Ozone Flux-Based Response Functions for Visible Foliar Injury and Photosynthetic Traits in a Bioindicator Species, Viburnum lantana L.

1. Introduction

Tropospheric ozone (O₃) is an oxidative atmospheric pollutant that negatively affects forest health. Despite regulatory efforts to reduce precursor emissions, O₃ concentrations across much of Europe continue to exceed the thresholds established for protecting ecosystems [1,2].

Upon entry into the leaves via the stomata, O₃ interacts with the mesophyll, triggering production of reactive oxygen species (ROS), which subsequently cause physiological impairment, visible foliar injury (O₃ VFI), and finally growth loss [3,4,5,6] while leaf phenotypic plasticity modulates plant sensitivity [7]. To investigate tree responses to O₃, several metrics have been developed, including the exposure-based index Accumulated Exposure Over a Threshold of 40 ppb (AOT40) [8]. However, flux-based indices, such as the phytotoxic O₃ dose above a flux threshold of Y nmol m^–2 s^–1 (POD_Y), have been proven to be more reliable [9] and account for the actual O₃ dose determined by cumulative O₃ uptake over the growing season [10].

O₃ VFI serves as a forest-health biomarker within monitoring initiatives, e.g., the International Co-operative Program on assessment and monitoring of air pollution effects on Forests (ICP Forests) [11]. This is because the O₃ VFI assessment is practical for evaluating O₃ impact on forests in the field without specialized equipment, making it suitable for long-term monitoring. Therefore, scientific efforts have been made to establish O₃ Critical Levels (CLs) for forest protection based on the assessment of O₃ VFI for some dominant tree species (e.g., Fagus sylvatica L., Pinus halepensis Mill., Picea abies L. Karst.) [12,13] and other forest tree or shrub species (e.g., Betula pendula Roth, Larix decidua Mill., Populus tremula L., Salix caprea L., Rubus sp., Vaccinium myrtillus L.) [14]. However, validation of O₃ VFI using manipulative experiments is still required to establish an appropriate CL for forest protection [15].

In the last decades, Viburnum lantana L. has been proposed as an in situ bioindicator of O₃ damage due to its foliar sensitivity and wide distribution in European forests. Field studies in the Italian Alps found “O₃-like” VFI (red stippling and leaf reddening), which was correlated with high O₃ levels (AOT40 > 30 ppm h), though symptom severity also reflected microclimatic conditions, indicating that the CL for O₃ VFI in this species is around 12 ppm h AOT40 [16]. Follow-up research confirmed these patterns, highlighting inter-annual variability likely linked to climate-driven differences in O₃ uptake [17]. Long-term monitoring through the ViburNeT network [18,19] further supported these findings, documenting a consistent relationship between regional O₃ concentrations and O₃ VFI frequency in V. lantana, with 20.7% to 50.6% of individuals exhibiting O₃ VFI frequency, depending on the year. More recently, Faralli et al. [20] demonstrated that symptom severity showed a significant association with lower specific leaf area (SLA) and higher trichome density, especially under high light exposure and in the upper canopy. These findings, although valuable, are based on field observations, where site-specific susceptibility of V. lantana to local environmental conditions can influence stomatal O₃ uptake and thus alter the response to O₃. To date, there has been no study evaluating O₃ damage to V. lantana on the basis of stomatal O₃ flux. Therefore, to fully understand and validate its biomonitoring potential for setting the biological O₃ standard, it is essential to investigate the species’ O₃ response under controlled conditions, such as through Free-Air Controlled Exposure (FACE) systems, aiming to establish a flux-based CL for this bioindicator species.

The present study induced O₃ VFI in the forest bioindicator species V. lantana plants exposed at the Free-Air O₃ eXposure (FO₃X) facility to compare these observations with those reported previously at natural forest sites (e.g., Gottardini et al. [16]), with the primary aim of validating field-based observations under controlled conditions.

In addition, the study aimed to determine which indices between exposure-based (AOT40) and flux-based (POD₁) more accurately capture the effects and incidence of O₃-related VFI and photosynthetic impairment in the bioindicator species V. lantana.

