Relationship between Vapor Pressure Deficit Condition and Water-stress Status in Strawberry

Abstract

To develop a novel highly robust humidification technology that requires minimal adjustment of operational methods, we investigated the vapor pressure deficit (VPD) conditions that cause water stress in strawberry (Fragaria × ananassa Duch.) plants. The relationship between the rate of relative change in fruit-stalk diameter (RCFSD_t) and VPD conditions, including VPD and rate of change in VPD (VPD′; hPa·min⁻¹) were analyzed by performing multiple regression. A negative RCFSD_t was defined as an indicator that the plant was experiencing water stress. Regression of RCFSD_t on VPD and VPD′ revealed that VPD′ greater than 0.069 hPa·min⁻¹ led to negative RCFSD_t when VPD was 8.0 hPa or higher. In addition, we investigated whether the rate of change in VPD′ (VPD″; hPa ·min⁻²) could be used as an indicator to cease humidification, employing both VPD and VPD′. Analysis of the relationship between RCFSD_t and VPD″ revealed that even if VPD and VPD′ exceeded 8.0 hPa and 0.069 hPa·min⁻¹, respectively, RCFSD_t could not be negative when VPD″ was less than −0.00426 hPa·min⁻². We measured the relative change in fruit-stalk diameter (CFSD_t) of strawberry under the conditions generated by a novel humidification system. The novel humidification system humidified the air when VPD, VPD′, and VPD″ exceeded each threshold for avoiding the condition, where VPD, VPD′, and VPD″ exceeded 8.0 hPa, 0.069 hPa·min⁻¹, and −0.00426 hPa·min⁻², respectively. In practice, there were periods when humidification control was not executed because while VPD exceeded its control threshold, VPD′ and VPD″ were below their respective control thresholds. In addition, the novel humidification system prevented a decrease in CFSD_t during the daytime under the non-humidification treatment.

Introduction

Recently, development of environmental control techniques has been required to increase strawberry production. It has been noted that low humidity during the daytime reduces plant growth and fruit yield under strawberry forcing culture in Japan. A vapor pressure deficit (VPD) ranging from 8 to 12 hPa is suitable for strawberry production (Lieten, 2002). High VPD causes stomatal closure and reduces the photosynthesis rate, resulting in high water stress for the plant, reduced plant growth and low fruit yield (Choi and Jeong, 2020; Jones, 1993; Lieten, 2002; Tagawa et al., 2016). Because the maximum VPD in the daytime in a greenhouse often exceeds 12 hPa throughout the cultivation period under strawberry forcing culture, humidification techniques have received increasing attention.

When introducing a humidification technique, however, several issues can be anticipated. Continuous high humidity reduced transpiration and caused tip burn in tomato (Ehret and Ho, 1986), lisianthus (Islam et al., 2004), and strawberry (Lieten, 2002). Continuous high humidity also reduced the fruit-set rate, fruit firmness, and fruit storability in strawberry (Lieten, 2002). Furthermore, continuous high humidity increased the incidence of gray mold disease in cucumber (Yunis et al., 1990) and lisianthus (Shpialter et al., 2009). If mist is sprayed such that it wets the plant body such as the leaves and fruits, the incidence of the above diseases may be promoted. Humidification controllers currently used in agricultural production sites can set the humidity threshold for starting mist spraying, as well as the interval times for mist spraying and stopping. To solve problems such as extreme high humidity and wetting of plant surfaces, growers need to keep adjusting the scan time of the humidity sensor, the threshold for starting mist spraying and the spraying mist interval time, creating a burden for growers. Furthermore, these adjustments require growers with a high level of experience and skill.

To reduce the burden required to adjust operational humidification methods, there have been some reports on humidification control methods using quantitative feedback theory (QFT) (Linker et al., 2011), a proportional-integral-differential (PID) controller (Su et al., 2020), a fuzzy neural network (FNN) (Jia, 2021), and model predictive control (MPC) (Ito and Tabei, 2021). In these reports, they simulated the techniques proposed for stably controlling temperature and humidity within the desired zone using heaters and humidifiers, either in a virtual space or a greenhouse. These techniques involve high implementation costs and low robustness due to their overly complex algorithms. Moreover, these techniques did not take plant growth into consideration (Linker et al., 2011). Because there are many small-scale growers in Japan, there is a need to develop a novel humidification control method that is low-cost, highly robust (with a simple algorithm), requires minimal adjustment of operational methods, and can steadily improve plant growth and yield.

