Balancing Nutrient Content and Nitrate Levels in Space Agriculture: Investigating LED Light and CO2 Effects on Space-Grown Leafy Green Vegetables

All seeds, growth media (arcilite, Turface ProLeague from Profile Products, LLC, Buffalo Grove, IL), 4” nursery pots, plant trays, wicking sheets, and 18-6-8 T180 controlled-release fertilizer (Nutricote from Florikan E.S.A., LLC, Sarasota, FL) and growth chambers were provided by the GBE program (https://fairchildgarden.org/science-and-education/science/gbe/, Fairchild Tropical Botanic Garden, Miami, FL). Experiments adhered to the GBE 2023 research protocol to simulate ISS conditions. Plants were cultivated in GBE growth chambers (16” W × 13” D × 17.9” H) designed to mirror the key features of the Vegetable Production System (Veggie) and Advanced Plant Habitat (APH) on the ISS. The Veggie is an open growth chamber designed as a “salad machine” to establish baseline salad performance under 1 g since 2014 (Massa et al., 2016; Khodadad et al., 2020), whereas APH is a closed chamber with advanced environmental sensors (Massa et al., 2016; Fritsche et al., 2024).

The GBE growth chambers used in this study had four reflective side panels to simulate APH's enclosed, controlled system, and six plant pots arranged to simulate six plant pillows used in the Veggie system (Massa et al., 2017). Each chamber also included an LED lighting panel, (a quantum board with red, green, blue, and white LEDs from Horticulture Lighting Group, LLC, Westerville, OH, custom order by GBE), an LED controller, a centrally mounted circulation fan running continuously (24/7), and environmental sensors including a digital hygrometer, a CO₂ monitor, a kilowatt meter, and an Apogee quantum meter for photosynthetic photon flux density (PPFD) measurement (Fig. 2). Each nursery pot was filled with approximately 450 mL of arcilite and supplemented with 3.3 grams of Nutricote controlled-release fertilizer (18-6-8 T180).

Figure 2.

GBE Growth Chamber Setup. (A) The chamber included an LED lighting panel (equipped with red, green, blue, and white LEDs), an LED controller, four reflective side panels, a circulation fan, and environmental sensors (CO₂ monitor, kilowatt meter, and quantum meter) to mirror key functions of Veggie and APH operated on the ISS. (B) The quantum meter was placed at the bottom center of the chamber to measure the chamber light intensity in μmol·m⁻²·s⁻¹ PPFD. (C) LED distribution pattern drawing (top) and a real-world view of the chamber fan (bottom).

For each trial, one cultivar was grown per chamber for 28 days under a 12-hour photoperiod. Six cultivars were tested: red romaine lettuce (Lactuca sativa cv. Outredgeous), mizuna (Brassica rapa var. nipposinica), mustard (JS) (Brassica carinata cv. Amara), hybrid leafy Asian green (Brassica rapa var. chinensis cv. Rosie F1), Scarlet Frills mustard greens (Brassica juncea), and Garnet Giant mustard greens (Brassica juncea) (Table 1). Two seeds of the same cultivar were planted per pot, and seedlings were thinned to one per pot upon the appearance of the first true leaf. Each chamber contained three pots of irradiated seeds and three pots of nonirradiated seeds of the same cultivar, enabling paired comparisons. Across all trials and schools, this setup yielded sample sizes ranging from 6 to 102 plants per cultivar, depending on the number of replicates and availability of chambers.

To simulate space radiation exposure, 50% of the seeds used in each trial were irradiated at the Colorado State University Neutron Radiation Facility using Californium-252, a neutron-emitting isotope leveraged by NASA researchers such as Alexander Meyers at the Kennedy Space Center to simulate aspects of spaceflight radiation for educational and experimental purposes in the GBE 2023–2024 program. The seeds were vacuum-sealed in a single layer and positioned at a calibrated distance from the neutron source at the facility to ensure uniform neutron exposure [Fig. 3(A) and Fig. 3(B)]. Although dosimetry data are pending, the radiation dose is estimated to be equivalent to approximately 10 years of exposure aboard the ISS (Meyers, 2023).

Figure 3.

