Two-floor bench bed systems consisting of upper and bottom beds (Figure S1) were used for cultivating strawberry plants in the two independent greenhouses. Plants in the bottom bed were found to receive only about 40% of the sunlight intensity incident on the upper bed because they were overshadowed by the upper bed. Fluctuations in photosynthetic active radiation of ambient lights on the two-floor bench bed in a greenhouse were continuously recorded with 10-min interval by using LI-190 quantum sensors (Licor, NE, USA). Quantum sensors were installed at the height of 20 cm above the surface of each bed and the measured values were stored in the CR 1000 data-logger (Campbell Sci. Inc., UT, USA).

Strawberry runners were planted in the two-floor bench beds when they are at the developmental stage of the 4th leaf and were cultivated from October 15, 2014 to March 31 of the following year. They were grown in coir medium hydroponically using a drip irrigation system by providing nutrient solutions of de Kreij et al. (1999) with EC between 1.0 and 1.2 dS·m⁻¹ for each 2 min at a frequency of five times per day.

We prepared two independent greenhouses experiencing two different regimes of temperatures. The temperature in each greenhouse was not allowed to fall below 5 or 10°C all day long, respectively, because the heaters called as complex environmental regulator (Woosung Hitec Co. Ltd., Yangsan, Korea) began to operate automatically every time when the room temperature drops below 5 or 10°C in the respective greenhouse. Every when each temperature was raised above the level of 5 or 10°C, heater was controlled to turn off automatically, recording the minimum temperatures of each greenhouse. The temperature sensor was set up at the height of 2.5 m above the ground in the greenhouse. In this way, four combinations of growth conditions were established, consisting of upper bed maintaining above 10°C (UT), bottom bed above 10°C (BT), upper bed above 5°C (UF), and finally bottom bed above 5°C (BF). Although light intensity and temperature inside each greenhouse varied continuously in the course of the experimental period depending on the ambient weather conditions, four different treatments named as UT, BT, UF, and BF were assumed to represent high temperature plus high light, high temperature plus low light, low temperature plus high light, and low temperature plus low light, respectively. Each group of 100 plants was subjected to four different kinds of treatment, respectively, for the subsequent analyses. The production was recorded as fruit yield harvested from December to next March. All the experiments were carried out in the greenhouses constructed in the Protected Horticulture Research Institute of Korea (latitude 35°19′N and longitude 129°22′E).

Leaves of 18 to 20-day-old plants that had been excised at 7 a.m. of January 17 were sealed in the dark bottle and swiftly moved to the laboratory to minimize water stress. Chlorophyll fluorescence (ChlF) images were immediately taken using a imaging fluorometer (FluorCAM FC800, Photon Systems Instruments, Drasov, Czech) equipped with the software of quenching analysis design protocols. ChlF induction curves upon illumination of dark adapted leaves were recorded and from these induction curves, the maximum quantum use efficiency of PS II was calculated in terms of F_V/F_M. Other parameters of ChlF such as non-photochemical quenching (NPQ), photochemical quenching (qP), and variable chlorophyll fluorescence decline ratio (R_Fd) were also measured by using the FluorCam system. From the light-induced curves of ChlF in dark-adapted leaves ChlF parameters were calculated (Figure S4). Changes in Fv/Fm ratio during daytime were measured in a greenhouse on the 18th of January after supplying light pulse of 3000 μmol·m⁻²·s⁻¹ by using a portable fluorometer FluorPen FP 100 (Photon Systems Instruments, Drasov, Czech).

When parameters such as photosynthetic rates (Pr), stomatal conductances (Sc), and transpiration rates (Tr) were measured, intact leaves attached to plants were subjected to in situ analysis using a hand-held photosynthesis system (LI-6400, Licor, NE, USA) on the 16th of January (under increasing irradiances up to 1500 μmol m⁻² s⁻¹), and on the 8th of February (under 1000 μmol m⁻² s⁻¹), respectively.

