Three seeds of lettuce (Lactuca sativa L.) ‘Rex’ (Johnny’s Selected Seeds, Winslow, ME, USA) were sown in 0.45 L plastic pots (8.8 cm x 8.8 cm x 8.9 cm; length x width x height) filled with a soilless substrate (BM6; peat-moss and perlite; Berger, Saint-Modeste, QC, Canada) in a glass-covered greenhouse. Seedlings were moved into growth chambers six days after sowing. The seedlings were thinned to one plant per pot based on uniformity. Plants were manually irrigated with a nutrient solution containing 150 mg L^-1 N and other essential mineral nutrients made with water-soluble fertilizer (21N-2.2P-16.6K; Peters 21-5-20; The Scotts Company, Marysville, OH, USA) throughout the experiment.

In this experiment, three walk-in growth chambers were used to establish three temperature setpoints: 20°C, 24°C, and 28°C (actual temperature was 21.2 ± 1.4°C, 24.4 ± 1.2°C, and 28.9 ± 1.0°C, respectively). This temperature range supports normal growth and development of lettuce ‘Rex’ without causing chilling or heat stress symptoms (Jeong et al., 2024). In each walk-in growth chamber, four sections (l x w x h; 60 x 60 x 70 cm) were created using growth-racks and reflective cardboards to accommodate four light treatments: two total photon flux densities (TPFDs; 400-800 nm) (150 or 300 μmol m^-2 s^-1) x two FR light levels (0 or 20% of FR light in TPFD). The corresponding light treatments were as follows: 0% FR (B₁₅G₁₅R₁₂₀) and 20% FR (B₁₅G₁₅R₉₀FR₃₀) of FR light at a lower TPFD of 150 μmol m^-2 s^-1 (TPFD₁₅₀) and 0% FR (B₃₀G₃₀R₂₄₀) and 20% FR (B₃₀G₃₀R₁₈₀FR₆₀) at higher TPFD of 300 μmol m^-2 s^-1 (TPFD₃₀₀). B stands for blue light (400-500 nm) and G stands for green light (500-600 nm) ( Figure 1 and Supplementary Table 1 ). The subscript after each waveband indicates its photon flux density in μmol m^-2 s^-1. The spectral treatments were created using an LED research lighting system (PHYTOFY^® RL, Osram, Munich, Germany) and the intensity of each light spectrum was adjusted using PHYTOFY^® RL software (version 1.0.8). The peak wavelengths were 450 nm for B, 521 nm for G, 660 nm for R, and 730 nm for FR LEDs. All treatments had a 24-h photoperiod. The daily light integral (DLI) of the treatments with a TPFD of 150 and 300 μmol m^-2 s^-1 was 13 and 26 mol m^-2 d^-1, respectively. The two light intensity levels were selected based on the recommended DLI range of 10 to 20 mol m^-2 d^-1 for indoor lettuce cultivation (Yan et al., 2019; Kelly et al., 2020; Matysiak et al., 2022). The FR fraction [FR/(R+FR); Kusuma and Bugbee, 2021] in the treatments with 20% FR light was 0.25.

Temperature in each growth chamber section was measured every 30 seconds and recorded every 20 minutes using type-E thermocouples connected to a data logger (CR1000; Campbell Scientific, Logan, UT, USA). Within each treatment area, photon flux density (30 cm distance from LEDs to the top of plant) was measured at fourteen locations (see Supplementary Table 1 for the standard deviation of the light intensity under each treatment) with a spectroradiometer (PS100; Apogee Instruments, Logan, UT, USA). To ensure consistent light intensity at the top of the plants, a constant distance between the LEDs and plant canopy was maintained by periodically lowering the shelves. To minimize any effects caused by spatial variation in light intensity, plants within each treatment were randomly rotated daily.

