Flux-Based Ozone Risk Assessment for a Plant Injury Index (PII) in Three European Cool-Temperate Deciduous Tree Species

: This study investigated visible foliar ozone (O 3 ) injury in three deciduous tree species with di ﬀ erent growth patterns (indeterminate, Alnus glutinosa (L.) Gaertn.; intermediate, Sorbus aucuparia L.; and determinate, Vaccinium myrtillus L.) from May to August 2018. Ozone e ﬀ ects on the timing of injury onset and a plant injury index (PII) were investigated using two O 3 indices, i.e., AOT40 (accumulative O 3 exposure over 40 ppb during daylight hours) and POD Y (phytotoxic O 3 dose above a ﬂux threshold of Y nmol m − 2 s − 1 ). A new parameterization for POD Y estimation was developed for each species. Measurements were carried out in an O 3 free-air controlled exposure (FACE) experiment with three levels of O 3 treatment (ambient, AA; 1.5 × AA; and 2.0 × AA). Injury onset was found in May at 2.0 × AA in all three species and the timing of the onset was determined by the amount of stomatal O 3 uptake. It required 4.0 mmol m − 2 POD 0 and 5.5 to 9.0 ppm · h AOT40. As a result, A. glutinosa with high stomatal conductance ( g s ) showed the earliest emergence of O 3 visible injury among the three species. After the onset, O 3 visible injury expanded to the plant level as conﬁrmed by increased PII values. In A. glutinosa with indeterminate growth pattern, a new leaf formation alleviated the expansion of O 3 visible injury at the plant level. V. myrtillus showed a dramatic increase of PII from June to July due to higher sensitivity to O 3 in its ﬂowering and fruiting stage. Ozone impacts on PII were better explained by the ﬂux-based index, POD Y , as compared with the exposure-based index, AOT40. The critical levels (CLs) corresponding to PII = 5 were 8.1 mmol m − 2 POD 7 in A. glutinosa , 22 mmol m − 2 POD 0 in S. aucuparia , and 5.8 mmol m − 2 POD 1 in V. myrtillus . The results highlight that the CLs for PII are species-speciﬁc. Establishing species-speciﬁc O 3 ﬂux-e ﬀ ect relationships should be key for a quantitative O 3 risk assessment.


Introduction
Tropospheric ozone (O 3 ) is one of the major concerns for forest health due to its phytotoxicity [1]. Despite the fact that peak O 3 concentrations have tended to decrease in the eastern part of United States and some European countries due to precursor emission controls [2], the global background O 3 concentration still remains high enough to cause negative impacts on tree physiology [3]. [4]. Hoshika et al., 2012a [5] reported that the percent of surface injury was negatively correlated with leaf gas exchange rate, highlighting a reduced photosynthesis and loss of stomatal control in poplar leaves with more than 5% injury. Ozone visible injury has been broadly investigated in native and exotic trees, shrubs, and herbs in Asia, Europe, and North America, and partly validated under controlled conditions [6][7][8].

