Introduction

The balloon flower (Platycodon grandiflorum [Jacq.] A. DC.), and lance asiabell (Codonopsis lanceolata), belonging to the family Campanulaceae, are native to Korea, Japan, and China; their roots are used as medicinal herbs and food ingredients (Lee et al., 2019; Lim, 1971). The root of the balloon flower is rich in saponins, such as platycodin A, C, D, polygalacin D, botulin, inulin, and phytosterol (Saeki et al., 1999). The roots of the lance asiabell also contain medicinal ingredients, such as saponin, inulin, phytoderin, and lecithin (Hwang et al., 2019). These ingredients are effective in detoxification, pulmonary phlegm, purulent discharge, expectoration, and sedation (RDA, 2019a). Balloon flower and lance asiabell were the most cultivated in hongcheon and jeju (Korea forest service, 2024). Although the domestic production volume of balloon flower and lance asiabell decreased by about 1.4 and 1.8%, respectively in 2023 compared to 2022, however the production value increased by more than 15.6 and 17.8%, respectively (Korea Forest Service, 2024). If the seedlings of these medicinal plants can be produced in large quantities, it could lead to increased income for farmers.

The medicinal plants have issues with inconsistent germination and low adaptability to environmental conditions, making it difficult to mass-produce uniform quality seedlings (Ghimire et al., 2006; Park et al., 2023; RDA, 2019b). Consequently, the practice of plug seedling production is receiving increased attention. Plug seedling production is a highly effective method for the mass production of high-quality standardized seedlings (Jeong, 1998). Plug seedling production offers several advantages, such as reduced seed consumption, improved space efficiency, and shorter seedling nursery periods. When applied to the cultivation of balloon flower and lance asiabell, plug seedling production can facilitate the large-scale production of superior seedlings (Calson et al., 1992). If high-quality standardized seedlings of lance asiabell and balloon flower are produced using advanced seedling cultivation technology and supplied to farms, it is expected that the quality of the produced medicinal plants could be standardized and the yield could be increased (Park et al., 2023). However, the narrow and densely packed cells of plug trays create a restricted growing environment. This density can result in shading between plants and inside of canopy, that leading to decreased photosynthesis and potentially causing overgrowth (Ruberti et al., 2012). Overgrown seedlings often have thinner stems than normal seedlings, making them more susceptible to mechanical damage, which leads farmers to typically avoid using them (Sun et al., 2010).

A closed-type plant production system (CPPS) enables year-round cultivation of uniform seedlings regardless of season or location through artificially modulated environmental conditions, including light, temperature, humidity, and CO₂ (Kozai et al., 2019). Furthermore, due to its ability to create optimal conditions for crops through artificial environmental control, which regulates their growth rate and suppresses overgrowth, active research is being conducted to apply CPPS for the seedling production of various horticultural crops, such as tomato, lettuce, and cucumber (Jeong et al., 2020; Ohyama et al., 2020; Yun et al., 2023). Furthermore, since research is being conducted on the seedling production and cultivation of medicinal crops such as centella and angelica in CPPS (Shawon et al., 2023; Lee et al., 2016), it is necessary to study the potential for seedling production and cultivation of balloon flower and lance asiabell in CPPS as well.

Light-emitting diodes (LEDs) are primarily and commonly used as artificial light sources for CPPS, greenhouses, and indoor farming (Kozai et al., 2019). LEDs can adjust the wavelength range to provide optimal light quality for specific crops (Morrow, 2008; Kozai, 2016). Furthermore, due to its high energy efficiency, minimal heat generation, and semi-permanent lifespan, LEDs are the most commonly used light source in a CPPS (Bula et al., 1991; Hwang et al., 2022). Plants use a wavelength range called photosynthetically active radiation (PAR, 400－700 nm) (McCree, 1981). Specific light quality can affect the seedling morphology, photosynthesis, flowering, and the production of secondary metabolites (Ouzounis et al., 2015). Red light (600－700 nm) is known to be a primary wavelength for photosynthesis, which can be applied to increase the efficiency of the photosynthesis process (McCree, 1981; Whitelam and Halliday, 2007). Green light (500－600 nm) is associated with low photosynthetic efficiency, inhibited growth, and closed stomata (McCree, 1981). However, when used in addition to red and blue mixed light, green light has been reported to be an important light that can increase photosynthesis in the lower leaves of the canopy and increase biomass because it is well transmitted and reflected (Nishio, 2000; Kim et al., 2004; Folta and Maruhnich, 2007). Similar to red light, blue light (400－500 nm) effectively supports photosynthesis (McCree, 1981; Zheng et al., 2018). However, blue light comprises high- energy photons and acts as a stressor on plants, increasing flavonoid and secondary metabolites in plants (Ebisawa et al., 2008; Whitelam and Halliday, 2007). Based on the wavelength, the impacts of light on plants vary; therefore, selecting the appropriate light quality in a CPPS that involves artificial light sources is important for optimizing plant growth, enhancing photosynthesis, and improving the quality and yield of crops.