To measure the amounts of chlorophylls and carotenoids, leaves were harvested three times on the each 10th of January, February, and March on a monthly basis. After punching six leaf disks (1 g fresh weight) with a cork borer, they were macerated in 15 mL of acetone (containing 100 mg of CaCO₃) with a homogenizer (PT-3100, Kinenatica AG, Switzerland). The homogenates were poured into a solvent-resistant microfuge tube, and after spinning for 5 min, the resulting supernatants were collected. The extracts were filtered, and the absorbance was measured at 661.6, 644.8, and 470 nm with a spectrophotometer (Evolution 300, Thermo Co., MA. USA) as described by Lichtenthaler (1987).

Fully matured strawberry fruits were harvested on the 10th of each month from December to March and each 1 kg fruits from 30 plants were provided for the three repetitions of measuring phytochemicals, sugar content, and organic acids. After homogenizing the fruits, the extracts were centrifuged with the speed of 16,000 g for 30 min at 4°C (64R Centrifuge, Beckman Coulter Inc., CA, USA). Filtered supernatants through Whatman No. 2 filter paper were immediately frozen to be stored at −70°C as referred by Choi et al. (2015).

Before measuring the soluble solid contents of fruit extracts, frozen samples were melted and filtered through 0.45 μm syringe filter. After diluting the filtrates with distilled water, they were analyzed with a HPLC system (YL9100, Younglin Co., Anyang, Korea) equipped with a Sugar-Pak (4.6 mm × 250 mm, Supelco, PA, USA) column and RI detector according to Choi et al. (2015).

Amounts of organic acids contained in the fruit extracts were analyzed with ion chromatography system (ICS 5000, Dionex, CA, USA) equipped with Ion-Pac column (9 × 250 mm ICE-AS6, Dionex, NY, USA) and a suppressor (AMMS ICE300, Dionex, NY, USA) according to Choi et al. (2015).

Anthocyanin contents in the fruit extracts were measured and calculated with reference to pelargonidin-3-glucoside as the standard according to the modified procedure of Kim et al. (2011). Briefly, fruit extracts were pretreated with methanol and 1% hydrochloric acid and filtered solutions were subjected to spectrophotometric measurements of absorbance at 530 nm. The content of total phenolic compounds in the fruit extracts was determined according to Slinkard and Singleton (1977) with reference to gallic acid equivalents as the standard.

The experimental data obtained from chlorophyll fluorescence measurements, photosynthesis parameters, photosynthetic pigments, sugar and organic acid, and phytochemical analysis were subjected to a randomized complete block design (RCBD) analysis of variance (ANOVA) using SAS (SAS Institute Inc., NC, USA). Values of P < 0.05 were regarded as significant.

Strawberry plants were cultivated in the high-bench bed system composed of upper and bottom bed (Figure S1). We recorded changes in the level of greenhouse daylight incident to plants growing in the upper or bottom bed during the cultivation period (Figure S2). Since heat insulation films were covered around the greenhouse during winter for the purpose of preserving the inner temperature, they were cleared away every day from 8 a.m. to 5 p.m. so that sunlight could penetrate into the greenhouse. Overshadowed by the upper bed, the light intensity of bottom bed was sharply dropped after 11 a.m. The respective monthly average of ambient daylight integral given to the upper and the bottom bed was 10.45 and 4.38 mol·m⁻²·d⁻¹ in December, 10.33 and 3.54 mol·m⁻²·d⁻¹ in January, 12.84 and 4.82 mol·m⁻²·d⁻¹ in February, and 19.54 and 7.35 mol·m⁻²·d⁻¹ in March (Figure S2). In overall, the light intensity incident to the bottom bed was about only 40% of the upper bed.

