Seedling Production of Cherry Tomato (Solanum lycopersicum var. cerasiforme) on Different Substrates Using Wastewater from a Recirculating Aquaculture System

Abstract

This study evaluated cherry tomato seedling production using pirapitinga RAS wastewater as the sole nutrient source in four substrate formulations: T1 (sand, gravel and coconut fiber), T2 (sand and gravel), T3 (gravel and coconut fiber), and T4 (sand and coconut fiber). No differences were observed for germination quality, germination percentage, seedling vigor index, germination vigor index, moisture content, total wet biomass, total dry biomass, or mortality. For small plants, leaf number (LN) was higher in T2 and lower in T4, while root length was greater in T3. The number of medium plants was higher in T3 and lower in T4; LN was higher in T1 and T2 and lowest in T3. For large plants, LN was higher in T1 and T2 and lower in T3; total length was higher in T1 and lower in T3 and T4. Visual differences in substrate water retention were observed: T4 exhibited rapid surface drying, T1 and T2 showed moderate moisture persistence, and T3 maintained surface water. Leaf yellowing was observed after 25 days, suggesting possible nutrient limitation or reduced nutrient availability at the measured pH. These findings indicate that substrate physical characteristics influence early seedling growth performance, whereas pirapitinga RAS wastewater can serve as a viable nutrient source.

1. Introduction

The cherry tomato (Solanum lycopersicum var. cerasiforme) is a species with small fruit that has attracted attention due to its higher market value compared to other tomato varieties [1]. In addition to its high value in national and international markets [1], it presents elevated contents of antioxidants, phytochemicals, nutrients, and vitamins [2], especially vitamins A and C [3]. Tomato cultivation in soilless systems and protected environments has demonstrated high adaptability, improving water and nutrient use efficiency and enabling more rational management of agricultural inputs [4]. Harvests obtained through this production method have shown superior results compared to conventional open-field systems [5], as they produce high-quality plants regardless of soil nutritional conditions [6].

Soilless cultivation systems, including substrate-based systems, have been increasingly used in vegetable production due to their potential to provide greater control over water and nutrient supply, resulting in improved plant uniformity and productivity [7,8]. However, despite their advantages, scientific information on these systems remains limited, particularly regarding their performance under different cultivation conditions and nutrient management strategies [9,10]. Among soilless cultivation methods, systems based on inert substrates require careful selection, since their physical properties determine water retention, aeration, and root support [11]. A wide range of natural and synthetic materials can be used as substrates, and their performance varies significantly among plant species, making comparative evaluations essential [12]. Previous studies have demonstrated that both substrate type and component combinations influence the early growth and quality of cherry tomato seedlings [13,14]. Because inert substrates do not supply nutrients, seedling development depends primarily on the composition and availability of the nutrient solution, similarly to hydroponic cultivation systems [15].

In soilless systems, all essential mineral nutrients are supplied via the nutrient solution, making its formulation a critical aspect of crop management [16,17]. Conventional hydroponic production relies heavily on chemically synthesized fertilizers to provide nitrogen, phosphorus, potassium, and other essential elements, representing a significant operational cost [18]. This dependence has encouraged the search for alternative nutrient sources capable of sustaining plant growth while reducing the use of synthetic fertilizers. In this context, the nutrient solution plays a central role in seedling development, particularly when inert substrates are used, as these materials do not contribute nutrients to plant growth [19,20]. Consequently, alternative nutrient sources capable of supplying essential macro- and micronutrients represent a promising option [21]. Among these, wastewater from recirculating aquaculture systems (RAS) presents a characteristic nutritional profile rich in nitrogenous compounds and other minerals derived from fish metabolism and feed residues [22], potentially favoring early plant growth stages such as tomato seedling production.

