The present experiments were carried out at the Department of Agricultural, Forest and Food Sciences (DISAFA), University of Turin, Grugliasco, Italy (45°03′59.73″N, 7°35′24.72″E). Two separate trials on biofortification of dill microgreens with selenium (Na₂SeO₄) through priming were conducted, under greenhouse (E1) and indoor farm (E2) conditions. The cultivation took place as follows, and as reported in a preliminary test described by Bulgari et al., 2022. Dill seeds from Cooperativa Ortofrutticola Srl, Albenga, Italy, underwent disinfection process using 5% of sodium hypochlorite for 5 min, followed by thorough washing with distilled water, repeated 10 times. The seeds were primed at room temperature (RT) for 8 h in selenium solution at 1.5 and 3 mg L^-1 concentrations. For hydro-priming, the seeds were immersed in distilled water for an equal duration. Unprimed dry seeds were considered as control. Following each treatment, the seeds underwent three washes with distilled water and were subsequently dried in an oven at 44°C for 15 h, until they reached the initial moisture level. Afterward, 1 g of dill seeds were uniformly sown in aluminum boxes (130x70x40 mm) filled with hemp-pad substrate, which was chosen because it is a typical product of the area where the cultivation took place and therefore also easy to find. The disinfection of the hemp was carried out using a 5% sodium hypochlorite solution and then rinsing with distilled water. The seeds were sown and immediately transferred to the experimental glasshouse, for E1, under monitored conditions (23°C, 35% RH, 11.3 MJ m^-2, and 12 h photoperiod, as average values during the experimental period). They were irrigated with distilled water until complete germination. Ten days after sowing, seeds were irrigated using a 40/60 N-NO₃ ⁻/NNH₄ ⁺ nutrient solution, composed of (all in mmol L⁻¹): 6 N, 2 P, 6 K, 2 Mg and 2.5 Ca. During the experiment, a total of 5.2 L of nutrient solution was used. Microgreens were harvested 25 days after sowing. They were collected at the base, approximately 5 mm from the growing pad.

In the E2, the seeds preparation and treatments were the same of the E1, except that the boxes were transferred to a small indoor farm (Microtype:<3 m³ volume). Temperature and relative humidity of the indoor farm were 24°C and 55%, respectively. The indoor farm was equipped with LED lights. The LED spectrum consisted of 24% blue, 15% green, 56% red and 11% far-red, with a PPFD of 225 µmol m⁻² s⁻¹ and a light/dark period of 14/10 h. The amount of nutrient solution used in this experiment was the same as the E1 (5.2 L). In the trial, microgreens were also harvested after 25 days, in the same way as previously reported.

For each experiment, 4 treatments were applied with 3 replications, and 2 aluminum boxes for replication were used. From each replication, 1 aluminum box was selected to measure the morphological traits, immediately at harvest. In order to perform biochemical assays, the remaining plant material was stored at -80°C, until the time of determination.

The growth parameters of dill microgreens, such as the plant height and the fresh weight (FW), were measured at harvest, which occurred 25 days after sowing. The yield of microgreens was indicated in g m^-2. After these measurements, the samples were transferred for 3 days into an oven with a constant temperature of 50°C, to calculate the dry weight (DW). After that, the dry matter percentage (DM%) was calculated.

For the mineral elements determination, dried microgreens were digested in wet conditions [HNO₃ 67% and H₂O₂ concentrated (30%)]. In detail, 0.25 g of sample were digested with a mixture of 8 mL of nitric acid and 2 mL of hydrogen peroxide, in microwave (method ISO 11466 - Milestone Ethos D, Sorisole (BG), Italy). The digested sample was then filtered with cellulose filters (Whatman Grade 41) and made up to volume to 50 mL with deionized water. The mineral elements (K, Mg, and Se) were then read, with appropriate dilutions, by ICP-MS (Perkin-Elmer NexION 350D, Waltham, MA, USA). The accuracy was checked using Standard Reference Materials for plant (NIST SRM 1572, tomato leaves, National Institute of Standards and Technology, Gaithersburg, MD, USA) and all recoveries of analyzed mineral elements were between 90% and 110%. All reagents used in the determination were of an ultrapure or analytical grade.

