pH Dynamics in Aquaponic Systems: Implications for Plant and Fish Crop Productivity and Yield

Abstract

The pH range of 7.0 to 7.2 is recommended to ensure reasonable nitrification rates in aquaponics; however, this range is conducive neither to nitrification, a critical process that occurs at pH 8.0, nor to plant growth. To determine the effects of pH in an aquaponic system, Swiss chard (Beta vulgaris L.), kale (Brassica oleracea L.), mustard green (Brassica juncea L.), cilantro (Coriandrum sativum L.), lettuce (Lactuca sativa L.), and arugula (Eruca vesicaria L.) were cultured with tilapia (Oreochromis niloticus) in 5-year-old coupled aquaponic systems at three pH levels, 6.0, 6.5, and 7.0, in comparison with hydroponics. Morphological and physiological growth parameters of vegetable and fish crops were measured regularly, and the ammonia-oxidizing bacteria (AOB) in the aquaponic system were analyzed by qPCR at the end of the production. This study found that feed conversion ratio, fish biomass, and copy number of AOB were not affected by different pH, but similar to hydroponic systems, lower pH in aquaponic systems increased fresh and dry mass and nutrient levels of all plant species tested. This study suggests that pH has a significant impact on plant performance and yield in both aquaponic and hydroponic systems and that, similar to hydroponics, a pH of 6 is desirable for aquaponic systems to improve plant crop yield without compromising nitrification activity and fish yield.

1. Introduction

Facing the challenge of food demand and sustainable agricultural practices, aquaponics has emerged as an important concept in global food production. Therefore, it is necessary to adjust a suitable environment for three organisms (aqua-organisms, microorganisms, and plants) to increase yield and quality in aquaponics. It has been reported that many aquaponics-grown fish release approximate 70–80% of nitrogen waste into the water [1,2]. This waste has the potential to serve as a valuable nutrient source for crop growth. Aquatic organism waste is filtered through microbial tanks in aquaponic systems, where ammonia (NH₄⁺) is converted to nitrate (NO₃⁻) for plant uptake [3,4]. While plants can assimilate nitrogen in either nitrate or ammonium form, excessive accumulation of NH₄⁺ has been proven to be toxic and can significantly reduce plant productivity [5]. Besides nitrogen, the presence of other essential macronutrients, such as phosphorus (P) and potassium (K), along with several micronutrients, significantly affects the yield of plants in aquaponic systems.

The major source of P in aquaponics is from fish feeds. Fish use only 15% of the P in fish feeds, and plants have different abilities to absorb P from recycling aquaculture wastewater based on different aquaponic designs [3,4]. Fish feeds also contain K and other micronutrients, but the amount of K and other micronutrients, such as iron (Fe), magnesium (Mg), manganese (Mn), and copper (Cu), is limited [5,6]. Therefore, some aquaponic systems apply synthetic salts into the aquaponic nutrient solution or apply a foliar spray to prevent the deficiency in K, Cu, Ca, Mg, Mn and Fe [5,6]. Moreover, it is common to observe some nutrient deficiency in plants when aquaponics relies on fish feed for plant nutrients [5,7,8]. Additionally, Quagrainie et al. [9] point out that it is hard for aquaponic growers to make a profit from vegetables compared to hydroponic growers due to lower yield. Therefore, it is necessary to find a balance among three organisms in aquaponics to have higher plant yields.

Aquaponic nutrient solutions are not as easy to maintain as hydroponic solutions. The aquaponic solution is affected by many factors, such as fish feeding rate, hydraulic loading rate, and pH [10,11]. Yang and Kim [2] analyzed three different feeding rates and found that a uniform feeding rate can enhance crop yield and/or quality in aquaponics by increasing initial nutrient availability and overall nitrogen-use efficiency of the system [2]. According to Yang and Kim [12], a flow rate of 3.3 m³/m²/day resulted in an increase of 50% and 80% in NO₃-N concentration, in comparison to flow rates of 2.2 m³/m²/day and 1.1 m³/m²/day, respectively. The pH is another important factor in regulating the nutrient level in aquaponics [10] and must be balanced for three different organisms in aquaponics at the same time. The critical step in aquaponics is nitrification, which converts toxic NH₃ to NO₃⁻ for plant uptake [10,13]. The optimal pH for ammonia-oxidizing bacteria (AOB) and nitrite-oxidizing bacteria (NOB) is 7.2 to 8.2, while their growth is restricted at pH 5.8 and 6.5, respectively [9,12,14]. The activity ceases typically below pH 5.5 in liquid pure culture [9,12,14]. As a result, aquaponic systems are set at pH 7.0 to get the best performance of nitrification. However, the recommended pH for hydroponics is 5.5–5.8 [15]. The available P for plants depends on the pH [16]. As pH increases above 7.0, most P converts to insoluble complexes and 30–65% of the P remains in solid fish sludge, which is unavailable to plants. Moreover, it is hard for plants to absorb Fe, Cu, zinc (Zn), boron (B), and Mn when pH is higher than 6.5 [16].

In this study, we established deep-water coupled aquaponic systems to investigate the interactive effects of pH and nutrient uptake on the growth of six plant species in comparison to hydroponic systems. The results led to a better understanding of how pH affects nitrification and plant yields and to a potential solution for limited nutrients in aquaponics.