RAS are closed fish-rearing systems in which water is continuously treated and reused, allowing intensive production with minimal discharge [23]. In RAS, biofilters support nitrifying bacteria that convert ammonia from metabolic excretion and feed residues into nitrite and subsequently nitrate, which can be taken up by plants [24,25,26]. Additionally, RAS wastewater accumulates other nutrients originating from animals and feed residues, such as phosphates [23], which can be removed through the integration of aquaculture and agricultural systems [27]. Among fish species adapted to this production model, pirapitinga (Piaractus brachypomus) is a hardy native species well suited to RAS [28,29]. RAS water has already been used to supplement the water supplied to soilless hydroponic systems [30].

Despite these advances, information remains limited regarding the direct use of pirapitinga RAS wastewater as the sole nutrient and irrigation source for seedling production, particularly concerning how its nutrient composition interacts with different inert substrates. Therefore, the present study pioneers the evaluation of cherry tomato seedlings grown on different inert substrates under irrigation with residual water from a pirapitinga RAS. Under a common nutrient and irrigation condition provided by pirapitinga RAS wastewater, we hypothesized that differences in substrate physical properties, especially water retention capacity and aeration, would significantly influence seedling germination, vigor, and biomass accumulation. We further expected that substrates providing a balance between moisture retention and root aeration would promote superior seedling performance. This study generates novel information supporting the sustainable reuse of aquaculture effluents and the optimization of integrated production systems.

2. Materials and Methods

2.1. Experimental Site

The experiment was conducted in Belo Horizonte, Minas Gerais, Brazil, in the Laboratório de Aquacultura (LAQUA) of the Universidade Federal de Minas Gerais (UFMG), during the period from March to April, which is autumn in the Southern Hemisphere.

2.2. Water Used for Irrigation

The water used for irrigation was sourced from a RAS with pirapitinga cultivation, following a previously described methodology [31]. The RAS consisted of a polyethylene tank with a useful volume of 0.8 m³, equipped with an air-lift system containing a mechanical acrylic filter and a biological gravel filter. The system operated at an average flow of 0.89 m³ h⁻¹, with a filtration cycle time of 54 min, resulting in 27 daily passages of water through the treatment system. The RAS was stocked with 15 pirapitinga, corresponding to a fish biomass of 8 kg m⁻³. Fish were fed three times a day until apparent satiety with 6–8 mm extruded commercial feed containing 32% crude protein (Aquos Tropical^®, ADM Animal Nutrition, Chicago, IL, USA), characterizing a variable feeding rate dependent on fish demand rather than a fixed percentage of biomass per day.

After 90 days of fish cultivation, 700 L of RAS wastewater were collected for use in the seedling experiment. This single batch of wastewater, obtained from the same RAS unit, was used exclusively for irrigation and nutrient supply in all treatments throughout the experimental period. The water was stored in a closed polyethylene box (useful volume of 0.8 m³) located near the cultivation benches in the outdoor area of the laboratory throughout the experimental period (34 days). The tank remained closed to prevent the entry of insects and rainwater and was opened only during fertigation and water quality analyses. Prior to each irrigation and water quality assessment, the stored water was manually homogenized to ensure uniformity of its physicochemical characteristics. Throughout the experimental period, no changes were observed in the physicochemical parameters of the stored water, ensuring that the same water characteristics were maintained across all treatments and over time.

2.3. Seedling Production System

The plant cultivation system was set up using white gutters measuring 20 cm in depth, 1 m in length and 30 cm in width, with no opening at the bottom for water drainage, resulting in the formation of standing water (water film). The following four treatments were employed in a completely randomized design, with three replicates each, totaling 12 gutters, a replication level considered adequate for detecting treatment effects under controlled conditions with homogeneous water supply and substrate volume. Treatment 1 (T1) = 7.2 kg of sand + 8 kg of gravel + 50 g of coconut fiber per gutter; Treatment 2 (T2) = 7.2 kg of sand + 12 kg of gravel per gutter; Treatment 3 (T3) = 12 kg of gravel + 50 g of coconut fiber per gutter; Treatment 4 (T4) = 7.2 kg of sand + 50 g of coconut fiber per gutter. The quantities used were sufficient to provide a substrate layer of 3 cm in height in all treatments. These substrates were chosen because they are inert materials, thus allowing the evaluation of nutrient input exclusively via RAS water.