We hypothesized that by basing the system on plant biological information, we could develop a humidification control method that meets the above requirements. In this study, we focused on the change in fruit-stalk diameter of strawberry. Shrinkage in the stem diameter in tomato (Oishi, 2002) and the fruit-stalk diameter in strawberry (Yamanaka et al., 2020) indicate that the plants are experiencing water stress. When VPD increases rapidly, the stomata first open and then close in response to the increase in transpiration to maintain the water content within the plant body (Grossiord et al., 2020). In strawberry, the fruit-stalk diameter shrinks when the stomata open and the transpiration rate is high (Johnson et al., 1992; Ninomiya et al., 2012; Yamanaka et al., 2020). Humidification to prevent the fruit-stalk diameter from shrinking is considered to prevent stomatal closure.

A change in fruit-stalk diameter of strawberry follows a rapid change in VPD (Yamanaka et al., 2020). With near-infrared spectroscopy, the second-derivative values are commonly used to analyze the correlation between changes in absorbance and variations in the concentration of a specific substances in detail. Using the second-derivative values may be helpful to investigate the relationship between changes in fruit-stalk diameter and changes in VPD. Therefore, a detailed analysis of the relationship between changes in fruit-stalk diameter and changes in VPD can be achieved by employing not only the absolute VPD value, but also the rate of change in VPD (VPD′; hPa·min⁻¹) and the rate of change in VPD′ (VPD″; hPa·min⁻²).

In this study, to develop a novel method of humidification that can reduce water stress in strawberry plants without generating excessive humidification, we investigated the relationship between the rate of relative change in fruit-stalk diameter (RCFSD_t; mm·mm⁻¹·h⁻¹) and VPD conditions, specifically VPD, VPD′, and VPD″.

Materials and Methods

Analysis of the relationship between VPD conditions and change in fruit-stalk diameter (Experiment I)

Correction for sensor system temperature drift

To measure microscopic changes in the fruit-stalk diameter of strawberry, temperature drift in the output value of the digital displacement sensor caused by the influence of the sensor support material and structure were corrected based on Yamanaka et al. (2020). In the present study, a contact-type digital displacement sensor (Model GT-H22L; Keyence Co., Ltd., Osaka, Japan) and a sensor support composed of acrylonitrile-butadiene-styrene resin were used (Fig. 1). In a growth chamber, the temperature was decreased in 5°C intervals every 90 min from 40 to 5°C, and then increased in 5°C intervals every 90 min from 5 to 40°C. In the growth chamber, the output value of the sensor at an arbitrary time (OS_t) was recorded at one-minute intervals in a state without anything being clamped. At the same time, the temperature at an arbitrary time (T_t) was recorded at one-minute intervals. The regression coefficient when regressing OS_t on T_t was defined as C (mm·°C⁻¹) (Equation 1). Next, OS_t and T_t were converted to five-minute averaged values (AOS_t and AT_t, respectively) (Equation 2). The correction value of the sensor output with a reference temperature (COS_t) was calculated from AOS_t and AT_t using Equation 3. In this study, the reference temperature was set to 19°C because the approximate median of the upper and lower temperature limits in a cultivation environment (32°C and 6°C) is approximately 19°C. Finally, COS_t and AT_t were smoothed using the Savitzky-Golay method (Savitzky and Golay, 1964), resulting in SCOS_t and SAT_t. To assess the correction, the difference between the OS_t at T_t and the average OS_t at 19°C (ΔOS_(Tt−19)) was compared with the difference between the SCOS_t at SAT_t and the average OS_t at 19°C (ΔSCOS_(SATt−19)).

  

$${\text{OS}}_{\text{t}}=C\times T_{\mathrm{t}}+\text{b}$$(Equation 1)

where OS_t respresents the sensor output values in a state without anything being clamped at an arbitrary time recorded at one-minute intervals (mm), C is the pseudo-thermal expansion coefficient of the sensor support (mm·°C⁻¹), T_t is the temperature at an arbitrary time recorded at one-minute intervals (°C), and b is the intercept (mm).

  

$$X_{\text{t}}=\left({\sum }_{k=\mathrm{t}-\mathrm{5}}^{\text{t}}{x}_k\right)/5$$(Equation 2)

where X_t is the five-minute averaged value of OS_t (AOS_t; mm) or T_t (AT_t; °C), and x_k is the raw value of the sensor output (OS_t) or T_t.