Visible foliar injury by O 3 (O 3 visible injury) is the first unequivocal visually detectable sign of O 3 damage and indicates an impairment of leaf physiological functions
An exposure-based index such as AOT40 (accumulated exposure over a threshold of 40 ppb) is used to assess O 3 risks to European forest trees [9,10]. Previous studies have reported an AOT40-based assessment of the first symptom onset of O 3 visible injury in field [11,12] or open-top chambers [13,14]. Those studies suggested that an O 3 critical level (CL) by 5 to 10 ppm·h AOT40 could protect the sensitive tree species from O 3 visible injury. However, it has been recognized that O 3 damage depends on stomatal O 3 uptake rather than only O 3 exposure [15]. To improve our quantitative assessment of O 3 effects on trees, a stomatal O 3 flux-based index such as POD Y (phytotoxic ozone dose above a flux threshold of Y nmol m −2 s −1 ) has been the focus. POD Y is estimated using the deposition of ozone and stomatal exchange model (DO3SE) [16]. Sicard et al., 2016 [17] estimated POD Y using the DO3SE model and analyzed field observation data in Southeastern France and Northwestern Italy. They proposed the stomatalflux-based standard to assess O 3 visible injury for two deciduous species (Fagus sylvatica and Fraxinus excelsior) and two conifer species (Pinus cembra and P. halepensis) as representative O 3 sensitive species (approximately 20 mmol m −2 of POD 0 corresponded to 5% injury). Many symptomatic species have been recorded in field monitoring campaigns [8,18,19]. Paoletti et al., 2019 [8] recently found O 3 visible injury in 23 tree species across forest sites in France, Italy, and Romania. Nevertheless, knowledge is still limited on species-specific model parameters to calculate stomatal O 3 uptake for establishing the O 3 flux-effect relationship in most symptomatic tree species.
The species-specific tree response to O 3 can be affected by the growth pattern (i.e., indeterminate or determinate) [20]. In elevated O 3 , the tree species with indeterminate pattern (e.g., poplar) can initiate new leaf formation to replace damaged older leaves [21]. This response can limit the development of O 3 visible injury at the plant level for those species. In this study, we selected three European cool-temperate deciduous tree species with different growth patterns, i.e., Alnus glutinosa (L.) Gaertn. (indeterminate) [22], Sorbus aucuparia L. (intermediate) [23], and Vaccinium myrtillus L. (determinate) [24]. These three species have often shown O 3 visible injury in field forest sites [8]. The aim of this study was to achieve the species-specific parameterization of the DO3SE model in these three tree species, examining both the onset date of O 3 visible injury and its expansion at the whole plant level using a plant injury index (PII) [4] in an O 3 free-air controlled exposure (O 3 FACE) experiment. The PII can be more closely related to the plant physiological status and especially whole plant carbon loss than to first symptom onset [14]. Three hypotheses were tested as follows: (i) Are species with higher stomatal O 3 uptake more sensitive to O 3 ? (ii) Does the tree growth pattern affect the expansion of O 3 visible injury at plant level? and (iii) Is the flux-based approach better than the exposure-based one to explain the PII?

Experimental Site and Plant Material
Measurements were conducted at an O 3 FACE facility located in Sesto Fiorentino, in central Italy (43 • 48' 59" N, 11 • 12' 01" E, 55 m a.s.l.). The details of the system are described in Paoletti et al., 2017 [25]. Five-year old saplings of A. glutinosa and S. aucuparia, and three-year old saplings of V. myrtillus were obtained from a nearby nursery in December 2017. Plants were transplanted into plastic pots (50 L for A. glutinosa and S. aucuparia and 25 L for V. myrtillus) containing a mixture of sand:peat:soil = 1:1:1 (v:v:v). In 2018, plants were exposed to the following three levels of O 3 concentration: Ambient air (AA), 1.5 times ambient O 3 concentration (1.5 × AA), and twice ambient O 3 concentration (2.0 × AA). Three replicated blocks (5 m × 5 m × 2 m) were set to each O 3 treatment Forests 2020, 11, 82 3 of 12 (n = 3) with three (A. glutinosa and S. aucuparia) or six (V. myrtillus) plants (total 27 plants for A. glutinosa and S. aucuparia, and 54 plants for V. myrtillus). Ozone concentrations in each treatment were recorded continuously by an O 3 monitor (Mod. 202, 2B Technologies, Boulder, CO, USA). All plants were irrigated to keep field capacity at 1-to 3-day intervals to prevent water stress. We monitored the light intensity, relative humidity, air temperature, precipitation, and wind speed above the O 3 FACE facility (2.5 m height) using a WatchDog meteorological station (Model 2000, Spectrum Technologies, Inc., Aurora, IL, USA).

Modeling of Stomatal Conductance
Leaf gas exchange was measured in fully expanded sun-exposed leaves (1 to 3 plants per replicated plot per each O 3 treatment) using a portable infrared gas analyzer (CIRAS-2 PP Systems, Herts, UK). Measurements were made on the days with clear sky in the morning (8 h to 10 h), afternoon (13 h to 15 h) and evening (16 h to 19 h) from May to October 2018. Natural illumination was used for the measurement. The CO 2 concentration in the chamber (Ca) was set to 400 ppm. The temperature and relative humidity in the chamber were adjusted manually to the ambient condition. Pooled data (210 for A. glutinosa, 217 for S. aucuparia, and 216 for V. myrtillus) were used to estimate the parameters of the DO3SE model [16], as follows: where g max is the maximum stomatal conductance, i.e., mmol O 3 m −2 projected leaf area (PLA) s −1 .
The other functions are all expressed as relative terms and are scaled from 0 to 1. The model accounts for the minimum stomatal conductance (f min ) and the variation in g s according to phenology (f phen ), photosynthetic photon flux density (PPFD) (f light ), temperature (f temp ), vapor pressure deficit (VPD) (f VPD ), and soil water content (f SWC ). The f SWC was not applied in this study (f SWC = 1) because the soil moisture was equivalent to field capacity. The g max and f min values were set as 95th and 5th percentile values recorded in the experiment. Parameterizations of other functions were carried out using a boundary line analysis [26,27]. Further details on f phen , f light , f temp, and f VPD calculations are provided in CLRTAP (2017) [16].