This experiment was replicated two times. In each replicate, four plants (subsamples) per treatment were used in each of the twelve treatments [two light intensities (150 or 300 μmol m^-2 s^-1) x three temperatures (20, 24, or 28°C) x two light spectra (0 or 20% FR)]. A split-plot block design was employed, where the main-plot factor was temperature and the sub-plot factor was light spectral quality. The chamber temperature set points and the locations of spectral treatments were randomized in each replicate. Three- and two-way analysis of variance (ANOVA) procedure was utilized to analyze the data using the Statistical Analysis System version 9.4 (SAS Inst., Inc., Cary, NC, USA). Subsamples were averaged before statistical analysis. Significant difference among the treatments was determined using Duncan’s multiple range test at p< 0.05. SigmaPlot software version 12.5 (Systat Software, Inc., Chicago, IL, USA) was used for regression analyses.

Overall, we found that FR light and temperature interactively regulated lettuce growth and morphology at a lower TPFD of 150 μmol m^-2 s^-1; however, increasing TPFD from 150 to 300 μmol m^-s2 s^-1 caused the interaction between FR light and temperature to diminish ( Figures 2 – 5 and Supplementary Figures 1 , 2 ).

Significant three-way interactions among light intensity, FR light, and temperature were also observed in plant biomass parameters, including the FW and DW of leaf, stem, root, and total plant mass ( Figure 6A and Supplementary Figure 3 ). Consistent with morphological parameters, FR light and temperature interactively regulated plant biomass accumulation and partitioning at low light intensity (TPFD₁₅₀), but the magnitude of the interactive effects diminished at high light intensity (TPFD₃₀₀). Specifically, the response of total plant DW to FR light followed a similar pattern as total leaf area ( Figure 6A ). At TPFD₁₅₀, increasing the FR light from 0 to 20% had no significant effect on total leaf area at 20°C but caused a 63% decrease at 24°C and a 75% decrease at 28°C ( Figure 6A ). However, at TPFD₃₀₀, substituting FR light for R light significantly increased shoot DW under all three temperatures, by 37% at 20°C, 29% at 24°C, and 14% at 28°C ( Figure 6A ). Regarding biomass partitioning, at TPFD₁₅₀, FR light did not affect percent leaf DW at 20°C, but significantly decreased percent leaf DW at warmer temperatures of 24°C and 28°C. In contrast, at TPFD₃₀₀, percent leaf DW was not affected by FR light regardless of temperature conditions ( Figure 6B ). In contrast to percent leaf DW responses, at TPFD₁₅₀, the stimulative effect of FR light on percent stem DW was greater at warm temperatures, with no significant increase at 20s°C, a 118% increase at 24°C, and a 179% increase at 28°C ( Figure 6C ). However, at TPFD₃₀₀, the magnitude of the interactive effects between FR light and temperature on percent stem DW diminished, and FR light was less effective on percent stem DW than those at a lower TPFD ( Figure 6C ). Percent root DW showed no significant interaction among TPFD, temperature, and FR light, but FR light significantly decreased percent root at lower TPFD ( Figure 6D ).

Neither three-way interaction among light intensity, temperature, and FR light, nor two-way interaction between FR light and temperature, was observed in any of the photosynthetic parameters, including Φ_PSII , P_net , and R_d ( Figure 7 ). At both TPFD levels, FR light significantly increased Φ_PSII , whereas temperature had no effect on Φ_PSII ( Figures 7A–D ). Both warm temperature (28°C) and FR light caused P_net to decrease under each TPFD level ( Figures 7E–H ). R_d significantly increased with increasing temperature from 20°C to 28°C but decreased under 20% FR light ( Figures 7I–L ).

Total DW was positively correlated with total intercepted photons and P_net , but it was negatively correlated with Φ_PSII ( Figure 8 ). Total intercepted photons showed much higher R² value of 0.94, compared to those for Φ_PSII (R² = 0.17), and P_net (R² = 0.34).

