Introduction 

Elevated CO₂ concentration has become a common phenomenon in the Earth’s atmosphere over the last half-century ([20]). The indiscreet use of fossil fuels and deforestation has raised the atmospheric CO₂ concentration, thus promoting photosynthesis and growth ([9], [1], [40]) and affecting the productivity of plants ([29], [32]), which is referred to as the CO₂ fertilization effect ([32]). However, there are conflicting opinions on whether such an effect will continue with increasing CO₂ concentration ([19]). In general, the productivity of terrestrial ecosystems is affected by nutrient availability in the soil ([46]). It should be taken into account that, if CO₂ fertilization promotes plant growth, the excessive nutrient use will result in insufficient nutrients in the soil, failing to meet the nutrient requirements needed for plant growth ([19], [51]) and decreasing the CO₂ fertilization effect. Therefore, the expected increase of ecosystem productivity and carbon storage due to CO₂ fertilization is still uncertain ([46]).

Early studies using pots and growth chambers hardly reflected the real forest ecosystem conditions ([7], [32]), therefore experiments with Open-Top chamber (OTC) and Free Air CO₂ Enrichment (FACE) techniques have been widely employed ([32]). Many OTC and FACE experiments reported an increased growth and photosynthesis under elevated CO₂ ([18], [45]). However, whether these effects would last for a long time is questioned. In particular, experiments on elevated CO₂ and soil nitrogen fertilization ([13]), which were conducted at the Duke-FACE and Oak Ridge-FACE, confirmed that nitrogen can directly limit the increase in forest productivity due to the carbon fertilization response, highlighting the relevance of nitrogen on forest productivity.

Nitrogen is an important component of plant photosynthetic organs and of functional and structural proteins ([21]). In particular, nitrogen in the leaves constitutes the photosynthetic enzymes, such as chlorophyll and rubisco (ribulose-1.5-biphosphate carboxylase/oxygenase), which determines the maximum carboxylation rate (V_Cmax - [27], [17]), with a decisive influence on forest productivity ([14], [15]). However, owing to increased productivity by long-term exposure to a high CO₂ concentrations, the consumption of nutrients may overcome the nutrients input, leading to a reduction in effective nutrients in the soil, and consequently to a reduction of plants’ photosynthesis ability. In particular, as the nitrogen concentration absorbed by plants decreases, its concentration in the leaves is lowered, thus decreasing photosynthesis. This in turn results in a progressive nitrogen limitation (PNL), which offsets the CO₂ fertilization effect ([9], [26], [4]). This phenomenon is more common under nitrogen deficiency in the soil ([35], [29]).

Under PNL, an increase in productivity due to the CO₂ fertilization effect is a short-term phenomenon, whereas in the long term, the ecosystem productivity might decrease again and return to its original state ([26], [13]). This suggests that increased forest productivity due to a high atmospheric concentration of CO₂ might be inhibited by soil nitrogen deficiency ([12]). Reich et al. ([37]) and Feng et al. ([12]) confirmed that the response of plants to CO₂ fertilization was limited by nitrogen deficiency. Moreover, PNL is expected to be more prominent when the high CO₂ concentration environment is maintained for a long time ([39]).

The main objective of this study was to determine the occurrence of nitrogen limitation under long-term CO₂ fertilization on three major species of the Korean temperate region: Korean red pine (Pinus densiflora), Chinese ash (Fraxinus rhynchophylla), and Korean mountain ash (Sorbus alnifolia). The effect of the increase in atmospheric CO₂ concentration on seedling diameter and height, as well as photosynthetic properties, such as maximum photosynthetic rate, maximum carboxylation rate, and maximum electron transfer rate, was evaluated over a period of 8 years. Moreover, we compared the nitrogen concentration in the leaves and soils to investigate whether the increased productivity due to CO₂ fertilization is a short-term effect owing to a progressive nitrogen limitation (PNL) in the soil.