The multispectral imaging technique is a technology that measures reflectance at various wavelengths, utilizing the characteristics that reflectance varies with chlorophyll, nutrient, and water content in plants (Gates et al., 1965). This technique allows non-destructive real-time investigation of crop conditions, water content, and nitrogen nutrient status, as well as the detection of fruit firmness, soluble solids content, and damage on the skin (Ban et al., 2023; Blasco et al., 2009; Lu, 2004; Wang et al., 2018). In particular, multispectral imaging has been primarily used in studies to assess field conditions or fruit quality by applying it to unmanned aerial vehicles (Blasco et al., 2009; Na et al., 2019). However, growth prediction using leaf area measurements of seedlings remains unclear.

The study was conducted to identify the light quality that can produce high-quality seedlings of balloon flower and lance asiabell by shortening the stem length and internode length in CPPS. Additionally, it seeks to evaluate the feasibility of applying multispectral imaging techniques to the seedlings of these medicinal plants. The findings are anticipated to make a significant contribution to the mass production of balloon flower and lance asiabell seedlings.

Materials and Methods

1. Plant materials and growth conditions

The balloon flower (Platycodon grandiflorum [Jacq.] A. DC. ‘Beak’, Asia Seed Co. Ltd., Seoul, Korea) and lance asiabell (Codonopsis lanceolata, Aram Seed Co. Ltd., Seoul, Korea) seeds were subjected to cold treatment at 4℃ for two weeks. Subsequently, seeds were disinfected by immersing them in 50 mg·L^-1 NaOCl for 10 minutes and rinsed four times using flowing tap water. Seeds were sown in 50-cell plug trays (cell size: 5 × 3 × 4.6 cm; Bumnong Co. Ltd., Jeongeup, Korea) filled with the commercial growing medium (Plant World, Nong Woo Bio, Suwon, Korea) on July 4, 2023. Seedlings were grown in the CPPS at 25 ± 0.6℃ under controlled conditions maintaining 55 ± 4.7% relative humidity, light intensity of 110 ± 5 μmol·m^-2·s^-1 photosynthetic photon flux density (PPFD), and photoperiod of 16/8 h (light/dark). Light intensity was measured at five points (the center and four edges) on the tray using a photometer (HD2102.2, Delta Ohm SrL, Caselle, Italy). Seedlings were grown for 64 days after sowing (DAS) under various light qualities. A standard greenhouse nutrient solution [(CaNO₃)₂·4H₂O 472.0, KNO₃ 202.0, KH₂PO₄ 272.0, NH₄NO₃ 80.0, MgSO₄·7H₂O 246.0, Fe-EDTA 15.0, H₃BO₃ 1.4, MnSO₄·4H₂O 2.1, ZnSO₄·7H₂O 0.8, CuSO₄·5H₂O 0.2, and Na₂MoO₄·2H₂O 0.1 mg·L^-¹] was prepared and adjusted to an EC of 0.8 dS·m^-¹ and a pH of 6.5 using a portable pH/EC meter. The solution was supplied daily to the seedlings through sub-irrigation. Plug trays were rearranged daily following irrigation to ensure uniform distribution of light wavelengths and intensities.

2. Light quality treatments

In this experiment, we used red (Red, peak: 660 nm), green (Green, peak: 520 nm), blue (Blue, peak: 450 nm), red:green (RG, 1:1, peak: 520 and 660 nm), green:blue (GB, 1:1, peak: 450 and 520 nm), red:blue (RB, 1:1, peak: 450 and 660 nm), red:green:blue (RGB, 1:1:1, peak: 450, 520, and 660 nm), and white (White, red:green:blue = 2:5:3, peak: 450, 580, and 600 nm) lights for corresponding treatments. The PPFD of each light source was determined by bandwidth integration (Table 1). The light spectral distribution was measured using a spectroradiometer (ILT950, International Light Technologies Inc., Peabody, MA, USA) at five points (center and four edges) on the top of the tray, with measurements taken at 1.4 nm intervals (Fig. 1).

Fig. 1.

The relative spectral distributions of LED light qualities used in the walk-in chamber. (A) red; (B) green; (C) blue; (D) RG (red:green = 1:1); (E) GB (green:blue = 1:1); (F) RB (red:blue = 1:1); (G) RGB (red:green:blue = 1:1:1); and (H) white (red:green:blue = 2:5:3).

