 and their declines have been interpreted as indicating reduced O3 damage to vegetation. While POD is known to be a better predictor of the physiological O3 dose than ambient O3 concentration, its use has been limited by data availability.
We use the SynFlux dataset to report decadal trends in POD for the first time and find that POD does not follow the same trends as the O3 concentration metrics commonly used to assess vegetation injury. POD trends have mixed increases and decreases across the United States and Europe, in contrast to the predominant decrease in concentration metrics at the sites we examined. Many sites have simultaneous decreasing AOT40 and W126 while POD is increasing. Using multiple statistical approaches, we show that the multi-year trends and temporal variability in POD differ significantly and systematically from the concentration metrics. The results are not affected by PODy threshold choices (y = 0 or 3 nmol m–2 s–1). Past work showed that concentration metrics have low spatial correlation with POD and, here, we add that there is also little temporal correspondence. Thus, AOT40 and W126 are not robust predictors of trends in plant injuries from O3. Rather, the widespread decreases of AOT40 and W126 in the United States and Europe in recent decades give an overly optimistic view of changing plant injury risk in recent years. If all else were equal, reduced concentrations would lead to less plant injury, but, in reality, stomatal conductance and its variability—driven by meteorology, hydrology, and climate—is an equally important control on POD. The analysis here further supports the recommendations of TOAR-Vegetation and others that future studies of plant damage and economic losses should avoid relying primarily on AOT40 or W126 and make greater effort to account for stomatal activity and stomatal flux. This is particularly important when considering the combined effects of climate variability and change in combination with evolving surface O3 concentrations.
The supplemental files for this article can be found as follows:
This work was supported by the Winchester Fund at Florida State University, NASA New Investigator Program grant NNX16AI57G, and NSF CAREER grant 1848372. JLS was funded by the Ubben Program for Climate and Carbon Science at Northwestern University.
The authors have no competing interests to declare.
Ainsworth, EA. 2017. Understanding and improving global crop response to ozone pollution. Plant J 90: 886–897. DOI: 10.1111/tpj.13298
Ainsworth, EA, Yendrek, CR, Sitch, S, Collins, WJ and Emberson, LD. 2012. The effects of tropospheric ozone on net primary productivity and Implications for climate change. Annu Rev Plant Biol 63(1): 637–661. DOI: 10.1146/annurev-arplant-042110-103829
Avnery, S, Mauzerall, DL, Liu, J and Horowitz, LW. 2011. Global crop yield reductions due to surface ozone exposure: 1. Year 2000 crop production losses and economic damage. Atmos Environ 45: 2284–2296. DOI: 10.1016/j.atmosenv.2010.11.045
Baumgarten, M, Huber, C, Büker, P, Emberson, L, Dietrich, HP, Nunn, AJ, Heerdt, C, Beudert, B and Matyssek, R. 2009. Are Bavarian Forests (southern Germany) at risk from ground-level ozone? Assessment using exposure and flux based ozone indices. Environ Pollut 157: 2091–2107. DOI: 10.1016/j.envpol.2009.02.012
Black, VJ, Black, CR, Roberts, JA and Stewart, CA. 2000. Impact of ozone on the reproductive development of plants. New Phytol 147(3): 421–447. DOI: 10.1046/j.1469-8137.2000.00721.x
Braun, S, Schindler, C and Rihm, B. 2014. Growth losses in Swiss forests caused by ozone: Epidemiological data analysis of stem increment of Fagus sylvatica L. and Picea abies Karst. Environ Pollut 192: 129–138. DOI: 10.1016/j.envpol.2014.05.016
