Ethylenediurea (EDU) effects on Japanese larch: an one growing season experiment with simulated regenerating communities and a four growing season application to individual saplings

Japanese larch (Larix kaempferi (Lamb.) Carr.) and its hybrid are economically important coniferous trees widely grown in the Northern Hemisphere. Ground-level ozone (O3) concentrations have increased since the pre-industrial era, and research projects showed that Japanese larch is susceptible to elevated O3 exposures. Therefore, methodologies are needed to (1) protect Japanese larch against O3 damage and (2) conduct biomonitoring of O3 in Japanese larch forests and, thus, monitor O3 risks to Japanese larch. For the first time, this study evaluates whether the synthetic chemical ethylenediurea (EDU) can protect Japanese larch against O3 damage, in two independent experiments. In the first experiment, seedling communities, simulating natural regeneration, were treated with EDU (0, 100, 200, and 400 mg L−1) and exposed to either ambient or elevated O3 in a growing season. In the second experiment, individually-grown saplings were treated with EDU (0, 200 and 400 mg L−1) and exposed to ambient O3 in two growing seasons and to elevated O3 in the succeeding two growing seasons. The two experiments revealed that EDU concentrations of 200–400 mg L−1 could protect Japanese larch seedling communities and individual saplings against O3-induced inhibition of growth and productivity. However, EDU concentrations ≤ 200 mg L−1 did offer only partial protection when seedling communities were coping with higher level of O3-induced stress, and only 400 mg EDU L−1 fully protected communities under higher stress. Therefore, we conclude that among the concentrations tested the concentration offering maximum protection to Japanese larch plants under high competition and O3-induced stress is that of 400 mg EDU L−1. The results of this study can provide a valuable resource of information for applied forestry in an O3-polluted world. Electronic supplementary material The online version of this article (10.1007/s11676-020-01223-6) contains supplementary material, which is available to authorized users.


BACKGROUND
Global change forced scientists to study effects of changing environment on vegetation (e.g. Matyssek et al. 2013). In the need to investigate tropospheric carbon dioxide (CO 2 ) and nearsurface ozone (O 3 ) effects on plants, several engineering design approaches have been developed (Kobayashi 2015). The first approach is by using closed systems, which are different kinds of closed facilities restricting interactions with the natural environment (Menser and Heggestad 1964;Hill 1967;Berry 1970;Rafarel and Ashenden 1991). The second approach is by utilizing semi-open systems, which are more realistic than closed chambers; the most commonly used method of semi-open systems is open-top chambers (OTCs, Barbee et al. 1973;D'Andrea and Rinaldi 2010). An important shortcoming of the first and second approaches is the isolation of plants from natural environment where many uncontrolled biotic (e.g. insects) and abiotic factors (e.g. temperature, moisture, light, wind speed) influence or interact with the plants.
Notably, plants are often grown in pots where the root system is restricted, imbedding, thus, authentication of the findings with what would be observed under unrestricted environment. In order to overcome such shortcomings, the third approach, open air systems, has been developed. This approach includes mainly Free Air Controlled Exposure (FACE) systems (Werner and Fabian 2002;Karnosky et al. 2008;Norby and Zak 2011;Tang et al. 2011;Matyssek et al. 2013;Koike et al. 2015 for a list with information on the global CO 2 FACEs) and, more rarely though, localized exposure systems with tubes wrapped around tree branches (Velikova et al. 2005). It should be noted that all methods have advantages and disadvantages (Unsworth et al. 1984;Macháčová 2010); for example, O 3 -FACE systems require much higher financial support.
O 3 -FACE is a method used by plant biologists, ecologists and ecophysiologists to enrich air with O 3 in a particular experimental area and allow the response of plant growth to be measured without restricting root growth and in interaction with the natural environment, in contrast to closed or semi-open systems. Severe negative effects of elevated mixing ratios of O 3 on vegetation urged dozens of scientists to engage O 3 research with annual or perennial plants (Agathokleous et al. 2015a(Agathokleous et al. , 2015b(Agathokleous et al. , 2016aAinsworth et al. 2012;Koike et al. 2013;Harmens et al. 2015;Chappelka and Grulke 2016;Sicard et al. 2016;Wang et al. 2016). However, opposed to the development of numerous CO 2 FACE systems around the globe , only few O 3 FACE systems for trees, with replicated experimental unit, have been operated -the most research has been carried out in closed or semi-open facilities (Wulff et al. 1992;Oksanen 2001;Nunn et al. 2002;Werner et al. 2002;Karnosky et al. 2003;Matyssek et al. 2013;Kitao et al. 2015;Paoletti et al. 2017).
O 3 impacts into vegetation are influenced by soil fertility (Agathokleous et al. 2016a;Shi et al. 2017). Thus, the influence of soil fertility and plant mineral status should be critically studied along with O 3 . For example, a large proportion of global potential arable soils are infertile in phosphorus, while severe phosphorus scarcity is expected in future (Zheng 2010;Cordell et al. 2011;Cordell and Neset 2014;Reijnders 2014). Despite the importance of soil, and obviously for technical difficulties, there is not hitherto an O 3 -FACE system for trees that employs soil as additional factor. Meanwhile, O 3 effects on plant may vary with growing substrate types, and fertility and properties of soil (Agathokleous et al. 2016a), something that is particularly important during the early stages of forest succession (Bazzaz 1996).
An O 3 FACE system that would allow the investigation of the effects of elevated O 3 on trees growing as a community in different kinds of soil was developed. This system was designed to simulate the early stages of the forest succession and thus is of particular importance to forest ecology research.