2. Materials and Methods

2.1. System Design

Experiments were repeated three times between February and July 2019. Three identical aquaponic systems and three identical hydroponic systems were assembled in a greenhouse at Purdue University in West Lafayette, USA (40°25′26.4″ N, 86°55′44.4″ W). Each aquaponic system unit consisted of a fish tank or a nutrient reservoir (350 L), a clarifier (20 L), a two-stage biofilter (40 L) [2,10], and a deep-water hydroponic grow bed (1.0 m² and 350 L) (Figure 1). The biofilter was connected to peristaltic pumps (MasterflexLive™ Cole-Parmer L/S Digital Drive, Vernon Hills, IL, USA) to enable the recirculation of the nutrient solution within the unit. The determination of water flow rate and nutrient management in the aquaponic and hydroponic systems in this study was made based on prior investigations conducted by the research team [2,17]. The photoperiod for the aquaponic and hydroponic systems in the greenhouse was established according to the previous design of Teng and Kim [2,17,18].

2.2. Plant and Fish Materials

Arugula (Eruca vesicaria L.), cilantro (Coriandrum sativum L.), kale (Brassica oleracea L.), mustard green (Brassica juncea L.), lettuce (Lactuca sativa L.) and Swiss chard (Beta vulgaris L.) were used in this study. Seeds (Johnny’s Selected Seeds, Winslow, ME) were sown in commercial growing media (Jiffy Preforma Plugs, Lorain, OH, USA). Seeds were irrigated with tap water, and the electrical conductivity (EC) was gradually increased with a half-strength fertilizer solution after germination. Seedlings were irrigated with full fertilizer after true leaves developed [19]. (The nutrient composition of the fertilizer is described in

Table 1. Nutrient composition and some macro- and micro-nutrient contents in hydroponic nutrient solutions and aquaponic fish feed.

Parameter	Hydroponic Fertilizer a	Aquaponic Fish Feed b
Macronutrient (%)
Total nitrogen (N)	0.043	>6.88
Phosphorus pentoxide (P)	0.093	>1.10
Potassium oxide (K)	0.035	0.99
Sulfate(S)	–	0.43
Calcium (Ca)	0.075	2.25–2.75
Magnesium (Mg)	0.039	0.23
Micronutrient (ppm)
Boron (B)	2	–
Copper (Cu)	1.05	10
Iron (Fe)	21	40
Manganese (Mn)	1.9	80
Molybdenum (Mo)	0.42	–
Zinc (Zn)	2.1	153

Data obtained from product description. The symbol “–” is used to indicate that no related information was available or was not included. ^a Sample collected was a 1:100 diluted solution of commercial fertilizer. ^b Information was calculated based on daily fish feeding rate.

). Then, 21-day-old seedlings were transplanted to the aquaponic and hydroponic experiment units. There were 24 plants per m² for each experiment unit [19].

Nile tilapia (Oreochromis niloticus L.) were sourced from the Animal Sciences Research and Education Center (ASREC) at Purdue University. The fish weights were measured and subsequently equitably allocated to three distinct fish tanks with a density of 20 kg m⁻³. In this study, the fish were fed a complete diet (AquaMax Sport Fish 500, Purina Mills, St. Louis, MO, USA) consisting of 4.8-mm floating pellets, which contained 41% protein and 1.1% phosphorus. The determination of the fish feeding rate was carried out according to the methodology proposed by Teng and Kim [2]. Water temperature was maintained at optimized level of 26–28 °C for tilapia by aquarium thermostat heaters (Eheim Jager TruTemp, Deizisau, Germany).

2.3. Measurement of Water Quality Parameters

The electrical conductivity was 1.5 mS cm⁻¹ and was maintained by adjusting the feeding rate in the aquaponic and hydroponic systems. The pH of the aquaponic and hydroponic systems was maintained at pH 6.0, 6.5, and 7.0 by 10% H₂SO₄ or a combination of base solutions (0.02 mM Mg(OH)₂ and 0.02 mM Ca(OH)₂ with 1:1 v:v). Air stones were supplied for each aquaponic system to maintain optimized dissolved oxygen (DO) concentrations at full saturation as per the recommendation. Water quality parameters such as EC, pH, water temperature, and DO were measured daily before feeding using the HQ40d portable meters (HACH Corp., Loveland, CO, USA).

2.4. Measurement of Water Quality Parameters Quantitative PCR

The abundances of AOB were quantified via SYBR Green chemistry qPCR using specific primers targeting amoA. The primer sequences and qPCR program utilized in this study followed the methodology described by Cristina et al. [20]. Duplicate qPCR runs were performed on an iQ5 real-time PCR thermal cycler, and the resulting data was analyzed using iCycler iQTM software (BioRad Laboratories, Hercules, CA, USA). To generate standard curves for qPCR, plasmid DNA was serially diluted by a factor of ten, and primer specificity was ensured.

2.5. Measurement of Fish Growth Rate and Feed-Conversion Ratio

At the commencement and conclusion of the experiments, fish specimens were obtained from each aquaponic system and weighed to determine the total fish biomass. The fish stocking density was determined by dividing the total fish biomass in each aquaponic system by the volume of the fish tank. The specific growth rate (SGR) and feed-conversion ratio (FCR) were calculated by the following formula [21]:

SGR = (ln final weight of fish − ln initial weight of fish) × 100/days.

FCR = total weight of fish feed applied/total fish biomass increase (wet weight).

2.6. Measurement of Plant Biomass and Photosynthetic Properties

All plant tissues were harvested at 30 days after transplanting into the aquaponic and hydroponic systems. The plants were then carefully divided into roots, stems, and leaves for fresh weight measurements. Then, the plant samples were subjected to desiccation in an oven set to 70 °C for a period of 72 h to eliminate all moisture content. Subsequently, the dry weight of the plant samples was assessed by measurement.

