Optimizing the Electrical Conductivity of a Nutrient Solution for Plant Growth and Bioactive Compounds of Agastache rugosa in a Plant Factory

: The objective of this study was to determine the proper electrical conductivity (EC) of a nutrient solution (NS) for accumulating bioactive compounds of Agastache rugosa without decreasing plant growth. Six-week-old seedlings were transplanted in a deep flow technique system with Hoagland NS with a 2.0 dS·m − 1 EC for the initial week. From eight days after transplanting, the plants were treated with six EC treatments of 0.5, 1.0, 2.0, 4.0, 6.0, and 8.0 dS·m − 1 for three weeks. Plant growth parameters, leaf gas exchange parameters, the relative chlorophyll value, and the ratio of variable to maximum fluorescence (F v /F m ) were measured, and the rosmarinic acid (RA), tilianin, and acacetin concentrations were analyzed at 28 days after transplanting. The results showed that almost all plant growth parameters were maximized at 2.0 and 4.0 dS·m − 1 and minimized at 8.0 dS·m − 1 compared with the other EC treatments. The relative chlorophyll and F v /F m values were maximized at 2.0 and 4.0 dS·m − 1 . Similarly, leaf gas exchange parameters were increased at 2.0 and 4.0 dS·m − 1 . The RA content exhibited significantly higher values at 0.5, 1.0, 2.0, and 4.0 dS·m − 1 compared with other treatments. The tilianin and acacetin contents exhibited the significantly highest values at 4.0 and 0.5 dS·m − 1 , respectively. These results suggest optimal EC treatment at 4.0 dS·m − 1 for increasing bioactive compounds in A. rugosa plants without decreasing plant growth. Excessively high or low EC induced salinity stress or nutrient deficiency, respectively. Furthermore, among the plant organs, the roots of A. rugosa contained the highest RA concentration and the flowers contained the highest tilianin and acacetin concentrations, which revealed a higher utilization potential of the roots and flowers for bioactive compounds.


Introduction
Agastache rugosa (Fisch. & C.A.Mey.) Kuntze is a popular species of the Lamiaceae family, which includes both herbal medicines and spices [1]. Secondary metabolites in A. rugosa plants include essential oils; phenolic compounds, such as rosmarinic acid (RA), ferulic acid, chlorogenic acid, and caffeic acid; and flavone glycosides, such as acacetin, tilianin, sesquiterpenes, triterpene, and diterpenes, which have the capability to prevent disease due to their anti-allergic, antifungal function, anti-inflammatory, antioxidant function, and antidepressant effects [1,2]. Rosmarinic acid is the major active component of A. rugosa. The rosmarinic acid activities consist of antioxidant activity, astringent features, an antimutagenic ability, an anti-inflammatory capacity, an antiviral capacity, and antimicrobial properties [3][4][5]. Previous studies [5][6][7] have indicated that acacetin and tilianin are important active compounds in A. rugosa, and have antiatherogenic, anti-inflammatory, vasorelaxant, and antihypertensive effects. Therefore, the bioactive compounds in A. rugosa are very important for medical applications. A. rugosa plants are mostly harvested from outside cultivation or wild regions. Nevertheless, the secondary metabolites of A. rugosa plants, as well as other medicinal plants grown outdoors, are very hard to control, so the concentration of secondary metabolites may vary widely under different environments and cultivation conditions, such as location, season, light intensity, photoperiod, temperature, and electrical conductivity (EC) [8,9]. The capacity to manage environmental conditions by plant factory technology has permitted producers to control their production systems to maximize plant growth and development, as well as quality [10]. An optimal nutrient supply is an essential factor for improving plant growth and development, plant physiology, and quality under a hydroponic culture system in a plant factory or greenhouse [8,9,11].
The plant growth of pak choi was demonstrated to be restricted by an imbalanced ion composition, an excessively low or high concentration of a nutrient solution (NS) (0.3, 0.6, and 1.2 dS·m −1 or 9.6 dS·m −1 , respectively) because of nutritional inadequacy or ion toxicity, and salinity stress [8]. The salt concentration and electrolyte concentration index of the solution are determined by EC. Moreover, the EC of the NS indicates the ion availability for the plants in the root zone [9,12]. A high EC of the NS has been shown to reduce plant growth, osmotic potential, and absorbing NS because of salt stress. In contrast, a low EC of the NS has been demonstrated to degrade the yield and plant growth because of nutrient deficiency [8,[11][12][13]. The total phenol, chlorogenic acid, and kaempferol contents were maximized at an EC of 2.0 dS·m −1 , whereas 3.0 dS·m −1 significantly increased the fresh and dry weights of carrot [14]. An increase in EC from 0.5 to 2.5 dS·m −1 was accompanied by a significant increment of the antioxidant capacity and total phenolic content per shoot of Crepidiastrum denticulatum; however, the number of leaves, fresh weight, and leaf area were significantly higher at 2.0 and 2.5 dS·m −1 than those at 0.5, 1.0, and 1.5 dS·m −1 [12]. When the EC of the NS was either excessively high or low, the antioxidant enzyme activities were significantly increased, whereas the fresh weights were significantly decreased at low and high ECs because of nutrient inadequacy and salinity stress, respectively [8]. Therefore, an excessively low EC of the NS can lead to nutrient inadequacy, while an excessively high EC can lead to salinity stress and ion toxicity.
The objective of this study was to clarify the effects of different ECs of the NS on the growth; leaf gas exchange parameters; chlorophyll content; and RA, tilianin, and acacetin concentrations and contents in plant organs and the whole plant of A. rugosa. Furthermore, we tried to determine the optimal EC of the NS to achieve the highest accumulation of RA, tilianin, and acacetin concentrations and contents in hydroponically-grown A. rugosa without plant growth limitations.