Calculation of Ozone Indices
AOT40 was calculated by using hourly O 3 concentrations during daylight hours (short wave radiation > 50 W m −2 ) according to CLRTAP (2017) [16]. It is given by: where [O 3 ] i is the ith measured hourly O 3 concentration (ppb) with i equal to 1 . . . n in the integral and n is the number of hours included in the calculation period. Stomatal O 3 uptake (F st , nmol m −2 s −1 ) was calculated as: where r c is the leaf surface resistance (= 1/(g s + g ext ), s m −1 ), g ext is the external leaf or cuticular conductance (s m −1 ) [16], and r b is the leaf boundary layer resistance, given as: where u is wind speed (m s −1 ) and L d is the species-specific leaf dimension (A. glutinosa 0.07 m, S. aucuparia 0.04 m, and V. myrtillus 0.04 m obtained as averaged value of 3 to 5 leaves of two plants in each block in each O 3 treatment) [16]. POD Y (mmol m −2 ) was estimated from hourly data as: where F st _ i is the ith hourly stomatal O 3 uptake (nmol m −2 s −1 ) and n is the number of hours included in the calculation period. Y is a species-specific threshold of stomatal O 3 uptake (nmol m −2 s −1 ). Exposure-or flux-based dose-response functions were determined from a linear regression between PII and AOT40 or POD Y over a threshold of Y (Y from 0 to 10, with an increment of 0.5 nmol m −2 s −1 ). Two criteria were applied to select the best dose-response function which included: (1) the confidence interval (C.I.) must include Y-intercept = 0, and (2) among the functions meeting criterion 1, the equation with the highest R 2 value was chosen. CLs were calculated as the level when PII reaches 5. In fact, a significant decline of physiological performance was found in leaves with more than 5 of PII [14].

Data Analysis
Statistical analyses were performed using SPSS (20.0, SPSS, Chicago, IL, USA). To assess the effects of O 3 on the number of attached leaves, a two-way analysis of variance (ANOVA) was applied. Data were checked for normal distribution and homogeneity of variance (Levene's test). Since the PII data were not normally distributed, the Kruskal-Wallis analysis of variance was applied to examine the effect of O 3 . The relationships between PII and O 3 indices were fitted using a simple linear regression. Results were considered significant at p < 0.05.

Ozone Visible Injury
The first O 3 visible injuries were observed in 2.0 × AA on 18 May for A. glutinosa, on 21 May for S. aucuparia, and on 26 May for V. myrtillus (Table 1). Ozone visible injury occurred as dark or reddish stippling on the upper leaf surface ( Figure 2) and was more severe in older than in younger leaves. The AOT40 corresponding to the onset of O 3 visible injury was 5.6 to 8. Ozone-induced increases of PII were also found in S. aucuparia on 20 August (1.5 × AA, +101% and 2.0 × AA, +182%) and V. myrtillus on 19 July (1.5 × AA, +65% and 2.0 × AA, +172%). Ozone stimulated the number of attached leaves in A. glutinosa on 20 August ( Figure 4); however, such an increase was not found in S. aucuparia and V. myrtillus.

Parameterization of Stomatal Conductance Model
The gmax value was 300, 240, and 140 mmol m −2 s −1 in A. glutinosa, S. aucuparia, and V. myrtillus, respectively (Table 2 and Figure 5). The fmin values were similar in all three species, i.e., 0.13 to 0.17. The response of gs to PPFD (flight) indicated that V. myrtillus had a higher a value (0.0104) relative to A. glutinosa (0.0024) and S. aucuparia (0.0043). The optimal temperature for stomatal opening was 20 to 30 °C in all species. A VPD higher than around 1 kPa induced stomatal closure regardless of the species. The fphen values peaked from June to August in all three species. Estimated gs values were in good agreement with the measured values as confirmed by the coefficient of determination (R 2 = 0.46 to 0.61) and root mean square error (RMSE = 31 to 57 mmol O3 m −2 s −1 ) ( Figure S1).

