The experimental tree species were Korean pine (Pinus koraiensis Siebold & Zucc.) and Japanese cypress [Chamaecyparis obtusa (Siebold & Zucc.) Endl.]. The P. koraiensis seedlings were 5 years old (1-1-3, indicating 1 year in the seedbed, 1 year in the first transplant bed, and 3 years in the field), with an average height of 102.3 cm (±6.32) and a root collar diameter of 14.63 mm (±1.72). The C. obtusa seedlings were 3 years old (1-1-1, indicating 1 year in the seedbed, 1 year in the first transplant bed, and 1 year in the field), with an average height of 100.3 cm (±9.69) and a root collar diameter of 11.03 mm (±1.61). Both species were cultivated at a nursery in Gyeonggi Province, South Korea, prior to the experiment.

On April 5, 2024, the seedlings were transported to Seoul, South Korea, where they were transplanted into 8 L plastic pots. The potting medium used was a commercial forest soil mixture comprising cocopeat (24.8%), peat moss (25%), perlite (25%), and vermiculite (25%). On April 16, 2024, 10 seedlings per species were moved into separate phytotron chambers (1.5 × 1.5 × 2.0 m; Koito Industries, Yokohama, Japan), which were natural light-type units designed to control temperature, humidity, and CO₂ levels, ensuring consistency throughout the experiment.

Environmental conditions were carefully regulated to meet experimental requirements. Temperature settings were adjusted based on seasonal variations: from April 26 to July 14, daytime (8:00–20:00) and nighttime (21:00–7:00) temperatures were maintained at 20°C and 15°C, respectively. From July 15 to August 10, the daytime and nighttime temperatures were raised to 25°C and 20°C, respectively, representing a 5°C increase. The relative humidity was maintained at 50–60% throughout the experiment, monitored using the phytotron control system and a HOBO MX2301A data logger (ONSET Computer Corporation, USA). Light intensity was measured using a quantum sensor (PQ-500, Apogee Instruments, Logan, UT, USA). Maximum light intensity, recorded between 12:00 and 14:00, reached ~1,500 μmol m⁻¹ s⁻¹ (PPFD).

The CO₂ concentration in the phytotron chamber was monitored and regulated using an integrated CO₂ control system (ZFU-DF23, Fuji, Tokyo, Japan) installed in the chamber control unit. High-purity CO₂ gas (99.99%) was supplied from 47 L cylinders and delivered through PTFE tubing to the control system. The ZFU-DF23 maintained CO₂ levels in real time using a mass flow controller (MFC). This closed-loop system minimized fluctuations, ensuring consistent 24-h control (Figure 1).

For the CO₂ treatment, the ambient CO₂ condition (aCO₂) was set at 400 ppm as the control, while the elevated CO₂ condition (eCO₂) was applied as the treatment. During the baseline acclimation period (April 16–May 26), both chambers were maintained at 400 ppm CO₂ for baseline acclimation. Starting in the 1st week of treatment (May 27), the eCO₂ chamber was set to 550 ppm, while the aCO₂ chamber remained at 400 ppm. From the 4th week of treatment (June 19), the CO₂ concentration in the eCO₂ chamber was elevated further to 900 ppm, while the aCO₂ chamber continued to be maintained at 400 ppm. Daily, the chamber doors were opened for 10–20 min at 8:40 a.m. to allow ventilation.

Irrigation was performed weekly with 3 L of water per pot, corresponding to the field capacity of the soil, to provide sufficient moisture while avoiding waterlogging. Pots were randomized weekly within the chambers to minimize positional effects. No fertilizers or additional supplements were applied during the experimental period.

Gas exchange measurements were conducted using a LI-6400XT Portable Photosynthesis System (LI-COR Biosciences, Lincoln, NE, USA). Net CO₂ assimilation rate vs. intercellular CO₂ concentration (A/Ci) response curves were generated by adapting the seedlings to a series of CO₂ concentrations (0, 30, 80, 100, 400, 550, 900, and 1,200 μmol mol⁻¹). The flow rate inside the leaf chamber, block temperature, chamber humidity, and photosynthetic photon flux density (PPFD) were maintained at 500 μmol s⁻¹, 25°C, 50–60%, and 1,500 μmol m⁻² s⁻¹, respectively. Three time points were selected for measurements: during the 3rd week, 6th week, and 10th week of the experiment, between 08:00 and 13:30.