  Materials and methods 

Experimental system and design

This study was carried out in an Open Top Chamber (OTC) of the Forest Biotechnology Division at National Institute of Forest Science in Suwon-si, Gyeonggi-do, Republic of Korea (37° 15′ 04″ N, 126° 57′ 29″ E). The OTC was made of a decagon structure of diameter 10 m and height 7 m (see Fig. S1 in Supplementary material). The angle of the roof was 45° in order to maintain an opening ratio of 75% or higher. A polyolefin film of thickness 0.15 mm with a light transmittance of approximately 88%, low specific gravity, and excellent chemical and water resistance was used as an external covering material ([23]).

The atmospheric CO₂ concentrations kept in the chambers during the experiment were: (i) the current atmospheric CO₂ concentration (Chamber 1, aCO₂); (ii) 1.4 times higher CO₂ concentration than the current concentration (Chamber 2, eCO₂1.4), which is expected to be the atmospheric CO₂ concentration by 2050, according to the IPCC scenario ([20]); and (iii) 1.8 times higher CO₂ concentration than the current concentration, which is expected to be the atmospheric CO₂ concentration by 2070 (Chamber 3, eCO₂1.8). The exposure of CO₂ were conducted for 10 hours (08:00-18:00) a day during the growing season (April-November). Pest controls were conducted every year and weed controls were conducted for first four years of the experiment. The annual average, minimum and maximum monthly temperature in chambers were 12.4 ± 0.2 °C, -1.3 ± 0.4 °C, 22.5 ± 0.2 °C, respectively. The annual rainfall range was 751.16 mm yr^-1 to 1975.96 mm yr^-1 over the 9 years of the experimental period.

In each chamber, the same three clones for each species were tested. Four-year-old seedlings of Korean red pine (Pinus densiflora), two-year-old seedlings of Chinese ash (Fraxinus rhynchophylla), and two-year-old seedlings of Korean mountain ash (Sorbus alnifolia) were planted in September 2009. The density within the OTC was 2547 seedlings ha^-1.

Growth measurements

Growth measurements (height and diameter) on trees started in May 2010 and were repeated every year until 2017 at the beginning (mid-April, mean daily temperature: ~10 °C) and the end (end of October, mean daily temperature: ~5 °C) of the growing season. The annual increase in growth was obtained as the difference in these two measures. Also, the lack of differences in diameter between the last measurement of the previous year and the first of the subsequent year was verified.

Tree height was recorded using a measuring rod (A-15, SENSHIN Industry Co., LTD., Osaka, Japan), while the tree diameter was measured twice in orthogonal directions 10 cm above the root collar using a digital Vernier callipers (CD-10CPX, Mitutoyo, Kawasaki, Japan), and the average value was used. Exceptionally, since 2016, Korean red pine and Chinese ash trees were measured using a diameter tape (F10-02DM, KDS, Malaysia) because of their large diameters.

Gas exchange measurements

The photosynthetic parameters were measured on tree leaves using a portable device LI-6400 (LI-COR Inc., Lincoln, NE, USA). From 2013 to 2017, except for 2015, the parameters were measured 1-3 times a year from June to August on 3 sunlit leaves for each species. The temperature of the leaves was set at 25 °C and the relative humidity at 55%-60%. The leaves were stabilized before measurements. The area of leaves was set at 6 cm² for all the species except Korean red pine. For this species the leaf area was recalculated by measuring the actual leaf area using a scanner after all measurements.

The light response curve was obtained using the photosynthetic rate recorded by sequentially varying the intensity of light irradiated on the leaves, using the following sequence in all the chambers: 1400, 1200, 1000, 800, 600, 400, 200, 100, 75, 50, 25, 0, and 1200 μmol m^-2 s^-1. The reference CO₂ supplied in each LI-6400 chamber was set at the atmospheric CO₂ in each OTC. The maximum photosynthetic rate (A_max) was estimated as the photosynthetic rate at light saturation (1200 μmol m^-2 s^-1) under the CO₂ concentration in the chambers, which was measured using the light response curve ([44]).

