Introduction 

Tropospheric ozone (O₃) is recognised as a phytotoxic gaseous air pollutant. Many experimental studies have shown the negative effects of O₃ on growth and physiological functions such as photosynthesis of tree species ([43], [44], [46]). O₃ concentration in East Asia has remarkably increased because of rapid increases in the emissions of main O₃ precursors, nitrogen oxides, and volatile organic compounds ([28]). This trend is likely to continue in the near future according to several estimations of gaseous emissions ([24], [47]).

Evergreen broad-leaved trees are the main component species of the laurel forest, which is typical in warm temperate climate zone in East Asia with high amount of precipitation in summer. Japan is located in the northern limits of laurel forest distribution. Castanopsis sieboldii, Quercus glauca, and Q. myrsinaefolia are representative evergreen broad-leaved tree species in the laurel forests of Japan ([25]). These species are also found in urban areas as roadside and park trees in Japan. Ozone sensitivity of evergreen trees is generally considered to be lower than that of deciduous trees ([48], [2], [19]). However, there is a great variation in O₃ susceptibility within a functional type. Actually, C. sieboldii is considered as sensitive tree species to O₃ and the sensitivity is similar to Fagus crenata, one of the most sensitive tree species to O₃ in Japan ([17], [39]).

Photosynthetic carbon assimilation is one of the vital processes essential for the growth and survival of forest trees. The understanding of photosynthetic responses of trees to O₃ is crucial to evaluate carbon absorption capacity of a forest under elevated O₃. Photosynthetic rate in leaves is regulated by CO₂ diffusion from the ambient air to chloroplasts and the biochemical assimilation capacity in chloroplasts ([4], [18]). A distinction between the process of diffusion and the biochemical assimilation by tree species with respect to the response to O₃ is useful for a deeper understanding of the effects of O₃ on forest production and for the future prediction of carbon sequestration and water balance of forest ecosystems under elevated O₃. Kitao et al. ([15]) reported that stomatal limitation of photosynthesis was the main factor behind the decrease in net photosynthetic rate while there was no effect of O₃ on the maximum rate of carboxylation (V_cmax), which is in situ activity of Rubisco, in the leaves of mature Fagus sylvatica trees. On the other hand, the opposite results were found in F. crenata saplings ([41]). To our knowledge, there are no current studies that separately evaluate the diffusion and biochemical processes in photosynthesis of East Asian evergreen tree species exposed to O₃. Thus, there is a large uncertainty in the risk assessment of the effect of O₃ on East Asian laurel forest ecosystems. In the present study, we conducted O₃ fumigation study by using three representative evergreen broad-leaved tree species native to Japan, C. sieboldii, Q. glauca, and Q. myrsinaefolia, to evaluate in detail the photosynthetic responses to elevated O₃.

  Materials and methods 

Growth condition, ozone exposure, and plant materials

Greenhouse-type O₃-fumigation chambers with natural light source located at the Field Museum (FM) Tamakyuryo of Tokyo University of Agriculture and Technology (35° 04′ N, 139° 02′ E and 144 m a.s.l., Hachioji, Tokyo, Japan) were used in the present study. We set three levels of O₃ treatment: charcoal-filtered air (CF, mean O₃ removal efficiency: ca. 60%), 1.0 time ambient O₃ concentration, and 1.5 times ambient O₃ concentration. The ambient O₃ concentration outside the chambers was used as the standard for regulating the O₃ concentration in the chambers. Three chambers were used for each O₃ treatment, with nine chambers in total. Further details of the fumigation and monitoring systems are described in Kinose et al. ([13]).

Two-year-old seedlings of C. sieboldii, Q. glauca, and Q. myrsinaefolia were individually planted in 1/2000 a Wagner’s pots (bulk: 12 L, width: 228-240 mm, depth: 259 mm) filled with brown forest soil (Cambisol according to international classification system - [10]) on 8-9 May 2014. The soil was collected from the A-horizon of the floor of a deciduous forest in the FM Mt. Kusaki of Tokyo University of Agriculture and Technology (36° 03′ N, 139° 02′ E Midori, Gunma, Japan). Before planting the seedlings, the soil was passed through a 5-mm sieve. Emergence of new leaves was observed about one week after planting. Therefore, we considered little plantation shock in the present experiment. On 15 May 2014, the seedlings were transferred into nine chambers and were grown for 166 days until 27 October 2014. In each species, 12 seedlings were assigned to each gas treatment (four seedlings per chamber), with 36 seedlings in total. Mean height and stem base diameter of the seedlings were 39 cm and 3.9 mm for C. sieboldii, 31 cm and 4.8 mm for Q. glauca, and 33 cm and 4.3 mm for Q. myrsinaefolia, respectively. Air temperature and relative air humidity in three of the nine chambers were continuously monitored at 10 min intervals using a TR-72U^® Thermo Recorder (T&D Corporation, Nagano, Japan). Light intensity in three of the nine chambers were monitored at 5 min intervals using a HOBO Pendant temperature/light data logger (model UA-002-64^®, Onset Computer Co., MA, USA), calibrated against a quantum sensor (model LI-190SA^®, Li-Cor Inc., NE, USA).

