Differential regulation of volatile emission from Eucalyptus globulus leaves upon single and combined ozone and wounding treatments through recovery and relationships with ozone uptake

Abstract

Both ozone and wounding constitute two key abiotic stress factors, but their interactive effects on plant constitutive and stress-elicited volatile (VOC) emissions are poorly understood. Furthermore, the information on time-dependent modifications in VOC release during recovery from a combined stress is very limited. We studied the modifications in photosynthetic characteristics and constitutive and stress-induced volatile emissions in response to single and combined applications of acute ozone (4, 5, and 6 ppm) and wounding treatments through recovery (0.5-75 h) in a constitutive isoprene and mono- and sesquiterpene emitter Eucalyptus globulus. Overall, the photosynthetic characteristics were surprisingly resistant to all ozone and wounding treatments. Constitutive isoprene emissions were strongly upregulated by ozone and combined ozone and wounding treatments and remained high through recovery phase, but wounding applied alone reduced isoprene emission. All stress treatments enhanced emissions of lipoxygenase pathway volatiles (LOX), mono- and sesquiterpenes, saturated aldehydes (C7-C10), benzenoids, and geranylgeranyl diphosphate (GGDP) pathway volatiles. Once elicited, GGDP volatile, saturated aldehyde and benzenoid emissions remained high through the recovery period. In contrast, LOX emissions, and total mono- and sesquiterpene emissions decreased through recovery period. However, secondary rises in total sesquiterpene emissions at 75 h and in total monoterpenes at 25-50 h were observed. Overall, acute ozone and wounding treatments synergistically altered gas exchange characteristics and stress volatile emissions. Through the treatments and recovery period, stomatal ozone uptake rate and volatile emission rates were poorly correlated, reflecting possible ozone-scavenging effect of volatiles and thus, reduction of effective ozone dose and elicitation of induced defense by the acute ozone concentrations applied. These results underscore the important role of interactive stresses on both constitutive and induced volatile emission responses.

1. Introduction

Tropospheric ozone (O₃) is a major oxidative pollutant and a key plant stress elicitor (Karnosky et al., 2007). The bulk of tropospheric ozone is formed in the reactions involving nitrogen oxides NO and NO₂ (NO_X) and reactive volatile organic compounds (VOCs) in the presence of sunlight (Ryerson et al., 2003). Elevated ozone generates oxidative stress in plants that leads to biochemical adjustments and metabolic shifts as a result of ozone-induced gene expression changes and acclimation responses or hypersensitive or necrotic responses associated with foliar injury, impairment of shoot and root growth, and accelerated organ senescence (Calfapietra et al., 2008; Gerosa et al., 2009; Heath, 2008; Peñuelas and Staudt, 2010).

The current surface ozone level is around 40 ppb in most parts of the world (Sicard et al., 2017). Although it varies with the geographical location, the present tropospheric ozone levels are capable of causing physiological damage in plants (Ashworth et al., 2013; Proietti et al., 2016). Furthermore, it is anticipated that tropospheric ozone concentrations will increase by 2-4 folds in the next two decades, primarily due to industrialization and burning of fossil fuels, implying that ozone stress is expected to become more severe in the future (Vingarzan, 2004).

Plants are a significant sink for atmospheric ozone both due to stomatal ozone uptake and non-stomatal ozone deposition (Fares et al., 2008). Stomatal uptake is the primary passage through which ozone enters the leaf intercellular spaces and generates physiological and oxidative damage in plant cells (Beauchamp et al., 2005; Gerosa et al., 2007). Apart from alteration of basic metabolic processes such as photosynthesis and plant growth, ozone exposure leads to major changes in plant volatile emission rates and emission profiles during the initial stress impact and through recovery. Upon acute ozone exposures, damaged plant cells release free fatty acids from their membranes, leading to an activation of lipoxygenase pathway (LOX) and rapid emission burst of LOX volatiles (also called green leaf volatiles, GLV) (Beauchamp et al., 2005; Copolovici et al., 2014; Portillo-Estrada et al., 2015). These early stress responses ultimately trigger the signal transduction pathways that activate defense reactions primarily through jasmonate (JA) and ethylene regulated transcription factors (Bailey et al., 2005). Typically, ozone-caused longer-term responses include elicitation of terpenoid and benzenoid emissions for hours to days following the initial stress impact (Beauchamp et al., 2005).

In nature, plants often encounter multiple stress factors simultaneously or in sequence. Interacting stresses can strengthen or weaken the effect of individual stress factors due to modification of overall stress severity or complex stress-priming responses (Copolovici et al., 2014; Ibrahim et al., 2008; Niinemets, 2010a, b). Leaf wounding is a key mechanical stress in the field that primarily results from herbivore feeding, but also from mechanical damage due to wind, falling debris and heavy precipitation (Benikhlef et al., 2013; Portillo-Estrada et al., 2015). Similarly to ozone, wounding leads to emissions of LOX volatiles that are quantitatively related to the degree of damage (Copolovici and Niinemets, 2015; Copolovici et al., 2017; Portillo-Estrada et al., 2015).

In fact, LOX emission is a very characteristic early stress response, followed by volatile isoprenoids, emissions of which have been detected upon many abiotic stresses such as heat shock in Solanum lycopersicum (Pazouki et al., 2016), drought and herbivory in Alnus glutinosa (Copolovici et al., 2014) and ozone in Nicotiana tabacum (Beauchamp et al., 2005), and biotic stresses such as feeding by larvae of common white wave (Cabera pusaria) in leaves of Alnus glutinosa (Copolovici et al., 2011) and leaf rust infection in Salix spp. (Toome et al., 2010) and Populus (Jiang et al., 2016).

Previous studies have revealed that the impacts of elevated ozone (Beauchamp et al., 2005; Llusià et al., 2002; Loreto et al., 2001; Loreto et al., 2004; Peñuelas et al., 1999; Velikova et al., 2005), and wounding (Brilli et al., 2011; Loreto and Sharkey, 1993; Portillo-Estrada et al., 2015) alter primary and secondary metabolic process of plants and especially, LOX, MEP/DOXP (2-C-methyl-D-erythritol 4-phosphate/1-deoxy-D-xylulose 5-phosphate) and MVA (mevalonate) pathways responsible for volatile emission. In all these studies, volatile emissions were quantitatively associated with the stress severity, characterized as the spread of damage or degree of biological infection, but it is unclear how combined application of two different stresses capable of causing cell-level injury can simultaneously affect plant volatile emission. Although both ozone and wounding stresses are known to induce LOX emissions, and ozone exposure further elicits isoprenoid emissions during recovery, it is also unclear whether wounding alone can also elicit volatile isoprenoid emission responses and whether wounding following ozone exposure amplifies the volatile emission responses during recovery. Furthermore, not only do the past studies lack the interaction effects, but they have typically analyzed immediate plant responses to stress rather than plant responses through recovery phase. Interaction studies are pertinent because in natural environments, occurrence of multiple stress factors with varying strength and duration is common and multiple stresses can significantly influence tree physiological responses through additive, synergistic, and antagonistic effects. In old-growth forest ecosystems, for example, low understory light levels combined with low soil nutrient and water availabilities can frequently limit understory regeneration (Bergh et al., 1999; Niinemets, 2010b). In addition to interacting stresses, an already existing stress can be superimposed by another stress factor, further complicating the stress situation. Due to stress priming and acclimation responses, the effects of superimposed and successive environmental stresses are often not additive, but different stresses can interact, leading to synergistic or antagonistic responses. The interactive effects on plant performance can either be negative, indicating an enhanced plant response to the given stressor due to an additional stress, or positive, implying a reduced plant response due to an additional stress (Niinemets, 2010b). Thus, the plant responsiveness to the given stress varies with the type and duration of the combined stresses imposed. In this study, wounding is a disturbance stress defined as a single episodic event or chronic impact that is associated with full or partial destruction of plant biomass, impairing plant physiological and functional activities such as photosynthesis and growth, whereas ozone is an oxidative stress that leads to a sustained deviation from optimal environmental conditions, causing reduced plant productivity and growth rates (Bansal et al., 2013). Therefore, ozone and wounding treatments are expected to lead to synergistic effects on plant functional activity. Especially for ozone, this is relevant as early and late stress responses can be qualitatively different (Calfapietra et al., 2013; Niinemets, 2010a).

Eucalyptus spp. are the dominant plant species in Australian forests and woodlands. Due to their high growth rates, they are considered economically highly valuable hardwoods for pulp industry (Külheim et al., 2015), and are therefore, widely grown in forest plantations around the world especially in tropical, sub-tropical and warm temperate regions (Loreto et al., 2000). Eucalypts are significant isoprene and monoterpene emitters under non-stressed conditions (Funk et al., 2006; Guenther et al., 1991; He et al., 2000; Loreto et al., 2000; Street et al., 1997; Winters et al., 2009). In this study, we used Tasmanian bluegum (Eucalyptus globulus Labill.) that is a classic species used in constitutive isoprenoid emission studies (Guenther et al., 1991; Loreto et al., 2000). However, much less is known of stress responses of volatile emissions in eucalypt species. We investigated how acute ozone and wounding stresses alone and in interaction change foliage photosynthetic characteristics, and volatile emissions through different recovery times in E. globulus leaves. Since plant-produced volatiles can quench ozone in ambient air and in leaf intercellular air spaces due to direct reaction with ozone (Fares et al., 2010), eucalypts are expected to be highly resistant to ozone stress, and we used acute ozone exposures in this study.