The gutters were placed on top of a 70 cm high metal support to facilitate treatment management. The gutters were protected with shade cloth (4 m × 5 m) that blocked 70% of the light and helped to block the entry of water and insects, in addition to providing protection for the seedlings during periods of intense heat and light. The gutters were protected with polyethylene covers when it rained.

Cherry tomato seeds were sown at a depth of 0.5–1 cm, with 3 cm spacing in width and 10 cm spacing along the gutter. A total of 50 seeds were sown per gutter. The seedlings were watered every other day at 5:00 p.m. with 1.5 L of RAS water per gutter during the first 15 days and 2 L per gutter thereafter. The initial irrigation volume was defined to meet the water demand of seedlings during early establishment while avoiding excessive water retention in the substrates. After 15 days, the irrigation volume was increased based on the observed increase in plant size, substrate drying rate, and water demand associated with seedling development. Substrate dynamics were monitored daily, considering water evaporation and the presence of apparent water layers.

Light intensity per unit area (lux) varied throughout the study, averaging 42,500.0 ± 39,867.9 lux, while room temperature averaged 28.4 ± 1.3 °C. Light intensity and ambient temperature were measured with a digital lux meter (0 to 200,000 lux INS-1381, Instrusul, Curitiba, Brazil) on alternate days.

2.4. Water Quality

Water quality parameters of the plant reservoir were measured on alternate days throughout the experimental period, with management carried out close to irrigation time (

Table 1. Physicochemical parameters of RAS wastewater (mean ± SD) monitored during the experimental period.

Parameters	Mean ± SD
pH	7.97 ± 0.72
Salinity	0.30 ± 0.01 g L−1
Electrical conductivity	0.59 ± 0.01 mS cm−1
Temperature	28.1 °C ± 1.62
Dissolved oxygen	5.5 ± 0.48 mg L−1
Alkalinity	120 ± 0.17 mg L−1 of CaCO3

). Data were collected on dissolved oxygen using a Hanna HI9146 portable DO meter (Hanna Instruments, Woonsocket, RI, USA), pH and water temperature with an AKSO V2 4234 pocket pH meter (AKSO, São Leopoldo, Brazil), and salinity and conductivity with a Hanna HI98130 portable meter (Hanna Instruments, Woonsocket, RI, USA). Alkalinity was measured according to standardized protocols [32].

Total ammonia [33], nitrite [34], and nitrate [35] were measured once a week. Mineral analyses of the water were performed every 10 days, for a total of three samples. Prior to sampling, the stored RAS water was manually homogenized to ensure uniform physicochemical conditions. The samples were then sent to a laboratory that analyzed potassium, calcium, magnesium, iron and manganese using the methods SMEWW 3120B and SMEWW 3125B (SMWW 23rd ed. 2017). The mean concentrations of these nutrients are presented in

Table 2. Nutrient composition of RAS wastewater used during the experimental period (mean ± SD).

Parameters	Mean ± SD
Ammonia	0.01 mg L−1
Nitrite	0.04 mg L−1
Nitrate	>55 mg L−1
Potassium	12.18 ± 0.23 mg L−1
Calcium	38.29 ± 0.59 mg L−1
Magnesium	8.83 ± 0.07 mg L−1
Iron	0.021 ± 0.02 mg L−1
Manganese	0.03 ± 0.02 mg L−1

.

