Experimental Design
The experiment was organized in a factorial design with 9 plants grown under each combination of light intensity (4) and photoperiod (2) for a total of 8 treatments. The experiment was conducted twice over time. At harvest, morphological data were collected on all 9 plants per replication. Plant-specific phytonutrient and mineral data were collected on all 9 plants per replication, except for plants grown at a 200 µmol·m− 2·s− 1 light intensity and a 16-hr photoperiod which did not provide enough tissue per plant for mineral nutrient extractions. Tissue from multiple plants was pooled to produce mineral concentration data from this treatment.
Plant Production
Seeds of Brassica rapa spp. nipposinica ‘Red Hybrid’ mizuna (Kitazawa Seed, Oakland, CA) were sown in 162-cell bound coco-peat cubes (Preforma plugs; Jiffy Growing Solutions, Lorain, OH) and placed in a walk-in growth chamber (TC2, Environmental Grow Chambers; Chagrin Falls, OH). Plant production was similar to Darby et al. [7] with treatment and other select differences. The temperature was 22.7 ± 0.3° C, and the relative humidity was 61.6 ± 7.3%, both measured and logged every 10 minutes by a shielded temperature and relative humidity sensor (Hobo MX2300; OnSet, Bourne, MA). Seedling trays were placed under broad-spectrum light-emitting diodes (Scorpion Diablo, Horticulture Lighting Group; Maynardville, TN) providing 21:36:43 blue:green:red (%) radiation ratios, with a red:far-red ratio of 13:1, and one of four light intensities applied at one of two photoperiods for a total of eight treatments (Table 1). Light intensity was monitored by quantum sensors (SQ-500S; Apogee, Logan, UT) for the duration of the experiment. Leaf surface temperature was monitored with ultra-narrow field of view infrared radiometers (SI-131-SS; Apogee, Logan, UT), averaging 21.8 ± 1.1° C across all treatments (Table 1). CO2 (Industrial Grade Carbon Dioxide CD 50S; Airgas, Knoxville, TN) was injected to provide 2,770 ± 180 µmol∙mol− 1. Dosing was monitored and controlled by an infrared gas analyzer (LI-850; LI-COR Biosciences, Lincoln, NE) connected to a relay switch controlling an electronic regulator. CO2 concentration was reported to a Raspberry Pi (Raspberry Pi 4 B; Raspberry Pi, Cambridge, England) every 20 seconds where it was recorded in a CSV file. Seeds were irrigated daily with reverse-osmosis water supplemented with 12N-1.8P-13.4K water-soluble fertilizer, providing (mg⋅L− 1) 100 nitrogen, 15 phosphorus, 112 potassium, 58 calcium, 17 magnesium, 2 sulfur, 1.4 iron, 0.5 zinc, 0.4 copper and manganese, and 0.1 boron and molybdenum, (RO Hydro FeED; JR Peters, Inc., Allentown, PA) and magnesium sulfate (MgSO4) to provide (mg⋅L− 1) 15 magnesium and 20 sulfur. The pH was adjusted to 5.8 with H2SO4 or KCHO3.
Table 1
Lighting treatments, their intensity, photoperiod, and daily light integral (DLI), as provided to the initial growing surface.
Target Light Intensity | Actual Light Intensity | Photoperiod | Actual Daily Light Integral | Leaf Temperature |
(µmol⋅m− 2⋅s− 1) | (µmol⋅m− 2⋅s− 1) | (hrs) | (mol⋅m− 2⋅d− 1) | (° C) |
200 | 202 ± 10 | 16 | 11.6 ± 0.6 | 21.6 ± 0.8 |
400 | 391 ± 24 | 16 | 22.5 ± 1.4 | 21.4 ± 1.0 |
600 | 583 ± 40 | 16 | 33.6 ± 2.3 | 21.9 ± 1.2 |
800 | 780 ± 30 | 16 | 44.9 ± 1.7 | 22.1 ± 1.5 |
200 | 203 ± 11 | 24 | 17.5 ± 1.0 | 21.5 ± 0.6 |
400 | 396 ± 16 | 24 | 34.2 ± 1.4 | 22.1 ± 0.9 |
600 | 603 ± 25 | 24 | 52.1 ± 2.2 | 22.0 ± 1.0 |
800 | 782 ± 38 | 24 | 67.6 ± 3.3 | 22.4 ± 1.1 |
Similar to Darby et al. [7], after thirteen days the seedlings were transplanted into 18-cm-deep by 61-cm-wide by 122-cm-long deep-water culture hydroponic systems (Premium Flood Table, Active Aqua; Petaluma, CA) with 61 by 122 cm sealed-surface foam rafts (36 ct lettuce raft; Beaver Plastics, Acheson, AB, Canada) and grown for 9 days. Plants were harvested 22 days after sowing to prevent over-crowding of the growing space. The environmental conditions after transplant were the same as the seedling stage. Eight systems were filled with 108 L of reverse-osmosis water and supplemented with 12N-1.8P-13.4K water-soluble fertilizer (RO Hydro FeED; JR Peters, Inc.) and MgSO4 providing twice the concentrations reported during propagation. Electrical conductivity (EC) and pH were monitored (HI9813-6N Portable Waterproof pH/EC/TDS Meter; Hanna Instruments, Woonsocket, RI) and pH was adjusted to 5.8 using H2SO4 or KCHO3. Air pumps (Active Aqua 110 L⋅min− 1 commercial air pump; Hydrofarm, Petaluma, CA) and air stones (Active Aqua air stone round 10 cm × 2.5 cm; Hydrofarm, Petaluma, CA) were used to provide dissolved oxygen to the nutrient solution.