We addressed the following questions: (1) how do individual and interactive effects of acute ozone and wounding influence foliage photosynthetic characteristics, overall emission amounts and blend of emissions through recovery? (2) how are the quantitative emission responses through recovery phase associated with stress severity, including ozone dose and stomatal ozone uptake rate? Both the plant responses to immediate ozone and wounding application, and recovery of those responses upon returning the plants to ambient environment after stress were analyzed to gain conclusive insight into the correspondence among photosynthetic characteristics, ozone and wounding treatments, stomatal ozone uptake rates and volatile emissions. We hypothesized that (1) single and combined applications of acute ozone and wounding stresses would result in a major reduction in foliage photosynthesis rate, stomatal conductance to water vapor, and reduction in maximum quantum yield of photosystem II (PSII, F_v/F_m); (2) application of ozone and wounding treatments alone would lead to a strong emission response of LOX and volatile isoprenoids, saturated aldehydes, benzenoids, and GGDP pathway volatiles and that these responses scale with ozone dose and are greater than the responses due to wounding; (3) the combined application of ozone and wounding leads to synergistic effects on gas exchange and VOC emission responses; and (4) as E. globulus is a strong isoprenoid emitter, a major proportion of volatile isoprenoids and antioxidants scavenges ozone in leaf intercellular airspaces and ROS in plant cells, eventually resulting in a poor relationships between stomatal ozone uptake vs. volatile emissions. To our knowledge, the sequence of events from elicitation to release of different volatiles upon application of individual ozone and wounding stresses through recovery and the interactive effects of combined ozone and wounding stresses on volatile emissions have not been studied yet, and it is also unclear how these responses are mediated by stomatal ozone uptake.

2. Materials and Methods

2.1. Plant material and growth conditions

Seeds of E. globulus were obtained from OMC seeds Ltd. (Lithuania) and sown in 5 L pots filled with a soil consisting of 1:1 mixture of quartz sand and commercial potting soil (Biolan Oy, Kekkilä group, Finland) that included essential macronutrients N (100 mg L^-1), P (30 mg L^-1), and K (200 mg L^-1) and micronutrients. The pH of the soil water was 6.2. Seedlings were grown for 3 weeks in a growth chamber (Percival AR-95 HIL, CLF Plant Climatics GmbH, Wertingen, Germany) under controlled environmental conditions of light intensity at leaf surface of 500 µmol m^-2 s^-1 provided for 12 h photoperiod, chamber temperature (day/night) of 28/25 °C, ambient CO₂ concentration of 380-400 ppm, and relative air humidity of 60-70%. The 3-week-old seedlings were transplanted into 10 L pots filled with the same potting mixture and the plants were further grown in controlled-conditions in a plant growth room under the light intensity of 400-500 µmol m^-2 s^-1 provided by HPI-T Plus 400 W metal halide lamps (Philips) for 12 h photoperiod, day / night temperature of 28/25 °C, ambient CO₂ concentration of 380-400 ppm, and relative air humidity of 60-70% until the completion of the experiment. Plants were watered every two days to soil field capacity and fertilized once a week with a liquid fertilizer (Baltic Agro, Lithuania; NPK content ratio: 5:5:6; and micronutrients B (0.01%), Cu (0.03%), Fe (0.06%), Mn (0.028%), and Zn (0.007%)). Each plant was supplied with 80 ml diluted liquid fertilizer (ca. 0.4% solution) to ensure optimum growth. One-year-old eucalypt plants were used in this experiment. In all experimental treatments, ca. 1 m tall plants with similar biomass, stem thickness, and fully mature leaves were used.

2.2. Experimental system for gas-exchange measurements, volatile collection and ozone fumigation

A temperature controlled, custom-made double-layered cylindrical glass chamber (1.2 L) with glass bottom was used for ozone fumigation, gas-exchange measurements, and volatile samplings (Copolovici and Niinemets, 2010, for details). The chamber temperature was maintained at 25 ºC by circulating water between the double layers of the glass chamber; the temperature of circulating water was controlled by a thermostat. Four 50 W halogen lamps were used for chamber illumination, and the light intensity at the leaf level was controlled by a regulatory unit. Air temperature inside the chamber was monitored by a thermistor (NTC thermistor, model ACC-001, RTI Electronics, Inc., St. Anaheim, CA, USA) and leaf temperature was measured by a thermocouple attached to the lower leaf surface. Ambient air was purified by passing through a charcoal-filled filter and a custom-made ozone trap. The chamber was equipped with a fan (Sunon Group, Beijing, China), yielding a high air turbulence at relatively low wind speeds, ensuring high leaf boundary layer conductance and no pockets of still air in the chamber. Chamber glass bottom had openings for plant stem or leaf petiole and for sensors. Upon plant enclosure, the site of leaf insertion was sealed with low-emitting modeling clay to prevent any gas leakage from the chamber. All chamber connections were made of stainless steel and Teflon® was used for all the tubing.

The chamber inlet and outlet ports can be switched between the reference and measurement modes for gas exchange measurements and for volatile sampling. CO₂ and H₂O concentrations at chamber in- and outlets were measured by an infrared dual-channel gas analyzer operated in differential mode (CIRAS II, PP-systems, Amesbury, MA, USA). Upon branch enclosure, steady-state foliage net assimilation and transpiration rates were usually achieved in ca. 10-15 min after leaf enclosure in the chamber. When required, ozone was produced by a Certizon C100 ozonizer (Erwin Sander Elektroapparatenbau GmbH, Germany), and ozone concentration in the chamber in- and outlets was monitored with a UV photometric ozone detector (Model-49i, Thermo Fisher Scientific, Franklin MA, USA). A detailed description of the gas-exchange system setup is provided by Copolovici and Niinemets (2010), see also Niinemets et al. (2011, for a comparison of different gas-exchange systems).

Stainless steel adsorbent cartridges were used to collect volatiles from the leaf chamber using a portable pump 210-1003MTX (SKC Inc., Houston, TX, USA) operated with a constant suction flow rate of 200 mL min^-1. A detailed description of adsorbent cartridges and the method of volatile collection can be found in Kännaste et al.(2014) and Niinemets et al (2010a). Air samples were also taken without the leaf to estimate chamber volatile background.

2.3. Experimental treatments

Mature leaves of E. globulus are comparatively thick, and this species is a very high isoprenoid emitter under control conditions. Preliminary experiments demonstrated that ozone was substantially scavenged and eucalypt leaves showed high resistance to acute ozone exposures. In fact, exposure to moderate to relatively high ozone concentrations of 0.3-2 ppm did not result in elicitation of LOX volatiles, and did not cause significant differences in photosynthetic characteristics (F_v/F_m, net assimilation rate and stomatal conductance to water vapour). Therefore, we decided to use higher acute ozone exposures. Four different sets of experiments were made: control (no treatment), wounding, ozone exposure (three different fumigations of 4, 5 and 6 ppm) and ozone exposure followed by wounding (the same three different ozone exposures). In all cases, a randomly selected branch consisting of six fully mature leaves was used, and standard environmental conditions maintained in the glass chamber were: light intensity at the leaf surface of 700-750 µmol m^-2 s^-1, chamber temperature of 25 °C (leaf temperature of 25-27 °C), ambient CO₂ concentration of 380-400 µmol mol^-1, and relative air humidity of 60-70%. In the case of wounding treatment, prior to leaf enclosure, four holes were rapidly (within 6 s) punched in the leaf lamina by a paper punch. The area of each disc removed was ca. 25 mm² and the total perimeter length of four discs (wound edge) was 7 cm per leaf (Portillo-Estrada et al., 2015, for detailed methodology). For ozone treatments alone, the branch was exposed to either 4, 5, or 6 ppm ozone for 3 hours. In the case of combined ozone and wounding treatment, the branch was taken out after ozone fumigation, and four punch holes were created in each leaf as for the treatment with wounding alone.

2.4. Gas-exchange and volatile measurement protocol and GC-MS analysis

Once the net assimilation and transpiration rates reached a steady-state, they were recorded and volatile sampling was carried out. After the initial measurement, the measurements for treated samples were repeated in 0.5, 3, 10, 25, 50, and 75 h after stress application. Separate measurements indicated that foliage gas-exchange and volatile emission rates of control samples were constant (variation less than 5%) during the measurement period.

The adsorbent cartridges were analyzed with a combined Shimadzu TD20 automated cartridge desorber and Shimadzu 2010 Plus GC–MS system (Shimadzu Corporation, Kyoto, Japan) for quantitative analysis of lipoxygenase pathway (LOX) volatiles, saturated aldehydes, volatile isoprenoids, benzenoids and geranylgeranyl diphosphate pathway (GGDP) volatiles (Kask et al., 2016 for definition of the compound classes). As lower molecular mass saturated aldehydes, pentanal (C5) and hexanal (C6) might partly reflect the LOX pathway activity (Heiden et al., 2003; Wildt et al., 2003), for the compound class “saturated aldehydes” we considered aliphatic saturated aldehydes with more than six carbon atoms (Kask et al., 2016). The volatile compounds were identified based on pure standards (Sigma-Aldrich, St. Louis, MO, USA) and NIST 05 library. For further details of volatile sampling and GC-MS analysis we refer to Copolovici et al. (2009) and Kännaste et al. (2014). A full list of compounds detected in control and treated leaves (wounding, 5 ppm ozone, and 5 ppm ozone and wounding) is provided in Table 1.