2.5. Agronomic Parameters

Counts of germinated seeds per gutter were made on alternate days, as were observations of mortality. These data were used to calculate the following:

Germination percentage (GP):

GP= (Number of germinating seeds/Total number of seeds) × 100

Germination Velocity Index (GVI):

$$\mathsf{\Sigma }(\mathrm{N}\mathrm{i}/\mathrm{T}\mathrm{i})$$

$$\mathrm{N}\mathrm{i}=\mathrm{C}\mathrm{u}\mathrm{m}\mathrm{u}\mathrm{l}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{v}\mathrm{e} \mathrm{n}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r} \mathrm{o}\mathrm{f} \mathrm{s}\mathrm{e}\mathrm{e}\mathrm{d}\mathrm{s} \mathrm{t}\mathrm{h}\mathrm{a}\mathrm{t} \mathrm{g}\mathrm{e}\mathrm{r}\mathrm{m}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{t}\mathrm{e} \mathrm{i}\mathrm{n} \mathrm{t}\mathrm{i}\mathrm{m}\mathrm{e} \mathrm{i};$$

$$\mathrm{T}\mathrm{i}=\mathrm{T}\mathrm{i}\mathrm{m}\mathrm{e} \mathrm{a}\mathrm{f}\mathrm{t}\mathrm{e}\mathrm{r} \mathrm{t}\mathrm{e}\mathrm{s}\mathrm{t} \mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}.$$

At the end of the 34 days of the experiment, the seedlings were carefully removed from the substrates so as not to damage the roots and counted per gutter to obtain the germination quantity (GQ). The following biometric measurements were performed on all plants: total length (TL), shoot length (SL) and root length (RL), measured using a tape measure (cm); leaf number (LN), counted; and total wet biomass (TWB), measured using a digital scale (Marte 135.0001.00 digital precision scale, resolution 0.2 g).

The plants were subsequently placed in an oven (Nova Ética/Ethink) to dry at 60 °C for three days and each replicate was weighed immediately afterward to obtain total dry biomass (TDB). These data were used to calculate the following:

Moisture content % (MC):

MC = [(Wet Weight − Dry Weight)/Wet Weight] ∗ 100

Seedling Vigor Index (SVI):

SVI = TL ∗ GP

Seedlings were classified into small (S) (0–10 cm), medium (M) (11–15 cm), and large (L) (16–25 cm) categories based on the distribution of biometric parameters observed during the experiment, particularly total length. This classification was adopted as an analytical strategy to group plants according to their relative size and developmental status, allowing a more detailed evaluation of treatment effects on early seedling growth and vigor. The number of seedlings in each category (NS, NM, and NL) and their respective proportions relative to the total number of plants (S%, M%, and L%) were subsequently determined. This approach facilitates the interpretation of treatment responses by distinguishing general growth patterns within the experimental population rather than relying solely on mean values.

2.6. Statistical Analysis

Data were first tested for normality using the Shapiro–Wilk test. Variables meeting parametric assumptions were analyzed by analysis of variance (ANOVA), followed by Tukey’s post hoc test at 5% probability, while non-normal data were analyzed using the Kruskal–Wallis test (5%). For large seedlings, SL, RL, NL, and L% were analyzed by ANOVA, TL, LN, TWB were analyzed using the Kruskal–Wallis test. For medium seedlings, SL, NM, and M% were analyzed by ANOVA, while TL, RL, LN, and TWB were analyzed using the Kruskal–Wallis test. For small seedlings, TL, SL, NS, and S% were analyzed by ANOVA, whereas RL, LN, and TWB were analyzed using the Kruskal–Wallis test. Germination-related variables (GP, MC, SVI, GVI) were analyzed using ANOVA, while TDB and mortality were analyzed using the Kruskal–Wallis test. Statistical analyses were performed using Infostat (version 2020) and R software (version 3.5.2).

3. Results

3.1. Seedling Performance

No differences were observed for GQ, GP, SVI, GVI, MC, TWB, TDB or mortality among treatments (p > 0.05) (

Table 3. Agronomic performance of germination and early development of cherry tomato seedlings grown on different substrates.