The A/C_i curve was achieved by varying the reference CO₂ concentration at light saturation point in the following order: (i) aCO₂: 400, 300, 200, 100, 75, 50, 25, 0, 400, 400, 600, 800, 1000, and 1200 μmol m^-2 s^-1; (ii) eCO₂1.4: 560, 400, 300, 200, 100, 75, 50, 25, 0, 560, 560, 800, 1000, and 1200 μmol m^-2 s^-1; and (iii) eCO₂1.8: 720, 600, 400, 300, 200, 100, 75, 50, 25, 0, 720, 720, 1000, and 1200 μmol m^-2 s^-1. The parameters V_Cmax and J_max were derived using the model proposed by Sharkey ([41]) and estimated in Excel™ spreadsheet version 2.0 (Microsoft, Redmond, WA, USA).

Leaf total nitrogen analysis

Three sunlit leaves were chosen on the same branch and used for photosynthetic measurements (n = 9 for each species, three repetitions for three species), and 1 cm² leaf disk was collected in 2017. The leaves were dried in a 70 °C in the lab for more than 72 h, then crushed using a FastPrep-24^® crusher (MP Biomedicals, Solon, OH, USA), homogenized, and finally analysed for nitrogen content using CHNS Analyzer Flash EA 1112^® (Thermo Electron Corporation, USA) at the National Instrumentation Center for Environmental Management (NICEM).

Soil nitrogen analysis

Before planting trees in 2009, the soil in the OTC was excavated to a depth of 1 m and replaced with forest soil to control the soil characteristics of all the treatment groups. Soil at depth < 30 cm was composed of forest soil and sand with a ratio of 1:1 ([23]). For all treatment groups, samples of soil from 0 to 5 cm depth were collected at five randomly selected points in September 2017, air-dried for 3 days at room temperature, and then analysed using the CHNS Analyzer Flash EA 1112 described above at the NICEM.

Statistical analysis

All statistical analyses were carried using the software R ver. 3.3.2 ([36]). Growth, photosynthesis variables and leaf total nitrogen over time were analyzed by repeated-measures ANOVA with CO₂ treatments as a fixed factor and the recording year as the repeated-measure. When significant CO₂ × Year interactions and CO₂ or Year effect were detected, means were compared using Tukey post-hoc comparisons. In addition, one-way ANOVA of individual species and Tukey post-hoc comparisons were conducted to assess the CO₂ effect for each year.

  Results 

To investigate the CO₂ fertilization effect over the period 2010-2017, diameter and height increments of trees of different treatments and species were recorded (Tab. 1, Tab. 2). The cumulative increments for height and diameter over the 8 years of the experiment are presented in Fig. 1.

Fig. 1 - Effects of elevated CO₂ concentration (aCO₂, Chamber 1; eCO₂1.4, Chamber 2; eCO₂1.8, Chamber 3) on average tree diameter (a, b, c) and height (d, e, f) growth per year from 2010 to 2017. There were no significant differences among CO₂ treatments (p > 0.05).

  Enlarge/Shrink   Download   Full Width  Open in Viewer

Diameter growth

Tab. 1 summarizes the annual diameter increase measured after CO₂ exposure. The results showed a significant increment in diameter growth under high CO₂ concentration due to the CO₂ fertilization effect (p <0.001 - Tab. 3). On average, the difference between aCO₂ and eCO₂1.8 was 47.7%, 25.3%, and 3.1% in Korean mountain ash (41.4 ± 6.2 vs. 61.1 ± 7.9 mm), Korean red pine (144.7 ± 15.1 vs. 181.3 ± 13.5 mm), and Chinese ash (70.9 ± 9.2 vs. 73.1 ± 5.5 mm), respectively.

