Effect of Training Method to Separate Sink and Source Organs on the Fresh Yield and Dry Matter Production of Cucumber (Cucumis sativus L.) under Greenhouse Conditions

Abstract

In Japan, greenhouse cucumber production is generally conducted as either a long-term, one cropping system or a short-term, two cropping system; short-term cultivation requires time for crop replacement, which may result in lower yields. It is important to increase plants’ intercepted light to increase yield, which requires maintaining a high leaf area index. In this study, a new training method (separating method), in which the sink and source are separated, was developed for the short-term cultivation of cucumber (Cucumis sativus L.) under greenhouse conditions. In the separating method, the main stem had only leaves and the lateral branch had only fruits. The main stem was trained vertically and pinched when it reached a 1.8 m high training wire, while the lateral branch continued to grow horizontally at approximately 40 cm above the floor. A hydroponic system was used from October 2020 to February 2021. During this period, we assessed the yield productivity of the new training method and compared it with that of the lowering method (training method for growing the main stem without pinching) based on yield components. We observed that the fresh yield of the separating method was lower than that of the lowering method owing to a decrease in the number of fruits as result of fruit picking from the main stem. Conversely, total dry matter under the separating method was higher than that under the lowering method, and this could be attributed to the greater increase in leaf area index under the separating method. In terms of photosynthetic rate, values for the upper leaves were lower than those for the lowering method in the separating method approximately one month after pinching. These results suggested that the separating method could produce yields comparable to those of the lowering method in short-term cultivation. In addition, this method may contribute to mechanical harvesting because the fruit is always in a fixed location.

Introduction

Cucumber (Cucumis sativus L.) is one of the most cultivated vegetables worldwide. In 2021, its global production was approximately 87.8 million tons (Food and Agricultural Organization, 2021) and 539,200 tons were produced in Japan. Data also shows that cucumber ranks second in the list of the most produced vegetable fruit globally in terms of total production, after tomatoes (Ministry of Agriculture, Fisheries, and Forestry, 2021a). In terms of productivity, the annual cucumber yield in Japan is 14.1 kg·m⁻², a much lower value compared to the 72.8 kg·m⁻² reported for the Netherlands. Moreover, cucumber production requires between 420 and 2,500 labor hours per 1,000 m² per year, and the majority of this labor time is dedicated to harvesting (Ministry of Agriculture, Fisheries, and Forestry, 2021b). Therefore, to maintain Japanese cucumber production, it is important to develop strategies to increase yield and reducing labor time. However, Japanese cucumbers have been produced using unique cultivation methods (e.g., pinching and steaming) and have strict standards in terms of fruit characteristics (e.g., length, 22 cm; weight, 100 g). These factors are possibly responsible for the low yields and increased labor time associated with cucumber production in Japan (Maeda and Ahn, 2021a). To address these issues, studies have been conducted in recent years on hydroponics and lowering methods (Higashide et al., 2012; Maeda and Ahn, 2021b). Notably, the hydroponically cultivated area for cucumber in Japan has more than doubled between 2016 and 2018 (Ministry of Agriculture, Fisheries, and Forestry, 2021c).

In Japan, cucumber cultivation involves two major cultivation methods, the pinching and lowering methods. The pinching method is predominantly used owing to the structural limitations associated with greenhouses, which have low eaves (Maeda and Ahn, 2021a), and is practiced particularly during the summer season. With this method, lateral branches developing from axillary buds are repeatedly pinched to increase their numbers; this plays an important role in increasing yield. Furthermore, the pinching is performed while controlling the growth vigor of the cucumber plant, and this practice requires a high skill level. Conversely, with the lowering method, which is practiced in the fall and winter seasons, the main stem or lateral branches (1 to 4) are maintained and trained on wires set at a height of approximately 1.5 to 2 m. Then, when the tips of the stems are within 0 to 30 cm of the overhead wires, they are continuously lowered (Samba et al., 2023). Furthermore, this cultivation technique is often used in industrial-scale production as in the case of large greenhouses given that it requires less expert knowledge in terms of training. Therefore, the lowering method has been increasingly utilized recently, and several studies on it have been conducted (Hirama et al., 2011; Isomura et al., 2001; Ota et al., 2005).