Table 1.

Average (±SE) emission rates (nmol m^-2 s^-1) of stress volatiles released from the leaves of E. globulus in control, wounding, 5 ppm ozone, and 5 ppm ozone and wounding treatments at 0.5 h of recovery since stress applications. Compounds were grouped according to compound biosynthesis pathways.

		Emission rate (nmol m-2 s-1) of volatiles at 0.5 h since stress applications
No	Compound	Control	Ozone 5 ppm	Wounding	Ozone 5 ppm and wounding
1	Isoprene	1.4 ± 0.8	1.2 ± 0.4	0.2 ± 0.1	14 ± 6
LOX volatiles
2	Hexanal	0.004 ± 0.001	0.06 ± 0.04	0.010 ± 0.001	0.200 ± 0.014
3	Pentanal	0.005 ± 0.004	0.02 ± 0.01	0.010 ± 0.002	0.030 ± 0.001
4	(E)-2-Hexenal	nd	0.010 ± 0.004	0.010 ± 0.005	0.400 ± 0.030
5	(Z)-3-Hexenol	nd	0.010 ± 0.004	0.020 ± 0.003	0.45 ± 0.03
Monoterpenes
6	Camphene	0.003 ± 0.002	0.13 ± 0.10	0.10 ± 0.02	1.6 ± 0.7
7	Δ3-Carene	0.004 ± 0.001	0.03 ± 0.02	0.06 ± 0.03	0.20 ± 0.05
8	1,8-Cineole	0.130 ± 0.001	15 ± 12	3.0 ± 0.4	65 ± 30
9	p-Cymene	0.003 ± 0.001	0.4 ± 0.2	0.02 ± 0.01	4.0 ± 2.4
10	α-Fenchene	0.010 ± 0.001	0.03 ± 0.02	0.010 ± 0.004	0.4 ± 0.2
11	Limonene	0.02 ± 0.01	3 ± 2	0.80 ± 0.04	25 ± 11
12	β-Myrcene	0.005 ± 0.001	0.5 ± 0.4	0.2 ± 0.1	4.4 ± 3.5
13	(Z)-β-Ocimene	nd	0.2 ± 0.2	0.006 ± 0.004	1 ± 1
14	α-Phellandrene	0.001 ± 0.001	0.2 ± 0.2	0.003 ± 0.002	3 ± 2
15	α-Pinene	0.03 ± 0.02	6 ± 4	1.0 ± 0.2	37 ± 15
16	β-Pinene	0.006 ± 0.002	0.2 ± 0.2	0.02 ± 0.01	4 ± 2
17	α-Terpinene	0.005 ± 0.003	0.01 ± 0.01	0.10 ± 0.02	0.5 ± 0.2
18	γ-Terpinene	0.004 ± 0.002	0.3 ± 0.2	0.10 ± 0.01	1.3 ± 0.7
19	α-Terpinolene	0.004 ± 0.004	0.010 ± 0.006	0.10 ± 0.02	2.0 ± 0.1
20	α-Thujene	0.001 ± 0.001	0.10 ± 0.05	0.020 ± 0.006	1.0 ± 0.6
Sesquiterpenes
21	Alloaromadendrene	0.001 ± 0.001	0.2 ± 0.2	0.01 ± 0.01	0.7 ± 0.4
22	Aromadendrene	0.003 ± 0.001	2.0 ± 1.3	0.103 ± 0.080	4 ± 2
23	(E)-β-Caryophyllene	nd	0.2 ± 0.2	0.010 ± 0.006	0.1 ± 0.1
24	α-Copaene	0.001 ± 0.001	0.10 ± 0.04	0.010 ± 0.003	0.3 ± 0.2
25	Epiglobulol	0.001 ± 0.001	0.10 ± 0.05	0.001 ± 0.001	0.24 ± 0.20
26	α-Gurjunene	0.001 ± 0.001	0.4 ± 0.4	0.04 ± 0.02	1.6 ± 1.3
27	β-Gurjunene	0.0003 ± 0.0002	0.020 ± 0.001	0.001 ± 0.001	0.13 ± 0.03
30	γ-Gurjunene	0.0002 ± 0.0001	0.10 ± 0.06	0.010 ± 0.001	0.13 ± 0.05
31	Isoledene	0.0004 ± 0.0001	0.1 ± 0.1	0.010 ± 0.003	0.4 ± 0.3
32	γ-Muurolene	0.0010 ± 0.0001	0.1 ± 0.1	0.010 ± 0.001	0.2 ± 0.1
33	Viridiflorene	0.002 ± 0.001	0.6 ± 0.5	0.10 ± 0.04	1 ± 1
Saturated aldehydes
34	Decanal	0.003 ± 0.001	0.3 ± 0.1	0.002 ± 0.001	0.3 ± 0.1
35	Heptanal	0.0010 ± 0.0001	0.05 ± 0.03	0.0010 ± 0.0004	0.04 ± 0.02
36	Nonanal	0.003 ± 0.002	0.3 ± 0.2	0.004 ± 0.002	0.30 ± 0.02
37	Octanal	0.002 ± 0.001	0.1 ± 0.1	0.002 ± 0.002	0.04 ± 0.03
Benzenoids
38	Benzaldehyde	0.0020 ± 0.0001	0.02 ± 0.02	0.010 ± 0.002	0.65 ± 0.25
39	o-Xylene	0.010 ± 0.001	0.01 ± 0.01	0.020 ± 0.001	0.001 ± 0.030
40	p-Xylene	0.010 ± 0.006	0.02 ± 0.01	0.020 ± 0.001	0.001 ± 0.010
GGDP pathway volatiles
41	6-Methyl-5-hepten-2-one	0.0020 ± 0.0001	0.1 ± 0.1	0.005 ± 0.001	0.3 ± 0.3
42	Geranyl acetone	0.030 ± 0.002	0.3 ± 0.1	0.10 ± 0.01	0.5 ± 0.3

The leaves were scanned at 300 dpi to measure leaf area using a custom made software tool. Foliage net assimilation rate (A), and stomatal conductance to water vapor were calculated according to von Caemmerer and Farquhar (1981), and volatile emission rates according to Niinemets et al. (2011).

2.5. Chlorophyll florescence measurements

After the completion of gas exchange measurements and volatile sampling, maximum dark-adapted (10 min darkening) quantum yield of photosystem II (PSII, F_v/F_m) of each treated and untreated leaf was measured with a PAM flourometer (Walz IMAG-MIN/B, Walz GmbH, Effeltrich, Germany) by providing a 1 s saturating (pulse intensity of 7000 µmol m^-2 s^-1) flash of blue light (460 nm).

2.6. Calculation of ozone uptake by individual leaves

The stomatal conductance for ozone was calculated as the ratio of the average stomatal conductance to water vapor divided by the ratio of the binary diffusion coefficients for ozone and water vapor (2.03). The ratio of the binary diffusion coefficients for ozone and water vapor was estimated from ozone diffusion coefficient in air of 1.267^.10^-5 m² s^-1 at 22.84 ºC (Ivanov et al., 2007) and water vapor diffusion coefficient in air of 2.569^.10^-5 m² s^-1 at the same temperature determined according to Chapman and Enskog (Niinemets and Reichstein, 2003). As the temperature effects on the collision integral for both water vapor and O₃ (Tucker and Nelken, 1982) were negligible (the ratio varied less than 0.1% for 25 ºC used for leaf measurements and 22.84 ºC), we used the non-modified ratio in our calculations (Li et al., 2017).

Average stomatal ozone uptake rate (Φ_O3,S) was estimated as the product of average chamber O₃ concentration and average stomatal conductance for ozone for the entire 3 h fumigation period (g_O3) assuming that ozone uptake through the cuticular layer was negligible (Kerstiens and Lendzian, 1989) and that the intercellular ozone concentration was zero (Laisk et al., 1989). We acknowledge that this latter assumption might give slightly overestimated ozone uptake as ozone concentrations in the leaf interior are somewhat higher than zero (Moldau and Bichele, 2002), but due to lack of intercellular ozone concentrations, we cannot correct for such effects and we consider this simplified estimate of ozone flux rate as a quantitative measurement of the severity ozone stress (Beauchamp et al., 2005; Li et al., 2017). Boundary layer resistance was neglected due to high air turbulence generated by a fan in the glass chamber. Stomatal ozone uptake rates were calculated for 4, 5, and 6 ppm ozone, and also for combined ozone and wounding treatments.

2.7. Data analysis

The measurements were replicated at least thrice for each treatment and six replicates were available for control samples. Correlation analyses were carried out between ozone uptake rate and isoprene, total LOX, and total mono- and sesquiterpene emissions for 0.5, 3, 10, and 25 h since ozone exposure. The individual and interactive effects of ozone and wounding, and recovery time on LOX, volatile isoprenoids, saturated aldehydes, benzenoids, and geranylgeranyl diphosphate pathway (GGDP) volatiles were statistically tested by generalized linear models (GLM) using maximum likelihood model fitting. Data were tested for normality of distribution and homogeneity of variance and then the data were log-transformed.

Correlation analysis was carried out using SigmaPlot Version 12.5 (Systat Software Inc, San Jose, CA, USA). All other statistical analyses were conducted using SPSS Version 24 (IBM SPSS, Chicago, IL, WA). All statistical effects were considered significant at P < 0.05.