Electrical Conductivity Experiment and Growth Conditions
Six-week-old seedlings with the same uniformity were selected and transplanted into six hydroponic culture systems (0.58 m (W) × 12.5 m (L) × 0.11 m (H)) in a plant factory which was maintained at the average temperature of 22 ± 3 °C and average relative humidity of 70 ± 10%. The plants were grown under white light-emitting diode (TL5 14W/865, Philips, Seoul, Korea) irradiation of 200 ± 10 μmol·m −2 ·s −1 PPFD with a 14/10 h (day/night) photoperiod for four weeks. Hoagland NS with 2.0 dS•m −1 EC was supplied for the initial week to avoid low or high EC transplant shock. From 8 days after transplanting (DAT) onwards, the plants were grown in six EC treatments of 0.5, 1.0, 2.0, 4.0, 6.0, and 8.0 dS•m −1 for three weeks. The EC was adjusted by the Hoagland NS. A portable conductivity meter and pH tester combo (HI98129, Hanna instrument, Woonsocket, RI, USA) was used to measure the EC and pH values of the NS. The tank had a volume of 50 liters. The six EC treatments were as follows:

Measurement of Growth Parameters
All plant growth characteristics of A. rugosa were determined at 28-DAT with six measurements per replication (n = 6). The number of leaves both longer than 1 cm and wider than 1 cm was counted, and the leaf area was measured by a leaf area meter (Li-3100, LiCor, Lincoln, NE, USA). The leaf width, leaf length, root length, and stem length were manually determined by a measuring tape. The root fresh weight and shoot fresh weight were measured by a digital scale (ARG224 OHAUS, Sigma-Aldrich. Co. LLC, Seoul, Korea). After that, samples were immediately placed in a drying oven (HB-502M, Hanback Sci, Suwon, Korea) at 70 °C for seven days. Afterward, a digital scale (ARG224 OHAUS, Sigma-Aldrich. Co. LLC, Seoul, Korea) was used to measure the root dry weight and shoot dry weight.

Relative Chlorophyll Value and Chlorophyll Fluorescence (Fv/Fm) Measurement
The relative chlorophyll value was measured at 28 DAT on the third intact leaves from the plant apex by a portable chlorophyll meter (502, Minolta Camera Co., Ltd., Tokyo, Japan). The ratio of variable to maximum fluorescence (Fv/Fm) was measured at 28 DAT on the third intact leaves from the plant apex by using a portable fluorometer (Fluorpen Pen FP 100, Photon System Instruments Ltd, Drasov, Czech Republic). Chlorophyll fluorescence and relative chlorophyll values were measured by three and six measurements (n = 3 and n = 6) in each replication.