Canopy structure manipulation based on appropriate spatial arrangements through training is recognized as an important management practice for maximizing marketable yields from greenhouse crops (Premalatha et al., 2006). For cucumber plants, canopy structure can be manipulated by controlling plant density (Bayat et al., 2021; Kapuriya et al., 2017), transplanting date (Raveena et al., 2023), cultivars (Bayat et al., 2021; Chacon-Padilla and Monge-Perez, 2020) and selecting training methods (Jadhav et al., 2023; Kapuriya et al., 2017; Kumar et al., 2018; Premalatha et al., 2006; Raveena et al., 2023; Samba et al., 2023). In general, changes in canopy structure affect the amount of light reception (Kaneko et al., 2015; Maeda and Ahn, 2022), number of fruits, and light use efficiency (LUE) (Higashide et al., 2012; Iwasaki et al., 2014; Ota et al., 2005). However, in these previous studies, the effects of canopy structure on yield were not quantitatively analyzed. Higashide et al. (2012) analyzed the effects of the lowering and pinching methods on dry matter production and yield in short-term cucumber cultivation in terms of yield components. They observed a higher yield for pinching cultivation than for lowering cultivation and attributed this difference to the higher proportion of fruit dry matter and total dry matter resulting from the higher leaf area index (LAI) under the pinching method. Furthermore, a comparison study of the pinching and lowering methods by Samba et al. (2023) indicated that the pinching method is suitable for short-term cucumber cultivation (3–4 months), while the lowering method is more suitable for long-term cultivation (5–6 months). However, contrasting findings have been reported. For example, a comparison of the harvest time between the lowering and pinching methods by Isomura et al. (2001) revealed that the time required per fruit was shorter for the lowering method than for the pinching method (8.0–11.7 and 8.8–12.3 s for the lowering and pinching methods, respectively), and the that lowering method was less labor-intensive. Thus, further studies are required.

In this study, we developed a novel training method (separating method) that integrates the advantages of the high-yield pinching method and the less labor-intensive lowering method, and investigated its performance in short-term cucumber cultivation. Two nodes were spread in one plant, one with only fruit and the other with leaves, so that the fruit always remained in the same position. Additionally, to evaluate the yield potential of the new training method, we compared it with the lowering method in terms of yield components.

Materials and Methods

Plant material and growth conditions

The experiment was carried out at the National Agriculture and Food Research Organization (Tsukuba, Ibaraki, Japan). On October 20, 2020, cucumber (Cucumis sativus L. S-27Z; Saitama Gensyu Ikuseikai Co., Ltd., Saitama, Japan) seeds were sown in 72-cell plug-trays filled with commercial substrate containing peat moss, vermiculite, and perlite (Takii Cell Media TM-1; Takii & Co., Ltd., Kyoto, Japan). After germination under dark conditions at 28°C for 1 d, the trays were placed in a growth chamber (NAE Terrace; Mitsubishi Chemical Agri Dream Co., Ltd., Tokyo, Japan) under a 14-h/10-h light/dark period, with the light/dark temperature, CO₂ concentration, and light intensity set at 25/19°C, 1,000 μmol·mol⁻¹, and 400 μmol·m⁻²·s⁻¹, respectively, for 14 d. The plants were supplied daily with a commercial nutrient solution (High-Tempo; Sumitomo Chemicals, Tokyo, Japan) with the following composition: 10.7 mM NO₃⁻, 6.3 mM K⁺, 5.4 mM Ca²⁺, 1.9 mM Mg²⁺, 7.2 mM H₂PO₄⁻, 3.8 mg·L⁻¹ Fe, 0.38 mg·L⁻¹ Mn, 0.26 mg·L⁻¹ B, 0.15 mg·L⁻¹ Zn, 0.05 mg·L⁻¹ Cu, and 0.07 mg·L⁻¹ Mo. At the end of this 14-d period, the seedlings were placed in rockwool slabs (1,000 × 150 × 75 mm³) in a greenhouse (7.15 m width, 16 m length, and 4.0 m height), covered with a polyolefin film, and supplied with a commercial nutrient solution (OAT House fertilizer with modified-SA prescription; OAT Agrio Co., Ltd., Tokyo, Japan), consisting of 12.2 mM NO₃⁻, 7.1 mM K⁺, 2.8 mM Ca²⁺, 1.0 mM Mg²⁺, 3.0 mM H₂PO₄⁻, 1.6 mg·L⁻¹ Fe, 0.8 mg·L⁻¹ Mn, 0.4 mg·L⁻¹ B, 0.06 mg·L⁻¹ Zn, 0.02 mg·L⁻¹ Cu, and 0.02 mg·L⁻¹ Mo with an electrical conductivity of 1.8 dS·m⁻¹. Four double rows (12 rockwool slabs per row) were divided into two blocks with the south side using the separating method and the north side using the lowering method. The planting density was 3.75 plants·m⁻² and the side row plants functioned as guard plants.

