Combining Diluted Seawater and Fertilizer in an Ion-Based Multivariate Approach as an Effective Assay of Salt Tolerance in Brassica juncea Seedlings

Abstract

Non-conventional water sources (saline and brackish water) are viable options for crop cultivation. Current salt-tolerance research largely focuses on Na⁺ and Cl⁻, while other ions in these waters remain ill-understood. Synthetic seawater was a representative of saline and brackish water in a Design of Experiments (DoE) treatment design used to evaluate the effects of factors [synthetic seawater (0, 15, 30, or 45%, v/v, Instant Ocean^®), total inorganic nitrogen (0, 14, or 28 mM; 1 NH₄⁺:8 NO₃⁻ ratio), potassium (0, 9, or 21 mM), calcium (0, 2, or 5 mM), silicon (0, 0.03, or 0.09 mM) and zinc (0, 0.05, or 2 mM)] on seedlings for two varieties of Brassica juncea [‘Carolina Broadleaf’ (CB) and ‘Florida Broadleaf’ (FB)] using a hydroponic assay. In 30–45% synthetic seawater, 0.09 mM of silicon or 2 mM of calcium alleviated salt stress. In FB, 0.04–0.06 mM of silicon was optimal for the production of new leaves. The CB variety showed greater production of new leaves with 0.09 mM of silicon and 28 mM of potassium. Potassium and calcium are components of seawater, and a sodium chloride assay would not account for their interactions without a multivariate approach to evaluate salt tolerance. The seedling assay identified factors and established criteria for larger-scale harvest experiments.

1. Introduction

The increasing scarcity of freshwater resources has driven global interest in non-conventional water sources, such as desalinated seawater and brackish water, for crop cultivation [1]. Seawater intrusion into groundwater exacerbates water quality degradation [2], necessitating innovative strategies for sustainable crop cultivation in saline environments [3,4]. While desalinated water offers potential alleviation for the freshwater required for irrigation, its high cost and the residual salinity pose significant challenges for large-scale agricultural use [5]. Understanding physiological saline stress and the nutritional impacts on crops is critical to addressing these challenges effectively.

Maintaining proper nutrition is a challenge for producing crops in saline and brackish water irrigation [6,7]. Hydroponic systems enable more highly controlled management of nutrient solutions and provide greater water availability than soil-based systems [8]. There are many interactions between toxic ions and nutrient ions in a saline milieu (e.g., Na⁺/K⁺) [9,10]. The colligative effect of ions are critical factors in crop cultivation [e.g., NPK (nitrogen, phosphorus, and potassium), pH, electrical conductivity (EC)], and these factors are in flux due to preferential nutrient uptake [11]. Crop salt tolerance research has historically focused on sodium chloride (NaCl) as a proxy to induce salinity stress. However, saline and brackish waters contain a variety of ions, such as potassium, calcium, magnesium, bicarbonate, sulfate, silicate, and zinc [3,12], which can interact with plant growth and salt tolerance in unpredictable ways. These ionic interactions can have synergistic, antagonistic, or neutral effects, significantly influencing nutrient uptake and plant health [13]. Consequently, traditional one-factor-at-a-time approaches are inadequate for capturing the interactive relationships between nutritive and non-nutritive ions. A multivariate approach to evaluating salt tolerance more comprehensively estimates the holistic effects of ionic constituents in hydroponic media on plant growth.

A multi-ionic approach to evaluating salt tolerance via Design of Experiments (DoE) with a fractional factorial treatment design integrates randomization, replication, and blocking in a single design where the critical response surfaces are mapped [14]. The data quality is maximized by varying more than one factor at a time, so interactions of important factors are estimated to capture some of the complexity of chemical solutions and biological responses [14]. Software is required to design and analyze DoE experiments, and many packages are readily available [14].

Experimental design is often constrained by materials, biological resources, facility space, and time. Smaller-scale experiments, referred to as “assay-scale” experiments, are characterized by a short-term treatment period, minimal resources (reagent salts, etc.), and minimal bench space. A hydroponic assay system utilizes seedlings and provides a high degree of control over the uniformity of vegetative growth conditions [15]. Using a chemically inert, soilless substrate, like phenolic foams submerged in treatment solutions, develops a hydroponic system, akin to a modified “Kratky method” passive deep water culture system [8,16,17].