Parameters	T1	T2	T3	T4
GQ (n)	26.3 ± 11.5	19.7 ± 7.6	29 ± 3.1	15 ± 2.9
GP (%)	52.7 ± 23	38 ± 15.3	58 ± 6.4	30 ± 5.8
SVI	520 ± 455.1	426 ± 152.1	661 ± 109.6	236 ± 37.5
GVI (index)	1 ± 0.4	0.7 ± 0.3	1.1 ± 0.1	0.6 ± 0.1
MC (%)	92.8 ± 6.5	91.8 ± 2.7	92.3 ± 2.6	92.1 ± 0.9
TWB (g)	10.6 ± 7.5	6.4 ± 0.4	5.2 ± 0.9	5 ± 1.1
TDB (g)	1.2 ± 1.6	0.6 ± 0.4	0.4 ± 0.2	0.3 ± 0.1
Mortality (n)	4 ± 1.5	3 ± 2.6	3 ± 0.6	5 ± 5.1

Legend: T1 = 7.2 kg of sand + 8 kg of gravel + 50 g of coconut fiber per gutter; T2 = 7.2 kg of sand + 12 kg of gravel per gutter; T3 = 12 kg of gravel + 50 g of coconut fiber per gutter; T4 = 7.2 kg of sand + 50 g of coconut fiber per gutter. GQ = number of germinated seeds; GP = germination percentage; SVI = Seedling Vigor Index; GVI = Germination Velocity Index; MC = moisture content; TWB = total wet biomass; TDB = total dry biomass. Values are expressed as mean ± standard deviation (p > 0.05).

). After 25 days of cultivation, visual symptoms of leaf yellowing were observed in some seedlings, characterized by a noticeable reduction in the green coloration of the leaves during routine monitoring.

Leaf number (LN) differed among treatments for small plants, being higher for T2 and lower for T4 (p < 0.05) (

Table 4. Growth performance of cherry tomato seedlings classified by size under different substrate treatments.

(a) Small Plants (S; TL 0–10 cm)
Treatment	NS	S%	TL (cm)	RL (cm)	SL (cm)	LN	TWB (g)
1	5.7 ± 4.7	24.7 ± 18.6	7.1 ± 2.0	1.7 ± 0.7 b	5.4 ± 1.7	9.1 ± 3.1 ab	0.2 ± 0.1
2	7.8 ± 3.5	36.2 ± 6.9	7.6 ± 2.3	1.9 ± 0.9 ab	5.7 ± 1.9	10.5 ± 3.9 a	0.2 ± 0.1
3	8.3 ± 3.2	31.4 ± 16.2	7.9 ± 2.1	2.5 ± 1.1 a	5.4 ± 1.9	7.8 ± 4.5 b	0.2 ± 0.1
4	7.3 ± 4.0	52.2 ± 21.7	6.6 ± 1.7	1.6 ± 1.1 b	5.0 ± 1.2	5.7 ± 4.0 b	0.1 ± 0.1
(b) Medium plants (M; TL 11–15 cm)
Treatment	NM	M%	TL (cm)	RL (cm)	SL (cm)	LN	TWB (g)
1	11.6 ± 3.1 ab	48.9 ± 17.3	12.3 ± 1.3	2.9 ± 1.5	9.3 ± 1.8	15.8 ± 3.3 a	0.4 ± 0.2 ab
2	11.0 ± 5.2 ab	52.6 ± 9.1	12.3 ± 1.4	3.1 ± 1.5	9.2 ± 1.4	16.8 ± 3.4 a	0.6 ± 0.2 a
3	15.7 ± 4.0 a	55.9 ± 8.8	12.1 ± 1.3	3.3 ± 0.8	8.8 ± 1.2	13.9 ± 2.6 b	0.4 ± 0.2 ab
4	5.0 ± 1.0 b	38.9 ± 11.7	12.5 ± 1.3	3.3 ± 1.8	9.3 ± 2.1	15.4 ± 2.2 ab	0.4 ± 0.1 b
(c) Large plants (L; TL 16–25 cm)
Treatment	NL	L%	TL (cm)	RL (cm)	SL (cm)	LN	TWB (g)
1	9.0 ± 13.9	26.6 ± 34.8	19.3 ± 2.6 a	5.2 ± 1.5	14.1 ± 2.3 a	21.7 ± 4.5 a	1.2 ± 0.6 a
2	1.7 ± 1.5	10.0 ± 8.7	17.8 ± 2.5 ab	5.2 ± 1.2	12.7 ± 2.2 ab	23.4 ± 4.6 a	1.4 ± 0.7 a
3	3.7 ± 2.5	12.6 ± 7.9	16.4 ± 1.2 b	4.7 ± 0.7	11.7 ± 1.2 b	17.1 ± 1.9 b	0.6 ± 0.1 b
4	1.0 ± 1.0	8.9 ± 10.2	16.5 ± 0.5 b	5.5 ± 1.3	11.0 ± 1.4 b	18.0 ± 2.0 b	0.9 ± 0.1 b