The Fv/Fm (maximum photochemical efficiency of PSII) is widely used for early stress detection in plants [22]. The Fv/Fm value of healthy C3 plants is above 0.8 [23,24,25]. The values were recorded by a Handy chlorophyll fluorescence meter (Handy PEA+, Hansatech Instruments, Norfolk, UK). Five leaves per plant were taken and averaged, and four replicate plants were taken and averaged from each treatment every 3 days. A young fully expanded leaf was shielded by dark adaptation leaf clips (PEA, Hansatech Instruments, Norfolk, UK) for 20 min. The SPAD-502 Chlorophyll Meter (Minolta Camera Co., Ltd., Osaka, Japan) was used to measure the SPAD value, which indicates the chlorophyll content per unit leaf area. SPAD values were determined using the method outlined by Dong et al. [26]. Gas exchange measurements were conducted in this study using a portable gas exchange system (LI-6400XT; LI-COR Biosciences, Lincoln, NE, USA) according to the methodology described by Chang et al. [27].

2.7. Nutrient Analysis

The oven-dried plant samples as described in Section 2.6 were ground and filtered through a 40-mesh sieve with a Wiley mini-mill (Thomas Scientific, Swedesboro, NJ, USA). Subsequently, two milligrams of each sample were placed into an empty 2-microliter tube and meticulously wrapped to prevent any loss of the sample. The dry samples were sent to Midwest Laboratories (Midwest Laboratories, Omaha, NE, USA) for nutrient analysis. Total nitrogen was measured by thermal conductivity/IR detection (LECO TruMac and TruSpec CN Analyzers), and other mineral nutrients were analyzed by ICP-AES Analysis [28].

2.8. Mediation Analysis

Mediation analysis is a statistic model to illustrate the observed relationship between an independent variable and dependent variable with a third hypothetical variable [29,30,31]. Figure 2 illustrates the relationship between pH effects, dependent variable (such as Fv/Fm, and shoot and root ratio), and plant yield (harvest index). The harvest index is the ratio of harvested shoot fresh weight to total plant fresh weight and can be used to measure the reproductive efficiency [32]. The function is shoot fresh weight/total plant fresh weight × 100% [32]. Considering the nonlinear version of the mediation model, as depicted in Figure 2, the corresponding structural equations would have the form [33,34,35]:

X = F1(e1) z = F2(x, e2) y = F3(x, z, e3)

(1)

where X, Y, Z are discrete or continuous random variables, F1, F2, and F3 are arbitrary functions, and e1, e2, e3 represent omitted factors. With the above formula, we next determine total effect, T E_0,1 which measures the change in harvest index produced by a unit change in pH effects, by the formula [33,34,35]:

TE_0,1(e2, e3) = F3[1, F2(1, e2), e3] − F3[0, F2(0, e2), e3]

(2)

Then, we determine the natural effects by the following formula [33,34,35]:

$$ND{E}_{x,x^{\prime }}\left(Y\right)={\displaystyle \sum }_z[E(Y\left|X=x^{\prime }, Z=z)-E(Y\left|X=x^{\prime }, Z=z)]P(Z=z \right.\left|X=x)\right. \right.$$

(3)

$ND{E}_{x,x^{\prime }}\left(Y\right)$is defined as the expected change in Y induced by changing X from x to x₀ while keeping all mediating factors constant at whatever value they would have obtained under X = x, before the transition from x to x₀ [33,34,35].

Figure 2. A schematic outline of mediation analysis.

Figure 2. A schematic outline of mediation analysis.

2.9. Data Analysis

The experiment was conducted using three aquaponic systems and three hydroponic systems. Each system has ability to provide space for 24 plants to grow. The pH was set as 6, 6.5 and 7 for three aquaponic systems and three hydroponic systems accordingly. The sequence of steps employed to assess statistical models is demonstrated in Figure 2. This study focused on the pH effects in plant yields associated with Fv/Fm in soilless production systems. For the analysis of water conditions, water quality parameters, and fish performance in aquaponic and hydroponic systems, randomized complete block design and two-way ANOVA were used. The plant biomass and nutrient concentrations were tested by using Tukey’s Honestly Significant Difference. The statistical analysis used post-hoc pairwise comparisons (R 3.6.1, Comprehensive R Archive Network, USA) at a significance level of 0.05.

3. Results

3.1. Water Quality

Solution pH was maintained at an average of 6.0, 6.5, and 7.0 across all aquaponic and hydroponic systems in the different pH treatments, which led to a significant difference in the usage of pH correction solution (

Table 2. Average values of water quality in plant growth bed for 4 weeks.

System	pH	pH Correction Solution (mL day−1)	Electrical Conductivity (mS cm−1)	Dissolved Oxygen (mg L−1)	Temperature (°C)
Aquaponic	6.0	39.5	1.57	7.51	22.8
	6.5	80.6	1.61	7.32	23.3
	7.0	91.3	1.59	7.25	23.1
Hydroponic	6.0	2.0	1.54	7.75	23.2
	6.5	0.8	1.50	7.80	23.0
	7.0	0.0	1.51	7.83	23.6
ANOVA
System		***	ns	ns	ns
pH		***	ns	ns	ns
System × pH		***	ns	ns	ns

Each value in the table is the mean. ns, *** mean no significant or significant at p ≤ 0.001, respectively. Significantly different based on ANOVA (α = 0.05).