2. Materials and Methods

2.1. Experimental Site and Plant Material

Ozone exposure experiments were conducted at the FO₃X facility situated within the experimental garden of the National Research Council of Italy (Sesto Fiorentino, Italy: 43°48′59″ N, 11°12′01″ E, 55 m above sea level). The fumigation system details are provided in Paoletti et al. [21]. In December 2023, three-year-old Viburnum lantana L. saplings, raised in Santa Giustina in Colle (Padova, Veneto, Italy), were obtained from a local nursery (Vivai Guagno). At the time of transplantation, the plants, grown in 18 cm pots, had an average height of 74.7 cm and a stem diameter of 13.5 mm. Each plant was transplanted into a 9 L plastic pot (24 cm diameter) containing a uniform mixture of sand, peat, and soil in equal parts (v:v:v). This substrate composition was selected to ensure adequate drainage, nutrient availability, and root aeration during the subsequent growth period. All plants were watered daily to prevent water stress. From 17th May to 16th October 2024 (152 days), three levels of O₃ concentration treatments (ambient O₃ concentration [AA], 1.5 times ambient O₃ concentration [1.5×], and doubled ambient O₃ concentration [2.0×]) were applied using three replicated plots (dimension of 5 m × 5 m × 2 m) and 1- 2 plants were assigned for each plot.

The average hourly O₃ concentrations in the AA, 1.5×, and 2.0× treatments were 38.0 ppb, 48.6 ppb, and 61.3 ppb, respectively. Figure 1 shows the seasonal patterns of solar irradiance, precipitation, and mean daily temperature, based on hourly meteorological data collected throughout the growing season.

2.2. Assessment and Quantification of Visible Foliar Injury

Visible foliar injury attributable to O₃ was monitored at 2- or 3-week intervals throughout the exposure period. Assessments were carried out independently by two trained observers with ICP Forests Intercalibration experience [11]. For each plant, we recorded the proportion of leaves showing symptoms (SL) and, within symptomatic leaves, the mean affected leaf area (SA), using a ×10 hand lens and published photographic references [22]. To quantify the injury at the plant level, the Plant Injury Index (PII) was calculated by combining the two parameters [22,23] as follows:

$$PII={\displaystyle \frac{\left(SL\times SA\right)}{100}}$$

(1)

To describe, classify, and standardize O₃ VFI, color composition was assessed based on the methodology suggested by Moura et al. [15]. Three pictures showing a range of injuries (expert visual assessment from 0 to 70% SL) were taken under natural lighting conditions, all on the same day (9–10 a.m.), using a digital camera (Cyber-shot DSC-H300, Sony, Tokyo, Japan). The photos were combined into a single RGB figure, subsequently processed using the indexed color mode. Following this, both a Local (Perceptual) and a 64-colour image were produced. Subsequently, all the tissue exhibiting greenish and the vein-related colors were extracted from the image, leaving only potential O₃ VFI colors visible [15].

2.3. Assessment of Photosynthetic Parameters

Measurements of photosynthetic parameters (i.e., SPAD, chlorophyll a fluorescence, and leaf gas exchange) were conducted three times (time zero—T0: 13th May [before O₃ exposure], T1: 11th July [55 days exposure], and T2: 8th September [114 days exposure]). Leaf gas exchange rates were measured using a portable photosynthesis measurement system (LI-6800, Li-Cor instruments, Lincoln, NE, USA). Measurements were carried out between 8 and 12 a.m. CET for one to two plants per plot per treatment. For each plant, three to five leaves with 4th to 6th order from the tip of the shoot were targeted. During the measurement, the parameters within the LI-6800 leaf cuvette were established as follows: CO₂ concentration (420 ppm), leaf temperature (25 °C), photosynthetic photon flux density (PPFD, 1500 μmol m^–2 s^–1, utilizing an LED light source of 10% blue and 90% red light), and relative humidity (50%). From these measurements, light-saturated net photosynthetic rate (A_sat) and stomatal conductance (g_s) were determined.

A SPAD meter (Konica Minolta, Tokyo, Japan) was utilized to measure leaf greenness as a proxy of chlorophyll content. In addition, a HandyPEA fluorimeter (Hansatech Instruments, Pentney, Norfolk, UK) was used to measure the chlorophyll a fluorescence. In order to measure the chlorophyll a fluorescence, after 40 min of dark adaptation, leaves were subjected to a 1 second saturating pulse of red light (peak wavelength: 650 nm) at an intensity of 3,000 μmol photons m^–2 s^–1 to determine the fluorescence yields in the dark (F₀) and those with a saturating pulse (F_m). The maximum quantum yield (F_v/F_m) was calculated as F_v/F_m = 1 − (F₀/F_m).