  

$${\text{COS}}_{\text{t}}={\text{AOS}}_{\text{t}}-C\times \left({\text{AT}}_{\text{t}}-19\right)$$(Equation 3)

where COS_t is the sensor output value corrected for the temperature drift caused by the sensor support (mm).

Fig. 1

Measurement of strawberry fruit-stalk diameter using a digital displacement sensor with a sensor support composed of acrylonitrile-butadiene-styrene (ABS) resin.

Plant materials and cultivation condition

The June-bearing strawberry ‘Koiminori’ was used to measure the change in fruit-stalk diameter (n = 4). The measurements were conducted on November 10, 2019 (Day 1), April 15, 2020 (Day 2), and April 16, 2020 (Day 3). The strawberry plants for measurement on Day 1 were transplanted on September 11, 2019. The strawberry plants for measurement on Days 2 and 3 were transplanted on November 12, 2019. The plants were grown in 18 cm diameter polyethylene pots filled with pH-adjusted peat moss from transplanting to measurement in a greenhouse at the Zentsuji campus, Western Region Agricultural Research Center, NARO (Zentsuji, Kagawa, Japan). The plants were provided with OAT House A nutrient solution (OAT Agrio Co., Ltd., Tokyo, Japan) with electrical conductivity (EC) of approximately 1.0 dS·m⁻¹, at the rate of 100–200 mL day⁻¹ per plant.

Measurement of the change in fruit-stalk diameter

The fruit stalk was fixed on the sensor, as illustrated in Figure 1, and the fruit-stalk diameter was measured at one-minute intervals. At the conclusion of measurements, the fruit-stalk diameter (FSD_e) was measured using calipers. The raw values of change in fruit-stalk diameter at one-minute intervals were converted to the averaged value at five-minute intervals using Equation 2. The five-minute averaged value was corrected to a reference temperature of 19°C using Equations 1 and 3. The fruit-stalk diameter at the initial time of the measurement (FSD_i) was calculated by subtracting the displacement of the sensor from the start to the conclusion of the measurement from FSD_e. The relative change in fruit-stalk diameter (CFSD_t; mm·mm⁻¹) was calculated using Equation 4. Then, RCFSD_t (mm·mm⁻¹·h⁻¹) was calculated using Equation 5. Finally, CFSD_t and RCFSD_t were smoothed using the Savitzky-Golay method.

  

$${\text{CFSD}}_{\text{t}}=\left({\text{FSD}}_{\text{t}}-{\text{FSD}}_{\text{i}}\right)\text{/}{\text{FSD}}_{\text{i}}$$(Equation 4)

where CFSD_t is the relative change in fruit-stalk diameter at an arbitrary time (mm·mm⁻¹), FSD_t is the fruit-stalk diameter at an arbitrary time (mm), and FSD_i is the fruit-stalk diameter at the initial time of measurement (mm).

  

$${\text{RCFSD}}_{\text{t}}=\left({\text{CFSD}}_{\text{t}}-{\text{CFSD}}_{\text{t}-\text{1}}\right)/5\times 60$$(Equation 5)

where RCFSD_t is the rate of CFSD_t (mm·mm⁻¹·h⁻¹).

Calculation of VPD, VPD′, and VPD″

Air temperature and relative humidity were recorded at one-minute intervals according to WARC/NARO (2021) using a thermo-hygrometer (RTR-576-S; T&D Corp., Nagano, Japan) covered with a shelter (CO-RS1; Climatec, Inc., Tokyo, Japan) with a DC fan (C4010H12BPLB-W-7; Misumi Group, Inc., Tokyo, Japan). The sensors were installed near the plants in the center of the greenhouse (but outside the canopy) at a height of approximately 1.5 meters from the bottom of the shelter, referencing the method of the Japan Meteorological Agency (1998). The VPD at an arbitrary time (VPD_t) was obtained by calculation from the air temperature and relative humidity (Buck, 1981), and converted to a five-minute averaged value using Equation 2. Next, VPD′_t and VPD″_t were calculated using Equations 6 and 7. The VPD_t, VPD′_t, and VPD″_t values were smoothed using the Savitzky-Golay method.