In the lowering method, the plants were trained with a single stem, and all lateral branches and lower leaves were pruned weekly (Fig. 1B, D, F). In the novel training method (separating method), the main stem was trained onto an overhead wire (1.8 m from the cultivation bed), and all the flowers were removed (Fig. 1A). On December 14, 2020, the main stem reached the overhead wire and was pinched (Fig. 1E). Only the yellowed lower leaves of the main stem were removed. The lateral branches were elongated from the third node of the main stem and trained horizontally so that the fruit was positioned approximately 40 cm above the floor (Fig. 1C), while all other lateral branches were removed. Subsequently, for each elongated lateral branch, all the leaves and lateral branches were removed, leaving only flowers and fruit. Harvesting was carried out every day, and fruits approximately 22 cm in length were harvested.

Fig. 1

Different cucumber training methods. (A), (C), and (E): The separating method on December 22, 2020. (B), (D) and (F): The lowering method on December 22, 2020.

The greenhouse environment was controlled using a ubiquitous environmental control method (DIY Environmental Control Method; WaBit Inc., Tokyo, Japan) (Fig. 2). The side window ventilation and air heater were operated when the temperature was higher than 30°C and lower than 15°C, respectively.

Fig. 2

Changes in environmental parameters. (A) Daily average temperature and integrated solar radiation and (B) Daily average relative humidity, and average daytime CO₂ concentration during the cultivation period.

Measurement of plant growth and yield components

Leaf area (LA) was calculated in two ways: the first was to estimate LA from the largest individual leaf area, which was used throughout the growing season for the lowering method. In the separating method, LA was calculated by measuring all individual leaves on the plant from January 4, because the area of individual leaves in the upper position increased after pinching. Until January 3, LAI for both training methods was estimated with reference to a previous study (Maeda and Ahn, 2021b) as follows:

  

$$\begin{array}{l}\text{LA at each node}\\ =\mathrm{a}(\mathrm{MLl}\times \mathrm{MLw})\left(\frac{0.896}{1+\left(14.811\mathrm{ }{\mathrm{e}}^{\mathrm{-0.51}\mathrm{n}}\right)}\right)\end{array}$$(1)

where “MLl” represents maximum leaf length and “MLw” represents maximum leaf width of each plant. The value of “a” (0.91 in this study) was calculated based on the values of MLl, MLw, and the measured individual leaf areas (n = 15). “n” represents leaf node position.

  

$$\mathrm{LAI}=\sum _{k=1}^n{\mathrm{LA}}_k·\mathrm{PD}$$(2)

where “PD” represents plant density. The LAI of the plants under the separating methods was determined from January 4, 2021, as previously described (Cho et al., 2007) by measuring all individual leaf areas as follows:

  

$$\mathrm{LAI}=\mathrm{a}\sum _{k=1}^n({\mathrm{LI}}_k\times {\mathrm{Lw}}_k)·\mathrm{PD}$$(3)

where “Ll” represents leaf length and “Lw” represents leaf width of each individual leaf. Daily LAI was obtained via the linear interpolation of the LAI calculated from the measurements.

Six plants (three plants per row) were removed five times on November 13, 2020, December 3 and 24, 2020, January 21, 2021, and February 18, 2021. Thereafter, stem, leaf, and immature fruits were collected and dried for one week at 95°C. Then, to determine fruit dry matter content, 20 fruits for each of the two training methods were dried for one week at 105°C. Total dry matter production (TDM) was then calculated by adding the dry weights from the destructive measurements and fruits harvested.