There are three main mechanisms for salt tolerance in plants: ion exclusion, ion compartmentalization, and water maintenance [6,18]. However, many salt tolerance experiments are performed on seeds during germination [19,20], and the lack of vegetative anatomy disallows the expression of these mechanisms. Furthermore, depolymerization of the seed’s vast reserves of storage polymers generates osmolytes to counterbalance the osmotic effects of non-nutritive ions and generates metabolic energy to acquire nutrient ions in a way that is not indicative of vegetative plant growth [21]. Alternatively, salt stress studies in mature crops cannot be conducted as quickly as seed studies and require more water, nutrients, and space [17,22,23]. Seedlings are in the same vegetative stage of growth as harvest-stage leafy vegetables, making seedling assays of salt tolerance more informative than the germination stage. The seedling stage is defined by the transition from heterotrophic metabolism via expending stored polymers for the development of autotrophic shoot and root systems [21,24]. Seedling assay-scale experiments can provide more confidence in observed salt tolerance than germination studies [24] while taking less time and conserving more resources than mature crops.

Specific ions play pivotal roles in mitigating salinity stress and augmenting hydroponic fertilizer media to deter salt stress. Potassium, a component of both hydroponic fertilizers and synthetic seawater, regulates stomatal conductance, enzymatic activation, osmoregulation, and turgor maintenance [25]. Furthermore, there is well-known antagonism between the uptake of potassium and sodium [9]. In a study conducted on sorghum (Sorghum bicolor L.) to investigate the ameliorative effects of potassium on water stress, supplemental potassium decreased leaf sodium content [26].

Nitrogen is a large contributor to dry biomass and is available to plants primarily in the inorganic ionic forms of ammonium and nitrate. Ammonium [6 mM of (NH₄)₂SO₄] supplemented to sandy soil maintained leaf and root growth in up to 150 mM of NaCl during a 60-day salt stress study in maize (Zea mays) [27]. Nitrate was also supplemented using Ca(NO₃)₂ in 25:75, 50:50, and 100:0 ratios to (NH₄)₂SO₄ in a total concentration of 6 mM [27]. Ammonium was beneficial to maize growth under saline conditions, but the proportion of nitrate to ammonium altered the leaf osmotic potential in response to NaCl [27]. Nitrogen species and their relative ratios impact how maize responds to NaCl stress [27]. It is hypothesized that modulation of salt stress from nitrogen supplementation is achieved by reducing oxidative stress and upregulating antioxidant production and photosynthesis [28]. Calcium, a secondary macronutrient also present in synthetic seawater, has also been reported to increase salt tolerance in plants. Calcium maintains the integrity and structure of cellular membranes and, under saline conditions, aids in antioxidant metabolism and ionic homeostasis maintenance [29,30]. Moreover, calcium signaling is a stress response that stimulates the Na⁺/H⁺ antiporter for sodium efflux [31]. Calcium application (5 mM, 10 mM, and 15 mM of CaSO₄·5H₂O) increased the yield of tomato (Solanum lycopersicum) fruit in the presence of salinity (12 dS/m NaCl) and improved the relative water content of the whole plant [29]. The results from these studies [27,31] suggest that modeling specific ions in solutions is important, as opposed to investigating the effect of fertilizer salts. When supplementing saline media with Ca(NO₃)₂, for instance, erroneous conclusions about the impact of specific nutritional ions may be made, as both calcium and nitrate independently impact salt tolerance in crops [27].

Silicon is not typically regarded as an essential plant element, although its status is under debate in plant sciences [32,33,34,35]. The role of silicon in salt stress tolerance is a relatively emergent field of study for most major crops. Silicon has been cited to alleviate salt stress in many ways, including regulating plant hormones, increasing antioxidant activity, and preventing sodium uptake [34,35]. Foliar silicon application (0 kg of Si/ha, 1 kg of Si/ha, and 2 kg of Si/ha) alleviated reductions in maize (Zea mays) biomass and increased water use efficiency and grain yields under salinity stress [17 mM (1000 mg L⁻¹), 34 mM (2000 mg L⁻¹), and 51 mM (3000 mg L⁻¹) of NaCl] [33]. Lastly, zinc is a micronutrient with broad implications for human health as well as plant health [36]. Zinc has been shown to improve water uptake, maintain membrane stability, improve photosynthesis, and improve uptake of potassium and calcium when plants are facing salt stress [36,37]. Foliar zinc application (1 mM of ZnSO₄·7H₂O) increased the uptake of mineral elements and maintained water content in the presence of salt stress (0 mM, 100 mM, and 200 mM of NaCl) for two varieties of Brassica juncea [37].

To explore these ion-based interactions, we employed a Design of Experiments (DoE) treatment design to evaluate the effects of six factors—synthetic seawater, total inorganic nitrogen, potassium, calcium, silicon, and zinc—on two varieties of Brassica juncea seedlings (Carolina Broadleaf and Florida Broadleaf) in hydroponic media. DoE provides a robust framework for identifying critical factors and their interactions while maximizing data quality [14]. By leveraging assay-scale hydroponics, this study offers a resource-efficient, controlled environment for evaluating salt tolerance, considering combined nutritive and non-nutritive ion concentrations at the seedling stage.