The annual diameter increment of individual species under eCO₂ was significantly enhanced in all the species in 2010-2011 (maximum p = 0.029 - Tab. 3). However, since 2012, there was no significant difference in diameter increment under eCO₂. Korean red pine showed a significant difference in diameter growth under eCO₂ in 2011 in the order of aCO₂ < eCO₂1.4 < eCO₂1.8, an increase of about 100.0% (aCO₂ vs. eCO₂1.8, p = 0.015 - Tab. 1). Korean mountain ash showed significant difference under eCO₂ in 2010, an increase of about 57.9% (aCO₂: 5.7 ± 0.7 vs. eCO₂1.8: 9.0 ± 0.5 mm; p = 0.02).

Height growth

Similarly to growth in diameter, annual increase in tree height was recorded each year from 2010 until 2017 (Fig. 1). For all the species, the increase in height was high under eCO₂ (p < 0.01 - Tab. 2). On average, the difference between aCO₂ and eCO₂1.8 was 22.01%, 18.3% and 14.3% in Korean red pine (470.7 ± 22.8 vs. 574.3 ± 72.5 cm), Chinese ash (585.3 ± 39.1 vs. 692.3 ± 48.1 cm) and Korean mountain ash (299.3 ± 40.5 vs. 342.0 ± 8.2 cm), respectively.

Tab. 2 summarizes the annual growth of tree height measured after CO₂ exposure. For all the species, results were similar to the diameter growth. Chinese ash showed a significant difference in height growth under eCO₂ in 2010 (p = 0.029). Differences were observed under eCO₂1.8 and aCO₂, which was increased about 57.9%. Korean red pine showed a significant increase under eCO₂ in 2011, an increase of about 110.7% (p = 0.006). However, after 2012, there was no difference in the increase in height due to CO₂ concentration for all the species.

Maximum photosynthetic rate (A_max)

The average A_max under eCO₂ measured from 2013 increased in the following order: aCO₂ (12.7 ± 0.7 μmol m^-2 s^-1) < eCO₂1.4 (13.6 ± 0.8 μmol m^-2 s^-1) < eCO₂1.8 (14.9 ± 0.8 μmol m^-2 s^-1) but did not differ significantly (p = 0.06 - Tab. 3). In terms of response by year, the average A_max decreased by about 28.9%, from 15.9 ± 0.8 μmol m^-2 s^-1 in 2013 to 11.3 ± 0.9 μmol m^-2 s^-1 in 2016 (p < 0.001). However, these results showed differences among species and year (Fig. 2, Tab. 4).

Fig. 2 - Effects of CO₂ concentration (aCO₂, eCO₂1.4, eCO₂1.8) on maximum photosynthetic rate (A_max), maximum rate of carboxylation (Vc_max), and maximum rate of electron transport (J_max) in: (a, d, g) Korean red pine (Pinus densiflora); (b, e, h) Chinese ash (Fraxinus rhynchophylla); and (c, f, i) Korean mountain ash (Sorbus alnifolia). The values shown are mean ± SE. Different lowercase letters above the bars indicate significant (p < 0.05) differences from multiple comparison among CO₂ treatments. (ns): not significant.

  Enlarge/Shrink   Download   Full Width  Open in Viewer

Korean red pine and Chinese ash did not show a significant difference in the average A_max (minimum p = 0.094 - Tab. 4), but Korean mountain ash showed significant CO₂ effect. Korean red pine in all the years under eCO₂ showed no specific tendency. The A_max of Korean mountain ash was 14.0 ± 1.3 μmol m^-2 s^-1 under eCO₂1.8, which showed an enhancement from 10.7 ± 0.7 μmol m^-2 s^-1 under aCO₂ (p= 0.014). The difference in the A_max due to CO₂ exposure in each year did not show significant differences (Fig. 2c). The A_max of Chinese ash decreased in the following order: aCO₂ (12.4 ± 1.2 μmol m^-2 s^-1) < eCO₂1.4 (15.0 ± 1.7 μmol m^-2 s^-1) < eCO₂1.8 (17.4 ± 1.6 μmol m^-2 s^-1), and also decreased steadily from 2013 (19.4 ± 1.1 μmol m^-2 s^-1) to 2017 (11.7 ± 0.8 μmol m^-2 s^-1) by about 39.7%, though not significantly (p=0.094, p=0.095, respectively - Fig. 2b). However, the enhancement of A_max due to elevated CO₂ was 27.7% at eCO₂1.8, and the increase was constantly maintained at 28.6% under eCO₂1.8 in 2013 and 27.5% in 2017 relative to the rate under aCO₂.

