Exploring a New O₃ Index as a Proxy for the Avoidance/Tolerance Capacity of Forest Species to Tolerate O₃ Injury

Abstract

Tropospheric ozone (O₃) is a detrimental air pollutant causing phytotoxic effects. Several O₃ indices are used to assess the risk for vegetation, e.g., the exposure-based AOT40 (accumulated ozone exposure over a threshold of 40 ppb) and the stomatal-flux based POD₁ (Phytotoxic Ozone Dose above a threshold of 1 nmol m⁻² s⁻¹). Leaf Mass per Area (LMA) is recommended as a simple index to explain the plant tolerance capacity to O₃. We therefore tested a new species-specific O₃ index (Leaf Index Flux—LIF: calculated as stomatal O₃ flux/LMA) as a proxy of the avoidance/tolerance capacity against O₃ stress according to datasets of visible foliar injury (VFI) in forest monitoring and a manipulative Free-Air Controlled Exposure (FACE) experiment. For the forest monitoring, AOT40, POD₁, and LIF were calculated from hourly O₃, soil moisture, and meteorological measurements at nine Italian forest sites over the period 2018–2022. The results were tested for correlation with the O₃ VFI annually surveyed at the same sites along the forest edge (LESS) or inside the forest (ITP) and expressed as relative frequency of symptomatic species in the LESS (SS_LESS) and Plant Injury Index per tree in the plot (PII_ITP). Based on VFI occurrence at ITP and LESS, Fagus sylvatica was considered the most O₃-sensitive species, whereas conifers (Pinus pinea and Picea abies) and other deciduous/evergreen broadleaf (Quercus petraea, Q. cerris, Q. ilex, and Phyllirea latifolia) showed rare and no O₃ VFI. Shrub species such as Rubus spp. and Vaccinium myrtillus were O₃-sensitive, as they showed VFI along the LESS. AOT40 did not show significant correlations with the VFI parameters, POD₁ increased with increasing SS_LESS (p = 0.005, r = 0.37) and PII_ITP (p < 0.001, r = 0.53), and LIF showed an even higher correlation with SS%_LESS (p < 0.001, r = 0.63) and PII_ITP (p < 0.001, r = 0.87). In the FACE experiment, PII was investigated for five deciduous and three evergreen tree species following one growing season of exposure to ambient and above-ambient O₃ levels (PII_FACE). Moreover, PII_FACE resulted better correlated with LIF (r = 0.67, p < 0.001) than with POD₁ (r = 0.58, p = 0.003) and AOT40 (r = 0.35, p = 0.09). Therefore, LIF is recommended as a promising index for evaluating O₃ VFI on forest woody species and stresses high O₃ risk potential for forest species with high stomatal conductance and thin leaves.

1. Introduction

Ground-level ozone (O₃) is a secondary air pollutant arising from the interaction among solar radiation, high temperature, and its precursors (CH₄, NO_x, CO, and Volatile Organic Compounds) emitted by natural and anthropogenic sources [1,2]. Although a decrease in precursors’ emissions led to a reduction of O₃ mean concentrations in North America, Europe, and Asia [3,4], an increase linked with high temperature is predicted in the near future due to climate change [5], reaching 50–55 ppb in Europe by the end of the 21st century [6,7]. In particular, Southern European countries are among the most affected areas due to the local climatic conditions, such as a hot and dry summer, that favor O₃ photochemical formation in the troposphere [8].

Ozone is a significant abiotic stressor for plants, affecting their health and vitality [9]. Furthermore, as O₃ concentrations tend to be higher in rural and semi-rural areas than in cities [10], forests are highly threatened, and their essential ecosystem services, such as carbon sequestration [11] and biodiversity preservation [12], might be affected.

Ozone is absorbed by trees through stomata and stimulates the formation of Reactive Oxygen Species (ROS) [13]. Its phytotoxicity promotes the decline of photosynthesis and the alteration of stomatal physiology [14,15], leading to a reduction in growth [16] and productivity [17,18]. Furthermore, specific O₃-visible foliar injuries (O₃ VFI), such as typical chlorosis, black spots, and interveinal necrosis, due to the oxidation of leaf tissues [19,20,21,22], have been detected in various tree species across forest sites and can be used as a biological indicator to assess O₃ impacts on vegetation in the field [23]. Moreover, O₃ monitoring in forests is crucial, and several indices were developed to probe forest responses to this pollutant. Currently, the European directive for forest protection from O₃ (EU Directive 2008/50/EC) [24] is based on AOT40 (accumulated ozone exposure over a threshold of 40 ppb), an exposure-based index that accumulates the hourly concentrations of O₃ exceeding the threshold of 40 ppb over daylight hours during the growing season. However, the damages caused by O₃ are mainly related to its stomatal flux rather than its atmospheric concentration [25]. Hence, the National Emission Ceilings (NEC) Directive, within the framework of the Convention on Long-range Transboundary Air Pollution (CLRTAP), recently decided to include the Phytotoxic Ozone Dose (POD_Y) to assess O₃ impact on plant ecosystems [26]. POD_Y is a flux-based index that estimates stomatal O₃ uptake during the growing season above a threshold Y of potential phytotoxicity [22] and considers species-specific stomatal conductance responses to environmental conditions.