The specific objectives of this study were to (1) validate the DoE treatment design as a method for identifying nutritive ionic relationships in saline hydroponic milieus, (2) demonstrate an assay-scale hydroponic screening system for salinity stress tolerance in seedlings, and (3) identify proper ionic concentration ranges of specific ions in hydroponic media to mitigate salt stress in Brassica juncea.

2. Materials and Methods

2.1. Seed Germination and Substrate Preparation

Seeds of two Brassica juncea varieties (Florida Broadleaf (FB) and Carolina Broadleaf (CB)), purchased from online retailers [Southern Exposure Seed Exchange (https://www.southernexposure.com/, accessed on 16 August 2023) and Eden Brothers (https://www.edenbrothers.com/, accessed on 16 August 2023)], were stored in dark conditions at 4 °C until use. Horticubes^® of the chemically inert polyphenolic foam hydroponic substrate “Aeromax™” (Smither Oasis Co., Kent, OH, USA; https://oasisgrowersolutions.com/collections/horticubes%C2%AE-aeromax%C2%AE, accessed on 3 July 2023) were prepared 24 h before seed sowing. The substrate was prepared by being rinsed with distilled UV-sterilized water three times, then placed into 1020 heavy-duty plastic trays with no holes (Greenhouse Megastore, Danville, IL, USA). Hoagland’s Modified Basal Salt Mixture H353 (PhytoTech Labs, Lenexa, KS, USA) was added to the flat of rinsed polyphenolic foam Horticubes^® in a concentrated solution. The fertilizer was applied so that the final rate of application to the prepared substrate was 0.815 g L⁻¹ (Hoagland’s full-strength H353 was applied at a rate of 1.63 g L⁻¹). The polyphenolic foam Horticubes^® were placed on a rocker table set to 5 oscillations per minute for 24 h to mix the liquid phase fertilizer solution within the substrate.

After the substrate was prepared, one seed per cell was sown, and the polyphenolic foam Horticubes^® were covered and placed under LED lighting (300 µmol/m⁻²s⁻¹). The supplemental lighting was set to power on from 4 pm until 8 am daily. After 48 h, the humidity dome was removed, and seeds were allowed to emerge for an additional three days in a climate-controlled positive-pressure clean laboratory. The shelving unit in the laboratory where the seeds germinated and emerged had temperatures of 23.4 ± 0.2 °C during periods without lighting and 30.4 ± 0.1 °C during lighting. The seedlings were then moved to a greenhouse space to acclimatize to greenhouse conditions for an additional two days before being placed into treatment solutions.

2.2. Design of Nutrient Treatments

The DoE matrix was computed using JMP (Version 17.1; SAS Corp, Cary, NC, USA) for response surface modeling (RSM) optimality. There were six factors: Instant Ocean^® synthetic seawater (SW) [0%, 15% (5.39 g L⁻¹), 30% (10.79 g L⁻¹), and 45% (16.18 g L⁻¹)], total inorganic nitrogen (NaNO₃ and NH₄NO₃; 0 mM, 14 mM, and 28 mM) (1 NH₄⁺:8 NO₃⁻ ratio), potassium (KCl; 0 mM, 9 mM, and 21 mM), calcium (CaCl₂; 0 mM, 2 mM, and 5 mM), silicon (CaSiO₃; 0 mM, 0.03 mM, and 0.09 mM), and zinc (ZnSO₄; 0 mM, 0.05 mM, and 0.2 mM). These factor-level concentrations were added as additions to the basal nutrient regime (Hoagland’s half-strength H353). Instant Ocean^® is a commercial product comprising salts made to substitute ocean water, and the concentration ranges are 0%, 15%, 30%, and 45% seawater, which will be referred to as no, low, moderate, and strong for the respective factor levels.

Table 1. Design of Experiments (DoE) treatment design optimized for response surface modeling. This table displays the combination of the six factors in the experiment (Instant Ocean^® synthetic seawater concentration (SW), total nitrogen (TN), potassium (K), calcium (Ca), silicon (Si), and zinc (Zn)). Rows with repeated information represent replications.