Maximum rate of carboxylation (V_Cmax) and electron transport (J_max)

Unlike the maximum photosynthetic rate, the V_Cmax was decreased of about 13.6% from eCO₂1.8 (49.6 ± 3.4 μmol m^-2 s^-1) to aCO₂ (57.4 ± 3.5 μmol m^-2 s^-1) but did not differ significantly (p = 0.194 - Tab. 3). Similar to the A_max, the V_Cmax value by year was 68.0 ± 5.2 μmol m^-2 s^-1 in 2013 and it significantly decreased to 46.7 ± 2.6 μmol m^-2 s^-1 in 2017 (p = 0.03).

In terms of response by species and year, Korean red pine showed no significant differences with CO₂ and year (p = 0.554 and p = 0.452, respectively - Tab. 4). Contrastingly, V_Cmax showed a decreasing tendency of about 46.4% from 2013 (98.1 ± 8.1 μmol m^-2 s^-1) to 2017 (52.6 ± 5.9 μmol m^-2 s^-1). Depending on the CO₂ concentration, it also showed same decreasing trend by about 14.8% from aCO₂ (70.9 ± 7.1 μmol m^-2 s^-1) to eCO₂1.8 (60.4 ± 7.5 μmol m^-2 s^-1) and the difference between eCO₂1.8 and aCO₂ was the highest and significantly different in 2017 (p = 0.002 - Fig. 2d). The annual V_Cmax of Korean mountain ash decreased significantly by 23.5% from 55.0 ± 4.5 μmol m^-2 s^-1 in 2013 to 42.1 ± 2.0 μmol m^-2 s^-1 in 2017 (p=0.015). For Chinese ash, no statistical differences were found. However, its annual V_Cmax decreased under eCO₂1.8 by 55.7% compared with that under aCO₂ in 2015 (p = 0.03, Fig. 2e), but there was no difference thereafter. All species showed no significant differences in V_Cmax due to CO₂ enhancement (minimum p = 0.094).

The J_max showed no significant differences under eCO₂ and year (Tab. 3, Tab. 4). However, it had a decreasing tendency of about 23.4% from 2013 (100.2 ± 7.2 μmol m^-2 s^-1) to 2017 (76.8 ± 3.1 μmol m^-2 s^-1) with no significance. Depending on the CO₂ concentration, there was no significant differences and tendency (minimum p = 0.210). In terms of response by species, all species showed no statistical differences with CO₂ and year (Fig. 2g, Fig. 2h, Fig. 2i).

Leaf and soil N

The leaf total nitrogen was significantly different under eCO₂ and year (maximum p = 0.026 - Tab. 3). The leaf total N exposed to CO₂ significantly (p = 0.026) decreased in the following order: aCO₂ (1.57 ± 0.05 %) > eCO₂1.8 (1.28 ± 0.04 %) > eCO₂1.4 (1.24 ± 0.03 %).

Korean red pine showed significant differences in leaf total N under CO₂ exposure (p < 0.001 - Tab. 4). The leaf total N decreased significantly by about 24.4% from 2013 (1.31 ± 0.08 %) to 2017 (0.99 ± 0.03 %). The leaf total N under eCO₂ decreased in the following order: aCO₂ (1.45 ± 0.15 %) > eCO₂1.8 (1.21 ± 0.09 %) > eCO₂1.4 (1.07 ± 0.04 %) but it was not significant (p = 0.247). The eCO₂1.4 treatment showed the lowest leaf total N and the difference between aCO₂ and eCO₂1.4 were 31.0% in 2013, 27.9% in 2015, and 10% in 2017 (Fig. 3a).