Leaf Gas Exchange Measurement
The leaf gas exchange parameters, which consisted of the net photosynthetic rate (Pn), intercellular CO2 concentration (Ci), stomatal conductance (gs), and transpiration rate (Tr) of A. rugosa, were measured at 28 days after transplanting by using a portable photosynthesis measurement system (LICOR 6400, Licor. Inc. Nebraska, NE, USA) that had previously been warmed for 30 minutes and calibrated in the ZERO IRGA mode. The optimal conditions of the leaf chamber were established as follows: 25 °C leaf temperature, 1500 μmol . m −2 s −1 PPFD, 60% relative humidity, 400 μmol . mol −1 ambient CO2 concentration, and 500 cm 3. s −1 air flow rate. Leaf gas exchange parameters were measured on the third intact leaves from the plant apex. The measurement process was conducted automatically. Four measurements were used in each EC treatment per replication (n = 4).

Analysis of Acacetin, Tilianin, and Rosmarinic Acid Concentrations and Contents
The whole plant, as well as the roots, stems, leaves, and flowers, of A. rugosa samples from each replication were separated and immediately put in liquid nitrogen at 28 days after transplanting, placed in a deep freezer at −70 °C, and dried at −50 °C in the dry freezer (TFD5503, IL Shinbiobase Co. Ltd, Gyeonggi-do, Korea) for 4 days. Acacetin, tilianin, and RA concentrations were analyzed by three measurements (n = 3) in each replication per EC treatment. Plant organs were subsequently milled to a fine powder with a blender. Afterward, the powder was sieved through mesh sieves. The freeze-dried powder (0.1 g) was blended with 80% (v/v) MeOH (1.5 mL) by vortexing the sample and it was then sonicated for 60 min. The solution was centrifuged at 12,000 × g for 10 min at 4 °C using a microcentrifuge (R17 Plus, Hanil Scientific Co., Ltd., Gimpo, Korea). Before high-performance liquid chromatography (HPLC) analysis, the supernatant was filtered by being passed through a 0.45 μm filter for HPLC analysis. The HPLC analysis was conducted at 30 °C by using a C18 column (250 mm × 4.6 mm, 5 μm; RS tech, Daejeon, Korea). Additionally, 0.20% (v/v) acetic acid (solvent A) and 100% methanol (solvent B) were used as the solvent systems. The gradient profile of solvent B was as follows: 0-15 min: 45%; 15-20 min: 45%-55%; 20-45 min: 55%-80%; 45-50 min: 80%; 50-52 min: 80%-45%; 52-60 min: 45% (total of 60 min). The injection volume was 20 μL, the flow rate was upheld at one mL/min, and the detection wavelength was 275 nm. The acacetin, tilianin, and RA from Sigma-Aldrich Corporation (Sigma-Aldrich, Co., Ltd., Seoul, Korea) were used as standard compounds. The tilianin, acacetin, and RA concentrations were determined by reversed-phase HPLC-ultraviolet analysis (1260 Infinity, Agilent Technologies, Santa Clara, CA, USA) [3,5]. The average values were determined from three independent replicates of each plant organ (root, leaf, stem, and flower). The acacetin, tilianin, and RA concentrations (mg·g −1 plant organ and whole plant dry weight (DW)) were determined. The RA, tilianin, and acacetin concentrations in each plant organ (root, leaf, stem, and flower) (mg·g −1 DW) were the mean mg RA, tilianin, and acacetin concentrations per gram plant organ dry weight (DW). The RA, tilianin, and acacetin concentrations in the whole plant (mg·g −1 DW) were the mean mg RA, tilianin, and acacetin concentrations per gram plant DW. The RA, acacetin, and tilianin contents of the whole plant (mg/plant DW) were the mean RA, acacetin, and tilianin concentrations (mg·g −1 plant DW) multiplied by the whole plant DW (g).

Statistical Analysis
The EC experiment was performed with two replications using completely randomized designs. The SPSS 20.0 soft program (SPSS 20, SPSS Inc., Chicago, IL, USA) was used to determine differences among the means of all EC treatments. One-way ANOVA was conducted. Significant differences among the means of treatment groups at the 5% level were compared using Tukey's multiple range test.