Treatment Combinations of Experimental Units Guided by DoE Matrix
		mM
Unit	SW (%)	TN	K	Ca	Si	Zn
1	45	14	0	5	0.09	0
2	45	14	0	5	0.09	0
3	30	0	9	0	0.09	0.2
4	30	0	9	0	0.09	0.2
5	0	28	21	2	0.09	0.2
6	45	14	21	0	0.09	0
7	45	14	0	0	0	0.2
8	45	0	21	5	0.03	0
9	45	28	0	5	0.03	0.2
10	0	28	21	0	0	0.05
11	0	14	0	5	0	0
12	45	28	0	0	0.09	0.05
13	0	0	21	2	0	0
14	45	0	0	0	0.03	0
15	0	28	21	5	0.03	0
16	45	0	21	0	0	0.05
17	30	28	21	0	0.03	0.2
18	45	0	9	5	0.09	0.2
19	0	28	9	5	0	0.2
20	0	14	0	0	0.09	0.2
21	0	28	0	2	0	0.05
22	0	0	21	0	0.09	0.05
23	0	14	0	5	0.09	0.2
24	0	0	9	0	0	0.2
25	30	14	21	5	0	0.2
26	45	14	21	2	0.03	0.2
27	0	28	9	0	0.09	0
28	15	14	0	0	0.03	0
29	15	28	0	5	0.09	0
30	0	0	9	5	0.09	0
31	0	0	21	5	0.03	0.2
32	45	0	0	5	0	0.05
33	45	28	21	2	0	0
34	45	28	0	2	0	0
35	15	0	0	2	0	0.2
36	45	28	21	5	0.09	0.05
37	15	14	9	2	0.03	0.05
38	0	0	0	2	0.09	0
39	30	14	9	2	0.03	0.05
40	30	14	9	2	0.03	0.05
41	15	14	9	2	0.03	0.05
42	15	14	9	2	0.03	0.05

displays the software-generated DoE experimental matrix generated, listing the concentrations of each factor. The data were modeled to the total ionic concentration of each factor rather than the factor-level concentration [38]. The concentrations of the ions in each treatment accounted for the basal fertilizer (0.815 g L⁻¹ of Hoagland’s Modified Basal Salt Mixture H353), Instant Ocean^®, and single salt additions for the remaining ionic factors (

Table A1. Concentrations (mM) of major ionic constituents in Hoagland’s H353 at half strength (0.815 g L⁻¹) and Instant Ocean^® synthetic seawater at 45% strength (16.18 g L⁻¹).

Concentration (mM) of Ions in Fertilizer and Simulated Seawater (Instant Ocean®)
	Ca2+	Mg2+	K+	Na+	Si4+	Fe2+	Zn2+	NH4+	NO3−	SO42−	H2PO4−	Cl−	HCO3−
-----------------------------------------------mM-----------------------------------------------------------
Hoagland’s Solution	2.0	1.0	3.0	-	-	2.2 × 10−3	4.0 × 10−4	0.5	4.0	1.0	0.5	2.3 × 10−3	-
Instant Ocean®	4.5	24.	4.8	211.0	-	-	-	-	-	12.5	-	244.0	1.50

).

Seven additional experimental units [units 39–45 (units 43–45 are not represented in Table 1)] were added to the RSM optimal design, all of which contained the same basal nutrient solution. Experimental units 39 and 40 were replications of 30% Instant Ocean^® strength and midpoint values for the remaining five factors. Experimental units 41 and 42 were replications of 15% Instant Ocean^® strength and midpoint values for the remaining five factors. Finally, experimental units 43, 44, and 45 were replications of 0% Instant Ocean^® strength and low points of the remaining five factors and served as variety “controls”. Due to each variety being run consecutively (

Table 2. Microclimatic factors [temperature (°C), humidity (%), and light intensity [Photosynthetic Proton Flux Density (PPFD) (µmol m⁻²s⁻¹)]] for trials 1–4.

		Temperature (°C)		Humidity (%)		PPFD (µmol m−2s−1)
	Dates	Day	Night	Day	Night	Day	Night
Trial 1	1/14/24–1/24/24	20.8 ± 0.2	15.7 ± 0.2	31.3 ± 0.7	33.3 ± 0.5	83.0 ± 5.0	5.0 ± 0.1
Trial 2	1/28/24–2/7/24	24.7 ± 0.3	17.8 ± 0.1	38.60 ± 0.3	43.7 ± 0.3	136.4 ± 9.4	9.4 ± 0.2
Trial 3	2/14/24–2/24/24	26.1 ± 0.4	17.89 ± 0.1	37.8 ± 0.5	44.2 ± 0.5	179.8 ± 12.5	12.5 ± 0.1
Trial 4	3/29/24–4/8/24	27.2 ± 0.3	19.2 ± 0.2	46.6 ± 0.9	58.6 ± 0.8	199.6 ± 10.5	10.5 ± 0.1

), variety controls for each trial were simultaneously grown in treatments numbered 43–45. Media formulations of the treatment stock solutions of each of the 41 total unique experimental combinations guided by the DoE matrix were created once, as shown in Table 1. Note that the number of rows for each treatment combination indicates the number of replications for each of the treatment combinations.

