*3.1. Yield*

Considering the average of the two stages of harvest, the most and the least productive species were *S. arvensis* (2.41 kg FW/m2) and *S. minor* (0.39 kg FW/m2), respectively. An intermediate yield was obtained in *T. o*ffi*cinale* (1.83 kg FW/m2). On average, the three species resulted in higher yield when they were harvested at the baby leaf stage (2.11 kg FW/m2) rather than as microgreens (0.99 kg FW/m2). A significant interaction species × stage of the harvest was observed (*F* = 24.66; *p* < 0.001), revealing that *S. minor* gave higher yield as microgreens than as baby greens, while the contrary occurred in *T. o*ffi*cinale* (Figure 1). In *S. arvensis*, harvesting at different stages resulted in comparable yields.

**Figure 1.** Yield of *S. minor*, *S. arvensis*, and *T. o*ffi*cinale* microgreens and baby greens grown in a hydroponic system. Data are means ± SD (*n* = 3).

#### *3.2. Chlorophylls, Carotenoids, Phenols, Anthocyanins and Nitrate Content*

Statistical analysis showed that chlorophylls concentration, considering both the total amount and the single chlorophyll types (Chl *a* and Chl *b*), as well as the phenol values (expressed as phenolic index), were not significantly different among the species and the stages of harvest (Table 2). For carotenoids, higher concentration was found in baby greens than in microgreens, while no differences were observed among the species (Table 2). On the contrary, the species, as well as the stages of the harvest, showed significant differences in anthocyanin concentration. Among the species, the highest anthocyanin amount was found in *S. minor* (0.19 mg/g FW); between microgreens and baby greens, the latter showed higher values (Table 2). However, the significant interaction species × stage of harvest highlighted that such difference did not occur in *T. o*ffi*cinale* (Figure 2A). The species did not differ in nitrate concentration, whose values, on average, ranged from 5205 mg/kg FW (*S. minor*) to 6833 mg/kg FW (*S. arvensis*) (Table 2). The comparison between the stages of harvest revealed a significantly higher nitrate concentration in baby greens than in microgreens. A significant interaction species × stage of harvest was found for nitrate content. Specifically, *T. o*ffi*cinale* microgreens showed much lower nitrate

values than baby greens, while for *S. minor* and *S. arvensis* nitrate concentration was similar in the two product types (Figure 2B).

**Table 2.** Chlorophylls (Chl *a*, Chl *b* and total), carotenoids, phenols, anthocyanins, and nitrate concentrations of *S. minor*, *S. arvensis*, and *T. o*ffi*cinale* grown in a hydroponic system and harvested at microgreen or baby green stage.


1 Cyanidin-3-glucoside equivalent. Means (± SD) in columns not sharing the same letters are significantly different according to Duncan's Test (*p* ≤ 0.05). ns = not significant; asterisk(s) = significant at 0.05 (\*), 0.005 (\*\*) or 0.001(\*\*\*) level of significance.

**Figure 2.** Interaction species × stage of harvest for anthocyanins (**A**) and nitrate concentration (**B**) of *S. minor*, *S. arvensis*, and *T. o*ffi*cinale* grown in a hydroponic system and harvested at microgreen or baby green stage. Data are means ± SD (*n* = 3).