Microgreens canopy colors (L*, a*, b*, Chroma, and hue angle) were measured, in vivo, at three different points on each aluminum box, using a Spectrophotometer CM-2600 (Konica Minolta Sensing Inc., Osaka, Japan). L* corresponds to lightness or brightness, while redness, greenness, and yellowness are expressed as +a*, -a*, and b*, respectively. Hue angle ( h o ) and Chroma (C*) were calculated using the following equations, according to McGuire (1992):

Regarding chlorophylls and carotenoids determination, 50 mg of frozen microgreen dill tissue were extracted with 100% methanol (v/v) and then kept at 4°C for 24 hours in a dark room. Absorbance for chlorophyll a, b, and carotenoids was read at wavelengths of 665.2, 652.4, and 470 nm, respectively, using a UV–Vis spectrophotometer (Cary 60 UV-Vis, Agilent Technologies, Santa Clara, CA, USA). Based on the FW of the tissue, Lichtenthaler’s formula (Lichtenthaler, 1987) were used to calculate the pigments concentration.

Phenolic index was determined from 30 mg FW of dill microgreens. Samples for each treatment were transferred to a tube containing 3 mL of methanol acidified with hydrochloric acid (1% v/v) and were kept in a dark room for 24 h at 4°C. Absorbance readings were determined with a spectrophotometer at 320 nm (Ke and Saltveit, 1989). Phenolic index was expressed as ABS320 nm g⁻¹ FW. For sugars, an aliquot of 1 g of fresh tissue was homogenized in 3 mL of distilled water and centrifuged at 4000 x g for 15 min at RT (Benchtop centrifuge - Hettich, model ROTANTA 460 R, Tuttlingen, Germany). The total sugar levels were assayed according to the anthrone assay (Yemm and Willis, 1954). Absorbance was read at 620 nm and the concentration was calculated referring to a glucose calibration curve. The soluble solids content was determined using a portable digital refractometer (Hanna HI 96801, Hanna Instruments, Smithfield, England), using the same aqueous extract prepared for the total sugars assay and previously described. Nitrates content was determined with the salicylsulphuric acid method (Cataldo et al., 1975). The aqueous extract used for the analysis was the same as the preparation for the determination of sugars. Twenty µL of sample were added to 80 μL of 5% salicylic acid in sulphuric acid and to 3 mL of NaOH 1.5 N. Samples were cooled at RT for 15 min and the spectrophotometer readings were performed at 410 nm. Nitrate concentration was calculated referring to a KNO₃ standard calibration curve (0, 1, 2.5, 5, 7.5, 10 mM KNO₃).

Statistical analysis was performed using SAS software version 9.4 (SAS Institute, Cary, NC, USA). The two experiments were carried out independently each as a completely randomized design with 4 treatments and 3 replications. The comparison of means was evaluated with the least significant difference test (LSD, p=0.05). The normality of the data was checked by the Kolmogorov-Smirnov test. The t-test was used to determine the significant difference between the means of E1 and E2. Using GraphPad Prism 9.5.1 (GraphPad prism, Prism for Windows, version 9.5.1), a heatmap was created for yield, phenolic index, nitrate, and Brix°, as well as mineral elements. Principal Component Aanalysis (PCA) and correlations among the traits were performed using the R Studio 2022 software version 4.2.1 (R Core Team, Development Core Team, 2020).