The A/Ci response curves were used to estimate key photosynthetic parameters, including the maximum carboxylation rate (V_cmax), maximum electron transport rate (J_max), triose phosphate utilization (TPU), and carboxylation efficiency (CE). V_cmax, J_max, and TPU were calculated using the biochemical model of photosynthesis described by Farquhar et al. (1980) and modified by Long and Bernacchi (2003). CE was determined following the method outlined by Tenhunen et al. (1984).

Additionally, at a CO₂ concentration of 400 μmol mol⁻¹, gas exchange parameters, including net assimilation rate (A_net), stomatal conductance (g_s), transpiration rate (Tr), and water use efficiency (WUE), were measured following the methods of Kwak et al. (2011). All gas exchange measurements were conducted on needle pairs positioned 3 to 5 nodes below the shoot apex. Six time points were selected for measurements: initial, 1st week, 2nd week, 4th week, 5th week and 10th week of the experiment, between 08:00 and 13:30.

To normalize gas exchange parameters, leaf area was measured for both species after gas exchange measurements. For P. koraiensis, needle width was set at 1 mm based on typical dimensions, and individual needle lengths were measured. Total leaf area was calculated by multiplying the width by the measured lengths of all needles placed inside the chamber. For C. obtusa, leaves were excised and photographed on a white background with a size reference scale. Photoshop software (Adobe, San Jose, CA, USA) was used to calculate total leaf area.

The actual leaf area enclosed within the Li-6400XT chamber was measured and used to normalize all gas exchange parameters, ensuring that V_cmax, J_max, TPU, CE, A_net, g_s, Tr, and WUE were expressed on a per-leaf-area basis. This normalization allowed for accurate comparisons of physiological responses between species and treatments.

The contents of Chl a (chlorophyll a), Chl b (chlorophyll b), Total Chl (total chlorophyll), and Car (carotenoid) were measured following the methods described by Arnon (1949) and Lichtenthaler and Buschmann (2001). Fresh leaves were harvested in the afternoon on the photosynthesis measurement days of the 3rd and 6th weeks. Samples were taken from the needle pairs located 3–5 below the shoot apex. 0.1 g of leaf samples were placed into 10 mL brown vials containing 10 mL of 80% (v/v) acetone. The samples were kept at 4°C for 2 weeks to extract pigments. Absorbance at wavelengths of 663 nm, 645 nm, and 470 nm was measured using a microplate reader (Synergy H1, BioTek, Winooski, VT, USA). The pigment contents were calculated using the following Equation 1.

Chlorophyll fluorescence was analyzed using a portable chlorophyll fluorescence meter (Pocket PEA, Hansatech Instruments Ltd., Norfolk, UK). Measurements were conducted between 2:00 p.m. and 3:00 p.m., immediately after photosynthesis measurements in the 3rd and 6th weeks, using needles located 3–5 below the shoot apex. Ten seedlings in each treatment group were selected for measurement, with leaves dark-adapted for 20 min using leaf clips. Chlorophyll fluorescence was induced by applying a saturating red-light pulse (627 nm) with an intensity of 3,500 μmol m⁻² s⁻¹ for 1 s. Fluorescence parameters, including the maximum quantum efficiency of PSII photochemistry (Fv/Fm), the performance index on an absorption basis (PI_ABS), absorption flux per reaction center (ABS/RC), dissipated energy flux per reaction center (DI_o/RC), trapped energy flux per reaction center (TR_o/RC), electron-transport flux per reaction center (ET_o/RC), electron flux reducing end acceptors at PSI per reaction center (RE_o/RC), absorption flux per excited cross-section (ABS/CS_o). dissipated energy flux per excited cross-section (DI_o/CS_o), trapped energy flux per excited cross-section (TR_o/CS_o), electron-transport flux per excited cross-section (ET_o/CS_o) and electron flux reducing end acceptors at PSI per excited cross-section (RE_o/CS_o) were calculated using Pocket PEA software (PEA Plus V1.10, Hansatech Instruments Ltd., UK) based on the equations described by Strasser et al. (2004).