2.4. Modeling of Stomatal Conductance

In order to estimate stomatal O₃ flux, it is essential to parameterize the g_s model. A measurement campaign of g_s for 4th to 6th ordered leaves of all target plants using an open flow-through differential porometer (LI-600, Lincoln, NE, USA) was conducted. Stomatal conductance was measured on 19 sampling days, scheduled at approximately monthly intervals and covering a range of environmental conditions. The resulting dataset, comprising 593 measurements, was used to parameterize the multiplicative g_s model [24,25], as described below:

$$g_s=g_{max}·f_{light}·max\left\{f_{min,}\left(f_{temp}·f_{VPD}\right)\right\}$$

(2)

where g_max is the maximum stomatal conductance (mol O₃ m^–2 Projected Leaf Area [PLA] s^–1). All other functions are expressed in relative terms and scaled from 0 to 1. The model incorporates minimum stomatal conductance (f_min), and it adjusts gs based on PPFD (f_light), temperature (f_temp), and vapor pressure deficit (VPD) (f_VPD). The details of the model functions are available in the Supplementary file (Appendix Method S1). In this study, the function of soil water content (f_SWC) was not included because plants did not receive any soil water stress. The model was parameterized using the boundary line approach [26,27]. The g_max and f_min values were set as the 95th and 5th percentiles of all g_s data, respectively, following the methodology of Bičárová et al. [28] and Hoshika et al. [25].

2.5. Calculation of Ozone Indices

For daylight periods, identified by shortwave radiation exceeding 50 W m^–2, AOT40 was obtained by summing the hourly exceedance of O₃ concentrations above the 40 ppb threshold, following CLRTAP [29]. It is defined as:

$$AOT40={\sum }_{i=1}^{n}max\left({\left[O_3\right]}_i-40,0\right)$$

(3)

where [O₃]i represents the measured hourly O₃ concentration (ppb), with i ranging from 1 to n in the integral, where n is the total number of hours in the calculation period.

According to CLRTAP [29], stomatal O₃ flux (F_st; nmol m^–2 s^–1) can be given as:

$$F_{st}=\left[O_3\right]·g_s·{\displaystyle \frac{r_c}{r_b+r_c}}$$

(4)

where [O₃] (ppb) is the hourly O₃ concentration, r_c is the surface resistance of leaf (= 1/(g_s + g_ext); s m^–1), g_s is the cuticular conductance (m s^–1), and r_b is the boundary layer resistance of the leaf (s m^–1) calculated as r_b = 1.3 ×150 ×(L_d / u) 0.5 where u is the wind speed (m s^–1), L_d is the cross-wind leaf dimension (0.05 m for broadleaves, [29]).

POD_Y (mmol m^–2) was calculated as the sum of hourly F_st data as:

$${POD}_Y={\sum }_{i=1}^{n}max\left(F_{st\_i}-Y,0\right)$$

(5)

where F_{st_i} is the hourly stomatal O₃ uptake (nmol m^–2 s^–1), n is the number of hours included in the calculation period. Y is a species-specific threshold of stomatal O₃ uptake (nmol m^–2 s^–1). In the mapping manual [29], Y = 1 nmol m^–2 s^–1 is recommended to assess the negative O₃ effects on woody plant species. Therefore, we utilized POD₁ to establish dose-response relationships with O₃ VFI and physiological parameters.

2.6. Data Analysis

To assess treatment effects, PII and physiological parameters (A_sat, g_s, SPAD, and F_v/F_m) were analyzed on absolute values by two-way ANOVA with O₃ treatment and time as fixed factors, including their interaction, after verifying normality and homogeneity of variances. When significant effects were detected (p ≤ 0.05), differences among levels were evaluated using Tukey’s HSD post hoc test.

Dose–response relationships were analyzed separately by expressing each response variable as a relative change from an operational baseline, defined by the mean value measured under AA at T0 (considered as the reference condition in which features were unaffected by O₃). Exposure- and flux-based dose–response functions were then fitted using AOT40 or POD₁ as predictors, comparing (i) a linear model and (ii) a non-linear polynomial model (proposed for O₃ risk assessment applications by Hoshika et al. [30]). Model selection was primarily based on the Akaike Information Criterion (AIC), which balances goodness-of-fit and model complexity; the coefficient of determination (R²) was used to quantify the proportion of variance explained.