Similar to the photosynthetic parameters, there was no significant three-way interaction among the three environmental factors, nor was there any significant two-way interaction between FR light and temperature in the chlorophyll contents and chlorophyll a:b ratio ( Supplementary Figure 4 ). At both TPFD levels, chlorophyll a content significantly increased with increasing temperature from 20°C to 28°C, but decreased at 20% FR light ( Supplementary Figures 4A–D ). Warm temperature of 28°C also significantly increased chlorophyll b content, although FR light did not affect chlorophyll b contents at either TPFD level ( Supplementary Figures 4E–H ). Chlorophyll a:b ratio significantly decreased by warm temperature (28°C) and FR light at both TPFDs ( Supplementary Figures 4I–L ).

For secondary metabolite contents and antioxidant capacity, there were no significant interactions among TPFD, temperature, and FR light ( Figure 9 ). Both higher TPFD and warm temperature of 28°C significantly increased the concentrations of carotenoids, flavonoids, phenolics, and antioxidant capacity, while FR light caused decreases in these parameters. Specifically, higher TPFD enhanced the concentration of carotenoids by 8%, flavonoids by 19%, phenolics by 42%, and antioxidant capacity by 31% ( Figures 9A, D, G, J ). Warm temperature (28°C) also increased the concentration of carotenoids by 32%, flavonoids by 49%, phenolics by 38%, and antioxidant capacity by 46% compared to a cool temperature of 24°C ( Figures 9B, E, H, K ). In contrast, FR light significantly reduced the concentration of carotenoids by 6%, flavonoids by 17%, phenolics by 18%, and antioxidant capacity by 17% ( Figures 9C, F, I, L ).

Previous studies have found that FR light and warm temperature synergistically promote hypocotyl elongation in Arabidopsis seedlings (Patel et al., 2013; Romero-Montepaone et al., 2020; Burko et al., 2022). However, the interactive effect between FR light and warm temperature on leaf expansion, which is crucial for photon capture and crop yield, showed opposite trends compared to hypocotyl elongation (Patel et al., 2013; Jeong et al., 2024). For example, in lettuce, FR light promoted the leaf expansion at cooler temperatures (20°C and 24°C) but not at warmer temperature of 28°C (Jeong et al., 2024). Similarly, we also found that FR light combined with warm temperature (28°C) synergistically enhanced stem elongation, while reducing leaf expansion of lettuce at TPFD₁₅₀ ( Figures 4A, B and Supplementary Figure 3 ). This indicates a trade-off between leaf expansion and stem elongation. Consequently, a negative correlation between total leaf area and stem length was observed at TPFD₁₅₀, particularly due to the excessive stem elongation under 20% FR light and warmer temperatures (24 and 28°C) ( Figure 5A ). The pronounced stem elongation induced by a combination of FR light and warm temperature (28°C) is likely an evolutionary adaptation to cope with heightened respiratory demands under elevated temperatures ( Figures 7I, K ) (Legris et al., 2017; Romero-Montepaone et al., 2021; Jeong et al., 2024). Such a strategy enables plants to escape shade by extending their stems, strategically positioning their leaves to gain better access to unfiltered sunlight for photosynthesis.

However, when light intensity increased from 150 to 300 μmol m^-2 s^-1, the magnitude of the interactive effect between FR light and temperature on leaf expansion and stem elongation diminished, with FR light consistently increasing total leaf area regardless of temperature ( Figures 4A, B ). Unlike the negative correlation between total leaf area and stem length observed at TPFD₁₅₀, a positive correlation between these two parameters was observed at TPFD₃₀₀ ( Figure 5 ). These findings consistently indicate that the interactive effects between FR light and warm temperature on plant morphology diminished as light intensity increased ( Figures 4 , 5 , and Supplementary Figure 1 ). The notion is further supported by root morphological data, where the interactive effects between FR light and temperature on total root length and average root diameter disappeared when light intensity increased from 150 to 300 μmol m^-2 s^-1 ( Supplementary Figure 2 ).