To evaluate the yield difference between the two training methods based on yield components (light extinction coefficient [k], intercepted light [IL], LUE), we used previously reported methods (Maeda and Ahn, 2022). The k value was measured on a sunny day on December 1, 2020, during the hours from 10:00 am to 12:00 pm using the following equation:

  

$$I=I_0{\mathrm{e}}^{-k·\mathrm{LAI}}$$(4)

where I represents the photosynthetic photon flux density (PPFD) at various horizontal heights in the plant canopy, measured using a line quantum sensor (LI-191SA; LI-COR Biosciences, Lincoln, NE, USA) at three different heights and with two different sensor insertion methods (insertion from the east and west side into the center of the plant canopy). I₀ represents the PPFD above the canopy, measured using a PPFD sensor (LI-190SA; LI-COR Biosciences).

For some destructive measurements (n = 6 per treatment), the cumulative IL up to the time of the destructive measurements and the TDM were plotted on the horizontal and vertical axes, respectively, and the slope of the line was defined as LUE (g·MJ⁻¹).

The daily cumulative photosynthetically active radiation (MJ · m⁻²) for each plant was calculated using the following equation:

  

$$\mathrm{IL}=\left(1-{\mathrm{e}}^{-k·\mathrm{LAI}}\right)\times \mathrm{PAR}$$(5)

where PAR represents indoor solar radiation determined using a facility light transmittance of 50%, calculated from the measured outdoor solar radiation using a pyranometer, and recorded at 1-min intervals. The PAR to solar radiation ratio was assumed to be 50% (Ohtani, 1997). Subsequently, daily LAI was obtained by linearly interpolating the LAI obtained from three destructive measurements.

The photosynthetic rates of the 7th and 14th leaves were measured twice from December 11 to 16, 2020 and January 18 to 22, 2021 using the LI-6800 portable photosynthesis method (LI-COR, Lincoln, NE, USA). The measurements were conducted in triplicate in a leaf chamber under the following conditions: air temperature, 28°C; PPFD, 1000 μmol·m⁻²·s⁻¹; CO₂ concentration, 650 μmol·mol⁻¹; relative humidity, 60–70%.

Statistical analyses

All statistical analyses were performed using the R software (R Foundation for Statistical Computing, Vienna, Austria). To analyze differences between groups, we performed the Student’s t-test. Statistical significance was set at P < 0.05.

Results

The environmental data collected in this study are shown in Figure 2. The average temperature and outside irradiation were 20.0°C and 9.49 MJ·m⁻², respectively, during the cultivation period (Fig. 2A). Furthermore, the relative humidity and CO₂ concentration in this period varied in the ranges of 68–80% and 570–780 μmol·mol⁻¹, respectively, except for November (Fig. 2B).

Fruit harvest started on December 10 for the lowering method and on December 16 for the separating method. The variations in cucumber growth with the training method are shown in Table 1. At the end of the cultivation period, the number of fruits and fresh fruit weight under the lowering method were significantly higher than those obtained under the separating method (P < 0.001). However, the separating method showed a TDM value that was 15% higher than that obtained under the lowering method (P < 0.001). Furthermore, regarding dry matter distribution, no significant difference was observed between the two training methods in terms of stem dry matter distribution; however, the lowering method showed 9.9% lower and 9.8% higher leaf and fruit dry matter distribution, respectively, than the separating method. The light extinction coefficient (k) for morphological characteristics did not differ between the training methods. Furthermore, the IL value calculated based on LAI and k values for the separating method was 4.4% higher than that obtained for the lowering method.

Table 1

Number of fruits per plant, fresh fruit weight, total dry matter (TDM), dry matter distribution in leaf, stem, and fruit, light extinction coefficient, and intercepted light under the different training methods.

In plants cultivated under the separating method, LAI increased to approximately 60 DAT and then decreased (Fig. 3). Additionally, before 40 DAT, the values were almost the same for both training methods, and thereafter, the values varied in ranges of approximately 3.0–4.0 and 2.5–3.5 LAI under the separating and lowering methods, respectively.

Fig. 3

Change in leaf area index (LAI) under the different training methods during the cultivation period. Black and white circles represent the separating and lowering method, respectively.