A sensitivity scenario analysis was also conducted to quantify how assumptions on stomatal regulation affect flux-based estimates. Specifically, when AOT40 reached 12 ppm h (as suggested by Gottardini et al. [16]), POD₁ was recalculated under a set of imposed stomatal-closure scenarios (0–50% closure, 10% increments) to simulate additional environmental stress (e.g., drought).

For VFI, CLs were defined as the exposure/uptake at which symptoms first occurred, operationally identified as the point at which PII became > 0 (PII = 0.01). For physiological parameters, the CL was defined as the dose corresponding to a 4% reduction relative to the baseline, following Hoshika et al. [30]. All statistical analyses were performed using OriginPro 2025 (OriginLab Corporation, Northampton, MA).

3. Results

3.1. Ozone Visible Foliar Injury and the Number of Leaves

The first O₃ VFI for the plants treated with 2.0× O₃ exposure were observed after 16 days of exposure (on 14th June). Subsequently, the symptoms emerged in the other two O₃ treatments (i.e., AA and 1.5×) 32 days after the beginning of the exposure (on 2nd July). The O₃ VFI are characterized by interveinal stippling on the adaxial leaf surface, with a reddish to dark brown coloration (Figure 2). These stipples were localized between the veins, often sparing the vein tissue itself, and did not show signs of tissue necrosis or insect damage.

Color-composition analysis identified 16 perceptual colors that captured the main chromatic features of O₃ VFI in V. lantana (Figure 2). The selected palette described a gradual shift from light to dark brownish tones, with reddish intermediates, consistent with the visual progression of injury.

The PII was significantly affected by the interaction between treatment and time (p < 0.05, Table 1, Table S1). Across all O₃ treatments, PII values increased over time, with O₃ VFI already evident but not statistically significant at T1. At this stage, mean PII values exceeded 1.5 in the two elevated O₃ treatments (2.03 ± 1.11 in 1.5× and 2.12 ± 1.41 in 2.0×), reaching maximum values of 0.61 in AA, 3.83 in 1.5×, and 4.79 in 2.0×. Injury became even more pronounced at T2, with a statistical difference under the 2.0× treatment. At T2, while the AA showed modest increases (mean values were 1.29 ± 0.71), the 1.5× O₃ treatment showed a decrease, justified by leaf senescence in this treatment (1.31 ± 0.62), and a significant rise in PII was observed in the 2.0× treatment at T2 (9.06 ± 3.24). Variability in PII was high across treatments; the lowest PII value was 0.1, recorded in the 1.5× at T1 (i.e., 1.00% of SL and 10% of SA), and the highest recorded value of 14.13 was in 2.0× T2 (i.e., 41.3% of SL and 34.21% of SA).

3.2. Ozone Effects on Photosynthetic Parameters

The A_sat was significantly affected by both treatment (p < 0.05) and time (p < 0.001) (Table 1, Table S1). Across treatment, a consistent decline was observed from AA to 2.0×, with AA presenting an average A_sat of 10.45 ± 1.10 µmol m^–2 s^–1, 1.5× an intermediate value of 9.07 ± 1.09 µmol m^–2 s^–1, and 2.0× an average of 8.16 ± 1.38 µmol m^–2 s^–1. Across time, A_sat at T1 (7.81 ± 0.45 µmol m^–2 s^–1) and T2 (6.34 ± 0.76 µmol m^–2 s^–1) were statistically lower than the baseline (T0: 13.54 ± 0.53 µmol m^–2 s^–1).

Regarding the g_s, there was only a statistically significant influence of time (p < 0.0001) (Table 1, Table S1), with mean values decreased from 0.172 ± 0.014 mol m ^–2 s^–1 at T0 to 0.091 ± 0.007 mol m ^–2 s^–1 at T1 and remained at a similar level at T2 (0.091 ± 0.006 mol m ^–2 s^–1).

SPAD values, indicative of relative chlorophyll content, were significantly affected by the O₃ treatment (p < 0.05, Table 1, Table S1). Across all timepoints, the AA plants showed the highest mean SPAD values (50.42 ± 1.76), which were significantly higher than those observed in the 1.5× treatment (42.22 ± 2.12). In contrast, the 2.0× treatment group exhibited intermediate SPAD values (42.82 ± 6.62), which did not differ significantly from either AA or 1.5× due to the high within-group variability.

