Effect of Nutrient Solution Concentration on Growth, Yield, and Fruit Quality of Tomato Grown Hydroponically in Single-Truss Production System

Experiments were conducted in 2015 and 2016 in a semi-controlled greenhouse at the Field Science Center of Ibaraki University, Japan. Tomato seeds of the ‘Momotaro York’ cultivar were sown in a 49-cell tray filled with a commercial peat substrate on March 17th and June 26th, 2015, for the spring and summer production cycles, respectively. The seeds were germinated at 28 °C for three days. Subsequently, for nearly one month, they were kept in a greenhouse where the temperature was maintained at approximately 25 °C. They were watered with tap water until planting. Seedlings were obtained from Suzuyo Shoji on December 18th for the winter production cycle. Seedlings were then planted in a closed hydroponic system 10 × 115 cm density on April 17th, July 24th, and December 24th for the spring, summer, and winter production cycles, respectively.

Plants were supplied with nutrient solutions at the following EC values: 0.8, 1.0, 1.2, and 1.4 mS·cm⁻¹ in the spring and summer production cycles and 1.0, 1.2, 1.4, and 1.6 mS·cm⁻¹ in the winter cycle. Such a wide choice of EC values of the nutrient solutions was necessary to determine the optimal range for this production system. The solutions used were prepared by a means of the stock solutions containing the following mineral fertilizers: OAT House 9 (P₂O₅–K₂O: 51–33) at a rate of 2.0 kg·100 L⁻¹ of water for the spring and summer production cycles or 1.55 kg·100 L⁻¹ of water for the winter cycle; OAT House 2 (N–CaO: 11–23) at a rate of 9.5 kg·100 L⁻¹ of water; OAT House 3 (N–K₂O: 13–46) at a rate of 8.1 kg·100 L⁻¹ of water; OAT House 5 (N–K₂O–MnO–B₂O₃–Fe–Cu–Zn–Mo: 6–9–2–2–5.7–0.04–0.08–0.043) at a rate of 0.5 kg·100 L⁻¹ of water; and OAT House 6 (MgO: 16) at a rate of 5 kg·100 L⁻¹ of water. A linear equation based on the measured EC values of the stock solutions was used to determine the initial solution concentrations for all EC treatments. Plants were fertigated for 10 minutes after every 30 minutes in the spring production cycle and after every 50 minutes in the summer and winter cycles. Regardless of the production cycle, the solution pH for all treatments was 6.3.

Tomatoes were trained opposite each other and twisted counterclockwise so that the flower clusters were left outside. The main stem was pinched 2 cm above the second leaf of the first flower cluster. The upper two leaves were also twisted counterclockwise to improve ventilation. Suckers were regularly removed, and fruits were thinned to four per cluster when their diameter reached approximately 40 mm. During the winter production cycle, greenhouse heaters were activated when air temperature dropped below 10 °C. In each treatment, there were 220 plants.

The daily temperatures of rooting media and plant canopy, the EC and pH of the nutrient solutions, and the weather conditions outside the greenhouse were recorded. The daily mean temperatures of the canopy and rooting media were monitored using data loggers (DS1921G-F50-thermochronic iButton, KN laboratory, Osaka, Japan). Twelve data loggers were used, six placed near the upper part of the plant and six within the media channels, which recorded temperature every ten minutes. EC and pH measurements of the nutrient solutions were taken using a pH meter (HI98129, Combo Hanna, Romania).

A random sample of 10 plants was assessed for each treatment for growth indices. The following growth indices were determined: plant height, stem diameter, leaf number, leaf chlorophyll content, and nitrate content in petiole sap. These parameters were determined every week until harvest. The plant height was measured from the medium surface to the growing point by a ruler, the stem diameter was determined by caliper, and leaf chlorophyll content was estimated using a portable chlorophyll meter (SPAD-502 P1us; Konica Minolta, Japan) on the three youngest fully developed leaves. Nitrate in the sap of petioles was measured on the recently matured leaves using a Nitrate meter (LAQUAtwin B-743, Horiba, Tokyo). To calculate the number of leaves per plant, only leaves longer than 2.5 cm were counted. Leaf area was determined using four harvested plants twice a month for each treatment. Fresh leaves with petioles were photographed, and the obtained pictures were analyzed for surface area using image analysis software (ImageJ 1.48v, Wayne Rasban, National Institutes of Health, Bethesda, Maryland, USA).