Harvest, Morphological Data Collection, and Tissue Processing
Plants were harvested at the substrate surface and morphological data were collected [7]. Height from the substrate surface to the tip of the tallest leaf, width at the widest point and perpendicular to the widest point, the number of fully expanded leaves, and fresh mass were collected for 9 plants per replication. After weighing, fresh tissue of 9 plants per treatment was immediately placed in a plastic sample bag and flash frozen in liquid N. Tissue was then stored in a -80°C freezer, freeze dried, and homogenized in liquid N. Once homogenized, the tissue was divided into separate aliquots for future extraction procedures and stored in individual 15 mL centrifuge tubes in the − 80°C freezer.
Anthocyanin Isolation and Quantification
Total anthocyanins were extracted and analyzed using a method derived from Islam et al. [17] and previously used in Darby et al. [7]. Briefly, 50 mg of ground tissue was used per sample, and the procedure occurred under red light. Samples were saturated with 5 mL of 95% ethanol/1.5 N HCl (85:15, v:v), placed on an orbital shaker for 15 minutes, and stored at 4°C for 24 hrs. Samples were then filtered and analyzed with a microplate reader (Agilent Technologies, Santa Clara, CA).
Carotenoid Isolation and Quantification
The carotenoids β-carotene, lutein, and zeaxanthin were extracted and analyzed using a method derived from Kopsell et al. [18] and previously described by Darby et al. [7]. Under red light, 50 mg samples of chilled tissue were extracted with purified water and tetrahydrofuran, and an internal carotenoid standard was used to quantify sample loss during homogenization. Samples were homogenized and tetrahydrofuran was used for a second extraction followed by additional homogenization. The resulting solution was centrifuged, and the eluent was dried down on a stream evaporator. After the addition of acetone, samples were filtered, and an aliquot was stored for subsequent identification and quantification on a 1200 Agilent Series HPLC unit equipped with a diode array detector. A reverse phase C30 column was used with an isocratic mobile phase composed of methyl tert-butyl ether, methanol, and triethylamine (11%, 88.99%, and 0.01%, v/v/v).
Fat Soluble Vitamin Isolation and Quantification
The fat-soluble vitamin phylloquinone (vitamin K1) was extracted using a method derived from Pokkanta et al. [19] and previously used in Darby et al. [7]. Extractions were performed under red light (peak wavelength: 654 nm) and samples were saturated with a series of three solvents (methanol, dichloromethane, and hexane). After being saturated with each solvent, the samples underwent orbital shaking, sonication, and were finally centrifuged. After removing the supernatant, the next solvent in the series was added and the samples again underwent the three previously mentioned procedures. An internal standard was added to the pooled supernatant, and the samples were dried completely, suspended in 5 mL dichloromethane, and filtered. A 1-mL aliquot was then collected for analysis with a 1200 Agilent Series HPLC equipped with a diode array detector, using a reverse phase C18 column and a mobile gradient composed of methanol and water. Phylloquinone was identified at 248 nm.