). Nutrient input was different in the aquaponic and hydroponic systems (Table 1), but the average EC was not significantly different between systems and pH levels (Table 2). DO and water temperature were maintained above 7 mg L⁻¹ and 22 °C, respectively, and there was no difference between the systems and pH levels (Table 2). Similarly, there was no significant difference in NH₄⁺-N, NO₃⁻-N, and NO₂–N in the aquaponic solution between different pH treatments. AOB was also not affected by pH treatment in the aquaponic system (

Table 3. Average concentrations of ammonia-oxidizing bacteria (AOB), ammonia, nitrite, nitrate in plant growth bed for 4 weeks.

System	pH	AOB (Copy Numbers g−1 Biomedia)	Ammonia (mg L−1)	Nitrite (mg L−1)	Nitrate (mg L−1)
Aquaponic	6.0	5.3 × 105	1.62	0.12	29
	6.5	1.1 × 106	1.45	0.09	30.4
	7.0	3.2 × 106	1.13	0.12	30.2
p		ns	ns	ns	ns

Abbreviations: AOB, ammonia-oxidizing bacteria. Each value in the table is the mean. ns means not significant. Significant difference is based on ANOVA (α = 0.05).

).

3.2. The pH Effect on Fish Production

Optimizing pH can prevent metabolic stress and avoid the mortality of stocked fish in ponds. The optimal pH range for Nile tilapia is between 5 to 8, and Nile tilapia can survive in acidic water as low as pH 4 [36,37,38]. In our study, we set up the aquaponic system at pH 6, 6.5, and 7, which was within the optimal pH range for Nile tilapia. Our results show that there were no significant differences in fish biomass gain, SGR, and FCR between different pH treatments (

Table 4. Fish production in aquaponics.

Treatment	Fish Feed Applied (g)	Initial Stocking Density (kg m−3)	Final Stocking Density (kg m−3)	Fish Biomass Gain (kg m−3)	SGR	FCR
pH 6	2100	19.9	27.0	2.5	8.3	0.84
pH 6.5	2100	20.3	27.5	2.5	8.4	0.83
pH 7	2100	20.4	27.9	2.6	8.7	0.80
p		ns	ns	ns	ns	ns

Abbreviations: SGR, Specific growth rate. FCR, Feed conversion ratio. ns means not significant. Significant difference is based on ANOVA (α = 0.05).

).

3.3. Plant Biomass

In this study, we repeated the experiments three times, and all the results showed a similar trend in plant biomass. The average biomass of most plant species was affected by the cropping system and pH treatments. Except for lettuce, plants had higher fresh weight and dry weight in the hydroponic system than in the aquaponic system regardless of pH treatment (

Table 5. Average biomass of arugula, cilantro, mustard green, kale, lettuce, and Swiss chard in aquaponic and hydroponic systems.

Treatments	Fresh Biomass (g plant−1)			Dry Biomass (g plant−1)
	Total	Shoots	Roots	Total	Shoots	Roots
	Arugula
System
Aquaponic	111.5 b	96.2 b	15.4 b	8.5 b	7.8 b	1.0 b
Hydroponic	270.4 a	270.8 a	22.8 a	19.5 a	20.2 a	1.9 a
pH
6	253.7 a	238.0 a	25.6 a	18.2 a	18.0 a	1.9 a
6.5	181.2 b	188.2 b	20.2 ab	14.1 b	14.0 b	1.7 a
7	137.9 c	124.4 c	11.5 b	9.6 c	9.9 c	0.8 b
Significance
System	***	***	*	***	***	**
pH	***	***	**	***	***	**
System × pH	ns	ns	ns	ns	ns	ns
	Cilantro
System
Aquaponic	72.5 b	55.0 b	23.0 b	6.6 b	4.8 b	2.0
Hydroponic	109.2 a	80.8 a	27.9 a	9.7 a	7.3 a	2.2
pH
6	124.7 a	87.7 a	37.0 a	11.3 a	8.1 a	3.2 a
6.5	87.6 b	65.0 b	22.6 b	7.7 b	5.8 b	1.9 b
7	60.3 c	51.0 b	16.9 c	5.4 c	4.4 b	1.1 b
Significance
System	***	***	ns	**	***	ns
pH	***	***	***	***	***	**
System × pH	*	ns	*	ns	ns	ns
	Mustard green
System
Aquaponic	218.0 b	195.6 b	17.5	17.9 b	16.4 b	1.1 b
Hydroponic	324.2 a	300.1 a	19.4	26.7 a	25.1 a	1.5 a
pH
6	323.9 a	296.2 a	23.2 a	27.8 a	25.9 a	2.0 a
6.5	275.6 b	249.3 ab	16.6 b	22.3 ab	20.9 b	1.4 b
7	213.7 c	198.1 b	15.5 b	16.9 b	15.6 c	0.7 c
Significance
System	***	***	ns	***	***	*
pH	***	***	*	***	***	***
System × pH	***	***	ns	***	**	*
	Kale
System
Aquaponic	177.3 b	160.3 b	19.1 b	16.0 b	14.5 b	1.6 b
Hydroponic	269.5 a	244.4 a	27.9 a	22.8 a	21.4 a	2.0 a
pH
6	247.1 a	223.4 a	22.6	22.4 a	20.3 a	1.9
6.5	246.9 a	228.2 a	25.6	20.9 b	19.7 a	1.9
7	176.1 b	155.4 b	22.3	15.0 c	14.0 b	1.6
Significance
System	***	***	***	***	***	*
pH	***	**	ns	***	***	ns
System × pH	**	**	ns	*	**	ns
	Lettuce
System
Aquaponic	303.9	286.5	17.2	10.4 a	8.9	1.3
Hydroponic	290.0	289.6	14.4	9.1 b	8.4	1.0
pH
6	334.8 a	315.4 a	20.4 a	11.6 a	9.7 a	1.5 a
6.5	270.4 b	277.9 b	13.6 b	8.8 b	8.3 b	1.1 ab
7	285.5 ab	270.9 b	13.5 b	8.8 b	7.9 c	0.9 b
Significance
System	ns	ns	ns	*	ns	ns
pH	*	*	**	**	**	*
System × pH	ns	ns	ns	ns	*	ns
	Swiss chard
System
Aquaponic	342.9 b	311.7 b	30.3 b	28.4	26.7	2.5
Hydroponic	521.9 a	419.2 a	50.0 a	33.5	30.4	3.0
pH
6	526.6 a	393.9 a	53.8 a	38.8 a	35.4 a	3.3
6.5	435.0 b	402.9 a	33.7 b	32.7 a	31.0 a	2.6
7	335.7 c	299.6 b	33.1 b	21.3 b	19.3 b	2.3
Significance
System	***	*	*	ns	ns	ns
pH	**	ns	ns	***	***	ns
System × pH	**	ns	ns	**	**	ns