Daily mean air temperature, relative air humidity, and accumulated photosynthetic proton flux density (PPFD) inside the chambers during the experimental period (15 May to 27 October 2014) were 21.7 °C, 87.0%, and 2881 mol m^-2, respectively. The mean light transmissibility of the chambers was approximately 80%. All seedlings were regularly irrigated to maintain soil moisture.

Growth measurement

We measured the height and stem diameter at 2 cm height from the soil surface and leaf number of all seedlings at the end of the experiment (27 October 2014). Leaves that emerged in 2013 and 2014 were separately counted. Stem volume index (square of diameter × height, D²H) was also calculated. We dug out the root of the several seedlings after the experiment to check the root condition in the pot. Although the roots reached the bottom of the pot and circled a little, we did not find intertwining roots. Therefore, we consider the restriction of root growth due to the limited soil volume was little.

Measurement of leaf gas exchange

The gas exchange rates of matured leaves in the upper canopy of the seedlings were determined during 22-30 July and 20-24 October 2014, using an open gas exchange system (LI-6400^®, Li-Cor Inc., Lincoln, NE, USA). Two (July) or three (October) seedlings were selected randomly in each chamber (i.e., six or nine seedlings in each gas treatment).

The PPFD during the gas exchange measurement was maintained at 1500 μmol m^-2 s^-1. Leaf temperature and leaf-to-air vapour pressure deficit were maintained at 28.0 ± 0.5 °C and 1.5 ± 0.2 kPa in July, and 25.0 ± 0.5 °C and 1.5 ± 0.2 kPa in October, respectively. To obtain the intercellular CO₂ concentration (C_i)-response curve of A, i.e., A/C_i curve, A was determined at ten CO₂ concentration levels in the chamber (C_a - [20]). Firstly, we measured the gas exchange rate under a stable condition of 400 μmol mol^-1 of C_a. Then, the C_a was decreased to 60 μmol mol^-1 in the following order: 300, 220, 140, and 60 μmol mol^-1. After the C_a was increased again to 400 μmol mol^-1 and the original A values were confirmed, we increased C_a in the following order: 600, 800, 1100, 1400, and 1700 μmol mol^-1. The A, stomatal conductance to water vapour, and ratio of C_i to C_a at C_a = 400 μmol mol^-1 CO₂ (A_sat, G_s, and C_i/C_a, respectively) were calculated from the gas exchange measurement under stable condition (i.e., the first measurement log of A/C_i curve). From the A/C_i curve, we calculated the stomatal limitation of photosynthesis at C_a = 400 μmol mol^-1 CO₂, the maximum rate of carboxylation (V_cmax), and the maximum rate of electron transport (J_max - [5], [20]). Values of the Rubisco Michaelis constants for CO₂ (K_c) and O₂ (K_o), and the CO₂ compensation point in the absence of dark respiration (Γ*) for analysis of the A/C_i curve were used according to Bernacchi et al. ([1]). Day respiration was considered as 2% of V_cmax ([36]). The values of V_cmax and J_max at leaf temperature of 25 °C were calculated using their temperature dependency ([36], [1]). Stomatal limitation (L_s) of photosynthesis was calculated as follows (eqn. 1):

where A_Ci400 is A at C_i = 400 μmol mol^-1 CO₂, and A_Ca400 is A at C_a = 400 μmol mol^-1 CO₂ (= A_sat - [20]).

Measurement of leaf nitrogen content and estimation of leaf nitrogen fraction in photosynthetic functions

After measurement of the gas exchange rate, we collected six leaf discs (diameter, 12 mm each) in order to determine the leaf mass per area (LMA) and nitrogen content. These leaf samples were dried in an oven for 5 days at 80 °C and weighed. The LMA was calculated as leaf dry mass divided by leaf area. The nitrogen content of the leaves per unit mass (N_mass) was determined using a C/N analyser (model MT-700^®, Yanaco, Tokyo, Japan). Four of six leaf discs were used for the determination. A calibration curve was generated using hippuric acid (N = 7.82% - Kishida Chemical Co., Ltd., Osaka, Japan). We calculated N_area as the product of LMA and N_mass.