Parameterization of Stomatal Conductance Model
The g max value was 300, 240, and 140 mmol m −2 s −1 in A. glutinosa, S. aucuparia, and V. myrtillus, respectively (Table 2 and Figure 5). The f min values were similar in all three species, i.e., 0.13 to 0.17. The response of g s to PPFD (f light ) indicated that V. myrtillus had a higher a value (0.0104) relative to A. glutinosa (0.0024) and S. aucuparia (0.0043). The optimal temperature for stomatal opening was 20 to 30 • C in all species. A VPD higher than around 1 kPa induced stomatal closure regardless of the species. The f phen values peaked from June to August in all three species. Estimated g s values were in good agreement with the measured values as confirmed by the coefficient of determination (R 2 = 0.46 to 0.61) and root mean square error (RMSE = 31 to 57 mmol O 3 m −2 s −1 ) ( Figure S1). g max is the maximum stomatal conductance; f min is a fraction of minimum stomatal conductance to g max ; f phen is the variation of stomatal conductance with season; f light , f temp , and f VPD depend on photosynthetically relevant photon flux density at the leaf surface (PPFD, µmol m −2 s −1 ), temperature (T, • C), and vapor pressure deficit (VPD, kPa), respectively; A start and A end are the year days for the start and end of the experiment; f phen_a and f phen_b represent the number of days of f phen to reach its maximum and the number of days during the decline of f phen to the minimum value, respectively; f phen_c and f phen_d represent maximum fraction of f phen at A start and A end , respectively; a is the parameter determining an exponential curve of stomatal response to light; T opt , T min , and T max denote optimal, minimum, and maximum temperature for stomatal opening, respectively; and VPD min and VPD max denote the threshold of VPD for attaining minimum and full stomatal opening, respectively.

Dose-Response Relationship for Plant Injury Index
In A. glutinosa, the first criterion (Y-intercept = 0 included in C.I.) was reached in the regressions between PII and AOT40 or POD0−7. Among these indices, POD7 had the highest R 2 (0.54) (Table S1). On the one hand, the CL corresponding to PII = 5 based on POD7 was 8.1 mmol m −2 ( Figure 6) and, on the other hand, the AOT40-based CL was 33 ppm·h. In S. aucuparia, the first criterion was achieved for AOT40 and POD0−5.5. The highest R 2 value was found in POD0 (0.87). The CLs in this species were found to be 22 mmol m −2 POD0 and 30 ppm·h AOT40. In V. myrtillus, the first criterion was achieved for AOT40 and POD1−4. POD1 had the highest R 2 value (0.82), while the exposure-based index AOT40 performed equally well (R 2 = 0.81). The CLs was 5.8 mmol m −2 POD1 and 20 ppm·h AOT40 in this species.

Dose-Response Relationship for Plant Injury Index
In A. glutinosa, the first criterion (Y-intercept = 0 included in C.I.) was reached in the regressions between PII and AOT40 or POD 0−7 . Among these indices, POD 7 had the highest R 2 (0.54) (Table S1). On the one hand, the CL corresponding to PII = 5 based on POD 7 was 8.1 mmol m −2 ( Figure 6) and, on the other hand, the AOT40-based CL was 33 ppm·h. In S. aucuparia, the first criterion was achieved for AOT40 and POD 0−5.5 . The highest R 2 value was found in POD 0 (0.87). The CLs in this species were found to be 22 mmol m −2 POD 0 and 30 ppm·h AOT40. In V. myrtillus, the first criterion was achieved for AOT40 and POD 1−4 . POD 1 had the highest R 2 value (0.82), while the exposure-based index AOT40 performed equally well (R 2 = 0.81). The CLs was 5.8 mmol m −2 POD 1 and 20 ppm·h AOT40 in this species.