Each column is a mean followed by a letter. The different letters show significant difference based on Tukey’s HSD test (α = 0.05). Each value in the table is the mean of 9 replicates of system and 6 replicates of pH treatment. “ns” means not significant, “*” means significant at a level of p ≤ 0.05, “**” means significant at a level of p ≤ 0.01, and “***” means significant at a level of p ≤ 0.001.

). In general, the total fresh weight of arugula, cilantro, kale, mustard green, and Swiss chard in the hydroponic system was significantly increased by 142.5%, 50.6%, 52%, 48.7%, and 52.2% compared to aquaponic plants (Table 5).

Furthermore, high pH levels had a negative impact on the fresh weight and dry weight of shoots across all plant species (Table 5). In both the aquaponic and hydroponic systems, plants grown in pH 7 conditions exhibited lower total fresh weight, shoot fresh weight, total dry weight, and shoot dry weight compared to those in other pH treatments (Table 5). Notably, the total fresh weights of arugula, cilantro, kale, lettuce, mustard green, and Swiss chard were substantially lower at pH 7 than at pH 6, with reductions of 45.6%, 51.6%, 28.7%, 14.7%, 34%, and 36.3%, respectively (Table 5). In addition, the total biomass of arugula, cilantro, mustard green, kale, lettuce, and Swiss Chard in a pH 6 environment were significantly increased by 184%, 207%, 152%, 140%, 117%, and 157% over those grown in pH 7 treatments (Table 5).

3.4. Leaf Chlorophyll and Photosynthetic Parameters

The effects of pH on SPAD and Fv/Fm values of plants grown in aquaponic and hydroponic systems were investigated in this study.

Table 6. SPAD value and Fv/Fm of crops grown in aquaponic and hydroponic systems with different pH. The values are photosynthetic parameters measured at day 7 after transplanting.

		Arugula		Cilantro		Mustard Green		Kale		Lettuce		Swiss Chard
Systems	pH	SPAD	Fv/Fm	SPAD	Fv/Fm	SPAD	Fv/Fm	SPAD	Fv/Fm	SPAD	Fv/Fm	SPAD	Fv/Fm
AQU	6.0	38.1	0.76	36.9	0.70	30.6	0.80	37.2	0.75	28.4	0.79	34.0	0.75
	6.5	31.8	0.71	29.5	0.62	31.7	0.74	32.7	0.70	26.5	0.75	33.2	0.73
	7.0	43.6	0.75	34.9	0.71	27.8	0.76	35.8	0.73	28.9	0.73	33.9	0.76
		ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns
HYD	6.0	45.9	0.80	35.4	0.76	32.7	0.80	37.9	0.75	27.2	0.77	31.4	0.76
	6.5	38.4	0.78	38.7	0.81	30.1	0.81	43.0	0.74	30.1	0.77	31.0	0.78
	7.0	39.7	0.76	34.3	0.69	30.4	0.75	32.4	0.69	26.2	0.71	32.8	0.77
		ns	ns	ns	ns	ns	*	ns	ns	ns	ns	ns	ns
Systems		ns	*	ns	*	ns	ns	ns	ns	ns	ns	ns	ns
pH		ns	ns	ns	ns	ns	*	ns	ns	ns	ns	ns	ns
Systems × pH		*	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns	ns

The value is the mean of 9 replicates of system and 6 replicates of pH treatment. “ns” means not significant, “*” means significant at a level of p ≤ 0.05, based on the ANOVA test.

shows that there were no significant differences in these values among the different pH treatments at day 7 after transplanting into the aquaponic or hydroponic systems. However, the pH affected the SPAD value of most plant species at day 21 and 28 after transplanting, except for lettuce and Swiss chard (Figure 3). Specifically, the SPAD value of arugula, cilantro, kale, and mustard green decreased in pH 7 aquaponics (Figure 3). In general, SPAD value was not affected by growing systems (Table 6). The Fv/Fm of six plant species decreased after transplanting into the aquaponic and hydroponic systems but most of them recovered by day 14 after transplanting (Figure 4). However, the Fv/Fm of arugula, cilantro, and mustard green in pH 7 aquaponic systems was significantly lower than other treatments at day 21 after transplanting (Figure 4). Nonetheless, photosynthetic parameters, such as photosynthetic rate (Pn), stomatal conductance (gs), transpiration rate (Tr), intercellular CO₂ concentration (Ci), were not affected by pH when measured at day 28 after transplanting (Figure 5).