Fig. 3 - Effects of CO₂ concentration (aCO₂, eCO₂1.4, eCO₂1.8) on leaf total nitrogen content (%) in (a) Korean red pine (Pinus densiflora), (b) Chinese ash (Fraxinus rhynchophylla), (c) Korean mountain ash (Sorbus alnifolia). The values shown are mean ± SE. Different lowercase letters above the bars indicate significant (p < 0.05) differences from multiple comparison among CO₂ treatments. (ns): not significant.

  Enlarge/Shrink   Download   Full Width  Open in Viewer

Chinese ash showed a significant decrease in leaf total N under eCO₂ and year (p = 0.021 and p < 0.001, respectively). The leaf total N under eCO₂ decreased significantly (p = 0.021) in the following order: aCO₂ (1.78 ± 0.04 %) > eCO₂1.4 (1.41 ± 0.04 %) > eCO₂1.8 (1.36 ± 0.05 %). In particular, the leaf total N was higher under eCO₂1.4 and eCO₂1.8 than under aCO₂ in 2013 (p < 0.001), but it became significantly lower under eCO₂1.4 and eCO₂1.8 than under aCO₂ over time (Fig. 3b). The difference between aCO₂ and eCO₂1.8 was 27.8% in 2016, and 18.8% in 2017 (maximum p = 0.04).

Korean mountain ash showed no significant differences under eCO₂ and year (minimum p = 0.522). But it decreased significantly by 15.4% under eCO₂1.4 compared with aCO₂ only in 2017 (p = 0.03 - Fig. 3c).

There was no significant difference in soil total N under eCO₂ in 2017 (p = 0.125). The absolute concentration of soil total N was very low, and therefore, the variation depending on CO₂ concentration was negligible (Fig. 4).

Fig. 4 - Effects of elevated CO₂ concentration (aCO₂, eCO₂1.4, eCO₂1.8) on the soil total nitrogen (%) in 2017. Values are mean ± SE. There were no significant differences among CO₂ treatments (p > 0.05).

  Enlarge/Shrink   Download   Full Width  Open in Viewer

  Discussion 

Relationship between the CO₂ fertilization effect and growth

Several studies have reported that elevated CO₂ concentration due to climate change directly affects photosynthesis, suggesting that eCO₂ promotes ecosystem production and biomass accumulation ([29], [32], [38]). For instances, the net primary production (NPP) increased by 22-32% in the Duke-FACE experiment, which conducted the CO₂ exposure for more than 10 years. In the Pop-FACE experiment, the diameter increased about 5% and the 12 species leaf area increased about 8% in seven FACE ([1], [13], [29]). In this study, the annual growth of the three speices was analysed, confirming a significant enhancement in height and diameter under eCO₂ for the first two years. However, the CO₂ fertilization effect was not sustained thereafter (Tab. 1, Tab. 2, Fig. 1). Temporary CO₂ fertilization effect was also reported by other studies. In the Oak-Ridge FACE and BioCON experiments, the reduction of CO₂ effect was offset after 6 and 3 years, respectively, and nitrogen fertilization showed an immediate increase in NPP ([37], [32]). In addition, there was no CO₂ effect in Picea abies growing on a very low-nutrient soil ([52]). Moreover, in the Duke FACE experiment, where the CO₂ fertilization effect remained for more than 10 years, soil N positively correlated with productivity increment under eCO₂ ([29]). Thus, soil nutrients and CO₂ fertilization effect appeared to be closely related. In particular, in our study site, the soil total N was very low because the nearby forest soil was mixed with sand and leaf total N decreased over time. For this reason, we argue that growth promotion through the CO₂ fertilization effect was not sustained.