Based on the results of the ANOVA analysis, it was found that the yield of dill was not significantly affected by seed priming with Se in greenhouse and indoor farm systems. It was observed that seed priming, especially hydro-priming, enhanced the mean value of yield of dill microgreens compared with seeds that had not been primed ( Table 1 ). Moreover, seed priming improved the biomass of dill microgreens in both experiments, although this difference was not statistically relevant ( Table 1 , Figure 1 ), with a more pronounced effect in indoor conditions, where yields were higher. The results revealed that Se seed nutri-priming positively affected the trend of the DM%, so its effect was greater than hydro-priming in terms of value. Under indoor farm conditions, nutri-priming with 3 mg Se L⁻¹ was more effective than 1.5 mg Se L⁻¹, increasing the DW and FW but, under greenhouse conditions, nutri-priming with 1.5 mg Se L⁻¹ was more effective. In both studied growing systems, seed priming with 3 mg Se L⁻¹ had a greater effect on DM% than 1.5 mg Se L⁻¹, although not significantly. There was also no significant difference in the height of dill microgreens in both growing systems in terms of its response to seed priming ( Table 1 ).

The t-test results showed a significant difference among these measured parameters under different growth systems ( Table 2 ). The plants grown in indoor conditions performed better than in greenhouse, concerning yield, FW, DW, and height. Based on the results of the present study, an increase of 91.66% in the yield of microgreens was found in indoor farm, with an average yield of 1255.6 g FW m⁻², compared to greenhouse (average yield of 655.1 g FW m⁻²) ( Table 2 ). Regarding FW and DW, the indoor farm produced microgreens with 91.61% and 72.72% increase in FW and DW, respectively, compared to those grown in the greenhouse, while their DM% decreased by 9.77% ( Table 2 ). The results of the t-test also showed that the height of the microgreens grown in indoor farm, with an average of 6.22 cm, was nearly twice the height of those grown in greenhouse, with an average of 3.48 cm ( Table 2 ).

As a result of the analysis of mineral concentration, to observe the effect of the treatments, there was no difference in the content of K and Mg in the dill microgreens after Se-biofortification through priming, but there was a significant increase in the amount of Se ( Table 3 and Figure 2 ). The Se added to the seeds was absorbed and accumulated in the aerial parts of the microgreens, even if the detected levels were very low. In greenhouse condition, nutri-priming with 3 mg Se L⁻¹ significantly increased the Se concentration of microgreens by 2.9 times compared to unprimed seeds. The difference is also significant compared to other treatments. The same conditions were also maintained in the indoor cultivation system, with the difference that the content of Se in microgreens treated with nutri-priming of 3 mg Se L⁻¹ was 3.6 times that of unprimed seeds. Also in this case, the treatment showed the highest average value. According to the findings of this study, Se can be enhanced in microgreens through priming to ensure that the level of Se in the plant tissue does not exceed the permissible limit. The t-test results showed that there was a significant difference between the greenhouse and indoor farm systems in terms of K and Mg contents, while the Se content did not show a significant difference between the two systems ( Table 2 and Figure 2 ). Interestingly, the K concentration of microgreens grown indoor was approximately 6 times that of those grown in greenhouse ( Table 2 ).

In the greenhouse system, dill leaves did not respond significantly to the color characteristics as a function of different Se-priming treatments ( Table 4 ). This trend was the same for microgreens grown in indoor conditions, although the a* value was significantly different among the treatments (p ≤ 0.05). The nutri-priming with 3 mg Se L⁻¹ in greenhouse conditions and the nutri-priming with 1.5 mg Se L⁻¹ in indoor farm conditions were most effective in increasing the value of greenness (−a*), and in the last one in a significantly way than unprimed seeds. Yellowness (b*), leaf saturation/vivid color (Chroma), and hue of dill leaves were not affected by Se-priming treatments. Based on t-test results, it was observed that L*, a*, b*, and Chroma of microgreens grown in the indoor farm system were significantly higher than those grown in the greenhouse system ( Table 2 ). The differences were visible by observing directly the plants: in indoor, they appeared to be of a more intense green color, a positive characteristic for the visual appearance of the product, which is more attractive to the consumer.