Dose-Response Relationship for Plant Injury Index
In A. glutinosa, the first criterion (Y-intercept = 0 included in C.I.) was reached in the regressions between PII and AOT40 or POD0−7. Among these indices, POD7 had the highest R 2 (0.54) (Table S1). On the one hand, the CL corresponding to PII = 5 based on POD7 was 8.1 mmol m −2 ( Figure 6) and, on the other hand, the AOT40-based CL was 33 ppm·h. In S. aucuparia, the first criterion was achieved for AOT40 and POD0−5.5. The highest R 2 value was found in POD0 (0.87). The CLs in this species were found to be 22 mmol m −2 POD0 and 30 ppm·h AOT40. In V. myrtillus, the first criterion was achieved for AOT40 and POD1−4. POD1 had the highest R 2 value (0.82), while the exposure-based index AOT40 performed equally well (R 2 = 0.81). The CLs was 5.8 mmol m −2 POD1 and 20 ppm·h AOT40 in this species. Figure 6. Dose-response relationships for plant injury index (PII) in Alnus glutinosa, Sorbus aucuparia ,and Vaccinium myrtillus using two O3 indices, PODY (phytotoxic ozone dose above a flux threshold of Y nmol m −2 s −1 ) and AOT40 (accumulated exposure over a threshold of 40 ppb). The critical levels (CLs) corresponding to PII = 5 were also shown. Simple linear regressions were applied. *** p < 0.001 and ** p < 0.01. Figure 6. Dose-response relationships for plant injury index (PII) in Alnus glutinosa, Sorbus aucuparia, and Vaccinium myrtillus using two O 3 indices, POD Y (phytotoxic ozone dose above a flux threshold of Y nmol m −2 s −1 ) and AOT40 (accumulated exposure over a threshold of 40 ppb). The critical levels (CLs) corresponding to PII = 5 were also shown. Simple linear regressions were applied. *** p < 0.001 and ** p < 0.01.

New DO3SE Parameterization in Three Deciduous Tree Species
An accurate parameterization of the g s model is essential to develop a flux-based approach for O 3 risk assessment [26,28]. The model performance with new parameterization was comparable to that in previous studies [26,29]. A comparison of the three target species showed that the g max value was relatively high in A. glutinosa (300 mmol m −2 s −1 ) as compared with the other species (S. aucuparia, 240 mmol m −2 s −1 and V. myrtillus, 140 mmol m −2 s −1 ). This value was within the range for a previous field observation of this species (170 to 380 mmol m −2 s −1 ) [30][31][32]. A high g s enhanced stomatal O 3 uptake, thus, leading to higher O 3 damage [15]. The level of g max in A. glutinosa was comparable to that of the other O 3 -sensitive species such as Oxford poplar clone (340 to 520 mmol m −2 s −1 ) [27,33] and Fagus crenata (315 mmol m −2 s −1 ) [34,35].
Interestingly, the parameter a, in the f light function, was relatively high in V. myrtillus among the three species, suggesting a lower light saturating point of g s . Karlsson, 1987 [36] and Gerdol et al., 2000 [37] reported a relatively low light saturating point of photosynthesis (200 to 300 µmol m −2 s −1 of PPFD) in this species. V. myrtillus is known as a shade tolerant species [38] while the other two species are light demanding [39,40]. In fact, a high a value in f light function was found in other shade tolerant species such as F. crenata (a = 0.0086) [34], while a lower a was obtained in a light-demanding poplar clone (Populus maximowiczii Henry x berolinensis Dippel, a = 0.0020) [27].
In the afternoon, a high VPD often closes stomata together with high air temperature [41]. This was supported by the parameters in f VPD and f temp for the three species. In fact, g s was decreased by 29%, 29%, and 49%, in A. glutinosa, S. aucuparia, and V. myrtillus, respectively, when VPD reached 3 kPa. In addition, g s was decreased by 13%, 38%, and 68%, in A. glutinosa, S. aucuparia, and V. myrtillus, respectively, when air temperature reached 35 • C. Since ambient O 3 concentrations were elevated in the afternoon, those functions were fundamental for the stomatal O 3 flux calculation [28].