Nevertheless, other foliar structural traits beyond stomatal conductance, such as leaf morphology, may play a prominent role in O₃ sensitivity [20]. In particular, species with thick and dense leaves, such as Mediterranean evergreens, are stated to be more resistant to biotic and abiotic stressors [27]. Feng et al. [28] suggested that Leaf Mass per Area (LMA) could explain O₃ sensitivity variation among plant species. In fact, thick leaves have a relatively low mesophyll surface area exposed to O₃ [29] and cellular defense substrates per unit of stomatal O₃ flux increase with increasing LMA, as reported in spruce, pine, and larch trees [30].

Therefore, in the present study, we proposed a simple index called Leaf Index Flux (LIF) to weigh stomatal O₃ flux by LMA and correlate with O₃ VFI better than POD₁ or AOT40.

To provide realistic evidence and a proper species-specific examination, VFI was assessed both in the field, i.e., along Light Exposed Sampling Sites (LESS) of the forest edge or inside the forest plots (ITP), and in an O₃ Free-Air Controlled Exposure (FACE) facility.

Our aim was to verify whether LIF was well correlated with O₃ VFI under field and semi-controlled conditions and thus can be used as a proxy to assess O₃ VFI in forests. In addition, we tested which forest species were more O₃-sensitive when evaluated by applying LIF.

2. Materials and Methods

2.1. Forest Monitoring Network

The evaluation occurred from 2018 to 2022 in nine Italian forest sites belonging to a European monitoring network (MOTTLES—MOnitoring ozone injury for seTTing new critical LevelS) (

Table 1. Features of the nine Italian sites belonging to the MOTTLES network.

Site Code	Geographic Coordinates	Altitude	Dominant Tree Species	Biome
ABR1	41.86064 N–13.57482 E	1500	Fagus sylvatica	Alpine
CPZ1	41.70423 N–12.35719 E	0	Quercus ilex	Mediterranean
CPZ2	41.70429 N–12.35732 E	0	Phyllirea latifolia	Mediterranean
CPZ3	41.68068 N–12.39084 E	0	Pinus pinea	Mediterranean
EMI1	44.71998 N–10.20345 E	200	Quercus petraea	Continental
LAZ1	42.82746 N–11.89817 E	690	Quercus cerris	Mediterranean
PIE1	45.68374 N–8.06994 E	1150	Fagus sylvatica	Alpine
TRE1	46.35825 N–11.49405 E	1800	Picea abies	Alpine
VEN1	46.06335 N–12.38810 E	1100	Fagus sylvatica	Alpine

). The sites covered distinct geographical areas of Italy and represent the Alpine, Continental, and Mediterranean biomes. The dominant forest species were conifers (Pinus pinea L. and Picea abies L.), deciduous broadleaf trees (Fagus sylvatica L., Quercus petraea (Matt.) Liebl., and Q. cerris L.), and evergreen broadleaf trees (Q. ilex L. and Phyllirea latifolia L.).

The ozone concentrations and meteorological parameters were monitored at field stations powered by photovoltaic panels and placed in open fields (OFD) near the forest cover in each site [25]. In detail, O₃ concentrations were measured in real time by an active monitor (Model 106–L, 2B Technologies, Inc., Boulder, CO, USA), while the environmental variables recorded by specific sensors (DeltaOHM, Selvazzano Dentro, Italy) were air temperature (T), relative humidity (RH), solar radiation (photosynthetic active radiation: PAR), precipitation (rainfall), wind direction, and speed. Each sensor was installed 2 m above the ground. Hourly average data were recorded and transmitted remotely via GPRS connection. In case of signal absence, data were stored in data loggers (Campbell Scientific, Logan, UT, USA; Loughborough, UK) and subsequently downloaded manually. Maintenance and calibration were periodically performed every year. The soil water content (SWC) was measured within the forest plots (ITP), using sensors (Campbell Scientific, Logan, UT, USA; Loughborough, UK) randomly placed at a depth of 10 cm.

2.2. Ozone FACE Facility

A new-generation tridimensional O₃ FACE facility located in Sesto Fiorentino (43°48′59″ N, 11°12′01″ E, 55 m a.s.l.) was used to artificially fumigate potted plants by O₃. Further details about the facility are reported in Paoletti et al. [31]. Three evergreen (Arbutus unedo L., Phillyrea angustifolia L., and Pinus pinea L.) and five deciduous species (Sorbus aucuparia L., Alnus glutinosa (L.) Gaertn., Vaccinium myrtillus L., Populus maximowiczii Henry X P. × berolinensis Dippel, and Populus x euramericana I-214) were exposed to three O₃ levels (Ambient Air, 1.5 $\times$ Ambient Air, and 2.0 $\times$ Ambient Air) during the growing season and well-watered every day to maintain field capacity. Hourly average data related to environmental parameters (T, RH, PAR, and wind) were recorded during the experimental period of each year by a Watchdog station (Mod. 2000; Spectrum Technology, Inc., Aurora, IL, USA) placed at 2.5 m above ground level.

2.3. Calculation of O₃ Indices: AOT40, POD₁ and LIF

The accumulated ozone exposure over a threshold of 40 ppb (AOT40; ppb h) was calculated by summing all hourly average O₃ concentrations exceeding the threshold of 40 ppb during daylight hours (shortwave solar radiation > 50 W m⁻²) from the beginning of the growing season to the time of the VFI survey, according to the following formula:

$$\mathrm{AOT}40\left(\mathrm{ppb}·\mathrm{h}\right)={\displaystyle \sum }_{i=1}^n\mathit{max}\left(\left[{\mathrm{O}}_3\right]i-40, 0\right)·\Delta t$$

This index assumes that vegetation does not suffer damage with O₃ concentrations lower than 40 ppb (AOT40 = 0) and that, during the night, the stomata are closed and prevent the entry of O₃. The critical level for forest protection based on accumulated ozone exposure over a threshold of 40 ppb (AOT40) is 5000 ppb h, corresponding to a 5% reduction in tree biomass [32] (CLRTAP, 2017).