  

$${\text{VPD}\prime }_{\text{t}}=\left({\text{VPD}}_{\text{t}}-{\text{AVPD}}_{\text{t}–\text{1}}\right)/5$$(Equation 6)

where VPD_t is the vapor pressure deficit at an arbitrary time (hPa) and VPD′_t is the rate of VPD_t (hPa·min⁻¹).

  

$${\text{VPD}″}_{\text{t}}=\left({\text{VPD}\prime }_{\text{t}}-{\text{VPD}\prime }_{\text{t}-1}\right)/5$$(Equation 7)

where VPD″_t is the rate of VPD′_t at an arbitrary time (hPa·min⁻²).

Calculation of the VPD′ and VPD″ threshold that led to a negative RCFSD

The RCFSD_t was regressed on VPD_t and VPD′_t during the daytime in a multiple regression using the ‘MASS’ package for R. To determine the threshold of VPD′ that leds to a negative RCFSD under the condition where the VPD exceeded 8 hPa, RCFSD = 0 and VPD = 8.0 were substituted into the above multiple regression equation. Under the condition where the RCFSD_t was less than 0 mm·mm⁻¹·h⁻¹ and the VPD_t and VPD′_t were greater than 8 hPa and the above threshold for VPD′, respectively, the distribution of VPD″_t was analyzed by generating a histogram using Statcel-the Useful Addin Forms on Excel, 4th ed. (OMS Publish, Tokyo, Japan). Class intervals in the histogram were calculated using Sturges’s rule (Sturges, 1926).

Effects of a novel humidification method based on VPD, VPD′, and VPD″ on change in fruit-stalk diameter of strawberry (Experiment II)

Plant materials and cultivation conditions

The cultivation method for strawberry plants was based on that of Yamanaka et al. (2024). The June-bearing strawberry cultivars ‘Koiminori’ and ‘Kaorino’ were used to measure changes in fruit-stalk diameter (n = 3–4). Nursery plants of ‘Kaorino’ and ‘Koiminori’ were transplanted on September 12, 2023, and October 2, 2023. The three nursery plants were transplanted into a bowl-shaped plastic pot of 30 cm diameter and 6 L volume (bowl-pot 10 go; VENTECH Co., Ltd.). The pots were filled with pH-adjusted peat moss and placed on the high bench in two greenhouses at the Zentsuji campus, Western Region Agricultural Research Center, NARO (Zentsuji, Kagawa, Japan). The plants were provided with OAT House A nutrient solution with EC ranging from 0.8 to 1.5 dS·m⁻¹. The two greenhouses were similarly ventilated, heated, CO₂-treated, and a nutrient solution was supplied using an environmental controller (YoshiMax; Sanki Keiso Co., Ltd., Tokyo, Japan) based on Yoshida and Yasuba (2020).

Humidification treatment

The humidification method was based on that of Yamanaka et al. (2024). A humidification system was installed only in one greenhouse (humidification treatment, HT); the other greenhouse was nontreated as a control (non-humidification treatment, NT). The humidification system consisted of a dry mist system, a programmable datalogger (CR300; Campbell Scientific, Inc., Logan, UT, USA), and a thermo-hygrometer (CVS-HMP60; Climatec, Inc.) covered with a shelter (CYG-41003; Climatec, Inc.). The humidification system was operated as indicated in Figure 2 (Yamanaka et al., 2022). Control parameters were set as follows; T₁ = 6:00, T₂ = 18:00, t₁ = 120 s, x = 7.0 hPa, x′ = 0.1 hPa·min⁻¹, x″ = −0.04 hPa·min⁻², and t₂ = 60 s. The scan time of the humidification system was 1 s.

Fig. 2

Flowchart showing control of the novel humidification system.

Measurement of changes in fruit-stalk diameter

The changes in fruit-stalk diameter were measured from March 1 to March 3, 2024, for ‘Koiminori’ and on March 9 and 10, 2024 for ‘Kaorino’. Four plants were used for each measurement, and the data of three or four accurately measured plants were averaged. However, the March 2 HT data was excluded because its trend differed from the ensemble average HT data from March 1 to March 3, and only two plants were accurately measured. The correction for the sensor output temperature drift and the calculation of CFSD_t were performed using the same methods as described for Experiment I.

Calculation of VPD, VPD′, and VPD″

Recording of the air temperature and relative humidity, and calculation of VPD-ave_t, VPD′_t, and VPD″_t were performed using the same methods as described for Experiment 1.