The diminished interactive effect between FR light and warm temperature on plant morphology at high light intensity could be attributed to the reduced effect of thermal reversion on PHY dynamics in two possible ways. Firstly, photoconversion rate increases with increasing light intensity (Sager et al., 1988; Mancinelli, 1988; Klose et al., 2015; Legris et al., 2016). Specifically, the photoconversion rates (both from Pr to Pfr and from Pfr to Pr) increased proportionally with increasing light intensity (Equations 2, 3; Sager et al., 1988; Mancinelli, 1988). However, thermal reversion rates (i.e., k_r1 and k_r2) are not influenced by light intensity (Jung et al., 2016; Klose et al., 2015; Legris et al., 2016; Sellaro et al., 2019). Thus, the effect of thermal reversion on the steady-state of PHYB could be overridden by the accelerated photoconversion rates under high light intensity, resulting in a diminished effect of temperature on PHYB activity and PHYB-mediated plant morphological responses (Klose et al., 2015; Sellaro et al., 2019). Secondly, high light intensity can decrease thermal reversion rate by stabilizing active PHYB through nuclear body formation. Specifically, the rate of thermal reversion of PHYB from D₁ to D₀ (k_r1) is significantly faster than those from D₂ to D₁ (k_r2), making the amount of P_fr-P_r heterodimer (D₁) a crucial determinant for thermal reversion rate (Klose et al., 2015). Previous studies have reported that high light intensity stimulates nuclear body association and increase the size of nuclear bodies, which serve as the storage site stabilizing and protecting PHYB from inactivation (Chen et al., 2003; Rausenberger et al., 2010; Trupkin et al., 2014; Van Buskirk et al., 2014; Legris et al., 2016; Sellaro et al., 2019). Given that P_fr-P_r (D₁) heterodimers are primarily located in nucleoplasm rather than nuclear body (Klose et al., 2015; Sellaro et al., 2019), a larger nuclear body size at high light intensity indicates a higher proportion of D₂ and a smaller D₁ pool. Therefore, the nuclear body association at high light intensity reduces the size of the D₁ pool and, consequently, reduces thermal reversion (Trupkin et al., 2014; Klose et al., 2015; Legris et al., 2016; Sellaro et al., 2019). Consistently, Sellaro et al. (2019) found that higher light intensities were necessary to reach PPE under warmer temperatures, suggesting that thermal reversion would have a bigger impact on phytochrome dynamics and plant morphology under warm temperatures and low light intensities. In this study, a significant interaction between the estimated PPE and temperature was observed in leaf expansion rate and stem elongation rate under a lower light intensity of 150 μmol m^-2 s^-1 ( Supplementary Figure 5 ). Specifically, leaf expansion rate decreased as the estimated PPE increased (i.e., when FR light was absent) at 20°C (α = -35.6), but increased as the estimated PPE increased at 24°C (α = 122.9) and 28°C (α = 173.2) ( Supplementary Figure 5A ). The effect of PPE on stem elongation rate became more pronounced as temperature increased from 20°C to 28°C (α = -0.22 at 20°C, -1.01 at 24°C, and -2.24 at 28°C) ( Supplementary Figure 5C ). In contrast, at high light intensity, the effect of PPE on leaf expansion rate and stem elongation rate were similar across the three temperatures, and there were no significant interactions between temperature and the estimated PPE ( Supplementary Figures 5B, D ). Taken together, our results, along with previous findings, suggest that at lower light intensity, light spectra and temperature interactively regulate the steady-state of PHYB and plant morphology. However, at high light intensity, PHYB activity and its effect on plant morphology are likely predominantly regulated by light spectra and are less affected by temperature.