3. Results

3.1. Effects of ozone and wounding treatments on foliar photosynthetic characteristics

Maximum dark-adapted photosystem II (PSII) quantum yield estimated by chlorophyll fluorescence (F_v/F_m) declined in response to all ozone treatments (4, 5 and 6 ppm O₃) and in response to the combination of ozone and wounding treatments (O₃ all + W) compared to control leaves (P < 0.001 for ozone, and combined ozone and wounding treatments). The reduction was particularly severe when the leaves were treated with 6 ppm ozone, and combined 6 ppm ozone and wounding at all recovery times (Fig. 1A). A partial recovery of F_v/F_m was observed since 25 h of recovery for all treated leaves. There was no significant change in F_v/F_m in leaves treated with wounding alone, compared with control leaves throughout all recovery times (Fig. 1A and Table S1).

Figure 1.

Changes in dark-adapted (10 min darkening) maximum quantum efficiency of PSII photochemistry estimated by chlorophyll fluorescence (A), and net assimilation rate (B) and stomatal conductance to water vapor (C) in control (0 ppm), wounded, ozone-exposed (4, 5, and 6 ppm), and ozone-exposed and wounded (first exposed to 4, 5, and 6 ppm ozone and then wounded) leaves of Eucalyptus globulus at different recovery times (0.5, 3, 10, 25, 50, and 75 h) after stress applications. Gas exchange measurements were conducted under stable conditions in the glass chamber with photosynthetic photon flux density at the leaf surface of 700 - 750 µmol m^-2 s^-1, leaf temperature of 25 - 27 °C, ambient CO₂ concentration of 380-400 µmol mol^-1, and relative air humidity of 60 - 70 %. All measurements were replicated at least thrice. Individual effects of ozone, and wounding, and the interactive effect of ozone and wounding were analyzed by generalized linear models (GLM) with maximum likelihood model fitting. Wald chi-squared test statistics (χ²) and its statistical significance demonstrated as: * - P < 0.05, ** - P < 0.01, *** - P < 0.001.

Generally, all treatments resulted in a reduction of net assimilation rate (A) (P < 0.001 for all comparisons), especially at 6 ppm ozone and wounding (6 ppm O₃ + W) (Fig. 1B and Table S1). Overall, after the initial response, net assimilation rate remained relatively stable for 10 h of recovery for all stress applications, and then it partly recovered at 25 h of recovery with further reduction observed at 75 h in leaves treated with 6 ppm ozone, and combined 6 ppm ozone and wounding (Fig. 1B). Generally, the reduction and recovery of net assimilation rate were associated with concomitant changes in stomatal conductance to water vapor (P < 0.001 for combined ozone and wounding treatments, and recovery time, and P < 0.05 for wounding; Fig. 1B, Fig. 1C, and Table S1). In fact, there was even almost a full recovery of g_s at 25 h for most treatments, except for leaves treated with 6 ppm ozone, and 6 ppm ozone and wounding (6 ppm O₃ + W) (Fig. 1C). However, g_s declined again for most treatments at 75 h, except for wounding, and 5 ppm ozone and wounding-treated (5 ppm O₃ + W) leaves (Fig. 1C).

3.2. Isoprene emission in response to ozone and wounding treatments and in dependence on time of recovery

Non-treated leaves constitutively emitted isoprene at a relatively low level of ca. 1.4 ± 0.8 nmol m^-2 s^-1, but isoprene emission rate achieved the highest level of ca. 17.6 ± 4.9 nmol m^-2 s^-1 in response to combined 5 ppm ozone and wounding treatment (5 ppm O₃ + W) at 10 h of recovery, followed by 4 ppm ozone treatment through all recovery times (Fig. 2).

Figure 2.

Average (+SE) emission rates of isoprene in control (0 ppm), wounded, ozone-exposed (4, 5, and 6 ppm), and ozone-exposed and wounded (4, 5, and 6 ppm ozone and wounding) leaves of E. globulus at different recovery times (0.5, 3, 10, 25, 50, and 75 h) after stress applications. Data presentation and statistical significance as in Fig. 1.

Generally, isoprene emission was strongly enhanced when the leaves were subjected to combined ozone and wounding treatments, followed by separate ozone treatments (P < 0.05 for impact of ozone, P < 0.001 for ozone and wounding, and P < 0.001 for wounding, Table S1). However, compared to controls, wounding alone resulted in reduced isoprene emission rate through all recovery times. In fact, there was no recovery of isoprene emission after stress, implying that isoprene emissions remained either high as in ozone, and combined ozone and wounding treatments or low in the wounding treatment (Fig. 2).

3.3. Emission of total LOX, mono- and sesquiterpenes, saturated aldehydes, benzenoids, and GGDP pathway volatiles

In non-treated leaves, the baseline emission rate of total lipoxygenase pathway volatiles (LOX) varied between ca. 0.001-0.002 nmol m^-2 s^-1. The maximum emission rate of average total LOX was 1.4 ± 0.5 nmol m^-2 s^-1, achieved in response to 6 ppm ozone and wounding treatment (6 ppm O₃ + W) (Fig. 3A). Total emission rate of LOX was the highest immediately after stress applications and sharply decreased through the recovery phase (Fig. 3A; P < 0.05 for ozone, and wounding treatments, P < 0.001 for ozone and wounding treatments (O₃ all + W), and recovery time, Table S1).

Figure 3.

Average (+SE) emission rates of total LOX volatiles (A), monoterpenes (B), sesquiterpenes (C), saturated aldehydes (D), benzenoids (E), and GGDP pathway volatiles (F) in control (0 ppm), wounded, ozone-exposed (4, 5, and 6 ppm), and ozone-exposed and wounded (4, 5, and 6 ppm ozone and wounding) leaves of E. globulus at different recovery times (0.5, 3, 10, 25, 50, and 75 h) after stress applications. Data presentation and statistical significance as in Fig. 1.

Eucalyptus globulus is a significant constitutive emitter of mono- and sesquiterpenes under non-stressed conditions. Average total monoterpene emission by control leaves was ca. 0.20 ± 0.04 nmol m^-2 s^-1. However, the rate of total monoterpene emission increased upon all stress treatments at 0.5 h since the stress application, and then it gradually decayed during the recovery phase (Fig. 3B). In particular, average total emission rate of monoterpenes increased non-linearly in response to stresses with the combined stress applications resulting in the highest elicitation rate (e.g., total monoterpene emission was increased by ca. 750-fold for combined 5 ppm ozone and wounding treatment, and by 650-fold for combined 6 ppm ozone and wounding treatment at 0.5 h of recovery), followed by ozone treatments alone. Wounding treatment alone resulted in the lowest enhancement, ca. 25-times greater monoterpene emission rate (Fig. 3B).

The average total sesquiterpene emission rate of control leaves was ca. 0.010 ± 0.002 nmol m^-2 s^-1 and the emissions non-linearly increased in response to all stresses applied (Fig. 3C). In contrast to isoprene and monoterpene emission responses, total sesquiterpene emission achieved the highest level of ca. 14.6 ± 9.5 nmol m^-2 s^-1 upon combined 4 ppm ozone and wounding treatment, followed by 6 ppm ozone and wounding (6 ppm O₃ + W) at 0.5 h since stress applications. The emission rate of total sesquiterpenes initially decayed through recovery phase with a further increase observed at 75 h of recovery from the leaves treated with 5 ppm ozone, 6 ppm ozone, and combination of 5 ppm ozone and wounding stress (Fig. 3C).

Average total emission rate of saturated aldehydes by control leaves was ca. 0.020 ± 0.004 nmol m^-2 s^-1. Emission of total saturated aldehydes remained almost similar through recovery period (Fig. 3D). Benzaldehyde, o-xylene, and p-xylene (data not shown) were the benzenoids emitted by E. globulus leaves in all stress applications (Fig. 3E and Table 1). Total benzaldehyde emission remained elicited throughout the entire recovery period with the highest emission observed in response to combined 5 ppm ozone and wounding stress, especially at 0.5 h since the treatment. Similarly, total GGDP volatiles, made up of geranyl acetone and 6-methyl-5-hepten-2-one, showed enhanced emission rates through all recovery period in all treatments; in particular, emission rates in combined ozone and wounding treatments (O₃ all + W) were higher (Fig. 3F and Table 1).

3.4. Emission of foliar LOX volatiles in response to ozone and wounding treatments at different times of recovery

LOX emissions primarily consisted of hexanal, pentanal, (E)-2-hexenal, and (Z)-3-hexenol (Fig. 4 and Table 1). The emission of LOX was substantially enhanced in all ozone treatments (O₃ all, O₃ all + W), and upon wounding (W), but ozone alone led to greater emissions than wounding alone, and ozone combined with wounding (O₃ all + W) led to the greatest emissions. Separate ozone and wounding treatments had a minor effect on the emissions of primary LOX volatiles such as (E)-2-hexenal and (Z)-3-hexenol, but combined ozone and wounding treatments enhanced the emission responses by more than 25 times compared to separate wounding and ozone treatments (Fig. 4). In all treatments, the initial LOX emission rate at 0.5 h was dominated by (Z)-3-hexenol, followed by (E)-2-hexenal, hexanal, and pentanal (Fig. 4). In particular, (Z)-3-hexenol and (E)-2-hexenal declined to close to background levels at 75 h, while substantial emission responses of hexanal and pentanal were observed throughout the recovery time (Fig. 4). Thus, a considerable fraction of total LOX emission remained at all recovery times due to the major contribution of hexanal and pentanal to the emission blend (Fig. 3A). We also observed (Z)-3-hexenyl acetate at trace levels in response to ozone (O₃ all), and combined ozone and wounding treatments (O₃ all + W).