Figure 4 shows changes in LUE values, which are indicative of the regression line between TDM and IL for the plants under the different training methods. The LUE values were 6.4 and 6.0 under the separating and lowering methods, respectively (R² > 0.99, P < 0.01). However, there was no significant difference between the two training methods in terms of 95% confidence intervals. Photosynthetic rates measured from December 11 to 16, 2020 (38–43 DAT) did not differ between the two training methods (Fig. 5). However, the values corresponding to the upper positions of the plants measured from January 18 to 22, 2021 (76–80 DAT) were higher under the lowering method than those under the separating method.

Fig. 4

Total dry matter as a function of cumulative intercepted light under different cucumber training methods. The slopes of the regression lines represent light use efficiency. Black and white circles represent the separating and lowering method, respectively.

Fig. 5

Photosynthetic rates of the 7th (upper) and 14th (lower) leaves under the different training methods. Values measured from (A) December 11 to 16, 2020 and (B) January 18 to 22, 2020. NS and *** indicate not significant and significant differences for P < 0.001, respectively, determined by Student’s t-test. Error bars indicate standard error. Gray and white bars represent the separating and lowering method, respectively.

Discussion

In this study, we observed that within the 3.5-month cultivation period, the productivity of the cucumber plants under the separating method in the low irradiance season was 12.9 kg·m⁻². This corresponds to at least 38.7 kg·m⁻² of year-round production (three cultivation times). The yield of the plants under the separating method was lower than that of plants under the lowering method. This could be primarily attributed to the lower dry matter distribution in fruit under the separating method owing to a decrease in the number of fruits. In fact, under the separating method, all the flowers in the main stem were removed, and this delayed the start of harvest by almost a week (data not shown) and reduced the number of fruits per plant by approximately 5.5 (Table 1). Furthermore, 7.7 fruits were harvested from the main stem under the lowering method until pinching under the separating method, suggesting that if the flowers on the main stem were not removed, the number of fruits may have been equal to or higher than that obtained under the lowering method. Additionally, under the separating method, there was a longer distance between the sink and source; however, the influence of the transport route on dry matter distribution may have been limited, as reported for tomatoes (Heuvelink, 1996; Slack and Calvert, 1977).

TDM can be expressed in terms of integrated IL and LUE values (Higashide et al., 2012). The k values for calculating IL have been treated as constants in several fruit vegetables in previous studies (Kaneko et al., 2015; Maeda and Ahn, 2021b, 2022; Watabe et al., 2021), and were treated in the same way in this study. Iwasaki et al. (2014) compared the effects of two training methods (pinching and lowering methods) on several cucumber cultivars and concluded that the TDM is correlated with IL at 40 DAT. Moreover, Higashide et al. (2012) reported higher TDM under the pinching method than under the lowering method due to a higher IL. In this study, we also observed higher TDM and LAI values under the separating than under the lowering method. However, there was no significant difference between the two methods in terms of LUE, suggesting that the observed difference in TDM was due to differences in IL between the two methods. Additionally, there was no significant difference between the two methods in terms of the light extinction coefficient, suggesting that the observed differences in IL were due to differences in LAI.

The photosynthetic rate in cucumber increases rapidly during leaf unfolding, reaching a peak when the leaves are fully expanded, then decreasing during the late developmental stage (Marcelis, 1991; Piñero et al., 2021). In this study, we also observed that the photosynthetic rate in the upper position under the separating method was lower than that observed under the lowering method when measured at 76–80 DAT (35–39 days after pinching), where the leaf stage differed between treatments; however, LUE did not differ between the two training methods. These results suggested that during the cultivation period in this study (approximately 3.5 months), the photosynthesis rate did not decrease to a level that resulted in a decrease in dry matter production capacity without leaf renewal. We also noted that during the cultivation period, LAI decreased from 60 DAT under the separating method. A decrease in the photosynthesis rate was also observed at 76–80 DAT, suggesting that an extended cultivation period resulted in a decrease in productivity. Therefore, this training method is more suited for short-term cultivation than long-term cultivation and can be used approximately three times a year to obtain high yields. In fact, repeated cultivation during the year also allows for weather-appropriate variety selection and reduces the risk of pests and diseases (Ayarna et al., 2021; Kinoshita et al., 2014; Maeda et al., 2020; Nakayama et al., 2021).