Plant Growth Parameters
The results showed that the variation in the ECs of the NS significantly influenced the plant growth parameters of Agastache rugosa ( Table 1). The leaf length at 8.0 dS·m −1 was 7.73 cm, which was the lowest value among all EC treatments. There was a decreasing trend in the leaf width and number of leaves at 6.0 and 8.0 dS·m −1 compared with the other treatments and these results were not significantly different from 0.5 to 4.0 dS·m −1 treatments. The leaf area was not significantly different from 0.5 to 4.0 dS·m −1 treatments and no significant differences were found among 0.5, 6.0, and 8.0 dS·m −1 , but the 4.0 dS·m −1 treatment was the highest and 8 dS·m −1 treatment was the lowest. The stem length of 4.0 dS·m −1 treatment was the longest compared to the other treatments. Stem lengths at 1.0, 2.0, and 4.0 dS·m −1 were longer than the other treatments. The shoot fresh weight (SFW) at 0.5 and 8.0 dS·m −1 was lower than the other treatments. The SFW was the lowest for 8.0 dS·m −1 treatment among all the EC treatments. A similar pattern was observed for the root fresh weight (RFW). The RFW was minimized with the 8.0 dS·m −1 treatment. The RFW at 2.0, 4.0, and 6.0 dS·m −1 was higher than the other treatments. The root length value under 2.0 and 4.0 dS·m −1 was larger than other EC reatments. In particular, the root length at 2.0 dS·m −1 was 17.6 cm longer than that at 8.0 dS·m −1 . The shoot dry weight (SDW) and root dry weight (RDW) at 8 dS·m −1 were the lowest among all the EC treatments. The overall growth parameters were minimized at 8 dS·m −1 treatment (Table 1). *** *** * *** *** *** *** *** ** *** L y *** *** NS ** ** NS NS * NS NS Q x *** *** * *** *** *** *** *** *** ***

Relative Chlorophyll Value and Chlorophyll Fluorescence (Fv/Fm)
The variation in the ECs of the NS had a significant influence on the Fv/Fm and relative chlorophyll values. The relative chlorophyll value was significantly increased at 2.0 and 4.0 dS·m −1 treatments, while the Fv/Fm value was higher from 2.0 to 6.0 dS·m −1 compared with the other treatments. The Fv/Fm and relative chlorophyll values were significantly minimized at 0.5 dS·m −1 (Figure 1A,B). .

Leaf Gas Exchange Parameters
The EC of the NS significantly affected the net photosynthetic rate (Pn), intercellular CO2 concentration (Ci), stomatal conductance (gs), and transpiration rate (Tr) (Figure 2). When the EC of the NS was increased to 8.0 dS·m −1 , Pn, Ci, gs, and Tr were significantly decreased. A higher photosynthesis rate was observed at 2.0 and 4.0 dS·m −1 , and there was no significant difference between 1.0 and 6.0 dS·m −1 ; however, the photosynthesis rate was minimized at 8.0 dS·m −1 ( Figure  2A). Furthermore, the gs at 2.0 and 4.0 dS·m −1 was significantly higher than that of the other treatments and there were no significant differences among 0.5, 1.0, 6.0, and 8.0 dS·m −1 ( Figure 2B). However, the Ci value from 1.0 to 4.0 dS·m −1 was higher than that of the other treatments ( Figure 2C). A similar pattern was observed for Tr, showing that Tr from 1.0 to 4.0 dS·m −1 was significantly higher than the other treatments ( Figure 2D). The leaf gas exchange parameters exhibited the significantly lowest values of 4.41 μmol CO2 . m −2 s −1 for Pn, 0.026 mol H2O . m −2. s −1 for gs, 107.07 μmol CO2 . mol −1 for Ci, and 0.49 mmol H2O . m −2. s −1 for Tr at 8 dS·m −1 (Figure 2A,D).