Figure 4.

Average (+SE) emission rates of individual lipoxygenase (LOX) pathway volatiles (LOX volatiles or green leaf volatiles) hexanal (A), (E)-2-hexenal (B), (Z)-3-hexenol (C), and pentanal (D) in control (0 ppm), wounded, ozone-exposed (4, 5, and 6 ppm), and ozone-exposed and wounded (4, 5, and 6 ppm ozone and wounding) leaves of E. globulus at different recovery times (0.5, 3, 10, 25, 50, and 75 h) after stress applications. Data presentation and statistical significance as in Fig. 1.

3.5. Emission of foliar monoterpenes upon ozone and wounding treatments and in dependence on time of recovery

Fifteen monoterpenes were observed in eucalypt emissions in this study (Fig. 5 and Table 1). Monoterpene emissions were dominated by 1,8-cineole (ca. 50% of total monoterpenes), followed by α-pinene (ca. 20% of total monoterpenes) and limonene (ca. 10% of total monoterpenes) in response to all treatments and in controls (Fig. 5). Apart from constitutively released monoterpenes, emissions of the characteristic stress-induced monoterpene, (Z)-β-ocimene, were detected in all treated leaves (Fig. 5H). All combined treatments of ozone and wounding had the highest (Z)-β-ocimene emissions followed by individual ozone, and wounding treatments. As with total monoterpenes, leaves treated with 5 ppm ozone and wounding demonstrated the highest level of (Z)-β-ocimene emission, 1.2 ± 1.0 nmol m^-2 s^-1 at 0.5 h since stress applications, but even at 75 h of recovery, a substantial quantity of (Z)-β-ocimene was detected in leaves subjected to combined 5 ppm ozone and wounding, and combined 4 ppm ozone and wounding treatments (Fig. 5H).

Figure 5.

Average (+SE) emission rates of monoterpenes (A-O) in control (0 ppm), wounded, ozone-exposed (4, 5, and 6 ppm), and ozone-exposed and wounded (4, 5, and 6 ppm ozone and wounding) leaves of E. globulus at different recovery times (0.5, 3, 10, 25, 50, and 75 h) after stress applications. Data presentation and statistical significance as in Fig. 1.

Regarding the recovery kinetics of individual monoterpenes, four different emission patterns were observed. For most of the monoterpenes, the emissions were the highest at 0.5 h after treatment. However, for Δ³-carene, once induced, its emission remained high throughout the recovery period in the case of most treatments, except for 6 ppm ozone treatment at 25 h and 5 ppm ozone treatment at 75 h (Fig. 5B). In addition, a secondary rise of (Z)-β-ocimene emission was detected in 10 h after exposure to 4 ppm ozone (Fig. 5H). For α-terpinene, the emissions remained high for most treatments at 3 h since the treatment, then the emissions declined (Fig. 5L). (Z)-β-ocimene was the most strongly elicited monoterpene followed by limonene, α-pinene, and 1,8-cineole. In addition, combined ozone and wounding treatments had statistically significant impact at P < 0.001 (Table S1) on the emission rate of all individual monoterpenes, except for α-fenchene, where P < 0.01 (Fig. 5 and Table S1).

3.6. Leaf sesquiterpene emissions in response to ozone and wounding treatments and in dependence on time of recovery

Eleven sesquiterpenes were detected in this study, whereas the emissions were dominated by aromadendrene (ca. 40% of total sesquiterpenes), followed by α-gurjunene (ca. 20%) and alloaromadendrene (ca. 10 %) (Fig. 6 and Table 1). These three abundant sesquiterpenes comprised ca. 70% of total sesquiterpene blend at 0.5 h of recovery (Fig. 6).

Figure 6.

Average (+SE) emission rates of sesquiterpenes (A-H) in control (0 ppm), wounded, ozone-exposed (4, 5, and 6 ppm), and ozone-exposed and wounded (4, 5, and 6 ppm ozone and wounding) leaves of E. globulus at different recovery times (0.5, 3, 10, 25, 50, and 75 h) after stress applications. Data presentation and statistical significance as in Fig. 1.

As with monoterpenes, the emission rate of all individual sesquiterpenes tended to peak for combined ozone and wounding treatments (O₃ all + W), followed by separate ozone (O₃ all) and wounding (W) treatments, reflecting the stronger sesquiterpene emission responses for more severe stress (Fig. 6). In addition, ozone treatment had a statistically significant impact on the emission rate of all other individual sesquiterpenes, except for epiglobulol and γ-muurolene (Fig. 6 and Table S1). Similarly to total sesquiterpene emissions, a secondary rise of sesquiterpene emissions at 75 h was observed for several individual sesquiterpenes with particularly pronounced increase observed for α-copaene (Fig. 6D), epiglobulol (Fig. 6E), β-gurjunene (Fig. 6G), and γ-gurjunene (Fig. 6H) emissions. Overall, the elicitation of (E)-β-caryophyllene was the highest, followed by epiglobulol, γ-gurjunene, and β-gurjunene, and there was little variation in elicitation of all sesquiterpenes at all recovery times.

3.7. Emission of saturated aldehydes as affected by ozone and wounding treatments and time of recovery

Four individual compounds contributed to the pool of saturated aldehydes: heptanal, octanal, nonanal and decanal (Fig. 7 and Table 1). The blend of saturated aldehydes was dominated by decanal, followed by nonanal, and the emission pattern of these two compounds broadly reflected the variation in total saturated aldehyde emission (Figs. 3D and 7). Initially, the emission rate of all saturated aldehydes was enhanced in all stress treatments, but the enhancement was the highest for leaves treated with combined 6 ppm ozone and wounding treatment (3.40 ± 1.60 nmol m^-2 s^-1 for combined 6 ppm ozone and wounding treatment at 0.5 h since stress application) compared to other treatments (Fig. 3D).

Figure 7.

Average emission rates (+SE) of saturated aldehydes (A-D) in control (0 ppm), wounded, ozone-exposed (4, 5, and 6 ppm), and ozone-exposed and wounded (4, 5, and 6 ppm ozone and wounding) leaves of E. globulus at different recovery times (0.5, 3, 10, 25, 50, and 75 h) after stress applications. Data presentation and statistical significance as in Fig. 1.

3.8. Emission of isoprene, and total mono- and sesquiterpene volatiles in relation to stomatal ozone uptake

Generally, the correlations between stomatal ozone uptake rate and isoprene, total LOX, and total mono- and sesquiterpene emission rates were poor for all ozone (O₃ all) exposures (Fig. 8). For total LOX emission, the correlation with stomatal ozone uptake was negative at 0.5 h since the treatments, and non-significant for other recovery times (Fig. 8). In all cases, the correlation of isoprene with ozone uptake rate was non-significant, but marginally significant negative correlations were observed between total mono- and sesquiterpene emissions and stomatal ozone uptake rate at 0.5 h since ozone exposure (Fig. 8).

Figure 8.

Correlation of total LOX, (A-D), isoprene (E-H), total monoterpene (I-L), and total sesquiterpene (M-P) emission rates with stomatal ozone uptake rate at different recovery times (0.5, 3, 10, 25 h) after the leaves of E. globulus were treated with ozone (4, 5, and 6 ppm) for 3 h. Each data point indicates an individual replicate. Data were fitted by non-linear regression for Fig. 8A and linear regressions for Figs. 8B-8P.

Similarly, there were poor correlations between stomatal ozone uptake rate and isoprene, total LOX, and mono- and sesquiterpene emissions for all combined ozone and wounding (O₃ all + W) treatments till 25 h since stress application (Fig. 9). However, marginally significant positive correlations were observed between stomatal ozone uptake rate and total LOX emission at 3 h and 10 h since the treatments, and a marginally significant negative correlation was observed between isoprene emission and ozone uptake rates at 0.5 h (Fig. 9). In contrast, significant or marginally significant positive correlations between stomatal ozone uptake rate and isoprene, and total mono- and sesquiterpene emission rates were observed only at 3 h since stress applications (Fig. 9). However, at 25 h since the combined ozone and wounding treatments (O₃ all + W), the correlations were negative and statistically significant between ozone uptake and total mono- and sesquiterpene emissions (Fig. 9).

Figure 9.

Correlation of total LOX, (A-D), isoprene (E-H), total monoterpene (I-L), and total sesquiterpene (M-P) emission rates with stomatal ozone uptake rate at different recovery times (0.5, 3, 10, 25 h) after the leaves of E. globulus were treated with combined ozone (4, 5, and 6 ppm for 3 h) and wounding treatments. Each data point indicates an individual replicate. Data were fitted by linear regressions.