Neutron Radiation Facility at Colorado State University. (A) Rodent radiation experiment setup, showing neutron radiation source in the center. (B) Plant seed radiation setup, illustrating the positioning of seed packets for uniform neutron exposure. Both images are from Dr. Alexander Meyers' 2023 presentation, used with permission.

Figure 4.

Custom-designed 3D-printed dry-ice CO₂ dispersion tube setup. (A) A computer-assisted design (CAD) drawing for the system and (B) a real-life image of the system.

To test the hypothesis that higher light intensity would reduce NO₃⁻ content (Gómez and Jiménez, 2020), Outredgeous red romaine lettuce was selected to grow under two LED light conditions. The control group was exposed to 230 μmol·m⁻²·s⁻¹ PPFD with a 1:1 red-to-blue LED ratio (25 μmol·m⁻²·s⁻¹ red, 25 μmol·m⁻²·s⁻¹ blue) and 180 μmol·m⁻²·s⁻¹ white LEDs. In the experimental group, light intensity was increased to 305 μmol·m⁻²·s⁻¹ PPFD, maintaining the same red and blue ratios (25 red, 25 blue) while raising the white LED output to 255 μmol·m⁻²·s⁻¹.

To simulate elevated CO₂ levels similar to those observed in the ISS cabin environment (Wheeler et al., 2024), a custom-designed dry-ice dispersion system was utilized. The device, positioned above the chamber fan with a mesh grating 2.5 inches above the airflow, allowed solid dry ice (Penguin Brand, Meijer, Midland, MI) to sublimate directly into circulation. The fan operated 24/7 to maintain consistent mixing and dispersal. Ambient CO₂ (i.e., classroom) served as the terrestrial control level, averaging approximately 500 ppm (500 ± 20 ppm) over a 28-day growing period. Elevated CO₂ treatments (~1000 ppm, ~1250 ppm, and ~1500 ppm) were established by adding ~500 g dry ice per application, repeated 9 to 15 times over during the experiment. CO₂ concentrations were monitored using a meter placed just in front of each growth chamber. When chamber temperature dropped to 16°C, a seedling heat map was placed under the plant tray to maintain adequate warmth. Table 2 summarizes the experimental design, including the photoperiods and the CO₂ treatment design alongside radiation and light intensity treatments.

Because CO₂ meters were placed outside the semi-enclosed chambers—each with gaps between panels and a 2-inch open base—internal CO₂ levels could not be measured directly. To estimate internal concentrations, we modeled daily gas accumulation over 28 days based on dry ice input, chamber volume (~ 60.9 L), and average temperature and relative humidity. Calculations used the ideal gas law, with a correction for water vapor pressure. A full description of the modeling procedure and worked example are provided in Supplemental Methods (see Supplemental materials). To account for CO₂ loss via diffusion and plant uptake, we applied a daily loss rate of 12.5% ±2.5%, consistent with NASA studies reporting 5~10% daily losses in closed systems (Wheeler et al. 1991, 2014) and reflecting the less sealed design of the GBE chambers. Table 3 outlines the modeled average daily internal CO₂ concentrations for each experimental group.

Plant growth data (plant dimensions and observations) and environmental data (chamber temperature, chamber humidity, and total water usage) were monitored daily during weekdays and recorded weekly. Table 4 summarizes the environmental data, including the average temperatures, relative humidity, water consumption, and CO₂ across all trials, along with the associated standard errors of the means (SEM). Following harvest, various growth parameters and nutrient concentrations of plant leaf tissue were measured to assess the effects of radiation, light intensity, and CO₂ treatments on plant growth and nutrient composition. Radiation effects were evaluated by germination rates, edible biomass, and plant volume of six plant types with data sourced from 85 GBE schools (n = 6 to n = 102 per plant type). Light intensity experiments focused on analyzing NO₃⁻ and nutrient content in red romaine lettuce cultivated by Dow High Space Farmers (DHSF), whereas CO₂ treatment effects were evaluated using hybrid leafy Asian green from Jefferson Middle School (JMS). Guided by the nutrient prioritization outlined in the Introduction, analyses targeted K, Mg, Ca, Fe, and NO₃⁻ levels. At harvest, plant leaf tissue was collected from each treatment group and submitted to a certified commercial laboratory (A&L Great Lakes Laboratories, Fairview, IL) for nutrient composition analysis. The laboratory quantified K, Mg, Ca, Fe, and NO₃⁻ concentrations using standard Association of Official Analytical Chemists (AOAC) protocols for plant nutrient tissue analysis. Nutrient values were used to generate normalized, weighted scores as described in the Spaceflight Suitability Matrix.