In both cultivation systems, the concentration of chlorophyll a, chlorophyll b, total chlorophylls, and carotenoids were not affected by the treatments. However, different Se-priming treatments significantly affected the chlorophyll a/b ratio and the chlorophylls/carotenoids ratio in greenhouse ( Table 5 ). More precisely, in greenhouse, nutri-priming with 3 mg Se L⁻¹ significantly reduced the chlorophyll a/b ratio and the chlorophylls/carotenoids ratio compared to unprimed seeds by 12.91% and 15.78%, respectively. According to the t-test results, there was no significant difference in the concentration of photosynthetic pigments of dill microgreens grown in either system, although the average chlorophyll a concentration of the microgreens grown indoor was 19.04% higher than those grown in greenhouse ( Table 2 ).

Under both the greenhouse and indoor farm system, the results showed that there was no significant effect of the Se-priming on the phenolic index, soluble sugars, and nitrate concentration, except for °Brix, which was affected by different levels of treatments under greenhouse condition, as reported in Table 6 and in Figure 3 . In addition, °Brix was not affected by Se-priming treatments under indoor farm system ( Table 6 ). In the greenhouse, 3 mg Se L⁻¹ caused a decrease in °Brix by 12.69%, 17.29%, and 12.69% compared to unprimed, hydro-priming, and 1.5 mg Se L⁻¹, respectively. According to the t-test ( Table 2 ), microgreens grown under greenhouse conditions had a higher value of phenolic index than those grown indoor farm. The trend of nitrate content and °Brix of microgreens grown in indoor farm system decreased compared to those grown in greenhouse system by 66.81 and 9.90%, respectively. The results of t-test also showed that there was no significant difference for soluble sugars in both cultivation systems ( Table 2 ).

The PCA was used for the analysis of the morphological and qualitative characteristics of dill microgreens in response to different levels of Se priming, in both greenhouse and indoor farm system, to provide a comprehensive overview and interpretation of the obtained results. Under greenhouse system (E1), the principal component (PC1) accounted for 57.9% of the cumulative variance, while PC2 explained 29.4% of the total variance ( Figure 4 ). The nutri-priming of 3 mg Se L⁻¹ was positioned on the positive side of PC1 in the upper right quadrant of PCA-Biplot ( Figure 4 ), as it produced plants with higher potassium (K), chlorophyll b (Chl b), phenolic index (PI), b*, chroma (c*), and L*. Moreover, the hydro-priming treatment was located on the negative side of PC2 in the lower left quadrant of PCA-Biplot ( Figure 4 ), as it produced plants with higher fresh weight (FW) and dry weight (DW).

Under indoor farm system (E2), the PC1 accounted for 40.7% of the cumulative variance, while PC2 explained 38.5% of the total variance ( Figure 5 ). Hydro-priming was placed in the positive side of PC1 in the upper right quadrant of PCA-Biplot ( Figure 5 ), as it produced plants with higher K and FW. The nutri-priming of 3 mg Se L⁻¹ was located on the negative side of PC2 in the lower left quadrant of PCA-Biplot ( Figure 5 ), as it produced plants with higher photosynthetic pigments and Se.

Under greenhouse system (E1), a* was negatively and significantly correlated with b*, L*, h°, and C*. There was a significant negative correlation between K and DM, as well as a significant negative correlation between Se and Mg, Car, and TSS (°Brix). In addition, it was found that NIT has a significant negative correlation with phenolic index (PI). It has been observed that Car had a significant positive correlation with Mg and TSS. Moreover, a similar relationship was recorded between SS and TSS ( Figure 6 ).

Under indoor farm system (E2), it was found that FW and DM, as well as K and Chl a, were significantly negatively correlated. The results showed that a* was negatively and significantly correlated with b*, L*, h°, and C*. Additionally, K and Mg, DW and Se, and SS and TSS demonstrated significant positive correlations in this study ( Figure 7 ).