4. Discussion

4.1. Changes in leaf photosynthetic characteristics due to ozone and wounding treatments

The reduction of maximum dark-adapted photosystem II (PSII) quantum yield estimated by chlorophyll fluorescence (F_v/F_m) upon ozone, and combined ozone and wounding treatments indicates that ozone induced perturbations in the reaction centers of PSII. This is consistent with previous studies where both energy dissipation activity and efficiency of PSII declined in ozone-exposed leaves (Guidi et al., 1999; Guidi and Degl'lnnocenti, 2008), and the level of reduction of PSII activity scaled with the quantity of ozone entering leaf intercellular air spaces (Gerosa et al., 2009). Furthermore, the reduction of PSII activity was also associated with sustained photoinhibition (Guidi and Degl'lnnocenti, 2008). In fact, the ozone-induced photoinhibition is a synergistic effect of light-dependent photodamage associated with net reduction of de novo synthesis of D1 protein, ultimately leading to a subsequent decline in the repair rate of PSII (Fiscus et al., 2005; Godde and Buchhold, 1992; Guidi and Degl'lnnocenti, 2008). In this study, wounding treatment had no apparent effect on F_v/F_m, suggesting that it did not reduce the number of open PSII centers (Fig. 1A).

Separate and combined applications of ozone and wounding substantially reduced net assimilation rate in a time-dependent manner (Fig. 1B). Elevated ozone can directly affect photosynthesis by reducing foliage biochemical potentials (non-stomatal inhibition) or by reducing stomatal conductance (stomatal inhibition) or both (Fusaro et al., 2016). The decline of net assimilation rate upon individual and combined applications of ozone and wounding treatments could be due to the impaired electron transport capacity and irreversible damage in chlorophyll reaction centers caused by oxidative stress (Guidi and Degl'lnnocenti, 2008) consistent with the reduction in PSII activity in our study. In addition to suppression of PSII, elevated ozone often inhibits photosynthesis through the suppression of Calvin cycle; particularly, it reduces the activity of ribulose-1, 5-bisphosphate (RuBP) carboxylase/oxygenase (Rubisco) (Guidi and Degl'lnnocenti, 2008; Zhang et al., 2011). In fact, Rubisco is the primary component affected by elevated ozone in plants (Noormets et al., 2001) and both the quantity and activation state of Rubisco are reduced in ozone-exposed plants, leading to a decline in maximum rates of carboxylation (V_cmax) (Fusaro et al., 2016; Niu et al., 2014). For example, upon chronic ozone exposures in Cinnamomum camphora seedings, both V_cmax and the capacity for photosynthetic electron transport (J_max) were substantially reduced, but there was no apparent change in J_max / V_cmax ratio. The constant value of J_max / V_cmax demonstrates that there was a strong correlation between reductions in RuBP carboxylation and light-driven electron transport (Niu et al., 2014). In addition, certain evergreen woody plants such as Quercus ilex L. and Arbutus unedo L. have the ability to maintain photosynthetic characteristics to a normal level even during salt stress episodes by removing harmful ions from effective shoot organs (Fusaro et al., 2014).

In our study, the time-dependent reduction of net assimilation rate was also associated with the reduction of stomatal conductance to water vapor (Figs. 1B and 1C) as observed in several other studies (Zhang et al., 2011). This is in agreement with the evidence that acute ozone exposure exerts an inhibitory effect on guard cell K⁺ channels, resulting in reduced stomatal conductance (Niu et al., 2014). However, in our study, there was even a certain overshoot in stomatal conductance at the end of the recovery period (Fig. 1C), implying that there might be compensatory responses (Copolovici et al., 2014; Copolovici et al., 2011; Niinemets, 2016).

4.2. Effects of ozone and wounding treatments on the emission of volatile isoprenoids, LOX, saturated aldehydes, benzenoids, and GGDP pathway volatiles

4.2.1. Effects of ozone and wounding treatments on isoprene emission kinetics

As biosynthesis and emission of isoprene occur simultaneously (Rasulov et al., 2009), the rate of isoprene emission is controlled by substrate availability and/or isoprene synthase enzyme activity. Eucalyptus globulus is a strong constitutive isoprene emitter, but stress can affect both substrate availability as well as the rate of isoprene synthase expression, thereby potentially resulting in major changes in isoprene emissions (Niinemets, 2010a). In our study, wounding reduced isoprene emission rate by ca. 85% at 0.5 h after stress application and by ca. 54% at 75 h of recovery, indicating that only a slight restoration of isoprene emission was observed (Fig. 2). However, wounding also initially reduced net assimilation rate by ca. 73% and by ca. 30% at 75 h of recovery; this is consistent with the previous report of Loreto and Sharkey (1993) showing that isoprene emission is highly sensitive to wounding stress and it responds more rapidly than photosynthesis and stomatal conductance. Although stomatal conductance declined in our study, stomata cannot control isoprene emissions (Fall and Monson, 1992), because the buildup of isoprene in the leaf gas phase upon stomatal closure compensates for reduced stomatal conductance (Niinemets et al., 2014; Niinemets and Reichstein, 2003). Indeed, isoprene emission is often even enhanced by moderate stomatal closure due to the positive effect of reduced intercellular CO₂ concentration on isoprene emission rate (Rasulov et al., 2016; Wilkinson et al., 2009). In our study, the reduction of isoprene emission upon wounding was likely associated with the downregulation of rate-limiting enzyme activities (Kanagendran et al., unpublished).

However, ozone stress, and ozone and wounding stress enhanced isoprene emission through the entire recovery period (Fig. 2), suggesting that when wounding was combined with ozone exposure, ozone might have primed for isoprene synthase gene expression and isoprene synthase enzyme activity, ultimately enhancing sustained de novo isoprene emission through recovery (Fig. 2). This suggestion is in line with the patterns of photosynthesis and isoprene changes in different treatments. Differently from the wounding treatment, where isoprene emission and photosynthesis rates declined together, in individual ozone, and in combined ozone and wounding stress treatments, isoprene emission rate was enhanced while net assimilation rate decreased through all recovery periods, demonstrating that enhanced isoprene emission was not associated with the changes in photosynthesis, particularly for individual ozone, and combined ozone and wounding treatments. Indeed, the enhanced isoprene emission was likely due to upregulation of isoprene synthase gene expression and enhanced rate-limiting enzyme activities through all recovery periods (Kanagendran et al., unpublished) (Figs. 1B and 2). Elevated isoprene emissions upon ozone exposure have also been observed in other studies (Fares et al., 2006; Velikova et al., 2005). This has been considered adaptive as isoprene can act as a volatile antioxidant scavenging reactive oxygen species (ROS) and thus, enhance membrane stability of leaves upon oxidative stress (Asada, 2006; Velikova et al., 2005).

However, chronic ozone exposure inhibited isoprene emission in field-grown trembling aspen (Populus tremuloides) (Calfapietra et al., 2008; Calfapietra et al., 2007). This was associated with downregulation of isoprene synthase gene expression and isoprene synthase enzyme level in the ozone-sensitive aspen clone (Calfapietra et al., 2008; Calfapietra et al., 2007). This difference from our study further supports the suggestion that different biochemical mechanisms are activated by acute and chronic ozone exposures.

4.2.2. Effects of ozone and wounding treatments on LOX pathway volatiles and saturated aldehydes

LOX compounds are synthesized from free polyunsaturated fatty acids in cell membranes (linoleic acid, 18:2 and linolenic acid, 18:3) as a part of stress-dependent hypersensitive reaction (Feussner and Wasternack, 2002). In plant cells, lipoxygenases are constitutively active and therefore, release of free polyunsaturated fatty acids due to membrane damage and consequent formation of volatile LOX products is a characteristic early stress response; thus, LOX volatiles are rapidly emitted upon acute stress treatments (Jiang et al., 2016; Li et al., 2017; Portillo-Estrada et al., 2015). In our study, the total LOX emission rate was quantitatively related to the severity of stress (Fig. 3A), consistent with previous studies of Beauchamp et al. (2005), Copolovici et al. (2012), Jiang et al. (2016), and Jiang et al. (2017).

The initial LOX emission burst usually lasts for some minutes to a few hours (Jiang et al., 2016; Li et al., 2017; Portillo-Estrada et al., 2015) and consists of classic LOX volatiles such as (Z)-3-hexenol and (E)-2-hexenal. In our study, the observation of sustained emissions of these classic LOX volatiles through the recovery period suggests that acute ozone and combined ozone and wounding exposures caused severe cellular damage, consistent with only a partial recovery of photosynthesis (Fig. 1B). These acute stresses could also have elicited overexpression of additional lipoxygenases and enzymes responsible for derivatization of primary LOX compounds, ultimately supporting the longer-term emission response of LOX compounds (Liavonchanka and Feussner, 2006; Porta and Rocha-Sosa, 2002). Thus, the regulation of LOX biosynthesis might have shifted from substrate level control at early stages of recovery to gene expression level control at later stages of recovery (Fig. 3A and Figs. 4A-D). A change of LOX biosynthesis regulation from substrate level to gene expression level through recovery phases was further suggested by Jiang et al. (2016) in Melampsora larici-populina-infected poplar leaves that had high emissions of derivatized LOX volatiles (Z)-3-hexenyl acetate and (Z)-3-hexenol. Possibly, the expression of genes responsible for derivatization of C6 aldehydes regulated the emission of (Z)-3-hexenyl acetate and (Z)-3-hexenol (Jiang et al., 2016). However, we cannot rule out the formation of (Z)-3-hexenol from (E)-2-hexenal (or from its unstable conjugate (Z)-3-hexenal) by a corresponding constitutively expressed alcohol dehydrogenase as reported by Davoine et al. (2006) and Farmer and Davoine (2007).