Relative increase in height and root collar diameter was derived from measurements of height and root collar diameter taken 1 week before the treatment began (May 20) and during the 7th week of the experiment (July 12). The initial measurement provided a baseline for growth comparisons, ensuring that any observed differences reflected the experimental conditions (Poorter and Navas, 2003). Plant height was measured from the soil surface to the tip of the tallest leaf using a folding ruler (No. 981012-12, Standardgraph, Stadtallendorf, Germany). Root collar diameter was measured just above the soil surface using a digital caliper (CD-15DC, Mitutoyo Corporation, Kawasaki, Japan). Relative increase in height and root collar diameter was calculated using the following formula Equation 2, where ln represented the natural logarithm:

Here, S₁ and S₂ represented the measured parameter (height or root collar diameter) at the initial (t₁) and final (t₂) time points (May 20 and July 12, respectively), and t₂ – t₁ = 53 days denoted the time interval in days (Hoffmann and Poorter, 2002).

Fresh and dry biomass were assessed on August 29, at the conclusion of the experiment. Five seedlings per treatment group were harvested and separated into leaves, branches, and roots. Fresh weight was recorded immediately after harvesting. The samples were then oven-dried at 85°C until reaching a constant weight to determine dry biomass. Leaf, branch, shoot, and root biomass were calculated, and the root-to-shoot ratio (R/S) was derived to assess biomass allocation between below ground and aboveground components.

Statistical analyses were performed using R software (version 4.4.1; R Core Team, Vienna, Austria). For all datasets, the assumption of variance homogeneity was assessed using Levene's test, and either independent t-tests or Welch's t-tests were applied accordingly.

Among chlorophyll fluorescence parameters, Fv/Fm and PI_ABS were analyzed using both appropriate t-tests (independent or Welch's) and two-way ANOVA to evaluate the effects of CO₂ treatment and measurement time. Other fluorescence parameters (e.g., ABS/RC, DIo/RC, TRo/RC, ETo/RC, REo/RC, ABS/CSo, DIo/CSo, TRo/CSo, ETo/CSo, REo/CSo) were analyzed using only independent t-tests or Welch's t-tests, depending on variance homogeneity.

For physiological and biochemical parameters (e.g., A_net, g_s V_cmax, J_max, pigment content, etc.), both two-way ANOVA and appropriate t-tests were conducted to evaluate treatment effects. Biomass-related data (relative increase in height and root collar diameter, fresh and dry biomass, shoot/root ratio) were analyzed using independent t-tests or Welch's t-tests only.

Sample sizes varied by trait and species. Fluorescence data were collected from 10 seedlings per treatment. Gas exchange and enzyme parameters (e.g., Vc_max, J_max) were measured on 4–5 replicates. Biomass traits were assessed from 5 individuals, and relative increase in height and root collar diameter was calculated using 6–9 replicates. Pigment concentrations were analyzed based on 17–20 samples per group. To enhance data reliability, sample sizes were progressively increased across experimental rounds, resulting in variation in the number of replicates per trait.

P. koraiensis showed minimal early response to eCO₂, with no significant differences in A_net, gs, or Tr during the initial weeks. However, by the 5th week, all parameters increased significantly under eCO₂ (p < 0.05 for A_net and gs; p < 0.01 for Tr), indicating a delayed photosynthetic response (Figures 2A–C). In contrast, WUE was initially higher under aCO₂ at the 2nd week (p < 0.001), but became significantly higher under eCO₂ by the 4th week (p < 0.01) (Figure 2D).

In C. obtusa, A_net and gs were both significantly higher under eCO₂ at the 5th week (p < 0.01 for A_net ; and p < 0.05 for g_s), accompanied by a peak in WUE (p < 0.05). All gas exchange parameters declined by the 10th week, implying seasonal acclimation or physiological downregulation (Figures 2E–H). Meanwhile, A_net in P. koraiensis significantly increased under both conditions by the 10th week but showed no treatment difference, possibly reflecting late-stage structural maturation.

P. koraiensis exhibited no treatment effect on V_cmax, J_max, TPU, or CE across the entire experimental period, indicating limited biochemical responsiveness to elevated CO₂. Nevertheless, a significant increase in all parameters emerged by the 10th week, regardless of CO₂ condition, suggesting a strong temporal effect (p < 0.001), likely driven by internal developmental progression (Figures 3A–D). This pattern points to delayed activation of photosynthetic machinery rather than a direct stimulation by CO₂ enrichment.