To reduce labor costs, automatic harvesting is increasingly being used, with the technologies applied facing challenges such as fruit detection and localization using sensor-based computer vision (Onishi et al., 2019). Several studies have been conducted to analyze fruit detection methods, such as image acquisition using sensors and cameras, and machine learning (Bulanon et al., 2002; Kurtulmus et al., 2014; Okamoto and Lee, 2009; Rakun et al., 2011). For cucumber in particular, studies have shown that automatic harvesting can be highly accurate (Mao et al., 2020; Van Henten et al., 2002; Zhang et al., 2007). However, these studies were conducted in high-wire cultivation systems, where the fruits are easily visible. Conversely, in Japan, most greenhouses have low-height eaves, and this may reduce the identification accuracy due to limited fruit visibility. Therefore, an inclined stem training system was developed for robotic harvesting of cucumber in Japan (Arima et al., 1996). However, information regarding robotic harvesting under the training methods used in this study is limited. Nevertheless, our results showed that the separating method may be more suitable for fruit detection, and thus robotic harvesting, as it resulted in fruits always growing in the same position (Fig. 1). Moreover, under the separating method, the fruit stems were horizontal to the blanch, making them suitable for harvesting with a robotic arm. In the near future, we plan to evaluate differences in harvest time between the separating and lowering methods using a robot.

In this study, our results showed a higher TDM and a lower yield for the separating method than for the lowering method. However, the yield of the separating method could have been equal to, or better than, that of the lowering method if the female flowers on the main stems of plants under the separating method had not been removed. We did not verify the recognition accuracy or speed of a harvesting robot for the separating method. Thus, further studies are required to measure the actual labor time associated with the separating method.