Statistical analyses were conducted using a custom Python 3.9 workflow to assess treatment effects and validate school-based findings against GBE data sets. Welch's t-tests were used to compare local and GBE-wide results, and two-way ANOVA and Tukey's HSD tested the effects of radiation, light intensity, and CO₂ enrichment on plant growth and nutrient composition. Pearson correlation and linear regression were used to evaluate relationships with NO₃⁻ levels. Cumulative sample sizes used in statistical comparisons are reported in figure captions. Graphical outputs included bar plots with standard error of the mean (SEM), along with statistical annotations such as p-values and grouping letters (e.g., A, B, C) at a significance threshold of p < 0.05.

To evaluate spaceflight suitability across CO₂ and radiation treatments, a composite scoring system was developed using trait-specific weights adapted from NASA's CRLs criteria (Massa et al., 2015; Fritsche et al., 2024). Table 5 outlines the assigned weights and physiological justifications for each trait. Plant-growth-related factors included germination rate (x1.5), edible biomass (x2), plant volume (x1.5, inverted), while nutritional factors included K (x2), Mg (x1.5), Ca (x1), Fe (x1.5, inverted), and NO₃⁻ (x1.5, inverted). All values were normalized to a 0–1 scale using min–max scaling. For traits where lower values are considered more favorable—such as plant volume, Fe, and NO₃⁻—normalized scores were inverted by subtracting each from 1. These normalized values were then multiplied by their respective weight, and the resulting weighted scores were summed to yield a composite score for each treatment. This scoring approach was first applied to nutrient-only traits (Table 6) and then extended to include growth and morphological parameters in the comprehensive treatment evaluations presented in Tables 7 and 8 (see Results section).

Neutron radiation did not significantly impact plant morphology, edible biomass, or plant volume across all six tested cultivars grown under 230 μmol·m⁻²·s⁻¹ PPFD conditions. Visual comparison of red romaine lettuce grown showed no observable differences in leaf count, coloration, or structure between irradiated and nonirradiated plants [Fig. 5(A)]. Statistical analysis confirmed these observations. Two-way ANOVA revealed no significant differences in edible biomass or plant volume across all six tested plant types that grew under 230 μmol·m⁻²·s⁻¹ PPFD [Fig. 5(B) and Fig. 5(C)]. Among plant types, mizuna exhibited the most significant variability in biomass and volume, with extreme outliers in nonirradiated groups. Hybrid leafy Asian green and Scarlet Frills mustard green showed the lowest median biomass, while red romaine, hybrid leafy Asian green, and mustard (JS) maintained compact growth profiles. Considering space agriculture suitability, irradiated hybrid leafy Asian green and red romaine lettuce ranked among the top performers, offering an optimal balance of high edible biomass and compact morphology [Fig. 5(D)].

Figure 5.

(A) Red romaine lettuce morphology under 230 μmol·m⁻²·s⁻¹ PPFD treatment. (B) Boxplot analyses of edible biomass across six cultivars. Data are presented as mean ± SEM, with varying sample sizes per cultivar (n = 42–102). Letters indicate statistically significant differences across plant types. (C) Boxplot analyses of plant volume across six cultivars. Data are presented as mean ± SEM, with varying sample sizes per cultivar (n = 42–102). Letters indicate statistically significant differences across plant types. (D) A scatter plot of average edible biomass versus plant volume for irradiated plants across six testing cultivars.

Independent two-sample t-tests confirmed that germination rates, edible biomass, and plant volume from DHSF and JMS did not differ significantly from the broader GBE data sets (all p > 0.05). As shown in Fig. 6(A) and Fig. 6(B), DHSF and JMS results fell within the variability observed across GBE schools, validating the use of their data in subsequent analyses of light and CO₂ treatment effects.

Figure 6.