Büker, P, Feng, Z, Uddling, J, Briolat, A, Alonso, R, Braun, S, Elvira, S, Gerosa, G, Karlsson, PE, Le Thiec, D, Marzuoli, R, Mills, G, Oksanen, E, Wieser, G, Wilkinson, M and Emberson, L. 2015. New flux based dose-response relationships for ozone for European forest tree species. Environ Pollut 206: 163–174. DOI: 10.1016/j.envpol.2015.06.033
Büker, P, Morrissey, T, Briolat, A, Falk, R, Simpson, D, Tuovinen, JP, Alonso, R, Barth, S, Baumgarten, M, Grulke, N, Karlsson, P, King, J, Lagergren, F, Matyssek, R, Nunn, A, Ogaya, R, Pẽuelas, J, Rhea, L, Schaub, M, Uddling, J, Werner, W and Emberson, L. 2012. DO3SE modelling of soil moisture to determine ozone flux to forest trees. Atmos Chem Phys 12: 5537–5562. DOI: 10.5194/acp-12-5537-2012
Chang, K-L, Petropavlovskikh, I, Copper, OR, Schultz, MG and Wang, T. 2017. Regional trend analysis of surface ozone observations from monitoring networks in eastern North America, Europe and East Asia. Elem Sci Anth 5: 50. DOI: 10.1525/elementa.243
Chappelka, AH and Samuelson, LJ. 1998. Ambient ozone effects on forest trees of the eastern United States. New Phytol 139: 91–108. DOI: 10.1046/j.1469-8137.1998.00166.x
Cieslik, SA. 2004. Ozone uptake by various surface types: A comparison between dose and exposure. Atmos Environ 38: 2409–2420. DOI: 10.1016/j.atmosenv.2003.10.063
Clifton, OE, Fiore, AM, Munger, JW, Malyshev, S, Horowitz, LW, Shevliakova, E, Paulot, F, Murray, LT and Griffin, KL. 2017. Interranual variability in ozone removal by a temperate deciduous forest. Geophsyical Res Lett 542–552. DOI: 10.1002/2016GL070923
CLRTAP. 2017. Mapping critical levels for vegetation, revised Chapter 3 of the Manual on Methodologies and Criteria for Modelling and Mapping Critical Loads and Levels and Air Pollution Effects, Risks and Trends. United National Economic Commission for Europe (UNECE). Convention on Long-range Transboundary Air Pollution (CLRTAP).
Colette, A, Tognet, F, Létinois, L, Lemaire, V, Couvidat, F, Alonso Del Amo, RM, Gonzalez Fernandez, IA, Rábago Juan-Aracil, I, Harmens, H, Andersson, C, Manders, A, Mircea, M and Tsyro, S. 2018. ETC/ACM Report 2018:15: Long-Term Evolution of the Impacts of Ozone Air Pollution on Agricultural Yields in Europe . Bilthoven: European Topic Centre on Air Pollution and Climate Change Mitigation.
Cooper, OR, Parrish, DD, Ziemke, J, Balashov, NV, Cupeiro, M, Galbally, I, Gilge, S, Horowitz, L, Jensen, NR, Lamarque, J-F, Naik, V, Oltmans, S, Schwab, J, Shindell, D, Thompson, A, Thouret, V, Wang, Y and Zbinden, R. 2014. Global distribution and trends of tropospheric ozone: An observation-based review. Elem Sci Anthr 2: 1–28. DOI: 10.12952/journal.elementa.000029
Council of the European Union, European Parliament. 2008. Directive 2008/50/EC of the European Parliament and of the Council of 21 May 2008 on ambient air quality and cleaner air for Europe. Off J Eur Union L152: 1–44. Available at https://eur-lex.europa.eu/eli/dir/2008/50/oj.
Ducker, JA, Holmes, CD, Keenan, TF, Fares, S, Goldstein, AH, Mammarella, I, William Munger, J and Schnell, J. 2018. Synthetic ozone deposition and stomatal uptake at flux tower sites. Biogeosciences 15(17): 5395–5413. DOI: 10.5194/bg-15-5395-2018
EEA. 2016. Air quality in Europe — 2016 report. European Environment Agency . DOI: 10.5194/bg-15-5395-2018
Emberson, LD, Ashmore, MR, Cambridge, HM, Simpson, D and Tuovinen, JP. 2000. Modelling stomatal ozone flux across Europe. Environ Pollut 109: 403–413. DOI: 10.1016/S0269-7491(00)00043-9
Emberson, LD, Kiwiroon, N, Beevers, S, Büker, P and Cinderby, S. 2013. Scorched Earth: how will changes in the strength of the vegetation sink to ozone deposition affect human health and ecosystems? Atmos Chem Phys 13: 6741–6755. DOI: 10.5194/acp-13-6741-2013
EPA. 2003. Ozone Transport Commission NOx Budget Program 1999–2002 Progress Report. United States Environmental Protection Agency .