#### *3.3. Mineral Content*

Significant differences in element concentration between the species were observed for Ca, Mg, P, Cu, Zn, Mn, Se, Mo, Cd, and Pb (Tables 3 and 4). *S. minor* was richer in Mg, P, Zn, Mn, Mo, and Pb than *S. arvensis* and *T. <sup>o</sup>*ffi*cinale*. The latter ones did not differ for these elements with the exception of Zn and Mn, which were higher in *S. arvensis* than in *T. <sup>o</sup>*ffi*cinale*. *S. arvensis* showed the highest concentration in Ca, but the lowest amount in Cu and Se, and *T. o*ffi*cinale* was richer in Cd. No significant differences between the species were noticed for the content in Fe, Cr, Co, Al, Ni, and As. As the average of the three species, baby greens were found to contain higher amounts of Ca, Mg, P, Mn, Mo, and Cd than microgreens, which, conversely, showed higher concentrations in Co, Al, and Pb (Tables 3 and 4). The interaction between species and stage of the harvest was significant for Ca, Mg, Fe, Cu, Zn, Mn, Mo, Co, Al, Cd, and Pb (Tables 3 and 4). *S. arvensis* was particularly reached in Ca and *S. minor* in Mg, Zn, Mn and Mo at the baby green stage (Figure 3A,B,E,F and Figure 4A) On the contrary, the high accumulation of Pb in *S. minor* occurred only in microgreens (Figure 4E). *S. minor* showed a higher concentration of Fe, Cu, Co and Al when harvested at the baby greens stage, while for *S. arvensis* and *T. o*ffi*cinale* microgreens were richer in these elements than baby greens (Figure 3C,D and Figure 4B,C). For Cd, the difference between microgreens and baby greens was observed in *S. minor* and *T. o*ffi*cinale* (Figure 4D).

**Figure 3.** Interaction species × stage of harvest for Ca (**A**), Mg (**B**), Fe (**C**), Cu (**D**), Zn (**E**), and Mn (**F**) concentration in *S. minor*, *S. arvensis*, and *T. o*ffi*cinale* grown in a hydroponic system and harvested at microgreen or baby green stage. Data are means ± SD (*n* = 3).


**Table 3.** Calcium, Mg, P, Fe, Cu, Zn, and Mn concentration in *S. minor*, *S. arvensis*, and *T. o*ffi*cinale* grown in a hydroponic system and harvested at microgreen or baby green stage. Means (± SD) in columns not sharing the same letters are significantly different according to Duncan's Test (*p* ≤ 0.05). ns = not significant; asterisk(s) = significant at 0.05 (\*). 0.005 (\*\*) or 0.001(\*\*\*) level of significance.

**Table 4.** Chromium, Se, Mo, Co, Al, Ni, As, Cd, and Pb concentration in *S. minor*, *S. arvensis*, and *T. o*ffi*cinale* grown in a hydroponic system and harvested at microgreen or baby green stage.


Means (± SD) in columns not sharing the same letters are significantly different according to Duncan's Test (*p* ≤ 0.05). ns = not significant; asterisk(s) = significant at 0.05 (\*). 0.005 (\*\*) or 0.001(\*\*\*) level of significance.

**Figure 4.** Interaction species × stage of harvest for Mo (**A**), Co (**B**), Al (**C**), Cd (**D**), and Pb (**E**) concentration in *S. minor*, *S. arvensis*, and *T. o*ffi*cinale* grown in a hydroponic system and harvested at microgreen or baby green stage. Data are means ± SD (*n* = 3).

#### *3.4. Principal Component Analysis (PCA)*

A PCA was carried out in order to investigate whether there were factors grouping correlated variables together and to identify clusters across species and stages of harvest. Two principal components (PCs) explaining a cumulative variance of 61.0% were identified based on a screen plot of eigenvalues (Figure 5). PC 1, which explained 35.1% of the total variance, was positively correlated with anthocyanins, Mg, Mn, Mo, and P, while PC 2 (25.9% of the total variance) was negatively correlated to carotenoids and Ca and positively to Fe, Cu, Co and Al. The loading plot reported in Figure 6A illustrates the relationships between the parameters considered in this study. Parameters located close to each other had a strong co-variance. Moreover, parameters far from the origin contributed more to the PCs than parameters close to it. In the rightmost part of Figure 6A, two clusters (the first with anthocyanins, Mo and Mg, and the second with P, Mn and Zn) suggested a strong co-variance between these variables, as well as a strong contribution to PC 1. The most important variables contributing to PC 2 were Ca and carotenoids and, on the opposite side, Al, Co, Fe and Cu. The relationship existing between the analyzed samples are shown in the score plot (Figure 6B). PC 1 and PC 2 discriminated species and stages of harvest in five groups. *S. minor* baby greens were positioned in the right half of the plot (the positive side of PC 1): they were characterized by the highest levels of anthocyanins, Mg, Mn, Mo, P, and Zn. *T. o*ffi*cinale* microgreens were included in the upper left quadrant (the positive side of PC 2): they were characterized by high Fe, Co and Al concentrations and low nitrate content. *S. arvensis* samples harvested at the baby leaf stage were included in the lower-left quadrant (the negative side of PC 2): they were characterized by high carotenoids and Ca content. Di fferently, *S. arvensis* microgreens were characterized by low anthocyanins and relatively high nitrate and Al contents. Finally, *S. minor* microgreens and *T. o*ffi*cinale* baby greens were closely clustered at the center of the scatterplot (Figure 6B).