Results and Discussion

We investigated the effect of VPD and VPD′ on the shrinkage in fruit-stalk diameter during the daytime to understand the VPD conditions that cause water stress in strawberry plants to develop a novel humidification technology for strawberry production.

First, we aimed to correct the effect on the temperature output values of the digital displacement sensor caused by the sensor support to measure the RCFSD. The ΔOS_(Tt−19) exhibited a negative trend with respect to temperature change and noise in hysteresis (Fig. 3). The noise in hysteresis was found to occur with a time delay of about 5 min in response to temperature change. To eliminate the noise in hysteresis, we converted OS_t to a five-minute averaged value (AOS_t). The AOS_t values were corrected for temperature drift and smoothed using the Savitzky-Golay method. As a result, the ΔSCOS_(ATt−19) did not show a distinct trend with respect to air temperature and little noise in hysteresis was observed (Fig. 3). Thus, the correction method enabled precise measurement using a digital displacement sensor.

Fig. 3

Effect of air temperature on the difference between the OS_t at T_t and average OS_t at 19°C (ΔOS_(Tt−19)) and the difference between the SCOS_t at SAT_t and average OS_t at 19°C (ΔSCOS_(SATt−19)). ΔX_(Tt−19) is ΔOS_(Tt−19) or ΔSCOS_(SATt−19). T_t and SAT_t are the temperatures at arbitrary times and the temperatures were averaged over 5 min and smoothed using the Savitzky-Golay method.

The RCFSD_t was significantly regressed on VPD_t and VPD′_t in a multiple regression, and Equation 8 was derived. The RCFSD_t data are shown in Figure 4.

  

$$\begin{array}{l}{\text{RCFSD}}_{\text{t}}={-7.90\times 10}^{-5}\times {\text{VPD}}_{\text{t}}{-1.86\times 10}^{-2}\\ \times {\text{VPD}\text{'}}_{\text{t}}{+1.92\times 10}^{-3}\end{array}$$(Equation 8)

Fig. 4

Diurnal change of the rate of relative change in fruit-stalk diameter (RCFSD_t) of ‘Koiminori’ on November 10, 2019 (A), April 15 (B) and 16 (C), 2020, respectively. The error bar indicates standard error (n = 4).

The regression coefficients of VPD and VPD′, and the intercept of Equation 8 were statistically significant. The root mean squared error between the RCFSD values calculated with Equation 8 and the measured values of RCFSD was 3.59 × 10⁻³.

The minimum VPD suitable for strawberry production is approximately 8 hPa (Kato et al., 2015; Lieten, 2002). When RCFSD is less than 0 mm·mm⁻¹·h⁻¹, strawberry plants are considered to experience water stress. Hence, by substituting RCFSD = 0 and VPD = 8 into Equation 7, we obtained VPD′ = 0.069. It was reported that strawberry plants may be under stress when VPD and VPD′ exceed 8.0 hPa and 0.069 hPa·min⁻¹, respectively (Yamanaka et al., 2020).

We next investigated the relationship between RCFSD_t and VPD″_t. The cumulative rate of frequencies in the VPD″_t histogram, leading to negative RCFSD_t under conditions in which the VPD_t and VPD′_t were greater than 8.0 hPa and 0.069 hPa·min⁻¹, respectively, is shown in Figure 5. The number of data points and class intervals in the histogram of VPD″_t were 95 and 0.00341, respectively. The minimum value of VPD″_t in the histogram was −0.011. The cumulative rate of frequencies conformed with a sigmoid-shaped curve. When the class values shifted from −0.00767 < VPD″_t ≤ −0.00426 to −0.00426 < VPD″_t ≤ −0.00085, the sigmoid curve increased. Even under conditions in which VPD and VPD′ exceeded 8.0 hPa and 0.069 hPa·min⁻¹, respectively, the results suggested that VPD″ less than −0.00426 hPa·min⁻² may not intensify the water stress on strawberry plants. In humidification systems, VPD″ is assumed to play the role of a breaking mechanism.

Fig. 5

Cumulative rate of frequencies in the VPD″_t histogram that led to a negative rate of relative change in fruit-stalk diameter (RCFSD_t) under the VPD condition in which VPD_t and VPD′_t exceeded 8.0 hPa and 0.069 hPa·min⁻¹ simultaneously in the daytime, respectively.