The pathway for saturated aldehydes is still uncertain (Hu et al., 2009), although the lower molecular mass compounds, C5 and C6 might partly share the same pathway with LOX (Heiden et al., 2003; Loreto and Fares, 2007; Wildt et al., 2003). In addition, ozone itself can oxidize membrane lipids and also cause oxidative degradation of lipids on the leaf surface as well (Senser et al., 1987), suggesting that some of the saturated aldehyde release might be non-biochemical. However, provided the oxidative burst is mainly confined to the initial phase of the recovery, the non-biochemical processes cannot explain the sustained release of the saturated aldehydes through the entire recovery period in our study.

In our study, emission of saturated aldehydes was ca. 2-3 higher than LOX (Figs. 3A, D and 7). Wildt et al. (2003) also suggested that LOX volatiles could have primed for the elicitation of saturated aldehydes.

4.2.3. Influence of ozone and wounding treatments on mono- and sesquiterpene emissions

A distinguishing feature of eucalypt leaves is the presence of numerous sub-dermal secretory cavities, oil glands, where mono- and sesquiterpenes are stored (Goodger et al., 2010; Goodger et al., 2016). In addition, leaf cuticle of E. globulus also stores multiple terpene products to a minor extent (Guzmán et al., 2014). As is common in terpene-storing species, there is a slow diffusion of these volatiles out of the storage compartments (so-called storage emissions, Copolovici and Niinemets, 2016; Grote et al., 2013; Grote and Niinemets, 2008). However, apart from the storage emissions, many storage-emitters also emit de novo synthesized mono- and sesquiterpenes in a light- and temperature-dependent manner (Grote et al., 2013; Holzke et al., 2006; Komenda and Koppmann, 2002). Ultimately, the sum of storage and non-storage emissions gives the constitutive terpene emission potential for the given species under specific ambient conditions. Furthermore, in all constitutive terpene emitters, further terpenoid emissions can be induced by abiotic and biotic stresses (Copolovici and Niinemets, 2016).

Treatment of E. globulus leaves with ozone, and combined ozone and wounding stress resulted in a substantial enhancement of emissions of all mono- and sesquiterpenes, albeit a strict stress dose-dependency of emission responses was not observed (Fig. 3B, Fig. 3C, Fig. 5, and Fig. 6). The wounding treatment alone enhanced mono- and sesquiterpene emissions to a minor degree compared to other stress treatments through all recovery periods (Fig. 3B, Fig. 3C, Fig. 5, and Fig. 6). This is surprising given that wounding leads to a direct exposure of terpene storage pools to the ambient air, while ozone stress is expected to affect less the thick-walled storage structures. Nevertheless, the highest emission burst of mono- and sesquiterpenes observed at 0.5 h of recovery could be due to the damage of epidermal cells and epithelial cells of storage glands as a result of hypersensitive response to ozone exposure (Bussotti et al., 2007), ultimately leading to increased terpene permeability of cell layers surrounding the storage structures. In our study, cellular damage was further confirmed by the burst of primary LOX emissions (Fig. 3A and Fig. 4). In addition, abrasion of leaf cuticle by acute ozone exposure could also have released some terpenoids. Association of the emission burst of mono- and sesquiterpenes with the damage of epidermal and epithelial cells of storage glands, and abrasion of cuticles at 0.5 h can explain why the emission responses were opposite to changes in net assimilation rate at the same recovery time (Fig. 1B, Fig. 3B, Fig. 3C, Fig. 5 and Fig. 6).

Once directly damaged, emissions of mono- and sesquiterpenes from free-air exposed storage structures continue for long time periods until the wounds are sealed due to terpenoid oxidization and polymerization (Loreto et al., 2000; Schuh et al., 1996). However, in our study, enhanced emissions of most mono- and sesquiterpenes were relatively short-lived (Fig. 3B, Fig. 3C, Fig. 5, and Fig. 6). This suggests that either the storage structures are sealed faster or these short-living initial emissions resulted from a rapid increase in substrate availability.

Biosynthesis of monoterpenes occurs via the 2-C-methyl-D-erythritol 4-phosphate/1-deoxy-D-xylulose 5-phosphate pathway (MEP/DOXP pathway) in plastids and sesquiterpene synthesis occurs via the mevalonate pathway (MVA) in cytosol. Activity of MEP/DOXP pathway in mesophyll is primarily driven by the continuous supply of NADPH and ATP generated by photosynthetic electron transport (Niinemets et al., 2002; Rasulov et al., 2009; Rasulov et al., 2016). Thus, monoterpene emission rates and photosynthesis are often correlated (Niinemets et al., 2002; Staudt and Lhoutellier, 2011) due to direct effects of the substrate availability on monoterpene synthesis, but no such association is expected for sesquiterpene emission. In our study, there was no clear relationship between the changes in photosynthesis and monoterpene emission rates. Given that emission rates of total monoterpenes were uncoupled from photosynthesis and were broadly correlated with sesquiterpene emission rate at all recovery points, it is unlikely that substrate level control played a significant role in regulating their emission responses. Furthermore, to cope with direct damage, photosynthetic intermediates and MEP/DOXP pathway products could primarily be allocated for the biosynthesis of larger isoprenoids such as chlorophyll and carotenoids (Pazouki et al., 2016) upon ozone and wounding treatments rather than for the biosynthesis of monoterpenes, thereby temporarily reducing the rate of monoterpene synthesis until enhancement of corresponding enzyme activities as the result of the onset of gene expression responses.

In addition, as volatility of sesquiterpenes is less than that for monoterpenes, sesquiterpene pools take longer to exhaust than monoterpene pools (Copolovici and Niinemets, 2015). This may be the reason why a considerable quantity of sesquiterpenes was present throughout all recovery phases. Interestingly, we did not observe a complete cessation of mono- and sesquiterpene emissions throughout the recovery phase. Possibly, de novo emission of mono- and sesquiterpenes could have contributed to the volatile blend and thus, stressed leaves continuously supported higher emissions than those in control leaves at all recovery periods. However, we cannot rule out that a part of mono- and sesquiterpenes was also removed within the leaf cells in the reactions with ROS generated by oxidative stress through recovery. Therefore, we suggest that mono- and sesquiterpene emission rates can reflect three simultaneous processes: depletion of storage pools, quenching by ROS, and de novo synthesis.

A secondary rise of mono- and sesquiterpene emissions, especially after 10 h since stress applications (Fig. 3B, Fig. 3C, Fig. 5 and Fig. 6), suggests that activation of gene expression and changes in rate-limiting enzyme activity upon ozone and wounding treatments might be responsible for the emission of individual mono- and sesquiterpenes at the later stages of recovery. In fact, several studies have demonstrated that enhanced emission of volatile isoprenoids hours to days after stress application is due to upregulation of the activity of terminal enzymes, terpene synthases (Frank et al., 2012; Pateraki and Kanellis, 2010). Given that temporal variation of volatile isoprenoid emission was associated with the emission of LOX volatiles in our study (Figs. 3A-C and Figs. 4-6), the stress-induced activation of LOX could also have induced changes in the transcript levels of lipoxygenase catalase, eventually triggering jasmonate biosynthesis. Jasmonate is a key regulator of terpene biosynthesis and can ultimately regulate the synthesis and emission of volatile isoprenoids (Schaller et al., 2004).

Among the induced responses, we observed enhanced emissions of monoterpene (Z)-β-ocimene (Fig. 5H) and sesquiterpene (E)-β-caryophyllene (Fig. 6C). Often, sesquiterpenes are stress-inducible, particularly (E)-β-caryophyllene (Lung et al., 2016; Toome et al., 2010). In addition, stress-induced emission of β-ocimene has been demonstrated in several studies. For example, enhanced emission of β-ocimene upon herbivory feeding of Medicago truncatula leaves was associated with greater transcript abundance of M. truncatula β-ocimene synthase (MtEBOS) (Navia-Giné et al., 2009) Analogously, enhanced β-ocimene emissions have been observed upon heat (Copolovici et al., 2012; Pazouki et al., 2016) and cold (Copolovici et al., 2012) shock in Solanum lycopersicum, biotrophic fungus Melampsora spp. infection of Salix spp. leaves (Toome et al., 2010), and feeding of Alnus glutinosa leaves by the larvae of Cabera pusaria (Copolovici et al., 2011). Taken together, these findings suggest that single and combined application of ozone and wounding treatments could have activated defense responses and also altered the expression rate of several terpene synthase genes and other genes in MEP/DOXP and MVA pathways, ultimately contributing to significant de novo emissions of (Z)-β-ocimene and (E)-β-caryophyllene.

4.2.4. Effects of ozone and wounding treatments on benzenoids and geranylgeranyl diphosphate pathway volatiles

Plant benzenoids, are biosynthesized through shikimate pathway and they are further esterified to other secondary metabolites or bound to cell walls (Zhao et al., 2010). In this study, we detected only three benzenoid compounds - benzaldehyde, o-xylene and p-xylene (Table 1) - that did not show dose-dependent emission responses (Fig. 3E). However, we did not detect the common stress-induced volatile methyl salicylate (MeSA) that triggers multiple biochemical pathways upon its release (Arimura et al., 2005; Zhao et al., 2010). To the best of our knowledge, there are no studies on the benzenoid emission from eucalypt foliage upon ozone and wounding exposure, although the impact of ozone on salicylic acid metabolism has been observed in numerous species, including tobacco (Nicotiana tabacum) (Beauchamp et al., 2005; Janzik et al., 2005), common bean (Phaseolus vulgaris) (Li et al., 2017), and beech (Fagus sylvatica) saplings (Betz et al., 2009; Olbrich et al., 2005), and trees (Betz et al., 2009; Jehnes et al., 2007).