(A) Comparison of germination rates, edible biomass, and plant volume between GBE schools (n = 102) and DHSF (n = 6) for irradiated and nonirradiated red romaine lettuce. Data are presented as mean ± SEM. (B) Comparison of germination rates, edible biomass, and plant volume between GBE schools (n = 42) and JMS (n = 6) for irradiated and nonirradiated hybrid leafy Asian green. Data are presented as mean ± SEM.

Similar to the effects observed with neutron radiation, variations in light intensity (230 μmol·m⁻²·s⁻¹ vs. 305 μmol·m⁻²·s⁻¹ PPFD) did not significantly influence germination rates, edible biomass, or plant volume across treatment groups [p > 0.05, two-way ANOVA; Figs. 7(A)–(C)]. Irradiated and nonirradiated plants displayed comparable growth metrics under elevated light conditions, indicating that neither increased light intensity nor its interaction with neutron radiation substantially impacted overall plant development. However, elevated light intensity (305 μmol·m⁻²·s⁻¹ PPFD) did induce distinct pigmentation changes in red romaine lettuce, specifically, an increase in purple pigmentation. Although not part of the original experimental objectives, pigmentation frequency was recorded on-site during plant care and later verified using standardized photographs, which were analyzed with ImageJ software (NIH), with four independent observers cross-checking the results. Irradiated plants exhibited nearly double the pigmentation frequency compared to nonirradiated controls, a relationship confirmed by chi-square analysis (χ² = 15.91, p = 0.00118), suggesting that anthocyanin expression was enhanced under combined high light intensity and radiation exposure (Massa et al., 2015) [Fig. 7(D) and Fig. 8].

Figure 7.

Effects of light intensity (230 vs. 305 μmol m⁻² s⁻¹ PPFD) and neutron irradiation on red romaine lettuce's germination (A), biomass (B), volume (C), and purple dot count (D). No significant differences were detected across treatments. Data are mean ± SEM (n = 6 per treatment).

Figure 8:

A representative image illustrates the purple pigmentation occurrence of red romaine lettuce.

CO₂ enrichment significantly influenced the growth of hybrid leafy Asian green, with the most pronounced effects observed at ~1250 ppm. Under ~1000 ppm CO₂, nonirradiated seeds exhibited a reduced germination rate (33%) compared to their irradiated counterparts (66%). This trend reversed at ~1500 ppm, where nonirradiated seeds achieved 100% germination [Fig. 9(A)]. For edible biomass, plants grown under ~1250 ppm achieved the highest yields (15.9 g for nonirradiated; 15.1 g for irradiated), significantly outperforming those grown under ~1000 ppm and ~1500 ppm CO₂ levels [Fig. 9(B)]. Plant volume followed a similar trend, with the largest mean volumes also observed under ~1250 ppm CO₂ [Fig. 9(C)]. Across all treatments, ~1250 ppm CO₂ treatment consistently enhanced germination, biomass, and volume compared to ~500 ppm (control) and ~1000 ppm. While seed irradiation improved germination at ~1000 ppm, it slightly reduced biomass production at ~1250 ppm.

Figure 9:

Effects of CO₂ enrichment and seed irradiation on harvested hybrid leafy Asian green. (A) Germination rate (%), (B) edible biomass (g), and (C) plant volume (cm³) are shown for four CO₂ levels: ~500 ppm (control, n = 15 per radiation group), ~1000 ppm, (n = 3 per group ), ~1250 ppm (n = 3 per group), and ~1500 ppm (n = 3 per group), with each group, including both nonirradiated (Non-IRR) and irradiated (IRR) plants. Data are presented as mean ± SEM.

Californium-252 neutron radiation did not significantly impact red romaine lettuce nutrient composition. As shown in Fig. 10, K levels remained relatively stable between 230 and 305 μmol·m⁻²·s⁻¹ PPFD treatments, with a slight decrease observed in irradiated plants at higher light intensity (76.9 mg/g). Mg and Ca concentrations were minimally affected. However, Fe concentrations increased significantly under elevated light conditions, reaching 164.6 μg/g in nonirradiated plants. These results suggest that while light intensity had limited effects on macronutrients, it may enhance Fe uptake, posing potential concerns for astronaut health.

Figure 10.

Nutrient composition analysis of harvested rred romaine lettuce plant leaf tissue under two light intensities. Data are mean ± SEM (n = 6 per treatment).

