Plant materials and greenhouse environmental conditions.

Green butterhead lettuce (Lactuca sativa ‘Rex’), red romaine lettuce (L. sativa ‘Outredgeous’), and sweet basil (Ocimum basilicum ‘Genovese’) were used during this study. Seeds (obtained from Johnny’s Selected Seeds, Fairfield, ME, USA) were directly sown in 1.8-L containers filled with a soilless substrate (85% peat and 15% perlite by volume; ASB Greenworld, Pointe-Sapin, New Brunswick, Canada) in a polycarbonate-covered greenhouse. These three crops were selected because of their economic importance, different growth habits, and contrasting leaf color (red and green). Ten days after sowing seeds, seedlings were thinned to one plant per container, and uniform plants of each species/cultivar were transferred to light spectral treatments in the same greenhouse (see experimental details). Plants were fertigated daily with a water-soluble fertilizer applied at 100 mg⋅L⁻¹ N (20N–8.7P–16.6K; 20–20–20 Peter’s Professional General Purpose; ICL Specialty Fertilizer, Summerville, SC, USA) throughout the experiment. During the growing period (6 Mar 2022 to 9 Apr 2022), the average daily temperature, relative humidity, and vapor pressure deficit in the greenhouse were 23.3 ± 2.1 °C, 52.6% ± 17.2%, and 1.5 ± 0.4 kPa, respectively.

Solar spectral modification treatments with QD films.

To characterize the responses of plant growth, morphology, and pigmentation to greenhouse solar spectral quality modifications, the following three types of films were examined during this study: regular-concentration CIS/ZnS QD film (reg QD film; UbiQD, Los Alamos, NM, USA); high-concentration CIS/ZnS QD film (high QD film; UbiQD); and PE film (to create a diffuse light environment with mildly reduced light intensity similar to that under the QD films). A control treatment without any light-modifying films was included as a fourth treatment (control). Nine cuboid structures (length × width × height: 1.5 m × 1.2 m × 0.9 m) were constructed using polyvinyl chloride (PVC) pipes to create the three types of film treatments (i.e., reg QD film, high QD film, and PE film; three structures per film type). The top of each PVC structure was completely covered with film, with three sides partially covered from the top (leaving uncovered openings 30 cm in height from the greenhouse benches) and the northwest-facing side open for air ventilation. The PVC structures were placed at least 2 m apart to minimize the shadows cast among spectral treatments. To quantify the spectral shifts of light passing through the films, we measured the light intensity and spectral distribution at 30 locations within the bench area (1.8 m²) enclosed by each PVC/film structure. The measurements were obtained on a sunny day around solar noon using a spectroradiometer (PS300; Apogee Instruments, Logan, UT, USA) placed 20 cm above the bench top. Each light measurement under a PVC/film structure was paired with a measurement obtained within seconds under unfiltered greenhouse sunlight, and the spectroradiometer remained in the same location while the PVC/film structure was removed. The proportions of ultraviolet-A (320–400 nm), blue (401–500 nm), green (501–600 nm), red (601–700 nm), and far-red (701–800 nm) wavebands of light filtered through each type of film and unfiltered greenhouse sunlight were determined by dividing the photon flux density within each waveband by the total photon flux density integrated from 320 to 800 nm (Fig. 1; Table 1). The phytochrome photostationary state (PSS) in each treatment was calculated according to the method described by Sager et al. (1988).

View Figure

• Download as Powerpoint
• Download Figure

Table 1. Proportion of different wavebands in unfiltered greenhouse sunlight and light transmitted through three types of films.

View Fullscreen

Experimental setup.

The four treatments were arranged in a randomized complete block design with three blocks (Supplemental Fig. 1). Eight uniform plants of each species/cultivar were placed under each treatment area (eight plants per treatment × three species/cultivar = 24 experimental plants per treatment). A total of 15 additional plants (five plants per species/cultivar) were placed along the edges as bordering plants in each treatment. A total of 156 plants per species/cultivar were grown in this study, including 96 experimental plants (eight plants per treatment × four treatments × three blocks) and 60 bordering plants (five plants per treatment × four treatments × three blocks).