2.3. Trasnfer to Microcosm

Each microcosm was constructed from a polyethylene terephthalate plastic container that was 12 cm × 12 cm × 15 cm in dimension (Figure 1A, WebstaurantStore, Lititz, PA, USA). Uniform seedlings were selected based on a similar hypocotyl length, cotyledon health, and color. Seedling preparation was determined by the number of days, not physiological age, so during acclimatization to greenhouse conditions, variations in the initial number of leaves between trials occurred. At the initiation of trials 1 and 2, the seedlings did not have expanded leaves (size > 2 mm). At initiation, trial 3 had 1.17 ± 0.02 leaves, and trial 4 seedlings had 2.04 ± 0.01 leaves.

Once seedlings were selected, 150 mL of each experimental solution was added to the microcosms, and the mass was recorded. Four uniform seedlings were then added to each microcosm which contained the treatment solution, and microcosm mass was recorded again. Experimental units were placed under a polyethylene tunnel on a bench in a greenhouse (34.6747° N, 82.8327° W) (Figure 1). Seedlings were in treatment conditions for 10 days, and data were recorded on days 1, 4, 7, and 10. To offset evapotranspirative loss and maintain initial treatment conditions (electrical conductivity and ion concentrations), deionized (DI) water was added back to the solution prior to data collection. The volume of water added to the microcosm was determined by the difference in microcosm mass between day 1 and the time of replenishment.

On days 1, 4, and 7, the recorded data were as follows: microcosm mass and number of leaves. At day 10 (final harvest), the previously mentioned data were collected as well as the length of the longest leaf, width of the longest leaf, hypocotyl length, surface area index (SAI) (Equation (1)), fresh mass, solution electrical conductivity, and pH. The tissue dry mass was recorded after 72 h in a biomass oven set to 70 °C.

$$SAI={\displaystyle \frac{WidthoftheLongestLeaf}{2}}\times LengthoftheLongestLeaf\times NumberofLeaves$$

(1)

2.4. Data Analysis

Data were analyzed using JMP (Version 17.1; SAS Corp, Cary, NC, USA). Higher-order interactions of factors were evaluated before lower orders to avoid violating the principle of marginality [39]. Model fitting began with response surface modeling (RSM) of the six factors (SW, TN, K, Ca, Si, and Zn), fit for both varieties separately. To evaluate any higher-order interactions within the experimental array, a forward stepwise elimination was conducted for the full factorial (7-degree factorial, including variety), including quadratic second orders for each variable in the factor design [i.e., Ca² × Si (which will be reported in the format Ca × Ca × Si henceforth)]. The stepwise threshold was informed by p-values of the model factors that were between 0.05 and 0.001. The parameters that passed through this initial elimination threshold were run as a model evaluating new leaves per day. The factors that were significant (p < 0.005) were added to the conventional RSM for each variety (fit separately) for the final elimination step and completion of the model.

3. Results

Growth of two Brassica sp. was reported (Figure 2) after being exposed to increasing strengths of synthetic seawater for 10 days. Higher concentrations of synthetic seawater (30% and 45%) were deleterious to fresh biomass accumulation in B. juncea seedlings. The large variation in fresh weight by the main effect, the concentration of synthetic seawater, is a consequence of the minerals in the multifactor design that are not displayed, conflating the experimental error (Figure 2). The figure below conveys a general trend of lower seedling fresh weights in higher concentrations of synthetic seawater, but the large variation in each synthetic seawater level supports the claim that salt tolerance was augmented by other ionic constituents in the hydroponic media. For example, one data point, highlighted in Figure 2 with a relatively high biomass (1.4269 g), was produced in 30% synthetic seawater in the presence of a combination of nutritive ions.

A Pearson correlation matrix (Figure 3) shows the relationship of new leaves per day to other important recorded responses (i.e., length of longest leaf, fresh weight, dry weight, and surface area index). Positive correlations (r > 0.4) between the five responses were statistically significant (p < 0.001; Figure 3). New leaves per day is a forward leading indicator, informing future growth, as the production of new leaves precedes other responses. Moreover, new leaves per day were calculated on each data collection date, whereas the surface area index, fresh weight, and dry weight were only measured once at termination. New leaves per day are discrete data, which contributed to the low R² values of the final modified response surface models (

Table 3. Summarized analysis of variance (ANOVA) of modified response surface model of new leaves per day for two varieties of Brassica juncea (Carolina Broadleaf (CB) and Florida Broadleaf (FB)). Only significant terms (p < 0.05) are reported in this table.