Legend: T1 = 7.2 kg of sand + 8 kg of gravel + 50 g of coconut fiber per gutter; T2 = 7.2 kg of sand + 12 kg of gravel per gutter; T3 = 12 kg of gravel + 50 g of coconut fiber per gutter; T4 = 7.2 kg of sand + 50 g of coconut fiber per gutter. NS, NM, NL = number of small, medium, and large plants; S%, M%, L% = percentage of small, medium, and large plants; TL = total length; RL = root length; SL = shoot length; LN = leaf number; TWB = total wet biomass. Values are expressed as mean ± standard deviation. Means followed by different letters in the same column differ significantly according to the Kruskal–Wallis test (p < 0.05).

). Root length (RL) of small plants was higher for T3 compared to the other treatments (p < 0.05). The number of medium plants (NM) was higher for T3 and lower for T4 (p < 0.05). LN of medium plants was higher for T1 and T2 and the lowest for T3, while TWB was higher for T2 and lower for T4 (p < 0.05). For large plants, TWB and LN were higher in T1 and T2 and lower in T3 (p < 0.05); TL was higher in T1 and lower in T3 and T4 (p < 0.05); and SL was higher in T1 and lower in T4 (p < 0.05).

3.2. Dynamics of Substrates

Even under the same irrigation regime and similar environmental conditions, differences in water retention among substrates were visually observed throughout the experimental period. T4 exhibited rapid surface drying, with the sandy substrate becoming dry and cracked. In contrast, T1 and T2 showed slower water loss and greater moisture persistence. In T3, a visible surface water layer was consistently observed during daily monitoring, indicating excessive water retention and continuous substrate saturation.

4. Discussion

Soilless agriculture has emerged as a sustainable and less polluting production strategy by enabling the reuse of nutrient-rich effluents, thereby contributing to resource efficiency and environmental protection by combining nutrient solutions with inert organic or inorganic substrates, this production approach allows for precise control of the root environment, enhancing productivity and supporting the production of diverse, high-quality crops. Within this context, evaluating substrate performance and nutrient dynamics becomes essential to optimize plant growth and ensure the effective use of alternative water and nutrient sources, such as those derived from aquaculture systems [7,36,37].

Germination occurred similarly across substrates, indicating that all treatments provided adequate water availability to support seed metabolic activation [38]. The absence of statistically significant differences among treatments demonstrates that substrate composition did not influence germination performance under the conditions evaluated, indicating equivalent physical conditions for water imbibition and early metabolic activation. Germination percentages remained within the range reported in the literature for cherry tomato seeds [39], reinforcing that all substrates were suitable for the germination phase. Previous studies suggest that substrates with higher proportions of coconut fiber may affect germination dynamics [40]; however, such effects were not statistically supported in the present study, indicating that differences in substrate composition did not result in measurable biological effects during germination. Overall, these results demonstrate that germination responses were statistically equivalent across substrates, confirming that substrate physical characteristics were not limiting factors at this developmental stage.