Acacetin, Tilianin, and Rosmarinic Acid Concentrations and Contents
The RA, tilianin, and acacetin concentrations of A. rugosa were separately analyzed in the leaves, flowers, stems, and roots ( Table 2). In addition, the RA, tilianin, and acacetin concentrations of A. rugosa were analyzed in the whole plant ( Figure 3). The RA concentration in leaves was significantly higher at 0.5 and 1.0 dS·m −1 compared with the other treatments. The RA concentration in flowers showed no significant difference among all EC treatments; however, the RA concentration in flowers was higher at lower EC (0.5, 1.0, and 2.0 dS·m −1 ) compared with higher EC treatments. The RA concentration in stems was significantly highest at 1.0 dS·m -1 compared with the other treatments. The RA concentration in roots was significantly highest at 0.5 dS·m -1 compared with the other treatments. These results revealed an increasing trend in the RA concentration of plant organs at low EC treatments. The RA concentration was maximized in roots compared with other organs of A. rugosa ( Table 2).
The tilianin concentration in leaves was significantly higher at 4.0, 6.0, and 8.0 dS·m −1 compared with the other treatments. However, the tilianin concentration in flowers had an increasing tendency as EC increased. The tilianin concentration in stems was lower at 4.0 and 6.0 dS·m −1 than the other treatments. The tilianin concentration in roots was maximized at 1.0 dS·m −1 and minimized at 4.0 dS·m -1 compared with the other treatments. The tilianin concentration was maximized in flowers compared with other organs of A. rugosa ( Table 2). The significantly highest acacetin concentration was observed at 2.0 dS·m −1 in leaves, 6.0 dS·m −1 in flowers, and 8.0 dS·m −1 in stems. The acacetin concentration was undetectable in roots. The acacetin concentration was maximized in flowers compared with other organs of A. rugosa ( Table 2).

Plant Growth Parameters
Plant growth and quality have been shown to be affected by the NS composition ratio and nutrient concentration in a hydroponic culture system [8,9]. In the present study, we found that almost all plant growth parameters were significantly decreased at 8.0 dS·m -1 compared to the other treatments, possibly because of salinity stress at 8.0 dS·m −1 treatment. A similar result was reported by Pecanha et al., [15] who reported that the higher EC of 10.0 dS·m −1 significantly decreased the biomass of lettuce because the high salinity concentration reduced the osmotic potential in the NS.
Notably, a high EC reduced the osmotic potential of the NS, which may reduce the water uptake and turgor pressure in the plant. Furthermore, a high EC may limit cell expansion due to the diminished water uptake and turgor pressure. A high EC may lead to the retention of toxic ions in the root zone, so ion imbalance may affect plant growth. Therefore, plant growth was decreased by a high EC [15,16]. Moreover, the increase in Cland Na + may limit the Ca 2+ and K + uptake of roots because the hydrated radii of potassium and calcium are higher than that of sodium. A low K + /Na + selectivity ratio may result from a high-affinity potassium transport system. Therefore, ion imbalance occurs under salinity stress [17,18]. According to Degl'Innocenti et al. [18], plant growth was significantly decreased in barley species because the osmotic potential was reduced and ion imbalance arose around the roots. We found that the effect of high and low EC stress of the NS on A. rugosa in this study was negligible because of the lower EC (0.5 dS·m −1 ) and higher EC (8.0 dS·m −1 ) shock in the root zone of the plant treated from 8 days after transplanting. In our study, the highest EC treatment at 8.0 dS•m −1 had the lowest biomass production, leaf size, leaf area, and root and stem lengths due to salinity stress, ion imbalance, toxicity, lower water uptake, and turgor pressure in the plant, and the variation in osmotic potential around the roots. In addition, the ion ratio and concentration of the NS may affect the uptake and translocation between each part of the plant. Therefore, the growth of Agastache rugosa was improved under the optimal EC treatments (2.0 to 4.0 dS•m −1 ).

Relative Chlorophyll Value and Chlorophyll Fluorescence (Fv/Fm)
In the present study, the relative chlorophyll and Fv/Fm values of Agastache rugosa decreased in response to a high electrical conductivity (8.0 dS·m −1 ) due to the effects of salinity stress and toxicity under the very high concentration of nutrient solution and low electrical conductivity (0.5 and 1.0 dS·m −1 ) caused by the nutrient inadequacy of nitrogen, magnesium, and iron, which are important factors for the synthesis of chlorophyll and maintenance of the chloroplast structure and function [19][20][21]. In addition, the leaf chlorophyll concentration and dark color in tomato leaves were increased at high EC (4.5 dS·m −1 ) of the NS [13]. Chlorophyll a and b, and the chlorophyll content were significantly decreased in nitrogen deficiency treatment [21]. A similar result was reported by Shah et al. [22], who revealed that the total chlorophyll was decreased at a low nutrient supply. Furthermore, Garriga et al. [23] found that the chlorophyll content was gradually reduced with an increasing salinity concentration. The relative chlorophyll value of cucumber was increased at a higher EC of the NS [24]. The maximum chlorophyll fluorescence (Fv/Fm) was increased by adding more nitrogen in winter wheat [25]. This agreed with our results, for which the relative chlorophyll value in A. rugosa was higher at 2.0 and 4.0 dS·m −1 treatments and Fv/Fm ratio was higher at 2.0, 4.0, and 6.0 dS·m −1 treatments, compared with the other treatments, probably because A. rugosa avoided EC shock right after being transplanted by conducting EC stress from 8 days after transplanting. These study results suggest that an excessively low EC (0.5 dS·m −1 ) may lead to nutrient deficiency and an excessively high EC (8.0 dS·m −1 ) may lead to salinity stress in A. rugosa. Therefore, nutrient deficiency and salinity stress could reduce the relative chlorophyll and Fv/Fm values in A. rugosa.