Experimental area
The O 3 -FACE system was established at Sapporo Experimental Forest of Hokkaido University, located at Sapporo, Japan (43 o .04' N, 141 o .20' E, 15 m a.s.l.), in the year 2014. This location falls in a transition zone from cool temperate to boreal forests which is part of the Asian boreo-nemoral ecotone with relative sensitivity to global climate change (Uemura 1992;Matsuda et al. 2002). The experimental forest stand is located in the campus of Hokkaido University, in a central area of Sapporo city.

Meteorological conditions
The snow-free period for 2014 and 2015 lasted from early May to mid-November. Air temperature, wind speed, relative humidity, sunshine duration and total precipitation (Table S1, see the bottom of this document) were monitored by a nearby station at Sapporo (WMO, ID: 47412, 43 o 03.6'N 141 o 19.7'E), which is operated by the Japan Meteorological Agency (2016).

Experimental design
The O 3 treatments were ambient (AOZ) and elevated (EOZ) O 3 , with three site replicates for each treatment. Each, approximately circular, plot (diameter=6.5 m) of AOZ and EOZ was surrounded by a metallic structure which consisted of six vertical metallic pipes (height=5m). The vertical metallic pipes were interconnected with horizontal metallic pipes at 1-m vertical interval. The architectural design of the free-air O 3-enrichment plots and the metallic structure of each plot are illustrated in Fig S1 and S2, respectively.

Fig S1.
Architectural design of a free-air O 3-enrichment plot. The peripheral thicker lines (black in color version) represent the outer metallic scaffolding of the plot; around it a Teflon tube (at 0.5 and 1.5 m above ground level) was releasing O 3 -enriched air in the plot. The fine lines (pink in color version), which form the two inner hexagons, represent the fixed nest on which the vertical Teflon tubes releasing O 3 -enriched air (filled circles of larger size; blue in color version) were spatially arranged. Each vertical tube was connected with a horizontal tube, fixed on the nest, which reached the edge of the plot and then became vertical to connect with the buffer tank. The 21 smaller size filled circles (red in color version) indicate the position of the poles on which passive samplers were fixed at the heights of 1 and 1.5 m above ground level.