We also recorded changes in the greenhouse temperature during the cultivation period (Figure S3). The minimum temperature of each greenhouse was controlled not to decline below 5 or 10°C during cold season. Air ventilation of the greenhouse was carried out by opening or closing the side window of the greenhouse when the inner temperature was over or below 25°C, respectively. As a result, the minimum temperature of each greenhouse was kept above 5 or 10°C, the upper temperatures were quite variable depending on the outside weather. It was noticeable that daytime temperature of the greenhouse increased significantly following the end of February.

For plants grown under four different conditions of light and temperature consisting of UT, BT, UF, and BF, chlorophyll fluorescence image was visualized from leaves of each plant (Figure 1). BT and BF leaves of the bottom bed under low light showed higher ChlF of Fo, Fm, Fp, and Fv/Fm than UT and UF leaves of the upper bed under high light. ChlF was most strongly emitted at leaf base close to the petiole and extended to the margin of leaves. Leaves exposed to different temperatures of 5 and 10°C showed little differences in the emission of ChlF.

Figure 2 shows time-dependent changes in several parameters of ChlF induction from leaves of strawberry when irradiated by actinic light at an intensity of 700 μmol·m⁻²·s⁻¹. NPQ, qP, and R_Fd values obtained from BF plants were significantly lower than those values from the other plant groups except for Fv/Fm. Figure 3 shows local-time dependent changes in Fv/Fm from morning till night from leaves of plants that had been cultivated under different regimes of light and temperature. All the plants showed the highest Fv/Fm toward the sunrise, while they recorded the lowest valued around 11 a.m. when the ambient light of the greenhouse was brightest, ranging from 0.79 to 0.80 (Figure S2A; Figure 3). Generally speaking, Fv/Fm ratios measured from plants that had been cultivated under different regimes of light and temperature ranged from 0.79 to 0.84. Plants of BF showed higher Fv/Fm ratios than those of UT. When each Fv/Fm ratio was compared in terms of growth temperature, plants under low temperature (UF and BF) showed higher values between 0.81 and 0.83 than those grown under high temperature (UT and BT) with values between 0.80 and 0.82.

Light-intensity dependent changes in photosynthetic rate (Pr) as well as related parameters such as stomatal conductance (Sc) and transpiration rate (Tr) were investigated with leaves of strawberry plants each under UT, BT, UF, and BF conditions (Figure 4). Each parameter of UT plants showed the highest level followed in turn by those of BT, UF, and BF plants. Differences in Pr between UT and other plants became clearly seen starting from the light intensity of 200 μmol·m⁻²·s⁻¹ that had been irradiated for the induction of photosynthesis. When the actinic light of 800 μmol·m⁻²·s⁻¹ was irradiated, UT plants showed the highest Pr value of 14.06 μmol CO₂·m⁻²·s⁻¹ as compared with 12.53 of BT, 11.77 of UF and 7.62 of BF plants, amounting to the increase of 11.2, 19.5, and 84.5%, respectively. Furthermore, BF plants recorded the lowest value of Sc and Tr, each showing 0.069 and 0.89 mmol H₂O·m⁻²·s⁻¹ as compared with other groups of plants. Changes in Pr, Sc, and Tr were investigated with the progress of local time during the daytime and the results were shown in Figure 5. At noon when the light intensity was at the peak, UT plants showed the highest value of Pr while their Sc and Tr were at the lowest level. In contrast to this, BF plants recorded the lowest levels in these parameters during the measurement, showing the least values at 3:00 pm (Figures S2, S3; Figure 5).

Figure 6 shows the amount of photosynthetic pigments in the leaves of UT, BT, UF, and BF plants that had been measured on a monthly basis from January to March of 2015. Amounts of chlorophyll a, b, and carotenoids were the highest, each showing 35.6, 9.7, and 18 g·kg⁻¹, in leaves of BF plants that had been grown under lower light and temperature. While Chlorophyll a was shown to continuously decrease from the start of January, contents of chlorophyll b and carotenoids were maintained at the initial level till the start of February and thereafter began to decline. In March when the strength of sunlight enhances, it was noticed that pigments in leaves of each plant groups were markedly reduced. For example in BT plants, levels of chlorophyll a, chlorophyll b, and carotenoids decreased by 15.9, 42.6, and 21.6%, respectively. Despite such outcomes, it was evident that plant groups under higher light condition (UT and UF) accumulated less amounts of photosynthetic pigments compared with those plant groups exposed to lower light (BT and BF; Figure 6). On the contrary, the ratio of chlorophyll a and b recorded a remarkable increase from February till March, with UT plants showing the highest value of 6.7 while BF plants the lowest of 5.0 during March.