Regarding the Phytotoxic Ozone Dose (POD_Y), the following formula was used to carry out the calculation:

$${\mathrm{POD}}_{\mathrm{Y}}\left(\mathrm{mmol}{\mathrm{m}}^{-2}\right)={\displaystyle \sum }_{i=1}^n\left({\mathrm{F}}_{\mathrm{st},\mathrm{i}}-\mathrm{Y}\right)·\Delta t$$

where F_st represents the hourly stomatal O₃ flux, n is the number of hours for the period considered, and ∆t = 1 h. We considered a Y threshold of 1 nmol O₃ m⁻² s⁻¹ (POD₁), as suggested by the UNECE Mapping Manual of CLRTAP [32].

Finally, the Leaf Index Flux (LIF) was calculated as the ratio between species-specific POD₁ and Leaf Mass per Area (LMA) and was expressed in mmol O₃ g⁻¹:

$$\mathrm{Leaf}\mathrm{Index}\mathrm{Flux}\left(\mathrm{LIF}\right)=\frac{\mathrm{POD}1}{\mathrm{LMA}}$$

The Leaf Index Flux was calculated ex novo for the forest sites (LESS and ITP) and by using published data [33] for the species subjected to artificial fumigation with O₃ in the FACE facility.

2.4. Stomatal O₃ Flux Modelling

The hourly stomatal O₃ flux F_st (nmol m⁻² s⁻¹), considered in the POD₁ calculation, was obtained as follows:

$${\mathrm{F}}_{\mathrm{st}}=\left[{\mathrm{O}}_3\right]·\left\{\frac{1}{R_b+R_c}\right\}·\left\{\frac{g_{sto}}{g_{sto}+g_{ext}}\right\}$$

where [O₃] is the hourly O₃ concentration (ppb), R_b is the boundary layer resistance (s m⁻¹), R_c is the leaf surface resistance (s m⁻¹), g_sto is the stomatal conductance (mmol O₃ m⁻² s⁻¹), and g_ext is the cuticular conductance (m s⁻¹). In detail, R_b = 1.3 × 150 × (L_d/u)^0.5, where L_d is the leaf size; and u is the wind speed, expressed in m s⁻¹.

The stomatal conductance (g_sto) estimation was based on the multiplicative model described by Jarvis [34] and modified by CLRTAP [32] according to the following equation [15]:

g_sto = g_max × f_phen × f_light × max {f_min, (f_temp × f_VPD × f_SWC)}

where g_max represents the maximum stomatal conductance, while f_phen, f_light, f_temp, f_VPD, and f_swc are functions that indicate the variation of g_sto in relation to phenology, photosynthetic active radiation (PAR, μmol m⁻² s⁻¹), air temperature (T, °C), vapor pressure deficit (VPD, kPa) and soil moisture (SWC, m³ m⁻³), respectively. Finally, f_min represents the minimum stomatal conductance. The functions used are detailed below.

$$f_{light}=1-\exp \left(-{\mathrm{light}}_{\mathrm{a}}·\mathrm{PAR}\right)$$

where light_a is a dimensionless coefficient for curvature of stomatal conductance response to PAR.

$$f_{temp}=\left(\frac{\mathrm{T}-{\mathrm{T}}_{\min }}{{\mathrm{T}}_{\mathrm{opt}}-{\mathrm{T}}_{\min }}\right)\left\{{\left(\frac{{\mathrm{T}}_{\max }-\mathrm{T}}{{\mathrm{T}}_{\max }-{\mathrm{T}}_{\mathrm{opt}}}\right)}^{\left(\frac{{\mathrm{T}}_{\max }-{\mathrm{T}}_{\mathrm{opt}}}{{\mathrm{T}}_{\mathrm{opt}}{-\mathrm{T}}_{\min }}\right)}\right\}$$

where T_max, T_min, and T_opt are the maximum, minimum, and optimal temperatures for the stomata opening, respectively.

$$f_{VPD}=\min \left[1, \max \left\{f_{min},\left(\frac{\left(1-f_{min}\right)·\left({\mathrm{VPD}}_{\min }-\mathrm{VPD}\right)}{\left({\mathrm{VPD}}_{\min }-{\mathrm{VPD}}_{\max }\right)}\right)+f_{min}\right\}\right]$$

where VPD_min and VPD_max are, respectively, the vapor pressure deficit to reach the minimum and maximum stomatal opening.

$$f_{SWC}=\min \left[1, \max \left\{f_{min},\left(1-f_{min}\right)\left(\frac{\mathrm{SWC}-\mathrm{WP}}{\mathrm{FC}-\mathrm{WP}}\right)+f_{min}\right\}\right]$$

where SWC is soil moisture, WP is the wilting point, and FC is the field capacity obtained from Anav et al. [35]. Finally, f_phen was set equal to 1 during the growing season. The start of the growing season was based on a phenological model in the forest [25] and by direct observations in the FACE experiment. The end of the accumulation window was the time of the VFI survey.

We applied the species-specific stomatal conductance model parameters in the target species reported in our previous papers [15,23,36] and the CLRTAP mapping manual [32]. The applied parameters are listed in Supplementary Table S1.