Although differences in water retention and structural support among substrates could be hypothesized based on their composition, physical properties such as porosity, water-holding capacity, and bulk density were not directly quantified in the present study. Importantly, the lack of statistically significant differences in germination and vigor indices indicates that potential variations in substrate physical behavior did not translate into differential biological responses. Therefore, interpretations regarding substrate physical properties should be considered contextual rather than indicative of contrasting performance among treatments. In addition, the absence of a control treatment using a conventional nutrient solution or commercial substrate limits direct comparisons with standard seedling production systems; however, this experimental design enabled the evaluation of inert substrates supplied exclusively with RAS wastewater, allowing substrate performance to be assessed under uniform water chemistry conditions.

The seed vigor index (SVI) did not differ among treatments, remaining close to values reported in the literature for cherry tomato seedlings grown with biofertilizers [41]. Similarly, the germination velocity index (GVI) showed no statistical differences and remained comparable to values reported for cherry tomato seeds germinated in washed sand [42]. Moisture content also remained stable across treatments, aligning with values previously reported for different cherry tomato varieties [43,44,45]. The statistical equivalence of these indicators confirms that all substrates provided comparable physical conditions for germination and early seedling establishment, ensuring stable physiological conditions regardless of substrate composition. These results indicate that, under the chemical conditions imposed by RAS wastewater, early growth was not constrained by substrate-related physical factors.

For larger seedlings, no statistically significant differences were observed among treatments for total length (TL), shoot length (SL), leaf number (LN), or total wet biomass (TWB). This statistical similarity indicates that all substrates supported comparable seedling growth when supplied with the same RAS wastewater, despite numerical variation among treatments. Seedling heights obtained across treatments exceeded the average values reported for alternative organic substrate mixtures used in cherry tomato production [46], and SL values were consistent with those documented in soil-based substrates enriched with organic amendments [47]. These findings demonstrate that substrate composition did not affect early biomass accumulation, indicating functional equivalence among treatments during this growth phase.

Across all treatments, LN values were statistically similar and higher than those reported for cherry tomatoes grown in latosol with chicken manure, organic compost, or Plantmax^® (Eucatex Agro, São Paulo, Brazil) [48], and within or above ranges previously observed using organic substrates [47]. Although higher LN values have been reported elsewhere [49], the values obtained in the present study are consistent with those typically associated with healthy early seedling development. The absence of statistically significant differences indicates that substrate composition did not influence leaf development at this stage, and that all treatments produced seedlings with morphological characteristics associated with adequate early growth and transplant potential [50,51].

The ambient temperature of 28.1 °C remained within the optimal range for cherry tomato cultivation (28–30 °C), supporting adequate physiological activity and nutrient uptake [52]. Electrical conductivity, an indicator of nutrient availability and organic matter decomposition [53], was 0.6 mS cm⁻¹, aligning with the 0.6–0.9 mS cm⁻¹ range recommended for aquaponic systems [54]. The stability of these environmental and chemical parameters supports the interpretation that observed plant responses were not influenced by external stressors, allowing substrate-related effects to be evaluated under controlled conditions.

The concentrations of nitrate (>55 mg L⁻¹), calcium (38.3 mg L⁻¹), and potassium (12.2 mg L⁻¹) determined in the RAS effluent characterize the nutrient profile available to the seedlings during the experimental period. This set of nutrients was associated with the occurrence of germination and the initial seedling growth observed in the study, as nitrogen availability and uptake efficiency are known to play a central role in early tomato development under soilless conditions [55]. Previous work has shown that the supply level of key ions such as potassium can significantly influence nutrient absorption and growth in tomato, especially as demand increases with plant development [56]. However, as plant nutritional requirements tend to increase throughout development, adjustments to the nutrient composition may become necessary at later growth stages.