Four plants were harvested twice a month for each treatment, to determine the dry weights of stems, leaves, clusters, and fruits. The dry weights of the plants were measured by drying them at 60 °C in a forced-draft oven (FS-60P, T.G.K, Tokyo, Japan) until a constant weight was reached.

For fruit production and analysis, four replications, each with 20 plants, were chosen for each treatment. Fruits were collected from each replication, and their total number was divided by the number of plants to determine the average number of fruits per plant. Similarly, the total fruit weight from each replication was divided by the number of plants to calculate the average weight per plant. To determine the average weight of an individual fruit, we divided the fruit weight per plant by the number of fruits per plant. Fruits were classified as regular and damaged. Damaged fruits were classified as cracked and affected by blossom end rot (BER). Titratable acidity (TA) and SSC were determined as fruit quality parameters using a saccharic acidity meter (Hybrid PAL-BX|ACID F5; Atago, Tokyo, Japan). These parameters were measured on eighteen randomly selected fruits per replicate three times during the harvest period (six fruits at the beginning, middle, and final harvest). The fruits were juiced in a juicer, and the obtained juice was filtered through a paper filter. The SSC was measured directly in the juice, whereas TA was measured by diluting the fruit juice by a ratio of 1 : 50 with deionized water. The results of TA were expressed as the percentage content of citric acid.

The absorbed amounts of nitrogen (N), phosphorus (P), potassium (K), magnesium (Mg), and calcium (Ca) by plants were estimated by the consumed amount of stock solution and planting density.

In order to investigate the isolated effects of EC treatments on tomato growth, yield, and fruit quality in each production cycle, data were subjected to one-way ANOVA, and Tukey's HSD test was employed for multiple comparisons (p = 0.05) using IBM SPSS Statistics 23.

In the spring production cycle, the daily mean plant canopy temperature remained stable, ranging between 18 and 28 °C (Fig. 1A). During the summer, temperatures rose to over 30 °C until the middle of August and then decreased during the flowering stage (Fig. 1B). In the winter production cycle, temperatures were below 15 °C until March 16th and gradually increased until the end of the cycle (Fig. 1C).

Figure 1.

In the spring production cycle, the EC of the nutrient solution rapidly increased at the 1.4 and 1.2 treatments and gradually increased at the 1.0 and 0.8 treatments after pinching the main stem (Fig. 2A). In this production cycle, the EC values reached 4.60, 2.72, 1.54, and 1.10 mS·cm⁻¹ by the first harvest, and 5.10, 3.10, 1.59, and 1.02 mS·cm⁻¹ by the third harvest for the 1.4, 1.2, 1.0, and 0.8 EC, respectively. In the summer cycle, until August 15th (two days before pinching the main stem) the EC values of solutions increased to 3.46 and 1.50 mS·cm⁻¹ for the 1.4 and 1.2 treatments, respectively, and decreased to 0.51 and 0.34 mS·cm⁻¹ for the 1.0 and 0.8 treatments, respectively (Fig. 2B). In winter, the EC of solutions at bud appearance (January 13) were 2.02, 1.81, 1.20, and 1.04 mS·cm⁻¹ for the 1.6, 1.4, 1.2, and 1.0 treatments, respectively. Then, during the fruit development period (until March 15th), the EC of solutions of all treatments decreased to the range of 0.30–0.68 mS·cm⁻¹ (Fig. 2C).

Figure 2.

In the spring production cycle, the final height of plants at the 1.4 and 1.2 EC was greater than that of the others (Fig. 3A). In the summer cycle, the tallest plants were found at the 0.8 EC. In contrast, in winter, plant heights were comparable among the tested EC (Fig. 3B, C). During the spring production cycle, from the third week of transplanting until the end of the experiment, the greatest stem diameter was found on plants at the 1.4 EC (Fig. 4A). In this cycle, no significant differences in stem diameter were found between the 1.0 and 0.8 treatments. In the summer cycle, the final stem diameter did not differ between plants from the tested EC (Fig. 4B). In winter, at the pinching time, the greatest stem diameter was observed at the 1.2 EC (Fig. 4C).

Figure 3.

Figure 4.

Across all production cycles, the final leaf number per plant did not differ between the tested treatments. The leaf number per plant in the spring cycle was 9, whereas in the summer and winter, it was almost 11 (Fig. 5A, B, C).

Figure 5.