Abridged ANOVA Modified Response Surface Model of New Leaves per Day
Model Term	Carolina Broadleaf	Florida Broadleaf
Intercept	***	***
Trial	***	***
Day	***	***
Y coordinate	*	*
Instant Ocean® strength (%) (SW)	***	**
K × Si	*
Si × Si		*
Ca × TN		*
SW × Ca × Si	*
R2	0.45	0.51
R2a	0.44	0.48

***: p-value < 0.0001; **: p-value < 0.001; *: p-value < 0.05.

and

Table A2. Analysis of variance (ANOVA) of modified response surface model of new leaves per day for two varieties of Brassica juncea (Carolina Broadleaf (CB) and Florida Broadleaf (FB)).

ANOVA Modified Response Surface Model of New Leaves per Day
Model Term	Carolina Broadleaf	Florida Broadleaf
Intercept	<0.0001	<0.0001
Trial	<0.0001	<0.0001
Day	<0.0001	<0.0001
X coordinate	0.9178	0.1704
Y coordinate	0.0203	0.0359
Instant Ocean® strength (%) (SW)	<0.0001	0.0009
Calcium mM (Ca)	0.0590	0.4910
Potassium mM (K)	0.6615	0.1041
Silicon mM (Si)	0.2208	0.5547
Zinc mM (Zn)	0.4657	0.3796
Total nitrogen mM (TN)	0.1493	0.8121
SW × SW	0.7856	0.6019
SW × Ca	0.5668	0.3985
Ca × Ca	0.4933	0.4344
SW × K	0.8413	0.5176
Ca × K	0.8161	0.3094
K × K	0.8066	0.6258
SW × Si	0.6364	0.4494
Ca × Si	0.6957	0.9300
K × Si	0.0500	0.6401
Si × Si	0.3715	0.0371
SW × Zn	0.6575	0.4839
Ca × Zn	0.2623	0.2984
K × Zn	0.3904	0.0663
Si × Zn	0.9857	0.3909
Zn × Zn	0.2187	0.1392
SW × TN	0.4484	0.1636
Ca × TN	0.2368	0.0379
K × TN	0.6372	0.1902
Si × TN	0.7855	0.9724
Zn × TN	0.4667	0.800
TN × TN	0.5099	0.4629
TN × TN × Ca	0.0572	0.8356
SW × Ca × Si	0.0207	0.0816
SW × Ca × Zn	0.4869	0.7067
Ca × Si × Si	0.8627	0.2756
SW × Si × Zn	0.4772	0.6149
SW × Ca × Si × Zn	0.8954	0.1366
R2	0.45	0.51
R2a	0.44	0.48

).

A modified (see Section 2.4) RSM model ANOVA (Table 3 and Table A2) shows that for both varieties, the trial and day were significant in the number of new leaves per day (p < 0.0001). Each microcosm was given an X and Y coordinate on the bench to account for any variation in microclimate (Figure A1) that impacted leaf production, denoted by the term “Y” in the ANOVA table. Increasing concentrations of synthetic seawater was also significant in the rate of new leaf production for CB (p < 0.0001) and FB (p < 0.001). Notably, calcium and silicon influenced new leaf production for both varieties (Table 3), and this nutrient relationship for CB was impacted by the concentration of synthetic seawater (see Section 3.2). The impact of silicon on CB was significantly impacted by potassium concentration (p < 0.05), and the impact of calcium was significantly impacted by the total nitrogen concentration (p < 0.05) for FB (Table 3).

3.1. Calcium and Nitrogen Milieu Interaction and Impact on New Leaf Formation

There were optimal calcium and total nitrogen concentration ranges for the production of new leaves (p < 0.05) (Figure 4; Table A2). For variety FB, calcium in concentrations between 4 mM and 6 mM was beneficial in producing leaves for no and low salinities, irrespective of the total nitrogen concentration (1 new leaf every 1.3 days) (Figure 4). Florida Broadleaf was insensitive to the total nitrogen concentration when calcium concentrations were optimal at high salinity (Figure 4).

Calcium was present in both Hoagland’s H353 (2 mM in half-strength, basal treatment nutrition) media and Instant Ocean^® (2 mM, 3 mM, and 4.5 mM in 15%, 30%, and 45%, respectively). At low calcium and high seawater combinations, the calcium and synthetic seawater factors are confounded, but no optima were suggested to be in that range of treatment factor combinations.