Mineral Isolation and Quantification
Nitrogen concentration was quantified via combustion of 0.25 g of tissue and detection with a Thermal Conductivity Sensor (vario MAX cube; Elementar, Ronkonkoma, NY). Other minerals, Ca2+, K+, Mg2+, and Fe3+ were extracted and quantified using a method derived from Jones [20]. A 0.5 g aliquot of tissue was dry ashed in a furnace at 500°C, and then wet ashed by the addition of dilute acid. The resulting solution was then analyzed on an ICP-OES to determine individual mineral concentrations (SPECTROBLUE; Spectro Analytical Instruments Inc., Kleve, Germany). Due to limited tissue production under the 200 µmol⋅m− 2⋅s− 1 by 16-hr photoperiod, mineral analysis on this treatment was performed on pooled samples (with two plants contributing tissue to each sample).
Water Soluble Vitamin Isolation and Quantification
The water-soluble vitamins (WSVs) ascorbic acid (vitamin C) and thiamine (vitamin B1) were extracted using a method derived from Seal and Chaudhuri [21] and Sun et al. [22], and previously used in Darby et al. [7]. Under red light, chilled samples of 200 mg were hydrated with purified water, hydrochloric acid, and a phosphate buffer. Samples were vortexed and supernatant was filtered. The process was repeated, and a 1 mL aliquot was collected for analysis on a 1200 Agilent Series HPLC equipped with a diode array detector. A reverse phase C18 column was used alongside an aqueous formic acid mobile phase. Both vitamins were detected at 255 nm.
Statistical Analysis
Statistical analysis was performed in R [23]. Data were organized and compiled using the ‘tidyverse’ package [24], and an initial two-way analysis of variance was performed with the ‘stats’ package to determine any interaction between the effect of light-intensity and photoperiod on the 14 parameters listed in Table 2 [23]; when interactions were not present, data were pooled. Linear or quadratic regression were performed with the ‘stats’ package when light intensity or the interaction of light intensity and photoperiod was significantly different [23]. For final scoring, treatment means of each parameter were calculated, min-max normalized, and multiplied by a weighting value, according to the values given in Table 2. Weighted means of each parameter were then added together per treatment to produce a final score for comparison of total treatment value. To determine energy use efficiency, the wattage used per treatment per day to power the LED lights was calculated and converted to megajoules. The mean total yield per metric was then divided by the energy used to produce said yield, further described in the following equation:
$$mg\bullet {MJ}^{-1} = \frac{\left(mean concentration \bullet mean dry mass yield\right)}{\left(fixture wattage utilized\bullet daily photoperiod hours\right)(0.0036 MJ\bullet {watt hour}^{-1})}$$
The exception to this equation was the calculation for fresh mass energy use efficiency, in which case there was no need to calculate a total yield from the concentration and the dry mass yield. Instead, the mean fresh mass yield was used as the numerator and the end result was described in g∙MJ− 1. Visualizations were created with ggplot [24].
Table 2
List of parameters of interest to NASA, their class, and the multiplier to be applied to the normalized mean. Multipliers were based on previous NASA research and reflect the importance of each parameter to astronaut health on future exploration-length missions (Massa et al, 2015). Negative values indicate undesired plant traits, while magnitude of value indicates importance.
Metric | Category | Multiplier | Justification |
Fresh Mass | Morphological | × 2.5 | Total phytonutrient content scales with this metric |
Dry Matter | Morphological | × 1.0 | Higher dry mass proportion is desirable |
Plant Volume | Morphological | × -1.5 | Smaller plants are desirable due to limited space |
Ca | Minerals | × 1.5 | Mitigates bone loss in microgravity [25] |
K | Minerals | × 2.0 | Currently deficient in space diet [6] |
Mg | Minerals | × 1.0 | Mitigates bone loss in microgravity |
Fe | Minerals | × -1.5 | Excessive iron can exacerbate bone loss in microgravity [6] |
Ascorbic Acid | Water-Soluble Vitamins | × 1.5 | Prevent scurvy; degrades quickly [2, 26] |
Thiamine | Water-Soluble Vitamins | × 1.5 | Prevent beriberi; degrades quickly [2, 26] |
Phylloquinone | Fat-Soluble Vitamin | × 1.5 | Promote proper blood coagulation; deficient in stored space diet [6, 27] |
β-carotene | Carotenoids | × 1.5 | Vitamin A precursor; degrades quickly [2, 28] |
Lutein | Carotenoids | × 1.5 | Preventing ocular photodamage [29] |
Zeaxanthin | Carotenoids | × 1.5 | Preventing ocular photodamage [29] |
Total Anthocyanin | Anthocyanins | × 1.0 | Potentially beneficial as an antioxidant |