Figure 7 shows the monthly production as well as the cumulative production of strawberry fruits measured until March from December of the previous year. The respective plants of each group showed a steady increase in the fruit yield with the progress of cultivation days from December to March of the following year, with UT plants showing the highest while BF plants the lowest yield. Although January had seen a sharp increase in the fruit production of BT and UF plants compared with December, there were no differences between January and February. In contrast to this, plants of UT and BF showed almost a linear increase in the fruit production with the passage of months during the growth period. Total cumulative productions of fruits from December to March were calculated for each plant group (Figure 7D). UT plants showed the highest fruit production of 425 g per plant, followed by 350 g for UF plants, 328 g for BT plants, and 244 g for BF plants, corresponding to 82, 77, and 57% of UT plants.

Effects of different light and temperature on the accumulation of soluble sugars and organic acids in the strawberry fruits are shown in Tables 1, 2. Total sugars of the fruits harvested in March declined by 29, 51, 47, and 64% lower than those in January for the plants of UT, BT, UF, and BF, respectively. On the contrary, organic acids of fruits harvested in March were elevated by 36, 36, 16, and 23% higher than those in January for the plants of UT, BT, UF, and BF, respectively. The decreased extent of fruit sugars was found to be much larger than the increased extent of organic acids in the fruits as the cultivation continues. In particular, fruits of UT and UF had higher capacity to build total soluble sugar and organic acids than those of BT and BF.

Effects of different regimes of light and temperature on the formation of phenolic compounds and anthocyanins in the strawberry fruits are shown in Figure 8. Fruits of UT and UF were found to accumulate much higher amounts of phenolic compounds and anthocyanins than those of BT and BF. When changes in the amounts of the compounds were investigated on a monthly basis from January to March, it was noted that phenolic compounds in the fruits decreased from January to February followed by a sharp increase from February to March. On the contrary, anthocyanin contents were found to be significantly elevated from January to February, followed by an equivalent decrease from February to March.

The correlation analysis was carried out between fruit production and a variety of photosynthesis-related parameters, by calculating Pearson Product Moment Correlation (Pearson, 1895). As shown in Table 3, we found that fruit productions were closely correlated with Pr (r = 0.962, p < 0.01) and R_fd (r = 0.952, p < 0.01), showing a little less correlation with Sc (r = 0.874, p < 0.01), Tr (r = 0.837, p < 0.001) and NPQ (r = 0.700, p < 0.001). In contrast, they were negatively correlated with Fv/Fm (r = −0.735, p < 0.01).

When correlation coefficients were calculated between photosynthesis rate (Pr) and related parameters, Pr showed a significant positive correlation with Sc (r = 0.929, p < 0.01) and R_fd (r = 0.902, p < 0.01), with little lower correlation with Tr (r = 0.869, p < 0.01) and NPQ (r = 0.644, p < 0.01). In contrast, Pr was negatively correlated with Fv/Fm (r = −0.650, p < 0.01). R_fd was also positively correlated with NPQ, Sc, and Tr showing the respective correlation coefficient of 0.830 (p < 0.01), 0.794 (p < 0.01), and 0.779 (p < 0.01). Correlation coefficient between Sc and Tr was 0.962(p < 0.01) and those between qP and Sc as well as Tr were 0.736 (p < 0.01) and 0.740 (p < 0.01), respectively. The results indicated that fruit yield of strawberry was strongly dependent on photosynthesis and other related parameters.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.