3.2. Silicon Milieu Interaction and Impact on New Leaf Formation

Variety CB displayed a third-order interaction of the factors synthetic seawater, calcium, and silicon. The effects of silicon on seawater tolerance for variety CB were also impacted by calcium concentration (Figure 5). No and low salinities had the fastest rate of new leaf production (1 new leaf every 1.8 days) in the presence of 9 mM of calcium and 0.09 mM of silicon. This combination produced leaves ~45% slower at moderate and strong salinities (1 new leaf every 2.9 days) (Figure 5). The production of new leaves was relatively constant with increasing seawater concentrations when there was 2 mM of calcium and 0 mM of silicon present in the milieu (1 new leaf every 2.5 days) (Figure 5). The antagonistic relationship between silicon and calcium is affected as salinity increases, suggesting that the addition of silicon to seawater is carefully balanced with any additional calcium, along with what is present in the milieu from seawater and fertilizers (Figure 5).

Significant impacts of silicon were identified by an analysis of variance (Table 3 and Table A2) for both varieties. The analysis of variance identified that there was a silicon and potassium synergism for the production of new leaves (Figure 6) (p < 0.05). In the presence of 0.09 mM of silicon, a greater number of leaves were produced (1 new leaf every 2 days) when there was 28 mM of potassium in the solution (Figure 6). Conversely, when there was 3 mM of potassium (1 new leaf every 2.5 days), there was a 20% slower rate of new leaf production (Figure 6). Moreover, when there was 0 mM of silicon in the solution, 3 mM of potassium was comparatively beneficial for the production of new leaves (1 new leaf every 3.6 days). This was 28% faster than when 28 mM of potassium was in the solution with no added silicon (1 new leaf every 4.4 days).

The spline for silicon on the response surface depicts a range of concentrations, and when high (0.09 mM), silicon is most impactful for the rate of new leaf production for variety FB. At 0% seawater, the effect of silicon is generally inconsequential (1 new leaf every 1.4 days at 0 mM of silicon and 1 new leaf every 1.25 days at 0.09 mM of silicon) (Figure 7). There is an inversion on the surface, suggesting silicon can prevent salt stress (Figure 7). At 45% seawater, no additional silicon generates ~45% slower new leaf production (1 new leaf every 3 days) compared to 0.09 mM silicon amendments to the solution (1 new leaf every 1.8 days) (Figure 7).

4. Discussion

4.1. Multivariate Approaches to Evaluating Salt Tolerance

We evaluated salt tolerance using an ion-based multivariate approach combining seawater and hydroponic nutrient solutions, wherein NaCl is not the only salt considered. Multifactor DoE treatment designs utilize response surface methodology, so optimal physiologically relevant ionic concentration ranges may be estimated for their main and interactive treatment effects. Multiple unique ionic interactions were determined to be significant in B. juncea seedling growth, which was resolved by a multivariate approach. This experiment, executed at the assay scale, identified critical ionic factors. The final model did not account for all variations in the new leaves per day response (R²_a = 0.44 CB; R²_a = 0.48 FB). While this treatment design is not optimal for screening all high-order interactions present within a multivariate array, it allows for targeted higher orders present within the array to be evaluated. Stepwise screening for higher order interactions that are estimable from the fractional factorial yielded a significant influence of SW × Ca × Si (p < 0.0207) and K × Si (p < 0.0500) in Carolina Broadleaf new leaf formation. In Florida broadleaf, Ca × TN (p < 0.0379) and Si × Si (p < 0.0371) were significant in the production of new leaves. Ions that are important in mitigating salt stress (Ca²⁺ and K⁺) can be common to both fertilizers and saltwater, and the treatment design must properly account for this.

4.2. Impact of Individual Factors and Interactions with Salinity Tolerance

Calcium, a constituent of complete fertilizers, as well as saline and brackish water, has shown ameliorative effects in salt stress up to moderate salinity for both varieties examined. In our study, nitrogen concentrations (15–25 mM) in media containing high concentrations of calcium (8 mM) maintained the production of new leaves in FB. Amendments of 5.6 mM of calcium improved the biomass production of B. juncea grown in soilless sand and vermiculite culture in the presence of 150 mM of NaCl [40]. The authors of one study were able to ameliorate the reduction in biomass production in the presence of 150 mM of Na⁺ and 150 mM of Cl⁻ (for comparison, 45% Instant Ocean^® synthetic seawater contains 210 mM of Na⁺ and 246 mM of Cl⁻). Similarly, additions of potassium nitrate (5 mM, 10 mM, and 20 mM) to Hoagland’s solution decreased the Na⁺ and Cl⁻ content in B. juncea leaves and increased the proline content when 100 mM of NaCl was applied [41]. These findings, in addition to those reported in the present study, support the idea that the totality of the ionic composition of the nutrient solution, including the form and concentration of nitrogen, may strongly influence how effectively calcium and other amendments mitigate salt stress.