3.5. The Nutrient Analysis

The study analyzed the impact of different production systems and pH treatments on the accumulation of macronutrients and micronutrients in arugula, cilantro, kale, mustard green, and Swiss chard. In general, both the pH levels and production systems had significant effects on the macronutrient concentrations in most of the crops tested (

Table 7. Average mineral nutrient concentrations of arugula, cilantro, mustard green, kale, lettuce, and Swiss chard in aquaponic and hydroponic systems.

Treatments	Macronutrient Concentration (%)						Micronutrient Concentration (ppm)
	N	P	K	Mg	Ca	S	Fe	Mn	B	Cu	Zn
	Arugula
System
Aquaponic	5.1	0.9 a	9.2 a	0.7 a	2.2 b	0.9 b	48.4 a	135.6 a	42.4 b	12.4 a	275.6 a
Hydroponic	5.7	0.8 b	8.0 b	0.6 b	2.5 a	1.1 a	42.1 b	62.1 b	44.5 a	7.8 b	164.8 b
pH
6	5.7 a	0.8 b	7.8 c	0.5 b	2.8 a	1.3 a	42.5 b	53.8 c	43.1	6.8	133.5 c
6.5	5.3 b	0.8 b	9.8 a	0.8 a	2.2 b	1.0 b	51.3 a	112.9 b	46.5	10.7	187.6 b
7	5.3 b	1.0 a	8.5 b	0.7 a	1.9 c	0.9 c	41.9 b	129.8 a	40.8	13	339.5 a
Significance
System	ns	**	***	***	**	***	***	***	*	*	***
pH	*	*	**	**	**	***	**	***	ns	ns	***
System × pH	*	ns	**	**	**	***	ns	***	*	**	**
	Cilantro
System
Aquaponic	4.0 b	0.6 b	10.1	0.4	2.1	0.4	43.7 b	751.1 a	78.5 a	15.4	284.8 a
Hydroponic	4.7 a	0.8 a	9.1	0.5	2.2	0.4	80.3 a	192.3 b	61.1 b	14.2	171.3 b
pH
6	4.9 a	0.8 a	8.7 b	0.4 b	1.8	0.4	95.3 a	533.0 b	70.4 b	14.5	214.7
6.5	4.2 b	0.7 b	10.1 a	0.5 a	2.3	0.4	56.5 b	619.8 a	58.5 c	15.3	263.7
7	3.9 b	0.6 c	10.0 a	0.5 a	2.3	0.4	34.2 c	262.3 c	80.4 a	14.7	205.8
Significance
System	**	**	ns	ns	ns	ns	***	***	**	ns	***
pH	**	**	**	*	ns	ns	***	***	**	ns	ns
System × pH	*	**	***	ns	ns	ns	**	***	ns	ns	ns
	Mustard green
System
Aquaponic	4.7	0.7	7.9	0.5	2.8	0.8 b	40.9	83.4 a	36.3	5.4	165.8
Hydroponic	4.8	0.7	8.3	0.5	3.1	1.2 a	35	62.0 b	40.3	6.3	156.8
pH
6	4.3	0.7	7.8	0.4	2.9	0.9 b	35.2	64.3	35.7	5	152.8
6.5	4.9	0.8	8.6	0.5	3.2	1.2 a	37.5	59.8	39	6.5	142.5
7	5.1	0.7	7.7	0.6	2.8	0.9 b	41.2	84.2	40.3	6.2	188.7
Significance
System	ns	ns	ns	ns	ns	*	ns	*	ns	ns	ns
pH	ns	ns	ns	ns	ns	***	ns	ns	ns	ns	ns
System × pH	*	ns	ns	*	ns	***	ns	**	*	ns	*
	Kale
System
Aquaponic	5.4	0.7	7.4 a	0.5	2.9	1.3 a	66.1 a	105.3 a	41.7 a	7.8	247.6 a
Hydroponic	5.5	0.7	7.0 b	0.6	3	1.1 b	51.0 b	71.1 b	38.3 b	6.1	116.4 b
pH
6	5.5 a	0.7 a	6.8 c	0.7 a	4.2 a	1.6 a	41.5 c	93.7 a	51.5 a	6.2 b	192.7 b
6.5	5.3 b	0.6 b	7.7 a	0.4 b	2.0 b	0.9 b	81.2 a	89.2 b	31.7 b	7.7 a	241.7 a
7	5.6 a	0.6 b	7.3 b	0.5 b	2.6 b	1.1 b	53.0 b	81.8 c	36.8 b	7.0 b	111.7 c
Significance
System	**	ns	***	ns	**	***	***	ns	*	ns	***
pH	**	***	***	***	***	***	***	***	***	*	***
System × pH	***	**	***	***	***	***	***	ns	***	*	**
	Lettuce
System
Aquaponic	4.5	1.2 a	9.9 a	0.5	2	0.4	253.1	386.9 a	33.8	13.6	156.9 a
Hydroponic	4.5	1.0 b	9.0 b	0.5	1.5	0.3	231.1	294.1 b	29.8	11.4	102.2 b
pH
6	4.4 b	1.2 a	8.8	0.5	1.7 b	0.4	265.2 b	392.5 a	29.5	12.5 b	143.8 a
6.5	4.2 c	1.1 b	8.5	0.5	1.5 b	0.4	334.2 a	362.3 b	30.7	14.0 a	106.2 b
7	4.9 a	1.0 c	11.1	0.5	2.0 a	0.4	127.0 c	266.7 c	35.2	11.0 b	138.7 a
Significance
System	**	***	***	ns	ns	***	***	***	ns	ns	***
pH	***	***	ns	*	***	***	***	***	ns	***	**
System × pH	*	***	***	***	***	***	***	***	***	***	***
	Swiss chard
System
Aquaponic	5.1 b	0.6 a	10.7 a	1.0 b	1.5 a	0.3	53.1 a	191.9 a	49.7 b	9.4	116.2 a
Hydroponic	5.4 a	0.5 b	9.5 b	1.1 a	1.1 b	0.4	49.8 b	34.9 b	51.1 a	6.1	49.9 b
pH
6	5.5	0.6 a	9.2 b	0.9 b	1.1	0.3	52.5	105.3 b	45.0 b	7.8 b	67.3 c
6.5	5.2	0.5 b	9.2 b	1.2 a	1.3	0.4	56	97.0 c	52.8 a	8.0 a	72.0 b
7	5.1	0.6 a	11.7 a	1.1 a	1.5	0.3	45.8	138.0 a	53.3 a	7.5 b	109.8 a
Significance
System	**	***	**	*	***	ns	**	***	**	ns	***
pH	ns	***	**	**	ns	ns	ns	***	**	**	***
System × pH	**	***	***	***	***	ns	ns	***	*	ns	***