We estimated fraction of leaf nitrogen in Rubisco (F_nr) and bioenergetics (electron carriers except for photosystems, coupling factor, and Calvin cycle enzymes except Rubisco, F_nb). The F_nr was calculated using the following equation ([27], [33] - eqn. 2):

where V_cr is the specific activity of Rubisco (the maximum rate of RuBP carboxylation per unit Rubisco protein), and the factor of 6.25 [g Rubisco (g N in Rubisco)^-1] converts nitrogen content to protein content. Here, V_cr is equal to 20.5 μmol CO₂ (g Rubisco)^-1 s^-1 at 25 °C for purified Rubisco enzyme from Spinacia oleracea ([11]). Because this method cannot involve inactivated Rubisco, the calculated value of F_nr is an underestimate ([37]). The F_nb was estimated from gas exchange characteristics according to the following equation ([16], [31] - eqn. 3):

We assumed that the nitrogen in bioenergetics is proportional to J_max, where the ratio of J_max to the cytochrome f content is 156 mmol mol^-1 s^-1 ([26]), and the nitrogen in bioenergetics per unit cytochrome f is 9.53 mol mmol^-1 ([6]).

Statistical analysis

Statistical analyses were undertaken using R software, version 3.4.0 ([30]). To test the effects of O₃ on leaf and growth parameters, we applied a generalized linear mixed model with average ozone concentration as a fixed-effect and difference of chamber as a random-effect. Log-transformation was applied to growth parameter before the analysis. Because we confirmed the normality for all parameters by Shapiro-Wilk test, response variables were assumed to follow Gaussian distribution in the model.

  Results 

The average concentrations of O₃ for 24 h and daylight hours during the experimental period were 10.7 and 13.0 nmol mol^-1 in CF treatment, 19.2 and 24.3 nmol mol^-1 in 1.0 time ambient O₃ treatment, and 27.6 and 35.8 nmol mol^-1 in 1.5 times ambient O₃ treatment, respectively (Tab. 1). Daylight AOT40 (accumulated O₃ exposure over a threshold of 40 nmol mol^-1 with solar radiation > 50 W m^-2) in CF, 1.0 time, and 1.5 times ambient O₃ treatments during the experimental period were 0.1, 4.1, and 14.1 μmol mol^-1 h, respectively.

Height and D²H of C. sieboldii seedlings was significantly decreased by exposure to O₃, while stem diameter remained unaltered (Tab. 2). There was a tendency of O₃-induced decrease in the number of 2013 leaves. We did not observe significant effects of O₃ on any growth parameters examined in Q. glauca and Q. myrsinaefolia seedlings.

There were no significant effects of O₃ on gas exchange parameters of all tree species in July, although the A_sat of Q. myrsinaefolia showed a tendency to increase after exposure to O₃ (Tab. 3). In October, we observed significant O₃-induced decreases in A_sat and V_cmax of C. sieboldii seedlings (Tab. 4). In addition, there was a tendency of increase in L_s of photosynthesis with increasing exposure to O₃. The exposure to O₃ did not affect any gas exchange parameters of Q. glauca and Q. myrsinaefolia seedlings in October. We found no significant effects of O₃ on LMA, leaf nitrogen content and leaf nitrogen fraction in photosynthetic functions in any of the tree species studied in July and October, except for N_mass of Q. myrsinaefolia in July and N_area of C. sieboldii in October (Tab. 5 and Tab. 6). There were tendencies of O₃-induced increase in N_mass of Q. myrsinaefolia in July and decrease in N_area of C. sieboldii in October.

  Discussion 

We observed a difference in O₃ susceptibility among the three tree species. According to the photosynthetic and growth responses to O₃, C. sieboldii showed higher susceptibility than the other two Quercus species. The higher susceptibility of C. sieboldii to O₃ is in agreement with the results of our previous studies ([17], [39]).

The difference of O₃ uptake into the leaves through the stomata is considered an important factor in determining the different susceptibilities to O₃ among the tree species ([12]). Stomatal O₃ uptake rate can be explained by stomatal conductance to O₃ (G_O3) when the meteorological condition and atmospheric O₃ concentration are the same. The G_O3 is proportional to G_s with a proportionality constant of 0.663 (i.e., G_O3 = 0.663 G_s - [23]). Thus, the difference of G_s among tree species is considered a good indicator of the difference of stomatal O₃ uptake. However, there was no great difference between G_s of C. sieboldii and that of the other two Quercus species (Tab. 3 and Tab. 4), indicating that the amount of O₃ uptake would not be largely different among these trees and would not serve as the main parameter to evaluate the difference in O₃ susceptibility.