Leaf Gas Exchange Parameters
A reduction of the photosynthesis rate often results in low assimilation production [26]. In our study, the net photosynthetic rate (Pn), intercellular CO2 concentration (Ci), stomatal conductance (gs), and transpiration rate (Tr) of A. rugosa were enhanced under medium EC treatments (2.0 and 4.0 dS•m −1 ) compared to the other EC treatments. In this respect, Ding et al. [8] indicated that an excessively high EC leads to salinity stress and an excessively low EC leads to nutrient deficiency. The salinity concentration was shown to decrease the water uptake and increase abscisic acid (ABA) in roots, which resulted in stomatal closure [27,28]. CO2 uptake is negatively affected by stomatal closure [29]. The photosynthesis rate of some plant species is decreased by stomatal closure due to the influence of salinity [8,28,30].
The leaf photosynthesis rate, stomatal conductance, leaf transpiration, and intercellular CO2 concentration of some plants have also been seen to be significantly decreased by a higher EC of the NS in a hydroponic culture system, such as pak choi [8], lettuce [30], and papaya [16]. Coinciding with these former studies, our results also showed that leaf gas exchange parameters (Pn, Ci, gs, and Tr) were significantly decreased at a high EC of 8.0 dS·m −1 because the salinity stress was increased with an increasing EC of the NS. The influence of salinity on A. rugosa resulted in a decreased water uptake and increased ABA, which reduced stomatal conductance and the photosynthesis rate. Moreover, the leaf photosynthesis response is mainly based on the leaf nitrogen content. Leaf nitrogen deficiency has been shown to reduce the chlorophyll and net photosynthetic rate [26,31]. In our study, leaf gas exchange parameters were reduced under an excessively low EC (0.5 dS•m −1 ) compared with the other treatments. These study results coincide with previous reports on pak choi [8], olive [31], and sorghum plants [26].