2.5. Calculation of Leaf Mass per Area (LMA)

To assess the species-specific LMA, five leaf discs of 0.8 cm in diameter were obtained by a leaf punch (Fujiwara Scientific Company Co., Ltd., Tokyo, Japan) for each sampled light-exposed leaf (n = 5 leaves × 3 plants per site for MOTTLES sites, for 3 leaves × 1 to 2 plants per plot × 3 replicated plots × 2 to 3 O₃ levels for the FACE experiments). The leaf discs were oven-dried at 70 °C for 72 h and weighed by an analytical balance (Sartorius, Goettingen, Germany). The projected needle area of the samples was assessed by Easy Leaf Area application [37]. The LMA, expressed in g m⁻², was finally calculated as the ratio between dry biomass and area.

2.6. Surveys of O₃ Visible Foliar Injury

As suggested by the ICP Forests Manual for O₃ injury assessment [38], in the MOTTLES sites, O₃ VFI was determined along the LESS and inside the forest (ITP), i.e., at around 600 m from the OFD. Every year, at the end of the growing season, a team of two expert evaluators carried out an estimate of O₃ VFI (interveinal black/reddish necrosis in the upper leaf surfaces, [38]) at each LESS and ITP forest site (Supplementary Figure S1). The evaluation was performed on current-year light-exposed leaves/needles. The ozone VFIs were validated for each species based either on the literature or on one-year exposure in the O₃ FACE facility [23]. Concerning LESS, a forest edge of 50 m in length and 1 m in depth was examined. The area was divided into 25 adjacent and non-overlapping quadrates of 2 m² each, randomly excluding 5 quadrates, as Schaub et al. [38] suggested. The plant species were cataloged in each quadrate, and the presence/absence of O₃ VFI was recorded. For each species in the LESS, the following relative frequency of symptomatic species was calculated:

$$\mathrm{SS}\_\mathrm{LESS}=\frac{\mathrm{number}\mathrm{of}\mathrm{LESS}\mathrm{quadrates}\mathrm{with}\mathrm{symptomatic}\mathrm{species}}{\mathrm{total}\mathrm{number}\mathrm{of}\mathrm{LESS}\mathrm{quadrates}\mathrm{where}\mathrm{the}\mathrm{target}\mathrm{species}\mathrm{is}\mathrm{present}}\times 100$$

Within each ITP, sampling was performed on five randomly selected trees belonging to the dominant species. Depending on the species, 5 branches exposed to the light, with at least 30 leaves or needles, were evaluated. The Plant Injury Index (PII) was obtained by combining the percentage of injured leaves per shoot (LA) and the percentage of injured leaf surface per symptomatic leaf (AA), as suggested by Paoletti et al. [39]. Finally, an average was made for the five branches, thus obtaining an average percentage value of O₃ VFI per tree (PII_ITP).

$$\mathrm{PII}\_\mathrm{ITP}=\frac{\mathrm{LA}\times \mathrm{AA}}{100}$$

The Plant Injury Index was assessed in light-exposed leaves that were grown during the season, as well as for the plants in the O₃ FACE facility (PII_FACE) in September, except for Alnus glutinosa and Vaccinium myrtillus, for which the PII was assessed in July (Supplementary Figure S1).

2.7. Statistical Analyses

For the MOTTLES datasets, only forest sites with at least 75% validated data of hourly environmental parameters and O₃ concentrations were considered. POD₁, AOT40, and LIF were related to SS_LESS and PII_ITP for each site, or to PII_FACE for the FACE facility. Normal distribution was tested with the Lilliefors test (Kolmogorov–Smirnov). The significance (p < 0.05) of the linear regression lines was verified using the Pearson’s correlation coefficient (r). Moreover, we performed a one-way ANOVA to assess if O₃ significantly affected species-specific LMA values during the FACE experiment. Statistical analyses were conducted with MS Excel (Microsoft ^®).

3. Results

3.1. Meteorological Conditions and Ozone Concentrations

3.1.1. MOTTLES Sites

The highest daily mean O₃ concentration was detected in Abruzzo (ABR1) and was equal to 52.13 ppb during the period considered (2018–2022,

Table 2. Recorded annual averages of air temperature (T), relative humidity (RH), vapor pressure deficit (VPD), photosynthetically active radiation (PAR), soil water content at 10 cm depth (SWC), annual rainfall and daily O₃ concentration at the 9 Italian forest sites. Data refer to the period 2018–2022 ± standard error.

Site	T (°C)	RH (%)	VPD (kPa)	PAR (μmol m−2 s−1)	SWC (%)	Rainfall (mm)	O3 (ppb)
ABR1	7.67 ± 0.27	75.90 ± 2.00	0.31 ± 0.04	374.44 ± 10.67	28.90 ± 1.30	893.81 ± 157.73	52.13 ± 2.08
CPZ *	16.30 ± 0.25	78.60 ± 1.00	0.47 ± 0.04	412.73 ± 14.63	13.01 ± 0.84	623.95 ± 125.79	29.58 ± 2.16
EMI1	13.11 ± 1.11	74.00 ± 1.00	0.51 ± 0.09	314.36 ± 14.75	15.06 ± 0.65	793.43 ± 101.91	36.50 ± 2.14
LAZ1	13.54 ± 0.25	73.50 ± 2.00	0.52 ± 0.06	369.75 ± 11.64	17.38 ± 0.71	1304.82 ± 328.57	45.59 ± 1.53
PIE1	8.52 ± 0.96	74.20 ± 1.00	0.31 ± 0.02	305.80 ± 11.38	27.31 ± 1.79	1765.56 ± 363.59	50.20 ± 1.09
TRE1	5.25 ± 0.55	70.50 ± 1.00	0.29 ± 0.03	361.84 ± 21.82	39.45 ± 4.41	814.65 ± 113.49	45.61 ± 2.45
VEN1	7.56 ± 0.39	86.80 ± 1.00	0.17 ± 0.02	329.30 ± 13.79	40.38 ± 0.43	1703.58 ± 291.60	33.96 ± 0.94

* Data are the same for CPZ1, CPZ2, and CPZ3, except for SWC, as it is equal to 16.72 ± 1.83 in CPZ3.