Some Quercus species emit biogenic volatile organic compounds (BVOCs), mainly isoprene, and the BVOC may act as a detoxifying substance for O₃ and reactive oxygen species ([21]). However, the BVOC emission varied among the Quercus species. According to Tani & Kawawata ([32]), Q. glauca and Q. myrsinaefolia, which are tolerant species to O₃ in the present study, do not emit BVOC. Li et al. ([19]) suggested that LMA is a potentially useful indicator for evaluating O₃ susceptibility based on an analysis using 29 tree species, both deciduous and evergreen. In the present study, however, we observed quite similar values of LMA for the three species. Although the reasons for the different O₃ susceptibilities between C. sieboldii and the two Quercus species are not clear, it is clear that not all evergreen broad-leaved tree species are tolerant to O₃ ([48], [2], [19]). Further screening studies must be conducted to find other “O₃-sensitive” evergreen broad-leaved tree species, especially in the genus Castanopsis.

The decrease in A_sat of C. sieboldii was accompanied by a significant decrease in V_cmax, marginal decrease in G_s, and increase in L_s (Tab. 4). This indicates that both biochemical and diffusion processes in photosynthesis are responsible for the decrease in A_sat under elevated O₃, although the contribution of the former may be higher to some degree. O₃-induced decrease in V_cmax was reported in several Japanese deciduous tree species ([8], [9], [41], [42]). The V_cmax represents in situ activity of Rubisco ([5], [18]), and a decrease in Rubisco content under elevated O₃ was also reported in deciduous trees ([45], [38]).

In contrast to V_cmax, O₃-induced decrease of J_max in C. sieboldii seedlings was not observed (Tab. 4), although decreases in both V_cmax and J_max under elevated O₃ were reported in several studies of deciduous trees native to Japan ([7], [9], [41]). Kinose et al. ([14]) indicated that the timing of significant O₃-induced decrease in V_cmax was earlier than that in J_max of F. crenata seedlings. There is a possibility that the first effect of O₃ on the biochemical process of photosynthesis occurs in carboxylation by Rubisco.

Nitrogen is a nutrient that is strongly related to the biochemical assimilation capacity of plants ([18]). A large fraction of nitrogen in the leaves is incorporated into proteins associated with the photosynthetic process ([3]), and the biggest destination of nitrogen investment is Rubisco ([27], [31], [40]). In the present study, we observed marginally significant decrease in N_area and no significant decrease in F_nr of C. sieboldii seedlings under elevated O₃. However, the reduction rates of N_area and F_nr are similar, 8.5% and 9.7%, respectively, in 1.5 times ambient O₃ treatment compared to CF treatment. Therefore, we consider both factors (i.e., nitrogen content and nitrogen use efficiency to Rubisco) contribute to the decrease in V_cmax under elevated O₃

We observed elevated O₃-induced increase of L_s in C. sieboldii seedlings (Tab. 4). A similar result was observed in mature F. sylvatica ([15]), but it is in contrast with the results of many other studies of deciduous tree species. Hoshika et al. ([8]) reported the avoidance of O₃ uptake by stomatal closure in F. crenata saplings in early summer, but not in late summer and autumn. Plant physiological studies have shown several mechanisms of O₃-induced stomatal closure such as direct modulation of K⁺ channels ([34], [35]), change of Ca²⁺ homeostasis in guard cells ([22]), and production of phytohormones ([29]).

  Conclusions 

To our knowledge, this is the first study to analyse photosynthesis with respect to diffusion and biochemical processes in the leaves of East Asian evergreen tree species under elevated O₃. High O₃ susceptibilities on photosynthesis and growth in C. sieboldii were confirmed. On the other hand, the two Quercus species showed high tolerance against O₃. The negative effect of O₃ on net photosynthesis of C. sieboldii was due to decreases in both diffusion (increase in L_s) and biochemical processes (decrease in V_cmax). These results indicate that not only the decrease in carbon absorption capacity but also the change in water balance due to stomatal closure should be considered as risks for C. sieboldii forest under elevated O₃.

  Acknowledgments 

The authors are greatly indebted to Mr. Yoshinobu Fukamachi, Ms. Hiroka Hiroshima, and Mr. Shigeaki Okabe (Tokyo University of Agriculture and Technology) for their technical support, and to Dr. Hayato Iijima for the advice of statistical analysis. This study was supported by JSPS KAKENHI, Young Scientists B (15K16136 to MW), and Challenging Exploratory Research (15K12217 to TI and MW). We would like to thank Editage (⇒ http:/­/­www.­editage.­jp/­) for English language editing.

  References