Acacetin, Tilianin, and Rosmarinic Acid Concentrations and Contents
The RA concentration was higher than the tilianin and acacetin concentrations in the whole plant of A. rugosa (Figure 3). A similar result was determined by Tuan et al. [5] who reported that the RA concentration was higher than the tilianin and acacetin concentrations of A. rugosa. Furthermore, RA is an important phenolic compound and was concentrated in the hairy roots of A. rugosa [3]. This also coincides with our results, for which the RA concentration was found to be the highest in the roots of A. rugosa compared with other plant organs (Table 2). Similarly, RA was maximized in the roots compared with other organs of A. rugosa [32]. In contrast, the RA concentration was maximized in flowers of A. rugosa [5], probably due to the different harvest times or types of varieties.
In our study, the RA concentration was significantly highest at 0.5 dS·m −1 in roots. Likewise, the RA concentration in the whole plant was the highest at 0.5 dS·m −1 among all treatments. Moreover, the results of our study indicated that A. rugosa under low EC treatments exhibited a trend of an increasing RA concentration because RA is a phenylpropanoid with a very strong antioxidant capacity [33]. A group of genes that are involved in the biosynthetic pathway of RA consisting of phenylalanine ammonia-lyase: PAL, 4-coumarate: CoA ligase: 4CL, and cinnamate 4-hydroxylase: C4H, was strongly produced under nitrogen deficiency [34]. These results are consistent with previous reports, such as those for the RA concentration in Prunella vulgaris, which was increased with a low nutrient supply [35]. The phenolic acid content in the plant was increased under nutrient deficiency (nitrogen, phosphorus, and potassium) [36]. The total polyphenols content in lettuce was maximized under nitrogen deficiency [37]. Furthermore, the RA of sweet basil was significantly highest at the low NO3 − concentration of 0.5 mol . m −3 compared with 5.0 and 10.0 mol . m -3 in the NS in a hydroponic system [38]. The RA concentration in green and red perilla was significantly highest at 1.0 dS·m −1 compared to at 2.0 and 3.0 dS·m −1 [9].
The highest tilianin and acacetin concentrations were detected in the flowers of A. rugosa compared with other plant organs (Table 2). Similarly, tilianin and acacetin concentrations were maximized in the flowers of A. rugosa compared with other plant organs [5]. Moreover, total flavonoids were maximized in the flowers compared to other organs of A. rugosa [32]. The results in our experiment indicated that the tilianin concentration in the whole plant increased when increasing the EC levels from 0.5 to 8.0 dS·m −1 ( Figure 3C). The tilianin concentration in the whole plant of A. rugosa was maximized at 8.0 dS·m −1 . The same result was reported by Ding et al. [8], who found that the treatment of 9.6 dS·m -1 increased the antioxidant enzyme activity and ascorbic acid content in pak choi compared to the other treatments because of toxicity and salt stress. The values of electrolyte leakage, lipid peroxidation, and some antioxidants (superoxide dismutase and peroxidase) were maximized at 8.0 dS·m −1 compared with the other treatments [39]. The antioxidant activity in artichoke was maximized at 6.9 and 6.45 dS·m −1 [40]. The phenol concentration, antioxidative capacity, total soluble solids, organic acids, and carotenoids in tomatoes were increased at a high EC (6.5 and 10 dS·m −1 ) [41] because the high EC of the NS induced salt stress [8]. In addition, the ascorbic acid, α-carotene, lutein, and polyphenol contents of pepper grown at high EC conditions were significantly increased in a hydroponic culture system [17]. Moreover, the lycopene concentration of red ripe tomato fruits was higher at a higher EC (4.5 dS·m −1 ) compared with a lower EC (2.3 dS·m −1 ) [13]. Secondary metabolites in Hypericum pruinatum were increased at a high nitrogen dose [42]. A high EC of the NS could lead to salt stress [8]. Salt stress enhanced the increase of hydrogen peroxide (H2O2) and superoxide radicals (O2) in cells, which may have induced the accumulation of some oxidative stress parameters, such as protein oxidation and lipid peroxidation [43], thereby increasing reactive oxygen species accumulation and the production of secondary metabolites in A. rugosa. This agreed with our results, for which tilianin was increased at high EC treatments.
The results in this study showed that 4.0 dS·m −1 increased both RA and tilianin contents, and 0.5 dS·m −1 maximized the acacetin content in A. rugosa ( Figure 3B,D,F). However, excessively low EC treatment (0.5 or 1.0 dS·m −1 ) restricted leaf gas exchange parameters, chlorophyll, and the tilianin content, probably due to nutrient inadequacy, while excessively high EC treatment (6.0 and 8.0 dS·m −1 ) limited chlorophyll, the photosynthesis rate, RA and acacetin contents, and plant growth, probably due to salinity stress and toxicity. The EC treatment at 2.0 dS·m -1 reduced tilianin and acacetin contents. Therefore, the optimal EC treatment was 4.0 dS·m −1 for plant growth and the accumulation of RA and tilianin contents in A. rugosa.

Conclusions
The optimal EC treatment is 4.0 dS·m −1 for growth and major bioactive compound accumulation because RA and tilianin contents were increased under 4.0 dS·m -1 and the amount of RA and tilianin contents in A. rugosa were higher than the acacetin content. A. rugosa grown under a hydroponic culture system in a plant factory could achieve high bioactive compounds by controlling the nutrient NS. Moreover, these results also provide quality information on each part of the A. rugosa plant (roots, flowers, leaves, and stems), according to the user's demands for RA, tilianin, and acacetin contents. Moreover, this study provides information about the effects of variation in the ECs of the NS on plant growth, photosynthesis, the chlorophyll content, and bioactive compounds in A. rugosa. Therefore, the study data are useful information for optimizing the NS to improve plant biomass and bioactive compounds.