**Figure 5.** Screen plot of eigenvalues in PCA analysis.

**Figure 6.** Loading plot (**A**) and scores (**B**) for each component (PC 1 and PC 2). Anthoc = anthocyanins; m and b correspond to microgreen and baby green stages, respectively.

#### *3.5. Contribution to Mineral Dietary Intake and Health Risk Assessment*

The potential contribution of the analyzed microgreens and baby greens to human mineral requirements was very different for the different elements (Table 5). With reference to a portion of 20 g of microgreens/baby greens, the EDI% ranged from very low values, even lower than 1% (Ca from *T. o*ffi*cinale* microgreens and Zn from *S. arvensis* and *T. o*ffi*cinale* regardless of the stage of harvest) to values higher than 100% in the case of Cr (*T. o*ffi*cinale* microgreens and *S. minor* baby greens), revealing a potential intake so far over the AI of this element. Values of EDI% over 10% were detected for Mg (*S. minor* baby greens), Fe (*S. arvensis* and *T. o*ffi*cinale* microgreens, *S. minor* baby greens), Mn (all the species, both the stages), Se (*S. minor* baby greens) and Mo (*S. minor* and *T. o*ffi*cinale* baby greens). Considering the average of the three species, the EDI% values from microgreens showed the following ascending order for the different elements: Zn (0.79%), Ca (1.14%), P (1.82%), Cu (3.48%), Mg (6.11%), Se (6.58%), Mo (8.56%), Fe (13.18%), Mn (15.16%) and Cr (109.90%). Similarly, the order of EDI% from baby greens was Zn (0.76%), P (2.45%), Ca (3.10%), Cu (3.46%), Se (7.40%), Fe (8.44%), Mg (9.52%), Mo (15.67%), Mn (48.95%) and Cr (112.21%).

Regarding the assessment of the health risk related to detrimental metals present in the micro/baby greens, all the EDIBW values (Fe, Cu, Zn, Mn, Cr, Se, Mo, Co, Ni, As, Cd), calculated with reference to a portion of 20 g, were smaller than the corresponding RFDs (US-EPA IRIS, 2013), and the HRIs were far below 1 (Table 6). For Al, for any species and stage of the harvest, weekly consumption of 20 g of product per day would bring to an element intake far below the TWI (1 mg/kg body weight/week) recommended by EFSA (2008) (data not shown). For Pb, the ML recommended by the FAO/WHO *Codex Alimentarius* Commission to be legally permitted in leafy vegetables (30 μg/100 g) was exceeded in *S. minor* microgreens (Figure 4E).


**Table 5.** Estimated dietary intake expressed as percentage (EDI%) of the recommended dietary intake (RDI) or adequate intake (AI) resulting from the consumption (20 g per day) of microgreens or baby greens of *S. minor*, *S. arvensis*, and *T. <sup>o</sup>*ffi*cinale*.

1RDI (bold) and AI (italic) according to SINU (2014).

**Table 6.** Estimated daily intake per kg of body weight (EDIBW, mg/kg body weight/day) and health risk index (HRI) resulting from the consumption (20 g per day) of microgreens or baby greens of *S. minor*, *S. arvensis*, and *T. <sup>o</sup>*ffi*cinale*.


R*f*D = oral reference dose (mg/kg/body weight/day) according to USEPA (2013).