EPA. 2011. EPA’s NOx Reduction Program and Clean Air Interstate Rule: 2009 Environmental and Health Results. United States Environmental Protection Agency . Available at https://www.epa.gov/sites/production/files/2015-08/documents/cair09_environmental_results.pdf.
EPA. 2016. Air Pollutant Emissions Trends Data. State Average Annual Emissions Trend. United States Environmental Protection Agency . Available at https://www.epa.gov/air-emissions-inventories/air-pollutant-emissions-trends-data.
Feng, Z, Büker, P, Pleijel, H, Emberson, L, Karlsson, PE and Uddling, J. 2018. A unifying explanation for variation in ozone sensitivity among woody plants. Glob Chang Biol 24(1): 78–84. DOI: 10.1111/gcb.13824
Feng, Z, Sun, J, Wan, W, Hu, E and Calatayud, V. 2014. Evidence of widespread ozone-induced visible injury on plants in Beijing, China. Environ Pollut 193: 296–301. DOI: 10.1016/j.envpol.2014.06.004
Fisher, RA. 1934. The combination of probabilities from tests of significance. In: Statistical Methods for Research Workers , 5th ed. Edinburgh: Oliver and Boyd.
Fishman, J, Creilson, JK, Parker, PA, Ainsworth, EA, Vining, GG, Szarka, J, Booker, FL and Xu, X. 2010. An investigation of widespread ozone damage to the soybean crop in the upper Midwest determined from ground-based and satellite measurements. Atmos Environ 44: 2248–2256. DOI: 10.1016/j.atmosenv.2010.01.015
Fleming, ZL, Doherty, RM, Von Schneidemesser, E, Malley, CS, Cooper, OR, Pinto, JP, Colette, A, Xu, X, Simpson, D, Schultz, MG, Lefohn, A, Hamad, S, Moolla, R, Solberg, S and Feng, Z. 2018. Tropospheric Ozone Assessment Report: Present-day ozone distribution and trends relevant to human health. Elem Sci Anth 6(1): 12. DOI: 10.1525/elementa.273
Fuhrer, J, Skärby, L and Ashmore, MR. 1997. Critical levels for ozone effects on vegetation in Europe. Environ Pollut 97(1–2): 91–106. DOI: 10.1016/S0269-7491(97)00067-5
Fuhrer, J, Val Martin, M, Mills, G, Heald, CL, Harmens, H, Hayes, F, Sharps, K, Bender, J and Ashmore, MR. 2016. Current and future ozone risks to global terrestrial biodiversity and ecosystem processes. Ecol Evol 6(24): 8785–8799. DOI: 10.1002/ece3.2568
González-Fernández, I, Calvo, E, Gerosa, G, Bermejo, V, Marzuoli, R, Calatayud, V and Alonso, R. 2014. Setting ozone critical levels for protecting horticultural Mediterranean crops: Case study of tomato. Environ Pollut 185: 178–187. DOI: 10.1016/j.envpol.2013.10.033
Gonzalez-Fernandez, I, Kaminska, A, Dodmani, M, Goumenaki, E, Quarrie, S and Barnes, JD. 2010. Establishing ozone flux-response relationships for winter wheat: Analysis of uncertainties based on data for UK and Polish genotypes. Atmos Environ 44: 621–630. DOI: 10.1016/j.atmosenv.2009.11.021
Grünhage, L, Pleijel, H, Mills, G, Bender, J, Danielsson, H, Lehmann, Y, Castell, JF and Bethenod, O. 2012. Updated stomatal flux and flux-effect models for wheat for quantifying effects of ozone on grain yield, grain mass and protein yield. Environ Pollut 165: 147–157. DOI: 10.1016/j.envpol.2012.02.026