Twelve quantum sensors (SQ-500; Apogee Instruments, Logan, UT) and 12 type-J thermocouples were connected to a datalogger (CR1000X; Campbell Scientific, Logan, UT, USA) and placed slightly above plant height at the center of plant canopies under each of the 12 experimental units (four treatments × three blocks) to measure PPFD and air temperature every 30 s (Table 1). The daily light integral (DLI; total amount of 400–700 nm photosynthetic photons received per square meter of area per day) was calculated from the PPFD measurements.

Growth parameters.

Each crop was harvested separately on three consecutive days 19 to 21 d after the start of treatments. ‘Rex’ lettuce was harvested first, followed by ‘Outredgeous’ lettuce and basil. The following growth parameters of each experimental plant were measured: number of leaves; total leaf area (using a leaf area meter; model LI-3100C; LI-COR, Lincoln, NE, USA); shoot fresh weight; and shoot dry weight (after fully dehydrated for 5 d at 80 °C in a drying oven). The leaf mass per area (LMA) was calculated by dividing the leaf dry weight by the total leaf area (g⋅m⁻²). The leaf chlorophyll contents (μmol⋅m⁻²) of green leaf lettuce and basil of mature leaves were measured three times per plant and averaged (MC-100 chlorophyll meter; Apogee Instruments, Logan, UT, USA). The plant height of basil was also measured, and leaves and stems were separated to determine the fresh and dry weights.

Projected leaf area.

A digital camera was installed 140 cm above the plant height to obtain top-down photos of ‘Rex’ and ‘Outredgeous’ lettuce plants at the end of the experiment. Each plant was placed at the center of a 60 cm × 60 cm whiteboard (background). Photos of ‘Rex’ lettuce (with green leaves) were analyzed to determine the projected leaf area using an open-source Python program (available at https://github.com/jakobottar/green-pixel-analysis). The Python program calculates the ratio of green pixels (plant tissues) to the total pixel count of the background area. The projected leaf area was calculated by multiplying the background area (3600 cm²) by the fraction of the green pixels (Klassen et al. 2004).

Anthocyanin content index.

The top-down photos of ‘Outredgeous’ lettuce (with red leaves) were analyzed using a Python program for the projected leaf area and using the normalized difference anthocyanin index (NDAI), which is an image-based index developed by Kim and van Iersel (2023), to estimate the anthocyanin concentration. The NDAI is determined based on the optical properties of anthocyanins, which show high absorptance in the green and low absorptance in the red region of the spectrum. The Python program separates the plant objects from the background and extracts the pixel intensity of the green and red color channels from the red-green-blue images to calculate the NDAI as (I_red − I_green)/(I_red + I_green), where I is the pixel intensity (see Fig. 2 for an example of a red-green-blue image of an ‘Outredgeous’ lettuce plant canopy, NDAI image, and its corresponding histogram).

View Figure

• Download as Powerpoint
• Download Figure

Statistical analysis.

This experiment contained four light spectral treatments arranged in a randomized complete block design with three blocks (i.e., three repetitions of each treatment). Data from each of the three species/cultivar were analyzed separately using an analysis of variance (Statistical Analysis Systems; SAS Institute, Cary, NC, USA). Mean separation was performed using Tukey’s honestly significant difference test to compare means among treatments, with statistical significance set at P < 0.05.

Light spectral quality and phytochrome PSS.

Significant light spectral shifts were observed under the two types of QD films (Fig. 1; Table 1). Compared with the control treatment, both the QD films resulted in partial conversion of ultraviolet-A and blue light into green, red, and far-red radiation. Under the high QD film treatment, the fraction of blue light relatively decreased by 33% (from 17.26% to 11.64% of the total photon flux density within 320 to 800 nm). In contrast, the fractions of green, red, and far-red relatively increased by 4%, 10%, and 6%, respectively, compared with the control. In the reg QD film treatment, the blue light fraction relatively decreased by 18%, whereas the green, red, and far-red radiation fractions relatively increased by 3%, 6%, and 3%, respectively (Fig. 1; Table 1). The greenhouse polycarbonate covering material poorly transmitted ultraviolet radiation (ultraviolet-A was only 0.21% of the total photon flux density in the control treatment), and the QD films efficiently converted 45% to 76% of ultraviolet-A radiation transmitted into the greenhouse. Compared with the control treatment, the PE film did not significantly alter the light spectral quality. Nonetheless, compared with those of the control and PE film treatments, the spectral shifts of the QD films did not cause significant changes in the phytochrome PSS, which was estimated following the work of Sager et al. (1988) (Table 1).