). The lowest values were recorded in Castelporziano (CPZ1, CPZ2, and CPZ3), corresponding to 29.58 ppb. The highest average daily mean temperature (16.30 °C) was recorded in Castelporziano, while the lowest was in TRE1 (5.25 °C). Moreover, the daily mean photosynthetically active radiation was higher in Castelporziano (412.73 μmol m⁻² s⁻¹), but the lower value was at PIE1 (305.80 μmol m⁻² s⁻¹). The relative humidity and soil water content showed the highest values at VEN1 (86.80% and 40.38%, respectively), while the lowest values of VPD were found at VEN1 (0.17 kPa), followed by TRE1 (0.29 kPa), PIE1 (0.31 kPa) and ABR1 (0.31 kPa). Finally, the wettest sites were PIE1 and VEN1, with an average of more than 1700 mm of annual rainfall.

3.1.2. FACE Experiments

Table 3. Recorded daily mean ± standard error of air temperature (T), relative humidity (RH), vapor pressure deficit (VPD), photosynthetically active radiation (PAR), and daily O₃ concentration at FACE experiment for years 2016–2020. Plants were subjected to three levels of O₃ (AA—Ambient Air, 1.5 $\times$ Ambient Air, and 2.0 $\times$ Ambient Air).

Year	T (°C)	RH (%)	VPD (kPa)	PAR (μmol m−2 s−1)	O3 (ppb)
					AA	1.5$\times$	2.0$\times$
2016	22.90 ± 0.30	57.00 ± 3.00	1.50 ± 0.08	578 ± 13	34.7 ± 0.60	51.2 ± 0.90	66.1 ±1.00
2017	24.50 ± 0.40	47.60 ± 1.30	1.89 ± 0.08	564 ± 16	40.3 ± 1.20	51.7 ± 1.80	63.8 ± 2.30
2018	22.80 ± 0.29	55.60 ± 0.83	1.48 ± 0.04	527 ± 13	35.2 ± 0.70	53.1 ± 1.10	65.2 ± 1.40
2019	23.50 ± 0.35	55.10 ± 0.94	1.61 ± 0.06	548 ± 14	38.8 ± 0.90	56.2 ± 1.50	68.6 ± 1.80
2020	22.10 ± 0.39	61.80 ± 1.12	1.31 ± 0.05	475 ± 17	37.5 ± 1.10	52.3 ± 1.70	73.3 ± 2.30

shows the daily means of meteorological data and O₃ concentrations in the FACE experiments. The highest temperature and VPD were recorded in 2017 (24.50 °C and 1.89 kPa, respectively), while the lowest ones occurred in 2020 (22.10 °C and 1.31 kPa, respectively). The relative humidity showed the highest value in 2020 (61.80%) and the lowest one in 2017 (47.60%). The photosynthetically active radiation was always higher than 500 μmol m⁻² s⁻¹ (maximum value 578 μmol m⁻² s⁻¹ in 2016), except in 2020 (PAR = 475 μmol m⁻² s⁻¹). Regarding O₃, the Ambient Air concentration was in a range of 35–40 ppb for the reference years (2016–2020).

3.2. Leaf Mass per Area

3.2.1. MOTTLES Sites

The highest LMA values were found for the conifers Picea abies (262.10 g m⁻²) and Pinus pinea (192.47 g m⁻²; Supplementary Table S2). Among the deciduous trees, Fagus sylvatica showed the lowest values (between 50.42 and 67.01 g m⁻²), while the evergreens Quercus ilex (142.65 g m⁻²) and Phillyrea latifolia (144.64 g m⁻²) showed the highest values. For LESS species, Rubus spp. and Vaccinium myrtillus LMA ranged from 55 to 78 and from 35 to 56 g m⁻², respectively.

3.2.2. FACE Experiments

As confirmed in the MOTTLES sites, evergreen species showed a relatively high LMA value relative to deciduous species (Supplementary Table S3). The lowest LMA values were found in a fast-growing deciduous species, Alnus glutinosa (approximately 48 g m⁻²). For most of the species, O₃ did not significantly affect LMA values, except for Oxford poplar clone (Populus maximowiczii Henry × P. berolinensis Dippel) in 2016 (one-way ANOVA, p = 0.047) and Phillyrea angustifolia in 2018 (one-way ANOVA, p = 0.025).

3.3. Ozone Visible Foliar Injury at LESS, ITP and FACE

At the MOTTLES sites, Fagus sylvatica showed O₃ VFI within the LESS during all the years at VEN1, PIE1, and ABR1 (

Table 4. Relative frequency of symptomatic species within the LESS (SS_LESS) for the period 2018–2022.