Harmens, H, Hayes, F, Mills, G, Sharps, K, Osborne, S and Pleijel, H. 2018. Wheat yield responses to stomatal uptake of ozone: Peak vs rising background ozone conditions. Atmos Environ 173(January): 1–5. DOI: 10.1016/j.atmosenv.2017.10.059
Holmes, CD. 2014. Air pollution and forest water use. Nature 507(7491): E1–E2. DOI: 10.1038/nature13113
Hoshika, Y, Katata, G, Deushi, M, Watanabe, M, Koike, T and Paoletti, E. 2015. Ozone-induced stomatal sluggishness changes carbon and water balance of temperate deciduous forests. Sci Rep 5(May): 1–8. DOI: 10.1038/srep09871
Hůnová, I, Livorová, H and Ostatnická, J. 2003. Potential ambient ozone impact on ecosystems in the Czech Republic as indicated by exposure index AOT40. Ecol Indic 3: 35–47. DOI: 10.1016/S1470-160X(03)00009-8
Iriti, M and Faoro, F. 2009. Chemical diversity and defence metabolism: How plants cope with pathogens and ozone pollution. Int J Mol Sci 10(8): 3371–3399. DOI: 10.3390/ijms10083371
Karlsson, PE, Uddling, J, Braun, S, Broadmeadow, M, Elvira, S, Gimeno, B, Le Thiec, D, Oksanen, E, Vandermeiren, K, Wilkinson, M and Emberson, L. 2004. New critical levels for ozone effects on young trees based on AOT40 and simulated cumulative leaf uptake of ozone. Atmos Environ 38: 2283–2294. DOI: 10.1016/j.atmosenv.2004.01.027
Kavassalis, SC and Murphy, JG. 2017. Understanding ozone-meteorology correlations: A role for dry deposition. Geophys Res Lett 44(6): 2922–2931. DOI: 10.1175/JCLI-D-12-00168.1
Keenan, TF, Hollinger, DY, Bohrer, G, Dragoni, D, Munger, JW, Schmid, HP and Richardson, AD. 2013. Increase in forest water-use efficiency as atmospheric carbon dioxide concentrations rise. Nature 499: 324–327. DOI: 10.1038/nature12291
Krupa, S, Mcgrath, MT, Andersen, CP, Booker, FL, Burkey, KO, Chappelka, AH, Chevone, BI, Pell, EJ and Zilinskas, BA. 2001. Ambient ozone and plant health. Plant Dis 85(1): 4–12. DOI: 10.1094/PDIS.2001.85.1.4
Lefohn, AS, Malley, CS, Simon, H, Wells, B, Xu, X, Zhang, L and Wang, T. 2017. Responses of human health and vegetation exposure metrics to changes in ozone concentration distributions in the European Union, United States, and China. Atmos Environ 152: 123–145. DOI: 10.1016/j.atmosenv.2016.12.025
Lefohn, AS, Malley, CS, Smith, L, Wells, B, Hazucha, M, Simon, H, Naik, V, Mills, G, Schultz, MG, Paoletti, E, Marco, A, Xu, X, Zhang, L, Wang, T, Neufeld, H, Musselman, R, Tarasick, D, Brauer, M, Feng, Z, Tang, H, Kobayashi, K, Sicard, P, Solberg, S and Gerosa, G. 2018. Tropospheric ozone assessment report: Global ozone metrics for climate change, human health, and crop/ecosystem research. Elem Sci Anthr 6: 28. DOI: 10.1525/elementa.279
Lefohn, AS and Runeckles, VC. 1987. Establishing standards to protect vegetation-ozone exposure/dose considerations. Atmos Environ 21: 561–568. DOI: 10.1016/0004-6981(87)90038-2
Lombardozzi, D, Sparks, JP, Bonan, G and Levis, S. 2012. Ozone exposure causes a decoupling of conductance and photosynthesis: implications for the Ball-Berry stomatal conductance model. Oecologia 169: 651–659. DOI: 10.1007/s00442-011-2242-3