F_v/F_m was significantly affected by the interaction between treatment and time (p < 0.01; Table 1, Table S1). Across all three O₃ treatments, F_v/F_m values remained stable throughout the experiment, with mean values of 0.72 ± 0.01, indicating limited temporal variation in PSII photochemical efficiency. A notable exception, however, was observed in the 2.0× treatment, which showed a significant decrease in F_v/F_m during mid-summer (T1: 0.64 ± 0.01), followed by a recovery at T2. This transient reduction was not observed at the other two exposure levels, which maintained consistent F_v/F_m values across all time points.

3.3. Parameterization of the Stomatal Conductance Model

The g_max value found for V. lantana was 0.158 mol O₃ m^–2 PLA s^–1, whereas f_min was set to 0.06. The limitations of g_s due to environmental factors are shown in Figure 3. With increasing light intensity, a sharp increase in g_s was observed, even under a relatively low PPFD (200 μmol m^–2 s^–1), as shown by f_light (f_{light_a} = 0.0163). The optimal temperature for stomatal opening (T_opt) was 24 °C, while T_max and T_min were set to 41 and 0 °C, respectively. In addition, a high air VPD observed in the afternoon caused stomatal closure, as confirmed by the f_VPD function (VPD_max: 1.8 kPa, VPD_min: 6.4 kPa).

3.4. The Relation of O₃ Indices, Plant Injury Index (PII), and Plant Physiological Responses

The results of the linear and polynomial regression models examining the relationships between PII and the physiological parameters (A_sat, g_s, SPAD, and F_v/F_m) and the two indices (POD₁ and AOT40) showed varying levels of model fit and explanatory power (Table S2). All the regressions were statistically significant except for SPAD and F_v/F_m, thus will not be discussed in the study.

For the PII, the polynomial model performed better than the linear one, and the AOT40 (Figure 4A inset) performed slightly better than POD₁ (Figure 4A). As it was clear that the PII response remained flat across time of exposure and then sharply rose over a narrow interval at the end of the exposure period, we considered both polynomial models (with AOT40 and POD₁) as suitable for representing the species response (Figure 4A).

The best-fit model for physiological parameters was the polynomial, performed with POD₁ (Figure 4 B,C). The A_sat declined rapidly at low POD₁, then approached a stable lower plateau at higher POD₁. However, A_sat remained well below its initial level across the higher POD₁ values, especially at the end of the exposure period, indicating reduced leaf photosynthetic capacity with increasing accumulated O₃ uptake (Figure 4B).

For g_s, the polynomial function using POD₁ also provided the best fit, showing a lower AIC compared to all the other tested functions (Table S2). The fitted curve was similar to that for A_sat, with g_s decreasing sharply at low POD₁, then stabilizing, when g_s had already declined substantially from its initial levels, and continued to decrease gradually as POD₁ increased (Figure 4C).

3.5. Critical Levels

The exposure- or flux-based CLs were calculated for all significant regressions (Table S2), and Table 2 reports only the CLs of the chosen best-fit functions.

Comparing the CLs calculated for the PII, the values were considerably higher with the polynomial function (7.82 mmol m^–2 for POD₁ and 4.42 ppm h for AOT40, Table 2) than with the linear regressions (0.54 mmol m^–2 for POD₁ and 0.81 ppm h for AOT40; Table S2).

In contrast, for the physiological parameters, the polynomial fits produced lower CLs than the linear fits, and g_s consistently showed lower CLs than A_sat, indicating an earlier response of stomatal conductance to O₃ (POD₁: 1.22 vs 1.61 mmol m^–2; AOT40: 1.31 vs 2.27 ppm h, respectively; Table 2).

4. Discussion

The manipulative experimental approach adopted in our study for the first time enabled the following new findings on O₃ effect on both O₃ VFI and the physiological features of the important bioindicator species, V. lantana, filling a significant knowledge gap regarding the integrated response of the species to O₃ stress.

4.1. Validation of Ozone Visible Foliar Injury Based on Free-Air Experiments

The appearance and progression of O₃ VFI observed in V. lantana under elevated O₃ treatments confirmed that this species is sensitive to O₃ exposure. Plants exposed to 2.0× developed VFI two weeks earlier than those at lower treatment levels, following a clear dose-dependent pattern. This observation aligns with previous experimental studies on O₃ VFI for this species, which found that higher AOT40 levels were associated with more frequent VFI [23]. Previous studies suggest that the onset timing of O₃ VFI shows species-specific dependency; for instance, the O₃-sensitive deciduous species, Alnus glutinosa and Vaccinium myrtillus, showed early-season symptom development, whereas the O₃-tolerant Mediterranean shrub, Arbutus unedo, exhibited a delayed onset of O₃ VFI, appearing only in the late growing season [11,15,31]. In our study, although the O₃ VFI onset was found in the early summer, PII values at 2.0× treatment became significantly higher than in the AA treatment only at T2, likely reflecting the slow development of O₃ damage in V. lantana throughout the growing season.