Year	LESS Site	Fagus sylvatica	Rubus spp.	Vaccinium myrtillus	Picea abies
2018	VEN1	55
	ABR1	20
	PIE1	56
2019	VEN1	90
	ABR1	45
	PIE1	50
	TRE1			44
2020	VEN1	70
	ABR1	25
	PIE1	100
	LAZ1		10
	CPZ3		100
	TRE1			63
2021	VEN1	90
	ABR1	47
	PIE1	100	9
	EMI1		20
	LAZ1		57
	TRE1			79
2022	VEN1	100
	ABR1	40
	PIE1	40	64.2	66.6
	EMI1		66.6
	TRE1			100	16.6

). Within all the LESS quadrates, O₃ VFI was found in F. sylvatica leaves at PIE1 in both 2020 and 2021, and at VEN1 in 2022 (SS_LESS = 100%). High frequencies for this species (70–90%) were also found at VEN1 in 2019, 2020, and 2021, while the lowest percentages (20–47%) were recorded at ABR1. Concerning Rubus spp., SS_LESS = 100% was quantified at CPZ3 during 2020. On the contrary, for 2021 and 2022, no plants of Rubus spp. showed O₃ VFI at CPZ3 (SS_LESS = 0%). Injured Rubus plants were also found at LAZ1, PIE1, and EMI1 with SS_LESS equal to 64.2 and 66.6% in 2022 for the latter two sites, respectively. Injured Vaccinium myrtillus was observed in the LESS of PIE1 (in 2018, 2021, and 2022) and TRE1 (period 2019-2022). In particular, SS%_LESS increased from 2019 to 2022 at TRE1, passing from 44 to 100%, while O₃ VFI was found at PIE1 (66.6%) only in 2022. No dominant species in the LESS (P. pinea, P. abies, Q. petraea, Q. cerris, and P. latifolia) showed O₃ VFI, except for P. abies (SS_LESS = 16.6%) in 2022, at TRE1. Regarding ITP, O₃ VFI was found only in Fagus sylvatica, Pinus pinea, and Picea abies (

Table 5. Average percentage value ± standard error of O₃ VFI per tree (PII_ITP) for each species inside the forest plot (ITP; n = 5).

Year	ITP Site	Fagus sylvatica	Pinus pinea	Picea abies
2018	VEN1	7.41 ± 2.15
	ABR1	6.90 ± 3.08
	PIE1	2.20 ± 0.91
	CPZ3		0.44 ± 0.35
2019	VEN1	6.40 ± 3.60
	PIE1	1.50 ± 0.49
2020	VEN1	8.90 ± 1.99
	ABR1	0.30 ± 0.06
	PIE1	3.98 ± 1.63
2021	VEN1	4.90 ± 1.05
	ABR1	1.48 ± 0.56
	PIE1	0.41 ± 0.13
	TRE1			0.32 ± 0.10
2022	VEN1	7.12 ± 1.50
	ABR1	0.68 ± 0.29
	PIE1	1.85 ± 0.80

). The highest value of PII_ITP for F. sylvatica was 8.90% at VEN1 in 2020, while the lowest value (0.30%) was recorded at ABR1 in the same year. Pinus pinea showed an O₃ foliar injury at CPZ3 only in 2018, with a PII_ITP equal to 0.44%, while Picea abies at TRE1 showed a PII_ITP of 0.32%. The other dominant species, such as Quercus petraea, Q. ilex, Q. cerris, and Phyllirea latifolia did not show O₃ VFI during the ITP surveys in any year.

At the FACE, a relatively high PII was shown in deciduous species (

Table 6. Plant Injury Index (PII) for species subjected to fumigation in O₃ FACE (years 2016–2020). For each O₃ treatment (Ambient Air, 1.5$\times$, and 2.0$\times$), the species-specific PII (mean ± standard error) is shown.

Year	Species	O3 Treatment	PII (%)
2016	Populus maximowiczii Henry X P. × berolinensis Dippel	Ambient Air	1.52 ± 0.74
		1.5$\times$	5.79 ± 1.99
		2.0$\times$	9.07 ± 1.96
2017	Arbutus unedo	Ambient Air	0.01 ± 0.01
		1.5$\times$	0.01 ± 0.01
		2.0$\times$	0.08 ± 0.05
2018	Phillyrea angustifolia	Ambient Air	0
		1.5$\times$	0
		2.0$\times$	0
2018	Vaccinium myrtillus	Ambient Air	1.76 ± 0.64
		2.0$\times$	9.14 ± 0.35
2018	Sorbus aucuparia	Ambient Air	3.99 ± 0.71
		2.0$\times$	11.27 ± 2.40
2018	Alnus glutinosa	Ambient Air	0.66 ± 0.51
		2.0$\times$	7.04 ± 1.88
2019	Pinus pinea	Ambient Air	0
		1.5$\times$	0
		2.0$\times$	0
2020	Populus maximowiczii Henry X P. × berolinensis Dippel	Ambient Air	1.52 ± 0.79
		1.5$\times$	6.29 ± 2.02
		2.0$\times$	10.02 ± 2.11
2020	Populus x euramericana I-214	Ambient Air	0.09 ± 0.02
		1.5$\times$	1.88 ± 0.72
		2.0$\times$	4.02 ± 1.33

). The PII increased with increasing O₃ levels. The highest PII was shown in Sorbus aucuparia exposed to 2.0 $\times$ Ambient Air O₃ treatment, which was more than 11%. On the other hand, evergreen species showed a very low PII (e.g., P. pinea and P. angustifolia showed PII_FACE = 0).