Matyssek, R, Karnosky, DF, Wieser, G, Percy, K, Oksanen, E, Grams, TEE, Kubiske, M, Hanke, D and Pretzsch, H. 2010. Advances in understanding ozone impact on forest trees: messages from novel phytotron and free-air fumigation studies. Environ Pollut 158(6): 1990–2006. DOI: 10.1016/j.envpol.2009.11.033
McDonnell, WF, Stewart, PW, Smith, MV, Kim, CS and Schelegle, ES. 2012. Prediction of lung function response for populations exposed to a wide range of ozone conditions. Inhal Toxicol 24(10): 619–633. DOI: 10.3109/08958378.2012.705919
Mills, G, Hayes, F, Simpson, D, Emberson, L, Norris, D, Harmens, H and Büker, P. 2011a. Evidence of widespread effects of ozone on crops and (semi-)natural vegetation in Europe (1990–2006) in relation to AOT40- and flux-based risk maps. Glob Chang Biol 17: 592–613. DOI: 10.1111/j.1365-2486.2010.02217.x
Mills, G, Pleijel, H, Braun, S, Büker, P, Bermejo, V, Danielsson, H, Emberson, L, Grünhage, L, González-Fernández, I, Harmens, H, Hayes, F, Karlsson, PE and Simpson, D. 2011b. New stomatal flux-based critical levels for ozone effects on vegetation. Atmos Environ 45: 5064–5068. DOI: 10.1016/j.atmosenv.2011.06.009
Mills, G, Pleijel, H, Malley, CS, Sinha, B, Cooper, OR, Schultz, MG, Neufeld, H, Simpson, D, Sharps, K, Feng, Z, Gerosa, G, Harmens, H, Kobayashi, K, Saxena, P, Paoletti, E, Sinha, V and Xu, X. 2018a. Tropospheric Ozone Assessment Report: present-day ozone distribution and trends relevant to vegetation. Elem Sci Anthr 6(47): 46. DOI: 10.1525/elementa.302
Mills, G, Sharps, K, Simpson, D, Pleijel, H, Broberg, M, Uddling, J, Jaramillo, F, Davies, WJ, Dentener, F, Van den Berg, M, Agrawal, M, Arawal, S, Ainsworth, E, Büker, P, Emberson, L, Feng, Z, Harmens, H, Hayes, F, Kobayashi, K, Paoletti, E and Van Dingenen, R. 2018b. Ozone pollution will compromise efforts to increase global wheat production. Glob Change Biol 24: 3560–3574. DOI: 10.1111/gcb.14157
Mills, G, Sharps, K, Simpson, D, Pleijel, H, Frei, M, Burkey, K, Emberson, L, Uddling, J, Broberg, M, Feng, Z, Kobayashi, K and Agrawal, M. 2018c. Closing the global ozone yield gap: Quantification and cobenefits for multistress tolerance. Glob Chang Biol 24(10): 4869–4893. DOI: 10.1111/gcb.14381
Morgan, PB, Ainsworth, EA and Long SP. 2003. How does elevated ozone impact soybean? A meta-analysis of photosynthesis, growth and yield. Plant, Cell Environ 26(8): 1317–1328. DOI: 10.1046/j.0016-8025.2003.01056.x
Musselman, RC, Lefohn, AS, Massman, WJ and Heath RL. 2006. A critical review and analysis of the use of exposure- and flux-based ozone indices for predicting vegetation effects. Atmos Environ 40: 1869–1888. DOI: 10.1016/j.atmosenv.2005.10.064
Parrish, DD, Law, KS, Staehelin, J, Derwent, R, Cooper, OR, Tanimoto, H and Gilge, S. 2012. Long-term changes in lower tropospheric baseline ozone concentrations at northern mid-latitudes. Atmos Chem Phys 11485–11504. DOI: 10.5194/acp-12-11485-2012
Pastorello, GZ, Papale, D, Chu, H, Trotta, C, Agarwal, DA, Canfora, E and Baldocchi, DD. 2017. A new data set to keep a sharper eye on land-air exchanges. Eos 98: 28–32. DOI: 10.1029/2017EO071597