The symptoms consisted of small reddish-brown spots distributed on the upper leaf surface, mainly in the interveinal areas. Veins were generally unaffected, and the leaves did not show necrotic patches or damage patterns attributable to insects or pathogens [11]. The color analysis (Figure 2) supported this pattern, showing a chromatic gradient from light grey to dark brown, indicative of increasing tissue alteration, similar to the colors observed in previous field assessments [17,31]. These colorimetric patterns not only reflect the severity of foliar damage but may also serve as a diagnostic tool for O₃ injury under field conditions. As demonstrated by Moura et al. [15], the color components observed in experimental manipulations are comparable in many species to those detected in field samples, confirming the comparability between controlled and natural environments. This suggests that the color profiles documented in our study could be effectively used for quality control and support field-based O₃ VFI assessments, enhancing the accuracy of visual diagnosis in biomonitoring applications.

Despite the clear development of O₃ VFI, the correspondence with systemic eco-physiological responses was only partial, highlighting the complexity of O₃–plant interactions. While PII increased significantly with O₃ exposure and A_sat showed a marked decline, the response of g_s was more limited. The decrease of A_sat in the absence of g_s reduction in O₃-exposed leaves was similarly found in Japanese Siebold’s beech [32] and I-214 poplar clone [33]. Since the rates of damage to photosynthesis and stomatal conductance do not always coincide, O₃ exposure can decouple g_s from A_sat [25,34]. In fact, O₃ impairs stomatal function, leading to reduced stomatal responsiveness to environmental stimuli, i.e., stomatal sluggishness [35,36]. The remaining open stomata reduce water-use efficiency, increasing the risk of other climate crisis factors, such as drought [37].

In our study, this partial decoupling between O₃ VFI and physiological impairment may be linked to ‘invisible’ damage and physiological modifications, including repair processes, before visible signs of injury appear [38]. A recent study applied Terahertz (THz) imaging analysis and revealed that the negative effect of O₃ on leaf water status was associated with cell damage before the onset of VFI in Carpinus betulus L. and Ostrya carpinifolia Scop. plants [39]. In fact, in our experiment, the marked reduction in A_sat was already found at T1, even though PII was still relatively low. Moreover, a significant decline in F_v/F_m was observed at 2.0× only at the early time point (T1), suggesting an initial impairment of PSII photochemical efficiency. The absence of sustained effects at later stages may indicate a partial recovery over time or a localized, non-systemic damage response. Similar dynamics were reported by Bussotti et al. [40] in a field-based study on V. lantana, in which chlorophyll fluorescence measurements under high O₃ exposure showed a marked reduction in parameters related to electron transport efficiency. These converging observations confirm the moderate physiological sensitivity of V. lantana to O₃, with photochemical alterations detectable even before O₃ VFI becomes widespread, consistent with findings from both field and controlled studies.

4.2. Flux-Based Assessment of Ozone Visible Injury and Physiological Parameters

This study also aimed to evaluate whether O₃-induced effects on V. lantana are better captured by the flux-based index (POD₁) rather than the exposure-based index (AOT40). Regression analyses confirmed that POD₁ was a superior metric for capturing the physiological response, with the polynomial model fitting better than the linear regression, specifically, for A_sat and g_s. While both O₃ indices were significantly related to PII, the polynomial model with POD₁ again showed a tendency toward a better fit and better represented the biological process (Figure 4A inset, Table S2).

The stronger association between POD₁ can be attributed to its physiological basis. Unlike exposure-based metrics such as AOT40, which rely solely on ambient O₃ concentrations, POD₁ estimates the actual O₃ dose absorbed through stomata. This flux-based approach is widely recognized as the primary driver of O₃-induced phytotoxicity [4,6], as it accounts for plant-specific factors such as stomatal regulation under varying environmental conditions (e.g., water availability). Furthermore, we recommend using non-linear models for dose-response relationships, as emphasized by Hoshika et al. [30], particularly when they outperform linear regression.