The pH of 7.97 measured in the present study exceeded optimal ranges reported for hydroponic (5.5–6.2) [57] and aquaponic systems (around 7.0) [54]. Tomatoes typically perform best at slightly acidic pH values (5.5–6.0), which maximize nutrient solubility and uptake [58]. Elevated pH levels can reduce the availability of essential nutrients such as phosphorus, iron, and manganese, potentially limiting plant growth and biomass accumulation [59,60]. Although this alkaline pH did not affect germination or early seedling establishment, the elevated pH likely limited the availability of essential micronutrients such as Fe and Mn, which play a key role in chlorophyll formation; this limitation may have disrupted normal chloroplast function and helps explain the leaf yellowing observed after 25 days [61,62]. Thus, the observed leaf yellowing is consistent with a pH-mediated restriction of micronutrient uptake rather than a limitation in total nutrient supply. These findings suggest that pH adjustment of RAS wastewater can enhance nutrient availability and overall plant performance. Previous studies have shown that organic acidifiers such as lemon juice may serve as a low-cost alternative for pH control in hydroponic systems, particularly for lettuce production, when compared to commercial products such as pH Down [63].

Mineral analyses from aquaponic systems cultivating tomato and tilapia [64] reported calcium, magnesium, and nitrate concentrations similar to those observed in the pirapitinga RAS wastewater used in the present study, while potassium, ammonia, and nitrite levels were higher. Leaf yellowing in tomatoes has been associated with calcium and magnesium deficiencies [64]. Although these nutrients were present at relatively high concentrations in the wastewater used here, their effective availability may have been limited by the alkaline pH, impairing nutrient uptake. Similar patterns have been reported in substrate-based systems, where elevated pH reduces micronutrient solubility and limits physiological performance [59]. This interaction between RAS water chemistry and the root environment mediated by inert substrates helps explain the chemical factors influencing early plant responses in the present system.

From an applied perspective, pH management in systems using RAS wastewater as an external nutrient source can be implemented through relatively simple and feasible strategies. The controlled addition of acids, such as phosphoric acid, is commonly adopted in soilless cultivation to reduce alkalinity and improve nutrient availability. In this context, the use of RAS wastewater as a reuse resource without direct fish–plant coupling allows greater flexibility in pH correction, as fish are not exposed to short-term chemical adjustments. This operational separation enhances the feasibility of chemical optimization while maintaining water reuse principles, increasing the applicability of RAS wastewater in nursery-scale systems.

Despite the promising results observed during the seedling stage, some limitations of the present study should be acknowledged. The experimental period was restricted to early seedling development, and post-transplant plant performance was not evaluated, limiting extrapolation to later growth stages or final yield. In addition, RAS wastewater was used as the sole source of water and nutrients, without supplemental fertilization, which may have constrained maximum growth compared to conventional nutrient management strategies. Furthermore, the alkaline pH observed during the experiment may have reduced nutrient availability. Future studies integrating direct measurements of substrate physical properties with controlled adjustments of RAS water chemistry are necessary to fully elucidate the interaction between these two core variables, optimizing the use of aquaculture effluents in soilless seedling production. Additionally, as RAS water chemistry depends on fish species and feed composition, the present results should be interpreted as system-specific, highlighting the need for validation under different aquaculture conditions.

5. Conclusions

The substrates tested in this study supported similar overall seedling development, although T1 and T2 produced large seedlings with superior leaf number and shoot growth, indicating more favorable physical conditions for advanced early development. The pirapitinga RAS wastewater supported germination and early seedling development under the experimental conditions evaluated in this study. However, leaf yellowing observed after 25 days suggests possible nutrient limitations or reduced nutrient availability at the elevated pH recorded. These findings indicate that substrate choice plays a central role in supporting early-stage cherry tomato seedling development and also show that RAS wastewater can serve as a viable nutrient source when paired with substrates that ensure adequate moisture balance and structural stability. Nonetheless, pH adjustment or nutrient supplementation may be required to support later developmental stages. Overall, these results represent a promising starting point, highlighting the need for further studies to characterize nutrient availability in RAS wastewater and better define the nutritional requirements of cherry tomato seedlings throughout subsequent growth stages.