3/14
Within each O 3 plot, three types of soil were employed: a) brown forest soil (BF: Dystric Cambisols); b) BF mixed with immature volcanic ash plus pumice soil at a rate of 1:5 v/v (VA; Vitric Andosols); and c) BF mixed with serpentine soil at a rate of 1:5 v/v. BF and VA parts covered an area of approximately 42 % each while serpentine soil covered an area of approximately 16 % ( Fig S3). Since BF is native to the experimental plots, soil in the surface 30 cm layer was removed from each soil site. For BF, the same soil was stirred and placed back. For VA and serpentine, the removed soil was mixed with volcanic ash or serpentine soil which was brought from Tomakomai Experimental Forest of Hokkaido University (Kim et al. 2011) and the eastern part of Teshio Experimental Forest (Watanabe et al. 2012), respectively, where they are native. BF is typical type of soil for deciduous broad-leaved forests which is fertile enough to support efficient plant growth (Kayama et al. 2011, Kim et al. 2011).

Fig. S3
Spatial arrangement of the soils and tree species in the experimental plots. The soil represents brown forest soil (BF; Dystric Cambisols)) or BF mixed with immature volcanic ash plus immature volcanic ash soil (VA; Vitric Andosols) which both have the same area. If this is BF, then VA is at the opposite site (right hand). A smaller area existed at the south (bottom of the illustration) which represented brown forest soil mixed with serpentine soil (serpentine). Serpentine differs from BF and VA in that there is no willow and only four individuals are planted per tree species. To separate run-off of serpentine to BF and VA, a plastic sheet was implanted into the soil, at a depth of 30 cm. The 2yr-old tree seedlings planted were Siebold's beech (Fagus crenata Blume), Japanese white birch (Betula platyphylla var. japonica), Mizunara oak (Quercus mongolica var. Volcanic ash soils are acidic, P deficient and N poor (Schmincke 2004;Kayama et al. 2011;Kam et al. 2015). Serpentine soils (derived from serpentine rock) have high pH and are characterized by high content in metals, such as Mg, that are harmful to plants and low content in essential elements for plants, such as Ca (Brady et al. 2005;Kayama et al. 2006;Kayama and Koike 2015). Earlier, for a decadal period of time (until 2013), these plots had been used for a free-air CO 2 -enrichment (FACE) system Agathokleous et al. 2016c). The plots were established on 13 th May 2014.
On the same day (after the preparation of the plots and before the plantation), the pH of the soils was measured using a pH-meter (B-712, HORIBA, Kyoto, Japan). Three pH measurements were taken from the water used. Then, twelve samples of 50 g each prepared for each soil (i.e. 36 samples in total), from six FACE plots (two samples per FACE). Five soil samples of 10 g each was taken randomly from each plot, and were combined into 50 g of soil samples. The pH was measured for a mixture of 1:2 soil and water. Each sample was stirred for 30 sec every 5 min, for 30 min, and then left for some times to stabilize before the measurement.
On 15 th May 2014, seedlings of five species were planted in the three soils: (a) Siebold's beech (Fagus crenata Blume); (b) Japanese white birch (Betula platyphylla var. japonica); (c) Mizunara oak (Quercus mongolica var. crispula); (d) Japanese larch (Larix kaempferi (Lamb.) Carr.); and (e) hybrid larch F 1 (Larix gmelinii var. japonica (Maxim. ex Regel) Pilg. × L. kaempferi). In mid. June, cuttings of willow (Salix udensis Trautv. & C.A.Mey.) were planted in BF and VA soils. Siebold's beech, Japanese white birch, Mizunara oak and the willow are deciduous broad-leaved species while the two larches are deciduous conifers. All the broad-leaved species are native to Hokkaido island, Japan. The spatial arrangement of soils and tree species in the plots are illustrated in Fig S3. Information on the properties of BF and VA soils can be found in Watanabe et al. (2013a) and Eguchi et al. (2005Eguchi et al. ( , 2008. Details on the soil properties of BF, VA and serpentine sites can be found in Shi et al. (2017). Overall, BF soil has lower Cr content than VA and serpentine soils; VA soil has higher Fe and Al content than BF and serpentine soils; serpentine soil has higher Mg, Ni, Cr and Mg/Ca ratio and lower K, Ca, Mn, Al and N content than BF and VA soils.