In the spring production cycle, from May 8 to June 19, the SPAD values on plants at the 0.8 EC were the lowest (Fig. 6A). However, at the last time point of SPAD readings (June 26), they did not differ between the tested treatments. In the summer cycle, significant differences in SPAD values between treatments were observed just after the first week of transplanting (August 2) (Fig. 6B). In the final week of measurement (October 11), the SPAD values on plants at the 1.4 and 1.0 EC were higher than those at the 1.2 and 0.8 EC. In the winter cycle, on February 27th and March 18th, the SPAD readings on plants at the 1.0 EC were lower than those of the other treatments (Fig. 6C).

Figure 6.

In the spring cycle, in most cases, petiole sap nitrate concentrations were comparable between the tested treatments (Fig. 7A). However, there was a tendency indicating that the plants at the 1.4 EC had increased nitrate concentrations. Throughout the summer cycle, plants at the 1.2 EC had more nitrates, while plants at the 0.8 EC contained less nitrate than the other treatments (Fig. 7B). In the winter cycle, at most time points, petiole sap nitrate concentrations were the greatest at the 1.4 EC and the lowest at the 1.0 EC (Fig. 7C). In the spring cycle, leaf area (LA) of plants at the 1.4 and 1.2 EC were larger than that of the other EC (Fig. 8A). In the summer, LA did not differ significantly between treatments, except on August 27th and September 12th, where it was larger at the 1.2 EC than at the 1.4 and 0.8 EC (Fig. 8B). In the winter, no significant differences in LA were found (Fig. 8C).

Figure 7.

Figure 8.

At most time points in the spring production cycle, total dry weight (TDW) did not differ between the tested treatments (Fig. 9A). In this cycle, plants at the 0.8 EC had the lowest TDW. During the summer, only on September 12th (the reproductive plant stage) was the TDW of plants at the 1.2 EC greater than that at the 1.4 and 0.8 EC (Fig. 9B). In the winter cycle, a significant difference in TDW was found only on March 17th (the reproductive stage), where this parameter was greater at the 1.2 EC compared to the 1.0 and 1.6 treatments (Fig. 9C).

Figure 9.

In the spring cycle, the marketable fruit yields at the 1.4 and 1.2 EC were higher than at the 1.0 and 0.8 EC (Table 1). Simultaneously, mean fruit weights at the 1.2 and 1.4 EC were higher than those of the 0.8 EC (Table 1). Only a small number of cracked fruits were found in our study, but no differences between the tested treatments were observed (Table 1). The incidence of fruits affected by BER was negligible (data not shown).

During the summer, the number and yields of marketable fruit at the 1.2 and 1.0 EC were greater than those of the 1.4 and 0.8 EC (Table 1). Cracked fruits were observed in all treatments; however, the number of cracked fruits per plant at the 1.0 EC was higher than at the 1.4 EC (Table 1).

In the winter cycle, the highest marketable yield was found at 1.4 EC and the lowest at the 1.6 EC (Table 1). Simultaneously, plants at 1.2 EC produced a higher marketable yield than those of the 1.0 EC (Table 1). Fruit from the 1.4 and 1.2 EC were bigger than those from the 1.6 EC (Table 1). There was no difference in the incidence of cracked fruits between the treatments (Table 1).

In the spring cycle, fruit at the 1.4 EC had higher SSC than the other treatments (Table 2). However, the fruit acidity in this treatment was similar to the other treatments. The SSC to acidity ratio values did not differ between the tested treatments (Table 2). Similarly, as in the spring cycle, in the summer, fruit at the 1.4 EC had the highest SSC (Table 2). Additionally, those fruits also contained more organic acids than in the other treatments (Table 2). However, fruit at the 1.2 EC had a higher SSC to acidity ratio than fruit at the 1.4 EC (Table 2). In winter, SSC, acidity, and the ratio of SSC to acidity in fruit did not differ between the tested treatments (Table 2).

In the spring cycle, the daily uptake rates of N, P, K, Mg, and Ca by plants grown at the 1.4 EC were greater than at the 1.0 and 0.8 EC (Table 3). During the summer, the lowest amounts of absorbed P, K, Mg, and Ca were found at the 0.8 EC (Table 3). However, the N uptake was higher at the 1.4 and 1.2 treatments compared to 0.8. In the winter cycle, uptake rates of all nutrients did not differ between the tested treatments (Table 3).