3.4. AOT40 and POD₁ versus Visible Foliar Injury

The AOT40 was higher at ABR1 than at the other sites for three out of the five years, with a maximum of 54,588 ppb·h recorded in 2018 (Supplementary Table S4 and Figure 1). The lowest values were at TRE1 (2019–2021) and CPZ (2018 and 2022), with a minimum of 8377 ppb·h at CPZ in 2022. POD₁ did not depend on site conditions or on the species. For F. sylvatica, the highest POD₁ values were at VEN1 (30.88 mmol m⁻² in 2018), and the lowest ones were at ABR1 (6.43 mmol m⁻²). High POD₁ values for this species at VEN1 were associated to the highest PII_ITP among all sites and years.

Concerning Rubus spp., the highest POD₁ (14.69 mmol m⁻²) and 100% SS_LESS were at CPZ3 in 2020. For V. myrtillus, POD₁ values were always high at TRE1 with a maximum of 28.47 mmol m⁻² in 2021, where SS_LESS values were also high. Interestingly, Picea abies showed a relatively high POD₁ value at TRE1 (18.12–33.01 mmol m⁻²). However, visible foliar injuries were rarely seen for this species, as confirmed by very low values of both SS_LESS and PII_ITP.

At the FACE, AOT40 values in Ambient Air conditions were 13,774 to 24,243 ppb·h (Supplementary Table S5) and increased with increasing enhanced O₃ levels. POD₁ values were species-specific, with a maximum in Oxford poplar clone (75.15 mmol m⁻²) and a relatively low value (2.79–6.13 mmol m⁻²) for the Mediterranean evergreen shrub Arbutus unedo.

The relationship between species-specific POD₁ calculated for LESS-species and their relative frequency of O₃ VFI (SS_LESS) was statistically significant (p = 0.005, r = 0.37; Figure 2A), while AOT40 and SS_LESS were not correlated (r = 0.11, p = 0.40). The percentage of ITP O₃ VFI (PII_ITP) increased with the POD₁ values calculated for all the dominant species at each site (r = 0.53, p < 0.001; Figure 2C), while AOT40 was not correlated with PII_ITP (r = 0.03, p = 0.85). Such a tendency was also confirmed in the FACE, where the correlation between PII_FACE and AOT40 was not significant, while it was for POD₁ (Figure 3A,B). Moreover, AOT40 did not correlate with POD₁ (r = 0.003, p = 0.97).

3.5. LIF versus Visible Foliar Injury

The LIF was calculated for each species and was significantly correlated to SS_LESS (r = 0.63, p < 0.001; Figure 2B) and PII_ITP (r = 0.87, p < 0.001; Figure 2D). Moreover, a significant strong positive correlation (r = 0.70, p < 0.001) between POD₁ and LIF was detected. Conversely, AOT40 did not correlate with the LIF (r = 0.03, p = 0.74). Interestingly, the Plant Injury Index (PII_FACE) calculated at the end of growing season for species exposed to O₃ in the FACE facility confirmed that the LIF had a higher correlation (r = 0.67, p < 0.001; Figure 3C) with PII_FACE than with AOT40 (r = 0.35, p = 0.09; Figure 3A) and POD₁ (r = 0.58, p = 0.003; Figure 3B).

4. Discussion

4.1. Different Sensitivity to O₃ among Species according to Visible Foliar Injury

For the dominant tree species inside the forest plots considered in this study (ITP), Fagus sylvatica was the most sensitive species, as it showed O₃ VFI at three sites (ABR1, PIE1, and VEN1) in the period of 2018–2022. Our results are in agreement with those of previous studies carried out in nurseries [40], FACE facilities [41], and open-top chambers [42], denoting that beech is characterized by low O₃ tolerance and the early appearance of O₃ VFI, mainly identifiable as browning and dark spots. Conifers were better at tolerating the abiotic stress due to O₃. In fact, O₃ VFI was found only in 2018 at CPZ3 for P. pinea and in 2021 at TRE1 for P. abies. The high tolerance for P. pinea was confirmed in the FACE, where O₃ VFI was not observed even under twice the ambient O₃ levels. The dominant broadleaf species present in the other forest sites located in Central–Southern Italy, such as Quercus ilex, Quercus cerris, Quercus petraea, and Phillyrea latifolia, did not show any O₃ VFI. First of all, those representative species of the Mediterranean area evolved in an environment constantly subjected to other oxidative stresses, such as long dry periods and high solar irradiation, and are likely endowed with an efficient antioxidant pool for the detoxification of O₃ inside the leaf parenchyma [43]. Experiments carried out in open-top chambers [44] also reported that Quercus ilex plants were asymptomatic even after two years of exposure to high O₃ concentrations (approximately daily mean of 40 ppb). In the LESS, Fagus sylvatica showed O₃ VFI, although young trees, usually present at forest edges, may be more sensitive to O₃ than adult ones [45] due to lower stomatal conductance of mature trees than juvenile trees caused by a greater resistance to height-dependent water transport [46,47]. In addition, shrub species (Rubus spp.) and underbrush plants (Vaccinium myrtillus) were symptomatic during LESS surveys. A relatively high sensitivity of V. myrtillus in terms of O₃ VFI was also confirmed in the FACE. A long-term exposure to high concentrations of O₃ also resulted in an alteration of the yield and nutraceutical quality of fruits, suggesting that Vaccinium myrtillus is sensitive to O₃ [48]. In addition to the above-mentioned species, according to the FACE experiments, Sorbus aucuparia and poplar species were considered to be O₃-sensitive, as they showed a high PII due to a high stomatal O₃ uptake.