2.4.
Ozone treatment This FACE system employs the method of O 3 exposure used at Kranzberg Forest in Germany (Nunn et al. 2002;Werner and Fabian 2002). This system is also similar to the O 3 FACE towers established in the same experimental forest (Watanabe et al. 2013b), however the latter tower-type system was lacking replicated experimental units and its operation was discontinued in 2016. Recently, Paoletti et al. (2017) also reported an O 3 FACE system in Italy which implements this O 3 exposure methodology.
Ambient O 3 concentrations were monitored continuously by an O 3 monitor (Model 202, 2B Technologies, Boulder CO, USA) of which the measurement principle is based on UV Absorption at 254 nm. The O 3 monitor was located at a distance of approximately 20 m from the AOZ plots and was recording O 3 data at a 1 m interval.
The O 3 treatments were ambient O 3 (AOZ) and elevated O 3 (EOZ). The efficiency of the system was assessed in the first two years of its operation. Plants were exposed to EOZ from August 15 th to October 26 th in 2014 and from April 24 th to October 26 th in 2015, during the daytime. This system was initially designed to maintain EOZ at a target mixing ratio of 70 nmol mol -1 , about two times the ambient O 3 mixing ratio at the experimental area. However, the attainment capacity of this system may deviate importantly under stronger winds which commonly occur at the experimental area (Watanabe et al. 2013b). Therefore, to approximate the target, the target mixing ratio was set to 80 nmol mol -1 . When the target level is lower than the target one, opening of the three-way control valve increases to supply more O 3 . When the target level exceeds 80 nmol mol -1 , the opening of the three-way control valve decreases (closing); then the O 3 level in the plot decreases.

5/14
Ozone generated from pure oxygen was diluted with pressurized ambient air and released into the rings by eighteen 85 % virgin Teflon tubes (length = 2.5 m, inner diameter = 6 mm) hanging down from a nest fixed above the plants (2.5 m above ground level), at fixed distances among them, and two Teflon tubes (length = 21 m, inner diameter = 6 mm) fixed horizontally around the plot at 0.5 and 1.5 m above ground level for each EOZ plot. However, the nest can be unfixed and fixed at a higher position, according to the plant height. Teflon is a known formula based on polytetrafluoroethylene (PTFE), a synthetic fluoropolymer of tetrafluoroethylene.
Each tube had one hole every 50 cm (i.e. one 2.5 m long tube had five holes), and an end-cap at the bottom. The end-cap had also a hole to avoid accumulation of humidity and maintaining pressure in the tubes. Each of the eighteen vertical tubes was connected at their top with a horizontal tube fixed on the nest (until the edge of the plot when it was again vertical) which was in turn connected to a buffer tank. All the holes had a diameter of 1.5 mm.
The blowing amount of air in the plot was 1.53 L min -1 . The spatial arrangement of the tubes in the EOZ plots is schematically represented in Fig S1. The buffer tank was connected, with two Teflon tubes, to the mixing tanks in which the air containing O 3 was diluted with ambient air before passes into the buffer tank and thereby into the horizontal and vertical air-releasing tubes of the experimental plot. Each EOZ plot had its own control station in which all the instruments were established.
An SM70 Fixed Ozone Monitor (Aeroqual Ltd., Auckland, NZ) was installed in the center of each EOZ plot, in an SM70 enclosure (Aeroqual Ltd., Auckland, NZ) which is pre-drilled and fitted with a cable gland for facilitating a cable connection to the SM70 relay and data output connectors. The SM70 enclosure is placed on a metallic pipe so as the measuring height to not be fixed and can be adjusted to plant height requirements. Air is sampled, and the SM70 response time is one minute. Calibration is done with zero gas (no O 3 ) and 1000 nmol O 3 mol -1 (as a span gas). The signal of the O 3 sensor is used to control O 3 at the target level through a proportional-integral-derivative (PID) control loop feedback mechanism (controller) algorithm.
The O 3 -enrichment control system was controlled with a Digital Indicating Controller SDC31 (Azbil Corporation, Tokyo, JP) which has an accuracy of ±0.2% FS and offers a standard PID control and an advanced neural/fuzzy PID. This PID performs process diagnostics and reduces overshoot. A Hioki LR5042-20 (HIOKI E.E. Corporation, Nagano, JP) recorded O 3 data on a two-minute interval. Data were wirelessly transferred and stored into a data logger (LR5092, HIOKI E.E. Corporation, Nagano, JP). A light sensor (EE4313, Panasonic Industrial Devices SUNX Co., Ltd., Aichi, JP) was installed outdoors the control room of each EOZ plot and connected with the control system in order to provide feedback on the hours of active O 3 enrichment. In a similar way, a wind speed sensor was installed on the top of the metallic structure of each EOZ plot and connected with a Digital Alarm Combined Wind Vane Anemometer (Isuzu Seisakusho Co. Ltd., Tokyo, JP) and the control system in order to regulate the opening/closing of the three-way control valve for O 3 supply. When the wind speed was higher than 5 m s -1 , the O 3 -enrichment system stopped. In the worst scenario that O 3 concentration in the plot reached much higher levels than the target concentration, the maximum closing time of the control valve was 3 min. The O 3 control system of each EOZ plot is autonomous, meaning that there was one control system for each plot.