The effect of silicon mitigating salt stress in Brassica juncea is influenced by genotype-specific responses and ionic interactions with the growth environment. Variety CB showed that the most beneficial application of silicon is dependent on the strength of synthetic seawater as well as the concentration of calcium. At moderate and high concentrations of synthetic seawater, it is most beneficial for a <4 mM calcium concentration and a 0.09 mM silicon concentration. This suggests that higher concentrations of silicon than those tested may enhance salt stress mitigation at moderate and strong salinities. Notably, calcium impacts the efficacy of silicon amendments for preventing salt stress. Thus, pre-assessing calcium concentrations in hydroponic systems is critical before silicon amendments are implemented for salinity management. Silicon (0.8 mM) has been shown to increase shoot/root length, shoot/root biomass, and leaf area when B. juncea is exposed to 150 mM of NaCl [42]. Coupled with the impact of silicon and potassium concentrations’ mitigation of salt stress, these ions need to be considered within the context of each other to assuage salt stress. Our data show that there is a synergistic relationship between potassium and silicon for CB seedlings. Potassium is a constituent of common fertilizers and is present in seawater in concentrations near 10 mM [43]. Silicon is not typically a constituent of commercial fertilizers; however, this study suggests that silicon amendments to media may mitigate salt stress when coupled with potassium amendments. The interaction of these two ions in solution, coupled with potassium in salinized water, supports the claim that evaluating B. juncea for salt tolerance using exclusively NaCl is not sufficient for concluding “salt tolerance”.

Silicon and calcium independently regulate toxic ion uptake under salinity. Silicon limits Na⁺ and Cl⁻ accumulation by forming apoplastic complexes [35], while Ca²⁺ competes with Na⁺ at root non-selective cation channels, reducing sodium influx [44]. Calcium also stabilizes photosynthetic membranes and enhances antioxidant pathways, whereas reactive nitrogen species (RNS), such as nitric oxide, modulate stress signaling alongside nitrogen metabolites [45]. Moreover, silicon amendments can increase calcium, potassium, nitrogen, and other mineral element contents in plant dry mass under salinity conditions [46]. These overlapping roles highlight the interconnectedness of Ca²⁺, Si⁴⁺, K⁺, and nitrogen in salinity responses.

4.3. Implications for Assay-Scale Hydroponic Systems and Future Work

The assay-scale hydroponic system served as a cost-effective, rapid screening tool to identify critical ionic factors and their interactions under salinity stress. By prioritizing efficiency, this preliminary approach minimized resource requirements while informing the design of subsequent large-scale experiments. Data from this assay revealed that all six tested factors, including synthetic seawater components and fertilizer ions, influenced plant responses either directly or through interactions.

The use of new leaves per day as a response variable provided a dynamic forward-leading indicator of plant health. This metric is linked to SAI (Equation (1)), as each new leaf contributes to the total canopy surface area, which is a critical factor for water use and photosynthetic capacity in B. juncea. Moreover, it is required for evaluating subsequent growth traits. For example, the length of the longest leaf may only be measured after a new leaf forms. By capturing this early-stage response, the assay provided critical insights for prioritizing key factors in subsequent larger-scale experiments.

The method of assay-scale screening in this experiment was quick and required few resources. The microcosms limited evaporative water loss but were too small to collect repeated measurements of leaf size throughout the experiment. Utilization of the assay-scale system can be optimized with larger vessels to better quantitatively observe plant material and treatment solutions. The use of indoor, electrically lit controlled environments (CEAs) may minimize the trial variation in these assays.

Variety is an important consideration for salt tolerance, as varieties may employ different phenotypic traits for salt tolerance. Breeding and phenotyping of B. juncea for salt tolerance traits is a burgeoning focus of research [47], and varied genetics should be screened.

5. Conclusions

This study demonstrates that a fractional factorial design, optimized through DoE, provides a robust platform for evaluating complex ionic interactions in dynamic saline environments and their impacts on early growth responses in B. juncea. The assay-scale hydroponic system proved highly effective as a rapid, resource-efficient alternative to traditional germination or mature plant studies. The critical synergies and antagonisms between the nutritive ions (e.g., Ca²⁺, K⁺, and Si⁴⁺) and saline components identified underscore the potential of tailored nutrient management to mitigate salinity stress in hydroponic systems and provide criteria for larger-scale harvest experiments. We are currently developing this methodology to harvest-scale hydroponic production of leafy mustard greens.