Changes in photosynthetic properties under elevated CO₂ concentration

Elevated atmospheric CO₂ concentration increases the difference in the partial pressure of CO₂ between the atmosphere and leaf mesophyll tissues, increasing the maximum photosynthetic rate ([7], [2]). Twelve FACE experiments showed that the average A_max of all the tested species increased by about 31%. Depending on species, it was increased by about 37% in Pinus australis ([31]), 67% in P. taeda, and 62% in Liquidambar styraciflua ([10]). However, over time, the increase of A_max under eCO₂ was reduced by the acclimation of photosynthetic properties ([2]). In our study, the reduction of A_max increase under eCO₂ varied depending on the species (Fig. 2). Korean red pine and Korean mountain ash did not show a significant difference since 2013, which was only three years after eCO₂ exposure. Chinese ash showed the enhancement in 2013, but the increase in A_max was reduced over time. A_max of Korean mountain ash under aCO₂ was maintained over time, but it was decreased in Chinese ash, which might have been due to the low soil nutrient (Fig. 2, Fig. 4). This decrease in the A_max resulted in the down-regulation of the Vc_max and J_max. In other words, the A_max is usually measured under the CO₂ concentration of each treatment, therefore no difference or a decrease indicates that photosynthesis enhancement is not maintained, rather lowered over time. Similar to previous studies ([9], [28]), we observed a reduction in the Vc_max and J_max under eCO₂. This reduction usually occurs after long term CO₂ exposure ([2], [32]). In P. ponderosa exposed to eCO₂ for 6 years, the V_Cmax and J_max were decreased by about 36% and 21%, respectively ([47]), and the V_Cmax was decreased by about 19% in P. abies ([48]). However, in some cases, the reduction of photosynthesic ability under eCO₂ did not occur even in the long term (more than 6 years - [3], [8], [53]). In an experiment on Fagus sylvatica, Quercus petraea, Carpinus betulus, Acer campestre, and Tilia platyphyllos in mature deciduous forests, the increase in the A_max was maintained even after 8 years of CO₂ treatment, and the down-regulation of V_Cmax and J_max did not occur. Similarly, in the Aspen FACE experiment, the increases in Vc_max (48%-85%) and J_max (23%-34%) were sustained for 11 years, and no reduction occurred. Warren et al. ([53]) also showed that the increase in photosynthesis was maintained during the first 8 years when the leaf N was 0.2 mg cm^-2 or more, but after a decrease of leaf N, the CO₂ fertilization effect disappeared. Thus, the increase in photosynthesis and the decrease in photosynthesic ability are largely affected by the nitrogen available in the environment. These studies were conducted in well-developed organic layer of forest floor, such as mature stands ([3]), or soil N of 3% or more ([8]), i.e., environments with considerably higher N than normal forest soil. On the contrary, the long-term exposure under eCO₂ leads to nitrogen limitation in common forest soil, which results in PNL that decreases photosynthesis and growth increment ([37], [32], [12], [39]).

Relationship between photosynthetic properties (Vc_max and J_max) and leaf nitrogen

In general, the leaf total N is positively correlated with the photosynthesis ability ([11]). Several studies have shown that the leaf N decreased in response to long-term eCO₂ condition ([10]). The elevated CO₂ concentration increases the capacity of the rubisco enzyme to adsorb carbon dioxide, thereby causing carboxylation and a decrease in the demand of nitrogen for rubisco ([34], [25]). Similarly, in this study, the leaf N was significantly lowered under eCO₂ over time (Fig. 3). In addition, studies have shown that the Vc_max and photosynthesis ability, decreased as the leaf N decreased ([1], [53]). Thus, fertilization at the time of nitrogen limitation increased the CO₂ fertilization effect again or the decreasing rate was reduced ([24], [6], [46]).