6/14
In order to check the efficiency of the O 3 -enrichment system, 21 poles were installed in an EOZ ring (Fig S1), each of which had two Ogawa O 3 passive samplers (Ogawa & Co., Ltd, Kobe, JP): one at the height of 1 m (around the lowest part of the canopy) and the other at the height of 1.5 m (around the top of the canopy). Each passive sampler was setup in an opaque plastic shelter (height = 14.5 cm, radius = 4 cm) which was further covered with aluminum foil; all shelters had the same size. Passive sampling lasted for 15,650±41 (hereafter mean ±S.D.) min. The data of the passive sampling were converted to mixing ratio by using the following conversion equation: where 3 is the nmol mol -1 mixing ratio conversion coefficient (nmol mol -1 min ng -1 ), 3 is the O 3 quantity (ng) converted from the NO 3 quantity collected and t is the exposure duration (min). The 3 was calculated using the equation: where T is the ambient temperature in degree Centigrade and t is the exposure duration in minutes. The equations are shown in Takeda and Komatsu (2007).
The 10-hour (07:00-17:00) mean O 3 levels of AOZ treatment were 20.66±5.52 and 34.7±7.55 nmol mol -1 in 2014 and 2015, respectively; S.D. indicates hour-to-hour fluctuations across the season. Notably, plants in AOZ condition are exposed to practically zero AOT40 (index of accumulated exposure to O 3 over the threshold of 40 nmol mol -1 , Mills et al. 2007  . In this FACE system, individual plants have been hitherto grown in pots Zhang et al. 2017). The third and newest one is the FACE system located in Yanqing district, Beijing, China (116.0°E,40.5°N), in an area of 4 ha, which includes a Denuder system for long-term atmospheric sampling (DELTA). This FACE system employs two O 3 levels, one at ambient O 3 concentration and another set at 1.5 times time ambient O 3 level. Each O 3 treatment has four plot replicates. The size of each plot is 16×16 m (Feng, personal communication). In this FACE system, communities of poplars are growing directly into the ground (no pot-grown).
The FACE system of Hokkaido University, Japan, is the first system in Asia (Koike et al. 2013) and the only currently exists in the global for exposing communities of tree saplings of different taxa to O 3 -enriched atmosphere. This system provides the minimum influence of the facilities on the natural environment, thus permitting studies on the interactions with the natural environment, such as plants and insect herbivores. Still, this is the first O 3 FACE employing soil as an additional factor and this is of benefit. The location of this FACE system is a further advantage: The technology which is available nowadays does not permit FACE systems to reduce the ambient O 3 levels which are elevated in several parts of the world. In contrary to the other available FACE systems, this FACE system is located in an area with relatively low ambient O 3 levels and thus provides a more realistic comparison in the sense that plants in areas with already highly elevated O 3 levels may be under background stress.

Table 1
Monthly mean air temperature, wind speed and relative humidity and monthly total sunshine duration and precipitation at Sapporo, Japan, for the months May-November of the experimental years 2014-2015. For each year, the median ±MAD of the monthly values are presented. MAD denotes the median absolute deviation, which is the median of the absolute deviations from the data median: = �� − � ���.