Each column is a mean followed by a letter. The different letters show significant difference based on Tukey’s HSD test (α = 0.05). Each value in the table is the mean of 9 replicates of system and 6 replicates of pH treatment. “ns” means not significant, “*” means significant at a level of p ≤ 0.05, “**” means significant at a level of p ≤ 0.01, and “***” means significant at a level of p ≤ 0.001.

). In this study, significant interactions were observed between the production system and nutrient levels in pH in most of the crops. However, the nutrient accumulation in mustard greens was less affected by the production system and the pH, and little interaction was observed between these variables (Table 7).

Likewise, the accumulation of micronutrients was significantly affected by the production system and the pH (Table 7). Mn concentrations were significantly higher when the crops were grown in the aquaponic system, but the pH had inconsistent effects on Mn accumulation based on each crop. Similarly, Fe concentration was higher in the aquaponic system in most of the crops except for cilantro, in which Fe concentration was higher in the hydroponic system, and a lower pH increased the concentration (Table 7). Zn concentration was also higher in the aquaponic system, except for mustard greens, but the pH effect varied with the crop species.

3.6. The Mediation Analysis of pH Effect on Nutrient Level in Plant Tissues and Total Fresh Weight

Mediation analysis is a commonly employed statistical modeling approach that aims to investigate the relationship between an independent and dependent variable. It does so by introducing a hypothetical third variable that acts as a mediator or an explanatory mechanism between the independent and dependent variable [29,30,31]. The mediation model is shown in Figure 2. The data should fit three linear regulation models to fit the mediation analysis. We used casual mediation analysis to examine the mediation effect of nutrients.

Total fresh weight = b0 + b1* pH effect + e

(4)

Dependent variable = b0 + b2* pH effect + e

(5)

Total fresh weight = b0 + b4* pH effect + b3* dependent variable + e

(6)

Except for Mn, the p-value of the average casual mediation effect (ACME) of other plant nutrients was not significant, which indicated that those variables were not mediation effects between pH treatment and plant yields. The p-value of the average direct effect (ADE) of all nutrients was lower than 0.05, which showed that pH treatment had direct effects on all nutrients and plant yields. Meanwhile, the p-value of both ACME and ADE of Mn were lower than 0.05, which indicated the partial effect of pH treatment on Mn and plant yield in the aquaponic and hydroponic systems (

Table 8. Causal mediation modeling results.

Dependent Variable	ACME (p-Value)	ADE (p-Value)	Result
N	0.79	<0.05	Direct effect
P	0.16	<0.05	Direct effect
K	0.94	<0.05	Direct effect
Mg	0.28	<0.05	Direct effect
Ca	0.91	<0.05	Direct effect
S	0.48	<0.05	Direct effect
Na	0.83	<0.05	Direct effect
Fe	0.45	<0.05	Direct effect
Mn	<0.05	<0.05	Partial mediation effect
B	0.67	<0.05	Direct effect
Cu	0.14	<0.05	Direct effect
Zn	0.38	<0.05	Direct effect

Abbreviations: ACME, average causal mediation effect. ADE, average direct effect.

).

4. Discussion

4.1. Water Quality of Aquaponic System

Water physical parameters, such as EC, DO, and water temperature, have direct impacts on plant yield in soilless culture systems. In this study, EC level was maintained above 1.5 mS cm⁻¹ for plant growth in the aquaponic and hydroponic systems [18,39,40]. That the EC level increased with time in the aquaponic system may be due to the accumulated mineral nutrient in the aquaponic solution [18]. The dissolved oxygen was maintained above 6 mg L⁻¹ to supported nitrification and fish growth in the aquaponic system [41,42]. The water temperature can also affect plant crop yield in soilless culture systems. Thompson et al. [43] indicated that root temperatures of around 24 °C can support better market quality and production. In this study, a water temperature of 22–25 °C was better for plant yield and fish feed conversion ratio [43,44,45].

4.2. The pH Effect on Nitrification Activity

In the process of nitrification, both AOB and NOB had great influences on the NH₄⁺ –N removal, NO₂^—N, and NO₃^—N production. The growth rate of nitrifiers was significantly affected by different temperatures and pH values, due to the energy requirement of cell maintenance. The most efficient nitrification activity in aquaculture biofilters was reported from pH 7.0 to 9.0 [12,14]. Both AOB and NOB are important in nitrogen fixation, but AOB are typically considered more important as they initiate the process of nitrification, which is a key step in the conversion of atmospheric nitrogen to a form that can be used by plants [46]. Without AOB, NOB would not have a source of nitrite to convert to nitrate, and nitrogen fixation would not occur efficiently [46].