We next examined the threshold of VPD, VPD′, and VPD″ for Experiment 1 using NT data in Experiment II. Because the high bench in Experiment II was structurally integrated with the greenhouse, the RCFSD_t had a negative value during the period when the change in RCFSD_t was very small under the low VPD condition. In Experiment I, however, the RCFSD_t did not have a negative value below 4.0 hPa for VPD_t. Therefore, the data in Experiment II with negative RCFSD_t at VPD_t less than 4.0 hPa was filtered out.

Using the same method as in Experiment I, we determined the thresholds of 0.068 hPa·min⁻¹ for VPD′ and −0.03256 hPa·min⁻² for VPD″, which tended to have a negative RCFSD_t at 8.0 hPa of VPD. This was based on data from the ‘Koiminori’ data in NT measured from March 1 to 3 (n = 3), and Equation 9 was obtained calculate the threshold of VPD′. The thresholds of 0.108 hPa·min⁻¹ for VPD′ and −0.06962 hPa·min⁻² for VPD″, which tended to a negative RCFSD_t at 8.0 hPa of VPD, were also obtained from the ‘Kaorino’ data in NT measured from March 9 to 10 (n = 4). In the ‘Kaorino’ data, Equation 10 was determined to calculate the threshold of VPD′. The thresholds of VPD′ in ‘Koiminori’ between Experiment I and II were almost the same. However, it was suggested that the VPD′ threshold could vary among cultivars, and the VPD″ threshold may fluctuate depending on the situation compared to the VPD′ threshold.

  

$$\begin{array}{l}{\text{RCFSD}}_{\text{t}}={-1.12\times 10}^{-3}\times {\text{VPD}}_{\text{t}}{-1.62\times 10}^{-2}\\ \times {\text{VPD}\text{'}}_{\text{t}}{+1.01\times 10}^{-2}\end{array}$$(Equation 9)

  

$$\begin{array}{l}{\text{RCFSD}}_{\text{t}}={-7.13\times 10}^{-4}\times {\text{VPD}}_{\text{t}}{-7.19\times 10}^{-3}\\ \times {\text{VPD}\text{'}}_{\text{t}}{+6.48\times 10}^{-3}\end{array}$$(Equation 10)

Using the thresholds of Experiment I, we developed a novel humidification system. The novel humidification system can be any humidification system that uses software to humidify with VPD, VPD′, and VPD″ as the threshold. In Experiment II, we investigated the effect of the novel humidification system on the water-stress characteristics of ‘Koiminori’ and ‘Kaorino’ strawberry plants. On sunny days, the CFSD_t showed almost the same pattern for each cultivar treatment. Therefore, the ‘Kaorino’ data on March 9, 2024 are shown in Figures 6 and 7 as representative values.

Fig. 6

Diurnal change in the relative change in fruit-stalk diameter (CFSD_t) and VPD_t (A), VPD′_t (B), and VPD″_t (C) in non-humidification treatment (NT) in ‘Kaorino’ measured on March 9, 2024.

Fig. 7

Diurnal change in the relative change in fruit-stalk diameter (CFSD_t) and VPD_t (A), VPD′_t (B), and VPD″_t (C) in the humidification treatment (HT) of ‘Kaorino’ measured on March 9, 2024.

The CFSD_t, VPD_t (Fig. 6A), VPD′_t (Fig. 6B), and VPD″_t (Fig. 6C) in NT are shown. The VPD′_t and VPD″_t exceeded 0.069 hPa·min⁻¹ and −0.00426 hPa·min⁻² simultaneously and VPD_t increased rapidly from 9:50 to 10:30. The CFSD_t was less than 0 mm·mm⁻¹ from 10:20 to 10:50. The VPD′_t and VPD″_t exceeded their thresholds again from 11:25 to 11:35 and then VPD_t exceeded 8 hPa firstly at 11:40. VPD_t exceeded 8 hPa from 11:40 to 14:25 and from 15:35 to 16:35. The CFSD_t greatly decreased from 11:35 to 13:15. The time lag of shrinkage in fruit-stalk diameter relative to the increase in VPD was estimated to be around 5 to 10 minutes. It was also assumed that the fruit-stalk diameter shrunk due to a rapid increase in VPD even if VPD was less than 8 hPa.