Daily light integral.

The average DLI was highest under the control treatment (24.3 ± 6.5 mol⋅m⁻²⋅d⁻¹; mean ± SD), followed by the reg QD film treatment (21.2 ± 5.7 mol⋅m⁻²⋅d⁻¹), PE film (20.2 ± 5.3 mol⋅m⁻²⋅d⁻¹), and high QD Film (18.9 ± 5.0 mol⋅m⁻²⋅d⁻¹) (Table 1). Compared with that of the control treatment, the average DLIs under the reg QD film, PE film, and high QD film treatments were reduced by ∼13%, 17%, and 23%, respectively. There were no significant differences in the DLIs among the three treatment blocks.

Air temperature.

The films did not significantly affect air temperature (including daily maximum, minimum, and average temperatures), and the average daily temperatures of the four treatments during the experiments were between 23.3 and 23.5 °C (Table 1). However, there was a significant block effect on air temperature, with the highest temperature in block 1 (23.9 ± 0.27 °C; farthest from the cooling pad of the greenhouse fan-and-pad cooling system), followed by block 2 (23.6 ± 0.25 °C) and block 3 (22.8 ± 0.60; closest to the cooling pad) (see Supplemental Fig. 1 for the block arrangement).

Effects of light environment modifications by the QD films on plant morphology and biomass accumulation.

The light environment (including light intensity and spectral quality) and temperature are crucial environmental factors that affect plant growth and development (Bayat et al. 2018; Hatfield and Prueger 2015). We found that both the average DLI and light spectral distribution varied among the four treatments (Fig. 1; Table 1). Specifically, the DLIs under the reg QD film, PE film, and high QD film treatments were reduced by 13% to 23% compared with that of the control, and there were no significant differences in the DLIs among the three film treatments. This is consistent with the findings of previous studies that reported that spectral conversion films cause reductions in light intensity (Hebert et al. 2022; Parrish et al. 2021; Salvador et al. 2008). Although the PE film only reduced light intensity without affecting the light spectral distribution, the QD films converted a portion of blue photons into green, red, and far-red photons, with a greater degree of spectral shift under the high QD film treatment (Fig. 1; Table 1). Despite the modifications of the light environment by the QD and PE films, air temperatures were similar across all four treatments (Table 1).

The reduction in the DLI under the PE film treatment without solar spectral modifications led to reduced shoot biomass production in all three crops compared with the control (Figs. 3A and 3B). This was expected because the photosynthetic rate and biomass production generally decrease as the DLI decreases (Dou et al. 2018; Huber et al. 2021; Modarelli et al. 2022; Sutulienė et al. 2022). In contrast with the responses under the PE film treatment, although the DLI under the high QD treatment was 23% lower than that under the control treatment (Table 1), shoot fresh weight, shoot dry weight, and total leaf area of red lettuce and basil grown under the high QD film treatment were similar to those grown under the control treatment (Fig. 3). This was most likely attributable to the greater spectral shift under the high QD film (i.e., a decrease in the proportion of ultraviolet-A and blue photons and a simultaneous increase in the green, red, and far-red regions) (Fig. 1). Hebert et al. (2022) reported that tomatoes grown under similar QD films showed a 10% increase in the vegetative growth rate, a 36% reduction in unmarketable fruit waste (because of fruit cracking), and a 5.7% increase in yield per unit production area despite a 14% decrease in the DLI compared with that of the control treatment without the QD film.

All three crops grown under the control treatment exhibited the highest LMA, indicating greater leaf thickness (Fig. 3D). Formisano et al. (2021) reported that higher light intensity resulted in several physiological and morphological adaptations, including increased leaf thickness and a more compact canopy. Another study reported that increasing the proportion of blue photons from 8% to 30% resulted in an increased LMA of cucumber and tomato under a high PPFD treatment of 500 μmol⋅m⁻²⋅s⁻¹ (Gommers et al. 2013). During our study, no significant differences in the LMA were detected in either lettuce cultivar among the PE film, reg QD film, and high QD film treatments, despite a lower blue light fraction under the QD film treatments. The LMA of basil plants grown under the PE film treatment was significantly higher than that of those grown under the high QD film treatment. This significant difference in LMA of basil suggests that basil has greater sensitivity to spectral shifts caused by QD films.