4.3. Critical Levels

The use of the PII offers an integrated measure of O₃ VFI damage by combining both the proportion of symptomatic leaves per plant (SL) and the average affected area within those leaves (SA). This dual-component approach provides a more realistic representation of total O₃ VFI than frequency-based metrics alone. In our study, we proposed that the CL for PII should be calculated when the PII value reaches 0.1, indicating the onset of damage (i.e., 1% SL and 10% SA), which, in the case of V. lantana, using the polynomial regression as the best fit (Table S2), occurred when POD₁ reached 7.82 mmol m^–2 (Table 2).

If calculated using the exposures-based index AOT40, the polynomial regression fit better, and the CL for the PII was determined to be 4.42 ppm h. In the past, a threshold of 12 ppm h was proposed for the species to indicate injury onset based on field observations in the Mediterranean Alps [18]. The discrepancy between the calculated CL presented here and previous literature may result from differences in environmental conditions affecting stomatal O₃ uptake between natural forests and the FO₃X experimental facility. In fact, during summer-time, drought often limits stomatal O₃ uptake for trees at the natural forest sites in the Mediterranean Alps (–31% due to soil water deficit, [14]), whereas, at the FO₃X, plants were well-irrigated and any water stress was not expected to reduce g_s for V. lantana and thus enhancing stomatal O₃ uptake. A sensitivity analysis revealed that, if the 30% hypothetical stomatal closure occurred at the FO₃X, the values of POD₁ at the field-observed threshold 12 ppm h AOT40 would be 5.9 to 9.9 mmol m^–2, which agrees with the flux-based CL (7.82 mmol m^–2) (Table S3). The results indicate that the AOT40-based CL may vary across the target regions due to insufficient consideration of the biological processes underlying O₃ uptake. Given that stomata are the primary interface for O₃ entry into plants, a stomatal flux-based approach should be recommended for a proper setting of CL in forest protection against O₃ pollution.

Interestingly, although a flux-based POD₁ CL for PII was identified as 7.82 mmol m^–2 POD₁, physiological thresholds for a 4% reduction in A_sat and g_s were observed at lower POD₁ values (1.61 mmol m^–2 and 1.22 mmol m^–2, respectively). This suggests that physiological alterations begin before the manifestation of visible injury, supporting the use of combined indicators for a comprehensive assessment. Therefore, a multi-indicator framework combining morphological and physiological indicators is recommended to detect early stress signals and assess cumulative O₃ impact more accurately [41]. This approach improves detection sensitivity, enhances inter- and intra-species comparisons in biomonitoring, and underscores the importance of flux-based metrics in O₃ risk assessment. Low POD values as CLs for physiological parameters were also found in other sensitive species, where the CLs for A_sat were suggested to be 1.9 to 3.5 mmol m^–2 POD₀ for O₃-sensitive deciduous poplars and hornbeam species [39]; therefore, V. lantana can be identified as a sensitive species to O₃.

5. Conclusions

This study examined O₃ effects on V. lantana under open-air experimental conditions. Results showed a dose-dependent pattern, with O₃ VFI appearing first in plants exposed to a higher O₃ environment (2.0× treatment). These visual injuries matched field symptoms and were confirmed by colorimetric analysis.

At the same time, physiological responses revealed a more complex pattern. While A_sat and F_v/F_m declined under elevated O₃, gₛ and SPAD were less responsive, suggesting that physiological alterations can occur with a different pattern. Notably, thresholds for a 4% reduction in A_sat and gₛ were reached at POD₁ levels below those needed to induce VFI, highlighting the importance of combining visual and functional indicators to avoid underestimating damages. The dose–response analysis also showed that the flux-based metric POD₁ better explained variability in physiological parameters than the exposure-based AOT40, supporting the actual consensus that stomatal uptake is a more accurate predictor of plant damage. Thus, we suggested CLs of 1.61 mmol m^–2 and 1.22 mmol m^–2 POD₁ for A_sat and g_s, respectively, and a CL of 7.82 mmol m^–2 for the onset of O₃ VFI.

In conclusion, this study provides the first detailed eco-physiological characterization of the O₃ effect in V. lantana under controlled and realistic exposure conditions. It confirms the species’ moderate sensitivity to O₃ and refines critical thresholds for O₃ risk assessment. V. lantana is confirmed as an auspicious bioindicator for integrated O₃ impact monitoring, particularly when physiological and morphological parameters complement visual assessments.