The CFSD_t, VPD_t (Fig. 7A), VPD′_t (Fig. 7B), and VPD″_t (Fig. 7C) in HT are also shown. Since the HT greenhouse was adjacent to the west side of the NT greenhouse, the VPD in HT was less likely to rise compared to NT in the morning. After 12:00, the VPD_t exceeded 8 hPa, but the VPD′_t decreased following sprayed mist application. The fluctuation in VPD″_t was also smaller in HT than NT. As a result, the CFSD_t was less than 0 mm·mm⁻¹ only from 14:50 to 14:55 and from 15:05 to 15:10.

It was assumed that the fruit-stalk diameter during the night was the steady state for fruit-stalk diameter. If the CFSD_t is considerably below 0 mm·mm⁻¹ during the daytime, the strawberry plants are considered to be under water stress (Yamanaka et al., 2020). When the fruit-stalk diameter shrinks, the plant is interpreted to experience water stress caused by a decrease in water potential along with a high transpiration rate (Johnson et al., 1992; Ninomiya et al., 2012; Sakurai, 1994; Yamanaka et al., 2020). Because the chances of CFSD_t dropping below 0 mm·mm⁻¹ were reduced, water stress in the HT plants was more alleviated compared with NT plants.

This novel humidification system is not a control for maintaining humidity within a certain range, as executed by complex QFT algorithms (Linker et al., 2011), a PID controller (Su et al., 2020), FNN (Jia, 2021), and an MPC (Ito and Tabei, 2021). This system uses a simple algorithm that references indirectly measured biometric information. It is very robust, allowing it to operate stably even in complex and fluctuating greenhouse environments. Indeed, Yamanaka et al. (2024) demonstrated that this system increased plant growth and fruit yield, but did disease incidence did not increase in some strawberry cultivars with the same thresholds and interval times throughout cultivation. Therefore, this system could contribute to labor saving for growers. We are planning to investigate the mechanism by which our novel humidification system enhances plant growth and fruit yield to optimize its management.

Acknowledgements

The authors thank the members of group 3, Western Region Operation Unit 2, Technical Support Center of Western Regions, NARO, Dr. Matsuda, Dr. Fujioka, Dr. Kishi, Mrs. Shikata, and Mrs. Ono for assistance with cultivation, research, and provision of valuable suggestions. We thank Robert McKenzie, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.