Even though the solar spectrum was also shifted from blue to longer wavelength regions under the reg QD film treatment, the magnitude of the spectral shift was much smaller than that under the high QD film. As a result, no significant growth differences were observed in all three crops grown under the reg QD film treatment compared with those grown under the PE film treatment, and reductions in the total biomass and leaf area compared with those of the control treatment were observed. Because the PE film, reg QD film, and high QD film reduced the DLI by similar amounts, these results suggested that the effects of QD films on plant growth and morphology are dependent on the magnitude of the light spectral shift. Unlike that under the high QD film treatment, the magnitude of the spectral shift under the reg QD film treatment was not sufficient to compensate for the reduction in the DLI during this experiment.

Spectral shifts under QD films regulate cryptochrome-mediated plant responses.

Light spectral quality is perceived by photoreceptors, such as phytochromes and cryptochromes, and regulates a wide range of plant responses mediated by these photoreceptors. Despite significant spectral shifts under the reg and high QD film treatments (compared with those under the control and PE film treatments), the PSS remained consistent across all four treatments. However, we observed distinct differences in plant growth and morphology among the treatments for all three crops. For basil plants grown under the high QD film treatment, significant increases in the plant height, stem fresh weight, stem dry weight, number of leaves, and total leaf area compared with those associated with the PE film treatment were detected (Figs. 3C and 4C–F). Because the DLI and PSS were similar under the PE film and QD film, this morphological change in basil was most likely attributable to cryptochrome deactivation because the blue light fraction within the 320- to 800-nm spectral range decreased from 16.9% under the PE film to 11.6% under the high QD film (Table 1). Prior studies reported that shade avoidance responses, which are primarily triggered by green and far-red photons, can induce increased stem elongation and leaf expansion and enhance light capture at the plant canopy level (Kim et al. 2004; Meng and Runkle 2019; Pierik and de Wit 2013; Robson et al. 1993; Smith et al. 2017; Zhen and Bugbee 2020a,b). Although basil grown under high QD film exhibited typical symptoms of the shade avoidance responses, and although green and far-red light were relatively enriched under the high QD film, the ratio of red to far-red light and PSS remained similar among all four treatments because the fraction of red light also increased along with that of far-red light under the QD films. Therefore, the increased stem elongation and leaf expansion of basil under the high QD film were likely not caused by the effects of phytochromes; instead, they were likely caused by cryptochrome deactivation. Ultraviolet-A and blue photons are widely known to activate cryptochromes, leading to a decrease in hypocotyl (and stem) elongation and leaf area, resulting in more compact plants (Meng et al. 2020; Ohashi-Kaneko et al. 2007; Son and Oh 2013).

Green photons, which stimulate the shade avoidance responses, can deactivate cryptochromes (Battle et al. 2020; Bouly et al. 2007; Kusuma et al. 2021; Müller and Ahmad 2011). During a short-term experiment, Bouly et al. (2007) reported that wild-type Arabidopsis seedlings exposed to blue photons (445 nm) for only 48 h showed reduced hypocotyl lengths and higher anthocyanin concentrations compared with those of seedlings grown under a mixture of blue and green photons. Meng et al. (2019) showed that decreasing blue photons from 60 to 0 μmol⋅m⁻²⋅s⁻¹ while simultaneously increasing green photon flux from 0 to 60 μmol⋅m⁻²⋅s⁻¹ (i.e., substituting green for blue photons) resulted in more leaf expansion and higher biomass production in green and red lettuces and kale during long-term crop cultivation under LEDs. In this study, we observed a slight increase in green photons but a significant decrease in the blue photon fraction under the high QD film treatment. Despite having a significantly lower DLI under the high QD film treatment, plant biomass was comparable to that under the control treatment, especially that of red leaf lettuce and basil (Fig. 3). This was likely attributable to enhanced leaf expansion and radiation capture because the total leaf area was highest under the high QD film treatment in all crops. Our findings of red and green lettuces and basil indicated that the reduced DLI under the high QD film treatment may be offset by the spectral shift that promotes cryptochrome deactivation.