Flux-Based Assessment of Ozone Visible Injury
The foliar symptoms of A. glutinosa, S. aucuparia, and V. myrtillus in this experiment were similar to those observed in the field at ambient O 3 levels [8]. The first foliar symptoms were observed in May in all species exposed to 2.0 × O 3 . The onset occurred at 5 to 9 ppm·h AOT40. This is supported by the findings in previous studies where 5 to 10 ppm·h AOT40 caused an emergence of O 3 visible injury for sensitive tree species such as F. sylvatica [11]. In addition, the present study found that approximately 4 mmol m −2 POD 0 was enough to cause the onset of O 3 visible injury regardless of tree species. A. glutinosa with a high g max quickly reached this critical point of POD 0 and showed the first symptom earlier than the other two species. In species with a high g s , the O 3 dose can easily exceed the metabolic capacity for detoxification, and therefore can quickly cause O 3 visible injury [42].
After the onset, the O 3 visible injury expanded to the plant level in all three species as confirmed by the increase in PII values. The increases of PII were well correlated with flux-based indices (POD Y ) in each species. The POD Y showed a higher R 2 than AOT40, suggesting that POD Y was better than AOT40 to assess PII. This is supported by the fact that O 3 impacts are more closely related to O 3 uptake than to external O 3 exposure [15]. Previous studies suggested a threshold Y as an assumed threshold below which stomatal O 3 flux by the plant may be detoxified [16]. The result shows that Y was relatively higher in A. glutinosa (Y = 7) as compared with other two species (Y = 0 to 1). This suggests that A. glutinosa can have a higher capacity for O 3 detoxification than the other two species, although this species had a high stomatal O 3 uptake.
Sicard et al., 2016 [17] indicated that 22 mmol m −2 of POD 0 corresponded to 5% visible injury in O 3 -sensitive deciduous F. sylvatica according to field measurements. Our results in A. glutinosa and S. aucuparia support their findings because the CLs corresponding to PII = 5 on the basis of POD 0 were 22 and 29 mmol m −2 in A. glutinosa and S. aucuparia, respectively. However, the CL in V. myrtillus was much lower than that of the other two species. On the basis of PII, V. myrtillus was more sensitive to O 3 than A. glutinosa and S. aucuparia. This is because V. myrtillus had a dramatic increase in PII from June to July. Although it had a relatively low g s and thus low stomatal O 3 uptake, this species was highly susceptible to O 3 in these months. V. myrtillus had a vegetative stage in May and then flowering and fruiting stages from June to July [43]. In fact, previous studies found that the capacity to detoxify O 3 was lower when the plants were flowering or producing fruits [44][45][46].
The seasonal dynamics of PII differed among the species. In S. aucuparia and V. myrtillus, the PII values showed a monotonic increase, while A. glutinosa had a rather constant PII during June to August. Novak et al., 2003 [13] reported that several species (Populus nigra, Prunus avium, and Salix alba) similarly had a leveling or even decreasing trend of total injured leaf area during the season. In general, O 3 visible injury usually appears on older leaves [4]. However, new leaf formation in A. glutinosa was significantly increased by elevated O 3 , while damaged old leaves were shed. This new leaf growth can alleviate the expansion of O 3 visible injury at the plant level in A. glutinosa. An accelerated leaf turnover can be considered as a compensation response to O 3 stress in plants with indeterminate growth pattern [20]. However, PII in 2.0 × AA was still significantly higher than that in AA in A. glutinosa, suggesting that such leaf growth did not fully compensate for the O 3 damage.

Conclusions
The present O 3 FACE experiment successfully confirmed O 3 visible injury in three cool-temperate deciduous tree species, A. glutinosa, S. aucuparia, and V. myrtillus. The onset of O 3 visible injury in these species required 4.0 mmol m −2 POD 0 and 5.5 to 9.0 ppm·h AOT40. The timing of the first symptom onset among the species was determined by the amount of stomatal O 3 uptake. The early emergence of O 3 visible injury in A. glutinosa was related to high g s ; however, PII was affected not only by stomatal O 3 uptake but also by other species-specific ecophysiological traits. The dynamics of PII suggest that an increased fructification (flowering, fruiting) can weaken the state of the V. myrtillus tree, then, finally the trees can be more sensitive to O 3 [47]. In addition, PII values in A. glutinosa were affected by its indeterminate growth pattern, and a new leaf formation alleviated the expansion of O 3 visible injury at the plant level in this species. Nevertheless, O 3 impacts on PII were better explained by the flux-based index, POD Y , than by the exposure-based index, AOT40, especially in A. glutinosa, although it changed in a complex manner. The CLs corresponding to PII = 5 were 8.1 mmol m −2 POD 7 in A. glutinosa, 23 mmol m −2 POD 0 in S. aucuparia, and 5.8 mmol m −2 POD 1 in V. myrtillus.
Forest trees also suffer from other climate change factors such as elevated CO 2 , nitrogen deposition, warming, the risk of flooding, drought, and forest fire [42]. The interactions between O 3 and other climate change factors are crucial to establish the species-specific O 3 flux-effect relationship for the O 3 risk assessment.
Supplementary Materials: The following are available online at http://www.mdpi.com/1999-4907/11/1/82/s1, Figure S1: Comparison between measured and estimated g s in Alnus glutinosa, Sorbus aucuparia and Vaccinium myrtillus. Table S1: Summary for R 2 and the information of confidential interval (C.I.) in exposure or flux-based dose-response functions for plant injury index (PII).