Tyson et al. [39] examined the effect of pH on nitrogen transformations in the biofilters in the aquaponic system. They found that no nitrification occurred at pH 5.5, and it took 12, 20, and 20–24 days at pH 8.5, 7.5, and 6.5 to decrease total ammonia nitrogen from 5 to 0 mg L-1, respectively [42]. Additionally, Wongkiew et al. [4] found there was higher total ammonia nitrogen and lower nitrate at a lower pH (6.0 and 5.2) aquaponic solution than in a neutral aquaponic solution. Even though lower pH decreases nitrification activity in aquaponic biofilters, the nitrogen utilization efficiency can reach a maximum at lower pH due to the ability of nutrient uptake in plants. Zou et al. [46] found that the maximum nitrogen utilization efficiency reached 50.9% at pH 6.0, followed by 47.3% at pH 7.5 and 44.7% at pH 9.0 in media-based aquaponics. Moreover, the nitrogen level of plant tissues was 34.8%, 30.3%, and 28.5% at pH 6.0, 7.5, and 9.0, respectively, which led to significantly higher plant yields at pH 6.0 [47]. They also found that the nitrogen level in water was not significantly different between pH 6.0, 7.5, and 9.0 [47]. The recommended pH for aquaponics is 7.0–7.5 for better nitrification activity [13,48]. However, there was no significant difference of AOB between aquaponic systems at different pH, which was similar to a previous study [49]. Day et al. [48] compared the effect of two different nitrifying bacteria sources, commercial nitrifying bacteria and their own biofilter medium, with nitrogen-limited aquaponics. They did not find a difference in nitrifying bacteria populations in aquaponics in both treatments, even though treating commercial nitrifying bacteria led to a higher plant yield [49].

Our results were similar to their finding. There was no significant difference of nitrogen concentration between different pH treatment aquaponics, but plant growth was significantly better at pH 6.0. Additionally, there was no significant difference in AOB populations between different pH treatments in our study.

4.3. The pH Effect on Plant Yield

Numerous studies have compared plant yields in aquaponic and hydroponic systems [50]. Due to different aquaponic designs, some studies showed aquaponics can have higher or similar plant yields compared to hydroponics, while some studies showed lower plant yields in aquaponics [18,51,52,53]. Roosta and Afsharipoor found that growing strawberries with 100% of perlite in an aquaponic system had higher fruit numbers and yields than in a hydroponic system, although strawberry accumulated less N, P, K in the aquaponic system [53]. Meanwhile, Yang and Kim compared nitrogen and phosphorous use efficiency for lettuce-, basil-, and tomato-based aquaponic and hydroponic systems, and found that the nitrogen and phosphorous use efficiency were significantly higher in the aquaponic than the hydroponic system even though plant yields were low in the aquaponic system [18]. Various environmental factors can affect plant yields in aquaponics, such as fish feeding rate, water flow rate and pH, since those factors can affect nitrogen transformation in aquaponics [10]. In our lab’s previous study, Yang and Kim found that nitrogen-use efficiency and crop production were increased in aquaponic systems with a uniform fish feeding rate [2]. Moreover, Yang and Kim compared the effect of the hydraulic loading rate in aquaponic systems, and they found not only improved plant performances from a higher flow rate (3.3 m day⁻¹) but also improved fish growth rate and biomass [17]. In this study, we grew six different plant species in aquaponic and hydroponic systems with different pH treatments and found that the plants had the lowest total fresh weight at pH 7 regardless of production system. This result is similar to previous findings. Anderson et al. grew butterhead lettuce at different pH in hydroponics, and pH 7 hydroponic systems produced 26% less shoot fresh weight than pH 5.8 hydroponic systems [51].

4.4. The pH Effect on Nutrient Level in Plant Tissues and Total Fresh Weight

The higher pH levels in aquaponics may lead to lower plant yields with lower nutrient uptake, as demonstrated in previous studies [2,51,54]. Meanwhile, the pH level of a hydroponic solution can significantly influence the availability of minerals such as P, Ca, and Mg [16]. In this study, we used mediation analysis to check the relation between pH, nutrient level in plant tissues, and plant total fresh weight. We found that most nutrients, except Mn, are directly affected by system pH and correlate with plant yield in aquaponic and hydroponic systems. Our findings are consistent with prior studies on plant yield in hydroponic systems, wherein low levels of nitrogen, copper, and molybdenum in lettuce grown under hydroponic conditions with a pH of 7 were found to be associated with reduced plant growth [51]. Mn is considered a crucial micronutrient for plant growth, particularly in the photosynthesis process [55,56]. However, the impact of Mn on plant yield is not direct, as the photosynthetic rate is more significantly influenced by N and Mg [57]. Based on our study, a lower pH level creates an optimal growth environment with increased iron levels, leading to higher crop yields. Hence, it can be inferred that Mn’s contribution to crop yield is less pronounced than N, Mg, and pH levels.

5. Conclusions

This study revealed that reducing the pH level from 7 to 6 enhanced the fresh and dry mass of all plant species in aquaponic systems, without any negative impact on fish yield. In addition, lower pH increased tissue nutrients in arugula, cilantro, lettuce, and Swiss chard in both the aquaponic and hydroponic systems. Findings from this study provide guidelines for aquaponic growers to increase profit margins by optimizing plant crop yields and quality through pH manipulation.