Reich, PB and Amundson, RG. 1985. Ambient levels of ozone reduce net photosynthesis in tree and crop species. Science 230(4725): 566–570. DOI: 10.1126/science.230.4725.566
Sandermann, H, Ernst, D, Heller, W and Langebartels C. 1998. Ozone: An abiotic elicitor of plant defence reactions. Trends Plant Sci 3(2): 47–50. DOI: 10.1016/S1360-1385(97)01162-X
Sarkar, A and Agrawal, SB. 2010. Identification of ozone stress in Indian rice through foliar injury and differential protein profile. Environ Monit Assess 161: 205–215. DOI: 10.1007/s10661-008-0738-z
Schnell, JL, Holmes, CD, Jangam, A and Prather, MJ. 2014. Skill in forecasting extreme ozone pollution episodes with a global atmospheric chemistry model. Atmos Chem Phys 14: 7721–7739. DOI: 10.5194/acp-14-7721-2014
Seabold, S and Perktold, J. 2010. Statsmodels: econometric and statistical modeling with Python. Proceedings of the 9th Python in Science Conference (SciPy 2010): 57–61. Available at http://conference.scipy.org/proceedings/scipy2010/pdfs/seabold.pdf.
Shindell, D, Faluvegi, G, Lacis, A, Hansen, J, Ruedy, R and Aguilar, E. 2006. Role of tropospheric ozone increases in 20th-century climate change. J Geophys Res 111: 1–11. DOI: 10.1029/2005JD006348
Sitch, S, Cox, PM, Collins, WJ and Huntingford, C. 2007. Indirect radiative forcing of climate change through ozone effects on the land-carbon sink. Nature 448: 791–794. DOI: 10.1038/nature06059
Solberg, S, Hov, O, Sovde, A, Isaksen, ISA, Coddeville, P, De Backer, H, Forster, C, Orsolini, Y and Uhse, K. 2008. European surface ozone in the extreme summer 2003. J Geophys Res 113(D7): D07307. DOI: 10.1029/2007JD009098
Tai, APK, Martin, MV and Heald, CL. 2014. Threat to future global food security from climate change and ozone air pollution. Nature Climate Change 4(9): 817–821. DOI: 10.1038/nclimate2317
Tang, H, Takigawa, M, Liu, G, Zhu, J and Kobayashi, K. 2013. A projection of ozone-induced wheat production loss in China and India for the years 2000 and 2020 with exposure-based and flux-based approaches. Glob Chang Biol 19: 2739–2752. DOI: 10.1111/gcb.12252
Turner, MC, Jerrett, M, Pope, CA, Krewski, D, Gapstur, SM, Diver, WR, Beckerman, BS, Marshall, JD, Su, J, Crouse, DL and Burnett, R. 2016. Long-term ozone exposure and mortality in a large prospective study. Am J Respir Crit Care Med 193(10): 1134–1142. DOI: 10.1164/rccm.201508-1633OC
Van Dingenen, R, Dentener, FJ, Raes, F, Krol, MC, Emberson, L and Cofala, J. 2009. The global impact of ozone on agricultural crop yields under current and future air quality legislation. Atmos Environ 43: 604–618. DOI: 10.1016/j.atmosenv.2008.10.033
Vingarzan, R. 2004. A review of surface ozone background levels and trends. Atmos Environ 38: 3431–3442. DOI: 10.1016/j.atmosenv.2004.03.030
Wittig, VE, Ainsworth, EA, Naidu, SL, Karnosky, DF and Long, SP. 2009. Quantifying the impact of current and future tropospheric ozone on tree biomass, growth, physiology and biochemistry: a quantitative meta-analysis. Glob Chang Biol 15: 396–424. DOI: 10.1111/j.1365-2486.2008.01774.x