Literature Cited

•  Arima,  S.,  N.  Kondo and  H.  Nakamura. 1996. Development of robotic system for cucumber harvesting. JARQ 30: 233–238.
•  Ayarna,  A. W.,  S.  Tsukagoshi and  G. O.  Nkansah. 2021. Effect of root restriction on the performance of three-truss cultivated tomato in the low-node pinching order at high-density cultivation system. Horticulturae 7: 60. DOI: 10.3390/horticulturae7030060.
•  Bayat,  B.,  M. H.  Ansari,  M.  Diyanat and  A. M.  Torkashvand. 2021. Evaluation of the effect of planting distance on row on growth and yield of different cultivars of greenhouse cucumber (Cucumis sativus L.). J. Veg. Sci. 9: 77–91.
•  Bulanon,  D. M.,  T.  Kataoka,  Y.  Ota and  T.  Hiroma. 2002. AE—automation and emerging technologies: a segmentation algorithm for the automatic recognition of Fuji apples at harvest. Biosyst. Eng. 83: 405–412.
•  Chacon-Padilla,  K. and  J. E.  Monge-Perez. 2020. Cucumber (Cucumis sativus L.) production under greenhouse conditions: comparison between cucumber types. Tecnología en Marcha 33: 17–35 (In Spanish with English abstract).
•  Cho,  Y. Y.,  S.  Oh,  M. M.  Oh and  J. E.  Son. 2007. Estimation of individual leaf area, fresh weight, and dry weight of hydroponically grown cucumbers (Cucumis sativus L.) using leaf length, width, and SPAD value. Sci. Hortic. 111: 330–334.
• Food and Agricultural Organization. 2021. http://www.fao.org/faostat/en/#data/QC (assessed on 16 May 2021).
•  Heuvelink,  E. 1996. Re-interpretation of an experiment on the role of assimilated transport resistance in partitioning in tomato. Ann. Bot. 78: 467–470.
•  Higashide,  T.,  I.  Gotoh,  K.  Suzuki,  K.  Yasuba,  K.  Tsukazawa,  D.  Ahn and  Y.  Iwasaki. 2012. Effects of pinching and lowering on cucumber yield and yield components. Hort. J. 11: 523–529.
•  Hirama,  N.,  H.  Mizusawa and  F.  Azuhata. 2011. Effects of different levels of greenhouse ventilation and training methods on cucumber growth and yield under forcing culture. Hort. Res. (Japan) 10: 499–505 (In Japanese with English abstract).
•  Isomura,  M.,  Y.  Matsugaki,  Y.  Funakoshi,  K.  Tokumaru,  K.  Tanaka and  Y.  Araki. 2001. Techniques for high yield and quality in cucumbers cultured in greenhouses by vertical training. Bull. Oita Prefectural Agricultural Research Center 31: 31–39 (In Japanese).
•  Iwasaki,  Y.,  D.  Ahn,  K.  Shimomura,  T.  Higashide and  A.  Nakano. 2014. Optimizing training method and environmental factors to improve the growth and yield of cucumbers. Bull. Natl. Inst. Veg. Tea Sci. 13: 65–75.
•  Jadhav,  A.,  R. D.  Pawar,  S. S.  Dhumal,  V. K.  Garande,  A. S.  Bagade,  S. A.  Sarvade and  U. S.  Shinde. 2023. Effect of different training systems on growth and yield of greenhouse cucumber. Pharm. Innov. J. 12: 2820–2824.
•  Kaneko,  S.,  T.  Higashide,  K.  Yasuba,  H.  Ohmori and  A.  Nakano. 2015. Effects of planting stage and density of tomato seedlings on growth and yield component in low-truss cultivation. Hort. Res. (Japan) 14: 163–170 (In Japanese with English abstract).
•  Kapuriya,  V. K.,  K. D.  Ameta,  S. K.  Teli,  A.  Chittora,  S.  Gathala and  S.  Yadav. 2017. Effect of spacing and training on growth and yield of polyhouse grown cucumber (Cucumis sativus L.). Int. J. Curr. Microbiol. App. Sci. 6: 299–304.
•  Kinoshita,  T.,  T.  Yano,  M.  Sugiura and  Y.  Nagasaki. 2014. Effects of controlled-release fertilizer on leaf area index and fruit yield in high-density soilless tomato culture using low node-order pinching. PloS One 9: e113074. DOI: 10.1371/journal.pone.0113074.
•  Kumar,  S.,  N. B.  Patel and  S.  Saravaiya. 2018. Influence of fertigation and training systems on yield and other horticultural traits in greenhouse cucumber. Indian J. Hortic. 75: 252.
•  Kurtulmus,  F.,  W. S.  Lee and  A.  Vardar. 2014. Immature peach detection in color images acquired in natural illumination conditions using statistical classifiers and neural network. Precis. Agric. 15: 57–79.
•  Maeda,  K. and  D.  Ahn. 2021a. A review of Japanese greenhouse cucumber research from the perspective of yield components. Hort. J. 90: 263–269.
•  Maeda,  K. and  D.  Ahn. 2021b. Estimation of dry matter production and yield prediction in greenhouse cucumber without destructive measurements. Agriculture 11: 1186. DOI: 10.3390/agriculture11121186.