4.2. Which Is the Best Index for Ozone Visible Foliar Injury

Concerning the O₃ indices investigated in this work, several studies suggest that AOT40 can provide misleading results if used as the only reference metric for forest protection from O₃ (e.g., Anav et al. [34]). Confirming this, although AOT40 values in our target sites were always much higher than the regulatory limit for the protection of forests set by CLRTAP [32] at 5000 ppb h, no significant correlation was found between AOT40 and O₃ VFI. On the contrary, POD₁ increased with the increase of both injury parameters, highlighting that a flux-based index is more appropriate than AOT40 for the definition of species-specific critical levels and the consequent protection of forests. Indeed, asymptomatic Phillyrea latifolia and Quercus ilex had relatively low POD₁ values in Castelporziano (CPZ1 and CPZ2); the values were lower than the critical level indicated by Sicard et al. [49] for hardwoods, i.e., 12 mmol m⁻², while Fagus sylvatica and Vaccinium myrtillus regularly exceeded this POD₁ critical level and showed VFI in Northern Italian sites (PIE1 and VEN1). Confirming this, the two poplar clones tested in the FACE experiment, i.e., Populus maximowiczii Henry X P. × berolinensis Dippel and Populus x euramericana I-214, characterized by very high stomatal conductance (g_max equal to 348 and 478 mmol O₃ m⁻² PLA s⁻¹, respectively), exceeded the critical threshold of 12 mmol m⁻² (POD₁ about 2–3 times higher in Ambient Air treatment), resulting in severe damage, as indicated by PII, while Arbutus unedo (g_max equal to 95 mmol O₃ m⁻² PLA s⁻¹) exhibited low POD₁ and limited symptoms (PII = 0.01–0.08%). However, several species showed very few symptoms, although they had a relatively high value of POD₁ (18.12–33.01 mmol m⁻² for P. abies, 7.73–16.51 mmol m⁻² for Q. petraea, and 8.89–15.66 mmol m⁻² for Q. cerris in the years considered) that was, on average, higher than the critical level of 5 and 12 mmol m⁻² recommended for conifers and broadleaves, respectively [49]. This study thus developed and tested the LIF, an index that explores the sensitivity to O₃ of forest species as a function of not only stomatal flux as a proxy of avoidance capacity but also LMA, a representative parameter of the leaf morphology as a proxy of defense capacity against O₃. Interestingly, in forest-monitoring data, we found a strong significant correlation (p < 0.001) of O₃ VFI (SS_LESS and PII_ITP) with LIF, even higher than with POD₁. Moreover, we confirmed that LIF fitted better than the other two indices also in experimental conditions with O₃-fumigated plants in a manipulative FACE experiment. Plant resistance to oxidative stress due to O₃ exposure can be explained not only by avoidance (restriction of O₃ entry via stomata) but also by tolerance (biochemical defense and repair) [50]. Since LIF considered both avoidance and tolerance against O₃, it is proposed as an interesting and effective new proxy for evaluating O₃ damages on forest species. Species characterized by high LMA values (and consequent low LIF), such as Picea abies, Pinus pinea, and the Mediterranean sclerophyllous Quercus ilex and Phyllirea latifolia may cope well with the oxidative stress caused by O₃. Several studies reported that leaves characterized by high LMA values possess high antioxidant capacities [30,44,51]. In particular, Mediterranean sclerophyllous plants, which are often suffering from drought, high irradiance, and/or O₃, have a well-developed mechanism of antioxidative protection to withstand oxidative stress [52,53]. For example, O₃-exposed Mediterranean oaks activated the phenylpropanoid pathways to counteract the accumulation of ROS and thus reduce the oxidative damage to physiological functions [54]. Moreover, other studies suggested that a leaf structure characterized by small, thick, and leathery leaves may contribute to resist O₃ stress [27,55]. Those leaves have a relatively low intercellular air space and, thus, a limited cell surface interaction with O₃ [29]. In addition, the presence of a dense layer of trichomes can often be found on the lower leaf blade, and it could increase the reaction surface with O₃, contributing to its degradation before O₃ entry into a leaf [44,56,57].

5. Conclusions

The O₃ VFI (SS_LESS, PII_ITP, and PII_FACE) increased with increasing POD₁ and did not vary with AOT40, suggesting that the new regulations for forest protection should be based on O₃ stomatal flux rather than O₃ exposure. Moreover, the new index (LIF) proposed here, which takes into account O₃ stomatal flux and leaf morphology, showed excellent results (i.e., the highest correlation coefficients with O₃ VFI both in forest sites and FACE facility) and deserves to be further explored to assess O₃ damage on forest species. To calculate the LIF index, species-specific stomatal conductance model parameters and LMA values are required. These parameters are still missing for many species, especially for shrub species such as Viburnum lantana [58] and Rosa canina [23], which often show O₃ VFI in the field. Therefore, for the practical use of LIF, further studies are needed to develop stomatal conductance models and calculate LMA for a larger array of species. Ozone is confirmed as a real and alarming threat to the health of Italian forests. Indeed, characteristic species such as Fagus sylvatica showed important O₃ VFI, which can translate into a decline in growth and productivity, with heavy repercussions at the ecosystem level. Conversely, conifers (Pinus pinea and Picea abies), as well as species belonging to the Mediterranean area (Quercus spp. and Phillyrea latifolia), did not show O₃ VFI, but it cannot be excluded that O₃ could affect other parameters, for example, canopy defoliation or growth reduction.