Literature Cited

•  Buck,  A. L. 1981. New equations for computing vapor pressure and enhancement factor. J. Appl. Meteorol. 20: 1527–1532.
•  Choi,  H. G. and  H. J.  Jeong. 2020. Comparison of chlorophyll fluorescence and photosynthesis of two strawberry cultivars in response to relative humidity. Hortic. Sci. Technol. 38: 66–77.
•  Ehret,  D. L. and  L. C.  Ho. 1986. Translocation of calcium in relation to tomato fruit growth. Ann. Bot. 58: 679–688.
•  Grossiord,  C.,  T. N.  Buckley,  L. A.  Cernusak,  K. A.  Novick,  B.  Poulter,  R. T. W.  Siegwolf,  J. S.  Sperry and  N. G.  McDowell. 2020. Plant responses to rising vapor pressure deficit. New Phytol. 226: 1550–1556.
•  Islam,  N.,  G. G.  Patil,  S.  Torre and  H. R.  GislerØd. 2004. Effects of relative air humidity, light and calcium fertilization on tipburn and calcium content of the leaves of Eustoma grandiflorum (Raf.) Shinn. Europ. J. Hort. Sci. 69: 29–36.
•  Ito,  K. and  T.  Tabei. 2021. Model predictive temperature and humidity control of greenhouse with ventilation. Procedia Comput. Sci. 192: 212–221.
• Japan Meteorological Agency. 1998. Kisho kansoku no tebiki. <https://www.jma.go.jp/jma/kishou/know/kansoku_guide/tebiki.pdf>
•  Jia,  Y. 2021. Design of an Intelligent Greenhouse Remote Control System Based on a fuzzy neural network. Int. J. Autom. Technol. 15: 243–248.
•  Johnson,  R. W.,  M. A.  Dixon and  D. R.  Lee. 1992. Water relations of the tomato during fruit growth. Plant Cell Environ. 15: 947–953.
• Jones, H. G. 1993. Drought tolerance and water-use efficiency. p. 193–203. In: J. A. C. Smith and H. Griffiths (eds.). Water Deficit. BIOS Scientific Publishers, Oxford.
•  Kato,  K.,  K.  Kobayashi,  C.  Shimamoto,  Y.  Nakamura,  H.  Kojima,  T.  Oyabu,  Y.  Ban and  Y.  Iwasaki. 2015. Effect of mist spray and long-time CO₂ enrichment on growth and yield of strawberry in forcing culture. Res. Bull. Aichi Agric. Res. Ctr. 47: 51–60 (In Japanese with English abstract).
•  Lieten,  P. 2002. The effect of humidity on the performance of greenhouse grown strawberry. Acta Hortic. 567: 479–482.
•  Linker,  R.,  M.  Kacira and  A.  Arbel. 2011. Robust climate control of a greenhouse equipped with variable-speed fans and a variable-pressure fogging system. Biosyst. Eng. 110: 153–167.
•  Ninomiya,  S.,  Y.  Mochizuki,  M.  Fuke,  K.  Okayasu,  M.  Yamada,  S.  Suzuki and  I.  Ogiwara. 2012. Improvement of dendrometer for monitoring diurnal changes in longitudinal diameter of strawberry fruit. Hort. Res. (Japan) 11: 545–551 (In Japanese with English abstract).
•  Oishi,  N. 2002. Development of irrigation control system in response to plant water stress in tomato hydroponics (1)—non-destructive evaluation of plant water stress by the change of stem diameter—. Environ. Control in Biol. 40: 81–89 (In Japanese with English abstract).
•  Sakurai,  N. 1994. Change in cell walls during growth. Chemical Regulation of Plants 29: 65–78 (In Japanese).
•  Savitzky,  A. and  M. J. E.  Golay. 1964. Smoothing and differentiation of data by simplified least squares procedures. Anal. Chem. 36: 1627–1639.
•  Shpialter,  L.,  D. R.  David,  I.  Dori,  U.  Yermiahu,  S.  Pivonia,  R.  Levite and  Y.  Elad. 2009. Cultural methods and environmental conditions affecting gray mold and its management in lisianthus. Phytopathology 99: 557–570.
•  Sturges,  H. A. 1926. The choice of class interval. J. Am. Stat. Assoc. 21: 65–66.
•  Su,  Y.,  Q.  Yu and  L.  Zeng. 2020. Parameter self-tuning PID control for greenhouse climate control problem. IEEE Access 8: 186157–186171.
• Tagawa, A., K. Hidaka, A. Okamoto, C. Ogata, M. Ehara and H. Nakajima. 2016. Kotonaru sotai-situdo joken-ka ni okeru ichigo ‘Sagahonoka’ no Kogosei-sokudo. Abstr. Japanese Society of Agricultural, Biological and Environmental Engineers and Scientists 2016 nen Kanazawa taikai: 276–277 (In Japanese).
•  Yamanaka,  R.,  T.  Wada,  H.  Furukawa,  M.  Tojo,  N.  Hirai and  Y.  Kitaya. 2020. Analysis of the property of fruit growth and the evaluation of sugar flux thorough a cut section of fruit stalk in strawberry grown under and artificial light. Trans. of the JSRAE 37: 1–10 (In Japanese with English abstract).
• Yamanaka, R., T. Yano, H. Yoshikoshi, M. Hikawa-Endo and T. Wada. 2022. Control system of water stress of plants. Unexamined Japan Patent 2022-018479.
•  Yamanaka,  R.,  T.  Yano,  M.  Hikawa-Endo,  H.  Yoshikoshi,  H.  Kawashima,  M.  Tojo and  T.  Wada. 2024. Effects of humidification based on vapor pressure deficit (VPD) on plant growth, fruit yield, and fruit quality traits in June-bearing strawberry. Hort. J. 93: 377–388.
• Yoshida, Y. and K. Yasuba. 2020. YoshiMax beginner’s manual. <http://www.okayama-u.ac.jp/user/vegetres/Beginers_for_Yoshimax_2020_Feb.pdf>
•  Yunis,  H.,  Y.  Elad and  Y.  Mahrer. 1990. Effects of air temperature, relative humidity and canopy wetness on gray mold of cucumbers in unheated greenhouses. Phytoparasitica 18: 203–215.
• WARC/NARO. 2021. Standard operation protocol for remote monitoring system for air temperature. <https://www.naro.go.jp/publicity_report/publication/files/SOP20-060K20210416.pdf>