•  Maeda,  K.,  M.  Johkan,  S.  Tsukagoshi and  T.  Maruo. 2020. Effect of salinity on photosynthesis and distribution of photosynthates in the Japanese tomato ‘CF Momotaro York’ and the Dutch tomato ‘Endeavour’ with low node-order pinching and a high-density planting system. Hort. J. 89: 454–459.
•  Maeda,  K. and  D.  Ahn. 2022. Analysis of growth and yield of three types cucumbers (Cucumis sativus L.) based on yield components. Horticulturae 8: 33. DOI: 10.3390/horticulturae8010033.
•  Mao,  S.,  Y.  Li,  Y.  Ma,  B.  Zhang,  J.  Zhou and  W.  Kai. 2020. Automatic cucumber recognition algorithm for harvesting robots in the natural environment using deep learning and multi-feature fusion. Comput. Electron. Agric. 170: 105254. DOI: 10.1016/j.compag.2020.105254.
•  Marcelis,  L. F. 1991. Effects of sink demand on photosynthesis in cucumber. J. Exp. Bot. 42: 1387–1392.
• Ministry of Agriculture, Fisheries, and Forestry. 2021a. https://www.maff.go.jp/j/tokei/kouhyou/sakumotu/sakkyou_yasai/attach/pdf/index-4.pdf (assessed on 16 May 2021).
• Ministry of Agriculture, Fisheries, and Forestry. 2021b. https://www.maff.go.jp/j/seisan/kakou/yasai_kazitu/attach/pdf/kakouyasai-2-.pdf (assessed on 16 May 2021).
• Ministry of Agriculture, Fisheries, and Forestry. 2021c. https://www.maff.go.jp/j/seisan/ryutu/engei/sisetsu/haipura/setti_30.html (assessed on 16 May 2021).
•  Nakayama,  M.,  S.  Fujita,  Y.  Watanabe,  T.  Ando,  M.  Isozaki and  Y.  Iwasaki. 2021. The effect of greenhouse cultivation under a heat insulation film covering on tomato growth, yield, and fruit quality in a subtropical area. Hort. J. 90: 304–313.
• Ohtani, Y. 1997. Effective Radiation, Micrometeorological Phenomena. p. 106–107. In: T. Maki, S. Iwata, Z. Uchijima, T. Oikawa, K. Omasa, K. Kurata, T. Kozai, E. Goto, E. H. Kon, I. Nouchi, et al. (eds.). Agricultural Meteorology Glossary; Society for Agricultural Meteorology of Japan. Tokyo, Japan.
•  Okamoto,  H. and  W. S.  Lee. 2009. Green citrus detection using hyperspectral imaging. Comput. Electron. Agric. 66: 201–208.
•  Onishi,  Y.,  T.  Yoshida,  H.  Kurita,  T.  Fukao,  H.  Arihara and  A.  Iwai. 2019. An automated fruit harvesting robot by using deep learning. ROBOMECH J. 6: 1–8.
•  Ota,  T.,  H.  Seino and  N.  Inoue. 2005. New method of training for cucumber in greenhouse. Bull. Saitama Prefectural Agriculture and Forestry Research Center 4: 79–83.
•  Piñero,  M. C.,  G.  Otálora,  J.  López-Marín and  F. M.  Del Amore. 2021. Nitrogen management under increased atmospheric CO₂ concentration in cucumber (Cucumis sativus L.): ameliorating environmental impacts of fertilization. Sci. Rep. 11: 1–8.
•  Premalatha,  M. G. S.,  K. B.  Wahundeniya,  W. A. P.  Weerakkody and  C. K.  Wicramathunga. 2006. Plant training and spatial arrangement for yield improvements in greenhouse cucumber (Cucumis sativus L.) varieties. Trop. Agric. Res. 18: 346–357.
•  Rakun,  J.,  D.  Stajnko and  D.  Zazula. 2011. Detecting fruits in natural scenes by using spatial-frequency based texture analysis and multiview geometry. Comput. Electron. Agric. 76: 80–88.
•  Raveena,  K.,  V.  Sharma,  A.  Pooja and  N.  Sarita. 2023. An examination of the effects of different transplanting dates, planting geometry and training techniques on the yield traits, productivity and quality of cucumber (Cucumis sativus L.) under naturally ventilated polyhouse. Int. J. Environ. Clim. 13: 4295–4301.
•  Samba,  N.,  O.  Nunomura,  A.  Nakano and  S.  Tsukagoshi. 2023. Effective training methods for cucumber production in newly developed nutrient film technique hydroponic system. Horticulturae 9: 478. DOI: 10.3390/horticulturae9040478.
•  Slack,  G. and  A.  Calvert. 1977. The effect of truss removal on the yield of early sown tomatoes. J. Hortic. Sci. 52: 309–315.
•  Van Henten,  E. J.,  J.  Hemming,  B.  Van Tuijl,  J.  Kornet,  J.  Meuleman,  J.  Bontsema and  E.  Van Os. 2002. An autonomous robot for harvesting cucumbers in greenhouses. Auton. Robots 13: 241–258.
•  Watabe,  T.,  M.  Homma,  D.  Ahn and  T.  Higashide. 2021. Examination of yield components and the relationship between dry matter production and fruit yield in greenhouse sweet pepper (Capsicum annuum). Hort. J. 90: 247–254.
•  Zhang,  L.,  Q.  Yang,  Y.  Xun,  X.  Chen,  Y.  Ren,  T.  Yuan,  Y.  Tan and  W.  Li. 2007. Recognition of greenhouse cucumber fruit using computer vision. New Zealand J. Agric. Res. 50: 1293–1298.