In general, there is a strong correlation between V_Cmax and J_max, with J_max/V_Cmax = 1.5-2.0 at 25 °C ([30], [50], [5]). Similarly, V_Cmax, J_max, and leaf N showed correlation under eCO₂ ([6], [53]). Therefore, when the leaf N and V_Cmax were decreased after long-term eCO₂ treatment, the J_max also decreased in some cases ([16], [30], [43], [6], [3]). As in our study, the decrease in J_max (Tab. 4) was relatively lower than that of Vc_max in some cases ([24], [53]). Moreover, these changes are different depending on species. Chinese ash and Korean red pine showed a decrease only in the V_Cmax, while Korean mountain ash did not show any decrease in both the above photosynthetic parameters (Fig. 2). These differences due to the change in intracellular nitrogen distribution under eCO₂ are dependent not only on the environment but also on the species. Further studies are needed to examine how species change the intercellular nitrogen distribution under elevated CO₂ ([11], [17]).

CO₂ fertilization effect and PNL

The enhanced productivity of forest ecosystem due to the CO₂ fertilization is heavily influenced by the nutrients available in the soil ([13]). PNL is a hypothesis that the increased forest productivity due to eCO₂ decreases over time because of the increased N accumulation in the biomass, resulting in a decrease of soil N availability, an increase of N immobilization and a decrease of N mineralization ([26]). In the long-term eCO₂ at Duke-FACE, Oak-Ridge FACE, and BioCON experiments, N fertilization led to an immediate increase in NPP. Therefore, there seemed to be an interaction between eCO₂ and N in terms of increased productivity ([37], [29], [32]). In particular, the soil N is important and N deficiency leads to a decline in production due to PNL even under eCO₂ ([33]). Such N deficiency changes the N distribution in the plants’ organs for effective nitrogen utilization, resulting in a decrease of the N used for above-ground photosynthesis ([35]). In addition, N distribution in photosynthetic apparatus such as Rubisco is greatly reduced. Therefore, nitrogen-use-efficiency is increased generally ([49], [42]), but the persistence of such an increase is controversial ([2], [22]). There is limited information about N distribution in the above- and below-ground under eCO₂. Therefore, it is necessary to determine and quantify the distribution of N in the above- and below-ground through follow-up studies.

  Conclusions 

In this study, we investigated the physiological and morphological characteristics of Korean red pine, Chinese ash, and Korean mountain ash, which are native tree species in Korea, under eCO₂ over a period of 8 years. We also examined the longevity of CO₂ fertilization effect in our study sites.

The CO₂ fertilization effect caused by eCO₂ led to an increase in growth in the early stage of exposure, but there was no significant difference thereafter. Photosynthetic properties showed a decreasing tendency in all species and a down-regulation of photosynthetic capacity increase with time, especially in Korean red pine and Korean mountain ash. The analysis of leaf and soil N to identify the cause revealed a significant decrease of leaf N under eCO₂. We argue that the enhancement of productivity might have decreased due to low soil N.

In conclusion, progressive nitrogen limitation (PNL) caused by the N reduction might have already started or is about to start in our study sites. Further investigation is needed to clarify N use efficiency and nitrogen distribution according to species.

  List of abbreviations 

(OTC): Open Top Chamber; (aCO₂): current atmospheric CO₂ concentration ; (eCO₂1.4): 1.4 times higher CO₂ concentration than the current concentration (Chamber 2); (eCO₂1.8): 1.8 times higher CO₂ concentration than the current concentration (Chamber 3); (A_max): maximum photosynthetic rate; (V_Cmax): maximum rate of carboxylation; (J_max): maximum electron transport rate.

  Acknowledgments 

This research was funded by the National Research Foundation of Korea, by the projects “Evaluation of climate change adaptation mechanisms of native temperate tree species and its ecological effects using open-top chambers” (2014R1A1A2055127) and “Testing progressive nitrogen limitation and nitrogen use efficiency increment through the quantification of leaf nitrogen allocation” (2017R1A2B2012605). We thank members of forest ecophysiology laboratory for their assistance with field sampling.

  References

  Supplementary Material

Fig. S1 - The Open Top Chamber (OTC) user in the experiment.