Photon capture and plant biomass typically exhibit a strong linear correlation (Monteith, 1977; Gifford et al., 1984). Consistently, we found a strong positive correlation between total DW and total intercepted photons (R² = 0.94) ( Figure 8A ). Therefore, enhancing leaf expansion and photon capture through the application of FR light can be an effective strategy to increase plant biomass in indoor vertical farming systems (Park and Runkle, 2017; Meng and Runkle, 2019; Zhen and Bugbee, 2020; Legendre and van Iersel, 2021). However, our results showed that the effects of FR light on leaf expansion, canopy photon capture, and biomass accumulation were dependent on temperature and light intensity ( Figures 4 , 6 , and Supplementary Figure 3 ). Specifically, FR light did not affect total DW at 20°C but caused reductions in plant biomass at warmer temperatures (24°C and 28°C) under low light intensity, while consistently increasing biomass accumulation across all temperature conditions at high light intensity ( Figure 6A ). The excessive stem growth induced by FR light and warmer temperature (24-28°C) at low light intensity (TPFD₁₅₀) led to decreased leaf area and, consequently, reduced photon capture and biomass production ( Figure 5 ). These results underscore the importance of preventing excessive stem growth via providing adequate light intensity when utilizing FR light to enhance leaf expansion and plant growth under warm temperature ( Figure 5 ).

We also found that total DW was only weakly correlated with the quantum yield of PSII and leaf-level photosynthetic rate (R² = 0.17 for Φ_PSII and R² = 0.24 for P_net ), compared to total intercepted photons (R² = 0.94) ( Figure 8 ). These results indicate that plant biomass accumulation is more dependent on canopy photon capture, compared to the single-leaf photosynthetic efficiency. Consistently, previous studies also reported that the photosynthetic activities at single-leaf level oftentimes do not predict crop yield (Evans, 1975; Elmore, 1980).

Enhancing health-promoting compounds, including photosynthetic pigments and secondary metabolites, is an important production goal in indoor farming systems (Poiroux-Gonord et al., 2010; Yang et al., 2018). Previous research has found that the FR light generally reduces the contents of photosynthetic pigments and secondary metabolites, resulting in lower nutritional quality (Li and Kubota, 2009; Stutte et al., 2009; Kong and Nemali, 2021). Similarly, we found that FR light decreased the contents of various phytochemicals, including chlorophyll a, carotenoids, flavonoids, and phenolics ( Figure 9 and Supplementary Figure 4 ). The reduced phytochemicals under FR light may be at least partially due to its direct effect on down-regulating the expression of genes involved in the biosynthetic pathways of these phytochemicals through the PHY signaling networks (Huq et al., 2004; Toledo-Ortiz et al., 2010). The reduction in phytochemical contents may also be attributed to the dilution effect due to increased leaf expansion induced by FR light (Casal et al., 1987; Li and Kubota, 2009; Zhen and Bugbee, 2020; Kong and Nemali, 2021).

In contrast to the responses under FR light, higher light intensity improved the levels of secondary metabolites and antioxidant capacity ( Figure 9 ). Similar results were also observed in several crop species (Tattini et al., 2014; Dou et al., 2018; Pérez-López et al., 2018). The increased concentration of phytochemicals at higher light intensity may be attributed to the increased availability of photosynthates ( Figure 7 ) (Mosaleeyanon et al., 2005). Similarly, warmer temperature also increased the levels of those antioxidants ( Figure 9 and Supplementary Figure 4 ) (Lefsrud et al., 2005; Shamloo et al., 2017). This enhancement of phytochemicals and antioxidant capacity may be due to the protective mechanism against increased oxidative stress under warm temperature of 28°C (Oh et al., 2009; Pospíšil, 2016). However, in this study, warm temperature of 28°C did not cause significant stress as plants grown under 28°C showed high F_v /F_m values between 0.81-0.82 (Maxwell and Johnson, 2000; Baker, 2008).

We found that the reductions in health-promoting compounds and antioxidant capacity induced by FR light could be compensated for by increasing light intensity and temperature. The treatment with higher light intensity, warmer temperature, and 20% FR light (i.e., TPFD₃₀₀ x 28°C x 20% FR light) increased crop yield [shoot (leaf + stem) FW] by 400% and antioxidant capacity by 60%, compared to the treatment with lower light intensity, cool temperature, and 0% FR light (i.e., TPFD₁₅₀ x 20°C x 20% FR light). This result indicates that the combination of high light intensity, warm temperature, and FR light can maximize not only plant biomass but also nutritional quality.