In this study, GGDP pathway volatile emission was affected by both ozone and wounding stress and their combination, and the initial increase in emissions resembled that for all other compound classes (Fig. 3F) through the recovery phase. GGDP pathway products are biosynthesized in plastids (Rajabi Memari et al., 2013) and thus, the emission pattern of these volatiles probably represents the carotenoid turnover during routine plant metabolism (Beisel et al., 2010). Among the GGDP pathway volatiles, we detected the emissions of geranyl acetone and 6-methyl-5-hepten-2-one that can be emitted as a result of oxidative cleavage of carotenoids (Buttery et al., 1988; Goff and Klee, 2006; Tieman et al., 2006). Thus, the alteration in the emission pattern of these volatiles can be associated with the corresponding changes in carotenoid synthesis upon ozone and wounding exposures.

4.3. Synergistic effects of combined ozone and wounding treatments on gas exchange and stress volatile emission responses in E. globulus leaves

The simultaneous actions of several stress factors on plants typically considerably exceed the simple additive effects of the individual stresses, i.e. the simultaneous stresses act synergistically (Alexieva et al., 2003; Niinemets et al., 2017). For instance, in nature, in mid-summer, high solar radiation often leads to a decline in net assimilation rate due to photoinhibition of reaction centers and photooxidation of plastid pigments, and such conditions are frequently exacerbated by concomitant heat stress, and drought (Niinemets, 2010a), implying a synergistic interaction of different stresses (Alexieva et al., 2003). However, the situation might be different for sequential stresses. Application of a short and mild stress treatment with a single stressor can alleviate the adverse effect of subsequent stress because the first stress can induce defense response (priming) (Frost et al., 2008a; 2008b), resulting in a certain resistance to the subsequent unfavorable factors (cross-adaptation). On the other hand, a similar treatment with a single stressor could result in an increased susceptibility or hypersensitivity of the overall physiological state of the plant to the same or different stressor, and this can result in irreversible damages (cross-synergism). As the tolerance to any given stress varies strongly among species, the physiological responses upon subsequent stress applications can be species-specific. For instance, a subsequent application of low temperature stress with UV-B resulted in cross-adaptation in pea (Pisum sativum) plants and cross-synergism in maize (Zea mays) plants (Alexieva et al., 2003). In addition, recovery from multiple stresses is a complex process that cannot be predicted from the recovery of single stress (Bansal et al., 2013).

In this study, enhanced reduction of the gas exchange characteristics and increased emission responses of stress volatiles through recovery phase in combined ozone and wounding treatments (Figs. 1-7), clearly demonstrate synergistic effects. This is in accordance with previous studies that have demonstrated interactions of stressors and their impacts on stress volatile emissions, sometimes leading to highly complicated and convoluted responses. For instance, in silver birch (Betula pendula), the combination of elevated temperature and ozone increased the emission blend of terpenes, LOX, and MeSA, but elevated ozone alone did not change the emission of mono-, homo- or sesquiterpenes and it decreased the emission of LOX and MeSA; furthermore, elevated temperature resulted in an enhancement of total mono-, homo- and sesquiterpene, and LOX and MeSA emissions (Hartikainen et al., 2012). Similarly, combined application of drought and herbivore feeding in Alnus glutinosa strongly amplified the emission responses of total LOX and monoterpenes compared with both stresses applied alone (Copolovici et al., 2014). In young hybrid poplars (Populus deltoides cv. 55/56 x P. deltoides cv. Imperial), combined application of elevated ozone and drought decreased isoprene emission, whereas drought increased the emission, and ozone decreased it (Yuan et al., 2016).

Synergistic effects of combined ozone and wounding treatments can differ among species with different physiology. Since plant-produced volatile isoprenoids can quench ozone in ambient air and in leaf intercellular air spaces due to their direct reaction with ozone (Fares et al., 2010), eucalypts are expected to be highly resistant to ozone stress. However, among eucalypt species, photosynthetic characteristics differ even under unstressed conditions (Battaglia et al., 1996). In addition, there is a significant variation of constitutive and induced isoprenoid emissions among eucalypt species (He et al., 2000). Further studies across a range of eucalypt species with different constitutive and induced volatile emission capacities are needed to gain an insight into the generality of the observed synergistic wounding and ozone stress responses. Moreover, we further suggest that synergistic effect of ozone and wounding will be more severe for plant species that are low-level isoprenoid emitters or non-emitter, since the reaction of ozone with volatile isoprenoids in intercellular airspaces, inside the leaf boundary layer, and outside the leaf boundary layer is marginal in low-level isoprenoid emitters (Loreto and Fares, 2007). Thus, the actual quantity of ozone exerting oxidative stress at given ozone exposure can be considerably higher. In strong isoprenoid emitters, the severity of oxidative stress caused by given level of ozone is lower due to the higher proportion of ozone reacting with volatile isoprenoids (Loreto and Fares, 2007; Loreto et al., 2001).

4.4. Relationship between stomatal ozone uptake and stress volatile emissions

Stomatal uptake plays the main role in elicitation and emission of stress volatiles irrespective of ozone concentration around a leaf surface (Beauchamp et al., 2005). However, the ultimate damaging effect of ozone inside the leaf is determined by the balance between the quantity of ozone taken up by stomata and that scavenged by the water-soluble antioxidants within the aqueous phase in cell membrane, and by volatile isoprenoids present in leaf gas and lipid phases (Fares et al., 2010). In our study, there were no strong correlations between stomatal ozone uptake and emission rates of isoprene, total LOX, and total mono- and sesquiterpenes in treatments with ozone alone and in combination with wounding at any recovery phases (Figs. 8 and 9). Lack of clear dose-dependencies of emission responses in our study can be explained by three factors: first, the concentrations of volatile isoprenoids in intercellular spaces are typically hundreds of times higher than those above leaf boundary layer (Loreto and Fares, 2007). Therefore, once ozone enters the intercellular airspaces, it quickly reacts with volatile isoprenoids, especially with mono- and sesquiterpenes and thus, the quantity of ozone molecules entering the mesophyll cells is much lower than that taken up through stomata; second, as soon as ozone enters the surface of leaf mesophyll cells, it is also scavenged by antioxidants such as putrescine and apoplastic ascorbate in cell wall water; and third, reactive oxygen species (ROS) generated by ozone could be removed through the reaction with volatile isoprenoids (Loreto and Fares, 2007). Therefore, the quantity of ozone molecules ultimately causing a physiological stress is expected to be much lower than that entering leaf intercellular space. This is supported by our findings that at 0.5 h after stress application, the highest negative correlation was achieved for total mono- and sesquiterpenes (Fig. 8 and 9). Another implication of the possible reaction of volatiles with ROS inside the leaves is that the emission rates of isoprene, LOX, and mono- and sesquiterpenes through recovery phase were lower than the rates of their biosynthesis, further contributing to the poor correlations between volatile emission and stomatal O₃ uptake.

To further complicate the emission responses, different ozone concentrations differently affected steady-state stomatal conductance, and possibly also the rate of change in stomatal conductance. In particular, stomatal ozone uptake at 6 ppm ozone was comparatively lower than at 4 ppm and 5 ppm ozone applied both separately and in combination with wounding. This was because of lower stomatal conductance to ozone at the highest dose of ozone, contributing to the lower induction of volatile emissions at this ozone concentration (Figs. 3B and 3C, and Figs. 2, 5 and 6). Kinetic studies through ozone exposure period are needed to understand whether differences in the rate of stomatal closure at different ozone concentrations could play a role in poor relationships between ozone uptake and volatile emissions in the steady state.

5. Conclusion

To our knowledge, the present study is the first report of combined application of acute ozone and wounding stress on foliage gas exchange characteristics, and differential regulation of key volatile emissions from stress application through recovery. The physiological and biochemical responses of plants to a combined effect of ozone and wounding through recovery were different from the responses in plants treated with ozone and wounding separately. In particular, the reduction of photosynthetic traits such as A, g_s and F_v/F_m, and enhanced emission responses were greater for combined ozone and wounding treatments than those for separate ozone and wounding treatments. The secondary rise of mono- and sesquiterpenes through recovery phase upon all stress applications was likely associated with the upregulation of terpene synthase gene expression and /or rate limiting enzyme activity.

Collectively these data indicated that combined acute ozone and wounding treatments resulted in synergistic effects on leaf gas exchange characteristics and volatile emissions. Generally, the cross-synergism was more severe in response to 5 ppm ozone and wounding treatments than for 4 and 6 ppm ozone and wounding treatments. LOX emissions were quantitatively associated with the severity of stress, but for other volatiles, there were no quantitative relationships between the severity of stress and emission responses. The absence of stress dose vs. emission relationship was likely due to scavenging of ozone within leaf intercellular air spaces, and ROS in mesophyll cells by volatile isoprenoids.

Abbreviations

BVOC

biogenic volatile organic compound

GGDP

geranylgeranyl diphosphate

GLM

generalized linear model

GLV

green leaf volatiles

LOX

lipoxygenase pathway

MEP/DOXP pathway

2-C-methyl-D-erythritol 4-phosphate/1-deoxy-D-xylulose 5-phosphate pathway

MVA pathway

mevalonate pathway

PSII

photosystem II

ROS

reactive oxygen species