CO₂ enrichment influenced nutrient composition in both irradiated and nonirradiated hybrid leafy Asian green [Fig. 11(A) and Fig. 11(B)]. Although data for the ~500-ppm group were unavailable, NO₃⁻ concentrations decreased as CO₂ levels increased, reaching the lowest values at ~1500 ppm. K and Mg peaked at ~1000 ppm but declined at higher CO₂ concentrations, especially in irradiated plants. Ca remained consistently higher in nonirradiated plants, while Fe concentrations increased across all CO₂ levels, with irradiated plants showing greater accumulation. These findings indicate that radiation amplifies CO₂-driven nutrient imbalances, accelerating Mg and Ca declines while promoting Fe uptake.

Figure 11.

(A) Nutrient composition analysis of harvested nonirradiated (Non-IRR) hybrid leafy Asian green plant leaf tissue under varying CO₂ levels. Data are mean ± SEM (n = 3 per treatment). (B) Nutrient composition analysis of harvested irradiated (IRR) hybrid leafy Asian green plant leaf tissue under varying CO₂ levels. Data are mean ± SEM (n = 3 per treatment).

To assess nutrient balance across CO₂ and radiation treatments, composite scores were calculated for hybrid leafy Asian green using normalized and weighted scoring framework described in the Spaceflight Suitability Matrix in the Methods section (Table 5). As shown in Table 6, the irradiated group at ~1000 ppm CO₂ achieved the most favorable nutrient profile (sum = 5.04), and the ~1250 ppm nonirradiated group ranked second (sum = 4.16). In contrast, the ~500 ppm CO2 treatments scored lowest for both nonirradiated and irradiated groups (sum = 1.32–1.71). The ~1500 ppm irradiated group exhibited the second-lowest score (sum = 1.93), due to elevated Fe and reduced Mg levels. These results indicate the ~1000 ppm irradiated group optimally maintains nutrient stability under simulated spaceflight conditions.

Light intensity and CO₂ levels significantly influenced NO₃⁻ metabolism in space-grown crops. Red romaine lettuce grown under elevated light (305 μmol·m⁻²·s⁻¹ PPFD) accumulated more NO₃⁻ than those under 230 μmol·m⁻²·s⁻¹ PPFD, with irradiated plants showing a steeper trendline (0.0107 versus 0.0013), suggesting radiation may accelerate NO₃⁻ accumulation [Fig. 12(A)]. While a moderate positive correlation (R = 0.503) was observed, it did not reach statistical significance (p = 0.309), indicating that more research is needed to confirm this trend. Conversely, CO₂ enrichment significantly reduced NO₃⁻ levels in hybrid leafy Asian green, with a strong negative correlation (R = −0.845, p = 0.049) [Fig. 12(B)]. This reduction was more pronounced in irradiated plants (slope = −4.35) than in nonirradiated ones (slope = −2.0), highlighting CO₂ as an effective strategy to mitigate NO₃⁻ accumulation in space-grown vegetables. These findings suggest that optimizing light and CO₂ conditions can improve astronaut dietary safety during long-duration missions.

Figure 12.

(A) Effects of light intensity on NO₃⁻ concentrations of harvest nonirradiated (Non-IRR) and irradiated (IRR) red romaine lettuce plant leaf tissue. Data are mean ± SEM (n = 6 per treatment). (B) Effects of CO₂ on NO₃⁻ concentrations of harvested (Non-IRR) and irradiated (IRR) hybrid Asian leafy green plant leaf tissue. Data are mean ± SEM (n = 3 per treatment).

Red romaine lettuce and hybrid leafy Asian green emerged as promising candidates for space agriculture, demonstrating compact growth and favorable nutrient profiles aligned with NASA's dietary targets (Massa et al., 2015; Fritsche et al., 2024). NO₃⁻ levels in irradiated red romaine were well below World Health Organization limits—requiring over 8.37 cups/day to exceed the threshold—and were further reduced by CO₂ enrichment. NO₃⁻ may also support cardiovascular health in space, though risks remain if combined with NO₂⁻-preserved foods. More concerning were elevated Fe levels under high CO₂ and light conditions, given the already Fe-rich space diet and its link to oxidative stress and bone loss. These findings underscore the need for precise control of growth conditions to make leafy greens safe for regular astronaut consumption.