By contrast, C. obtusa responded more rapidly to elevated CO₂, showing early upregulation of photosynthetic biochemical traits. At the 3rd week, V_cmax, J_max and TPU were all significantly higher than in the control (p < 0.01, p < 0.05, and p < 0.01, respectively), reflecting a swift biochemical adjustment to the elevated environment (Figures 3E–G). However, this stimulation was transient. All parameters tended to peak at the 6th week and declined thereafter, implying a short-lived enhancement and potential early acclimation. CE did not show treatment differences, but exhibited significant temporal variation across sampling points (p < 0.01) (Figure 3H).

P. koraiensis exhibited an initial reduction in PSII performance under eCO₂, with significantly lower Fv/Fm and PI_ABS values observed at the 3rd week compared to the control (p < 0.01 for both). These results suggest early photochemical stress or delayed acclimation of PSII to the elevated CO₂ environment. However, both parameters showed recovery by the 6th week regardless of treatment, indicating temporal stabilization of photochemical efficiency as leaf development progressed (Figures 4A, B).

In contrast, C. obtusa maintained stable Fv/Fm values throughout the experimental period, with no significant treatment or temporal effects detected (Figure 4C). This stability suggests that PSII efficiency in C. obtusa was not strongly perturbed by elevated CO₂. However, PI_ABS displayed a distinct pattern: it was significantly lower under eCO₂ at the 3rd week (p < 0.01), followed by partial recovery by the 6th week, at which point no significant difference remained between treatments (Figure 4D). This transient reduction may indicate short-term stress responses rather than sustained suppression of PSII function.

Further fluorescence analyses revealed that P. koraiensis under eCO₂ exhibited significantly higher values of ABS/RC, DIo/RC, and TRo/RC at the 3rd week (p < 0.01, p < 0.001, and p < 0.05, respectively), along with elevated ABS/CSo and DIo/CSo (p < 0.05 for both). In contrast, ETo/CSo was significantly lower under eCO₂ (p < 0.01), suggesting that excess energy input was dissipated rather than used for electron transport (Figure 5). These patterns point to increased excitation pressure and non-photochemical quenching activity. Meanwhile, C. obtusa showed no significant differences across any of the fluorescence parameters, reinforcing its relatively stable photochemical profile under varying CO₂ conditions.

P. koraiensis exhibited pronounced changes in pigment composition in response to elevated CO₂ conditions. Chl a content remained consistently higher in the treatment group at both time points, contributing to a significant increase in Total Chl by the 6th week. In contrast, Chl b did not differ between treatments but showed a temporal increase (p < 0.001), indicating ongoing pigment accumulation over time. Notably, Car levels spiked significantly under eCO₂ at the 3rd week (p < 0.001), but declined thereafter, suggesting a short-term stress response that was not maintained through the experimental period. This pattern may reflect early photoprotective activation followed by physiological acclimation under elevated CO₂ conditions (Figures 6A–D).

C. obtusa also exhibited pigment-related responses under elevated CO₂. Chl a content was significantly higher in the eCO₂ treatment at both the 3rd (p < 0.001) and 6th weeks (p < 0.05), indicating sustained stimulation during the early phase Similarly, Car content remained significantly elevated under eCO₂ across both time points (p < 0.001), despite a slight decrease over time. These findings suggest that pigment composition in C. obtusa was responsive to CO₂ enrichment during the early growth stages (Figures 6E–H).

P. koraiensis exhibited no measurable growth or biomass response to elevated CO₂. No significant treatment effects were detected in any component, including leaf, branch, shoot, and root dry weights. Total biomass also remained statistically unchanged between eCO₂ and aCO₂ groups. Similarly, relative increase in height and root collar diameter for height and root diameter, as well as the R/S, showed no differences, suggesting that elevated CO₂ had limited influence on either allocation patterns or overall growth in this species (Supplementary Table S1).

In contrast, C. obtusa showed a notable suppression in growth rate under eCO₂. Relative increase in height and root collar diameter for both height and root diameter were significantly lower compared to the control (p < 0.001), indicating that elevated CO₂ hindered vertical and belowground extension. However, this did not translate into differences in biomass accumulation, as dry weight, fresh weight, and R/S values were comparable between treatments. This decoupling between relative increase in height and root collar diameter and biomass suggests compensatory allocation or stress-related growth modulation in C. obtusa under elevated CO₂ conditions (Supplementary Table S1).