Pigmentation was affected by both light intensity and spectral quality.

We observed the highest chlorophyll and anthocyanin contents in the control treatment with unconverted greenhouse sunlight (Fig. 4), which might have resulted from the combined effects of the higher DLI and higher blue photon fraction compared with those of the PE and QD film treatments. The chlorophyll content tends to increase under higher light intensity; for example, Snowden et al. (2016) found that a higher light intensity of 500 μmol⋅m⁻²⋅s⁻¹ resulted in an increased chlorophyll concentration in seven species/cultivars, including green lettuce, compared with that associated with a lower light intensity of 200 μmol⋅m⁻²⋅s⁻¹.

Under high blue light, active cryptochromes inhibit leaf expansion and stem extension but promote pigments, including chlorophylls and anthocyanins (Bouly et al. 2007; Kusuma et al. 2021; Park and Runkle 2023; Snowden et al. 2016). A reduction in the blue light fraction under the QD films likely led to reduced cryptochrome activities. Among PE and both QD film treatments (which had similar DLIs), green leaf lettuce grown under the reg QD film and high QD film treatments (with reduced blue light fractions) tended to have reduced leaf chlorophyll contents compared with those of plants grown under the PE film (Fig. 5A). The leaf chlorophyll contents of green and red leaf lettuces and kale grown under LEDs decreased as blue photons decreased from 60 to 0 μmol⋅m⁻²⋅s⁻¹ (Meng et al. 2019). Similarly, Hogewoning et al. (2010) reported that the leaf area-based chlorophyll content in cucumber decreased with the decreasing blue light fraction (from 50% to 0%) under a red background light. Even though these previous studies were conducted in climate-controlled growth rooms using LEDs rather than in a greenhouse, the trends of the chlorophyll responses mirrored the responses of green lettuce to spectral shifts under QD films. No significant differences in the chlorophyll contents were observed in sweet basil grown under the PE, reg QD film, and high QD film treatments (Fig. 5B), which indicated varying sensitivity to blue light among species.

The anthocyanin index in red leaf lettuce exhibited a pattern similar to that of the responses of the chlorophyll content. In red leaf lettuce, the anthocyanin index was highest in the control group, followed by the PE film, reg QD film, and high QD film treatment groups (Fig. 5C). Our findings aligned with those of a previous study that used a Laboratory color space analysis as an indicator of anthocyanin accumulation and found that ‘Rouxai’ red lettuce exhibited reduced anthocyanin accumulation as the blue photons decreased from 60 to 0 μmol⋅m⁻²⋅s⁻¹ (Meng et al. 2019). Other studies reported that blue light promotes anthocyanin production by stimulating cryptochromes (Giliberto et al. 2005; Li and Kubota 2009; Meng et al. 2004). Meng et al. (2004) found that blue photons promoted the gene expression of chalcone synthase and dihydroflavonol-4-reductase, which are key regulators of the anthocyanin pathway. Both the lower DLI and reduced blue photon fraction under the high QD film treatment may result in reduced activation of cryptochromes compared with that of the control treatment. A reduction in the anthocyanin content in plants grown under QD films should be considered because higher anthocyanin accumulation in red leaf lettuce and other red-colored leafy vegetables is generally desirable.

Implications for the development of novel greenhouse glazing materials.

Our results suggested that a sufficient magnitude of spectral red shift can overcome DLI loss and lead to leaf expansion and yield increases. However, the DLI loss of the QD film itself (and that of the control PE film) is clearly responsible for the initial yield loss observed with the PE film and reg QD film during this study. For the greenhouse industry, direct incorporation of QDs into the façade materials of greenhouses would lead to minimal intensity loss if applied as an added layer in an extruded PE cover, rather than being applied as an additional internal layer within an existing greenhouse. This may lead to the realization of the benefit of spectral red shifting without the DLI loss. Additionally, QDs could potentially be applied as an inner layer of a glass, acrylic, or polycarbonate greenhouse glazing.