3. Growth characteristics

To compare the growth characteristics under various light qualities, the leaf length, leaf width, number of leaves, stem diameter, dry weight of shoot and root, and leaf area of balloon flower and lance asiabell seedlings were measured. Leaf length and leaf width were measured on the fully expanded fourth leaf. The stem diameters of the seedlings were measured 1 cm above the medium surface using digital a vernier caliper (CD-20CPX, Mitutoyo Co. Ltd., Kawasaki, Japan). The dry weight was measured using samples dried for 72 hours in a constant temperature drying oven (Venticell- 222, MMM Medcenter Einrichtungen GmbH., Planegg, Germany). The dry weight was measured using an electronic scale (EW220-3 NM, Kern & Sohn GmbH., Balingen, Germany). Leaf area was measured using a leaf area meter (LI-3000, LI-COR Inc., Lincoln, NB, USA). Root morphological parameters, such as the total root length, root surface area, average root diameter, and root volume were measured using the image analysis system (WinRhizo Pro 2020, Regent instruments Inc., Sainte-Foy, QC, Canada) coupled with a professional scanner (Epson Expression 12000XL, Seiko Epson Co. Ltd., Nagano, Japan). The number of nodes, plant height, and internode length were measured to evaluate the shoot morphology. Internode length and compactness were calculated following the methods reported by Kim et al. (2013) and Jeong et al. (2020), respectively. The formula is as follows:

Internodelength(cm)=Plantheight(cm)Numberofnodes

Compactness(g·cm-1)=Dryweightofshoot(g)Plantheight(cm)

4. Predicting leaf area

We evaluated the applicability of multispectral imaging for leaf area estimation based on multispectral images. Linear regression analysis was conducted between the leaf area predicted based on the spectral images and the actual leaf area obtained using a leaf area meter to evaluate the applicability of multispectral imaging for leaf area estimation. A multispectral camera (FS-3200T-10GE-NNC, JAI, Copenhagen, Denmark) was used to capture images of the seedlings across the spectral range of 400－1,000 nm. The camera detects reflectance at wavelengths of 450, 550, 650, 750, and 830 nm, covering the visible (RGB) and near-infrare (NIR) regions. The regions of interest (ROIs) were extracted from the seedlings using ENVI software (ENVI 5.3, L3Harris Geospatial, Broomfield, CO, USA) (Fig. 2). During extraction, normalized difference vegetation index (NDVI) values of 0.3－1.0 were selected. Multispectral images were captured at 35, 50, and 64 DAS. NDVI was calculated using the formula as reported by Rouse et al. (1974):

NDVI=R750-R650R750+R650

Fig. 2.

Captured leaf regions using a multispectral camera at 35 (A), 50 (B), and 64 (C) days after sowing (DAS).

5. Statistical analysis

The experimental treatments were laid out in a completely randomized block design. Each treatment consisted of 150 plants. The 50 plants were used for each replicate. Experiments were performed in triplicate. Statistical analyses were performed using Statistical Analysis System software (SAS 9.4; SAS Institute Inc., Cary, NC, USA). Duncan’s multiple range test was performed. Statistical significance was set at p ≤ 0.05. Graphs were plotted using SigmaPlot (SigmaPlot 14.5, Systat Software Inc., San Jose, CA, USA).

Results and Discussion

1. Growth characteristics

Among the monochromatic lights, blue light significantly increased the leaf length, stem diameter, dry weight of the root, and leaf area of the balloon flowers, whereas, the green light showed the lowest leaf length, stem diameter, and dry weight of the shoot (Table 2 and Fig. 3). RG light significantly decreased the leaf length, leaf width, number of leaves, stem diameter, and dry weight of shoot and root and leaf area compared to their counterparts recorded under other mixed lights. The dry weight of shoot were higher under RGB light than under RB light. The leaf area was significantly higher under RGB and RB lights than under other lights. Under monochromatic blue light significantly enhanced the leaf length, leaf width, stem diameter, and dry weight of the shoot and root in lance asiabell compared to the red and green light (Table 3 and Fig. 4). Growth was significantly lower under RG light than under the other mixed light. Shoot growth characteristics were significantly higher after GB and RB light treatments than after RG, RGB, and white lights. Similar to balloon flower seedlings, lance asiabell showed a trend of enhanced growth under mixed lights compared to that induced by monochromatic lights. The effects of different light qualities on plant growth have been extensively studied. Red light is associated with high photosynthetic efficiency of horticultural crops (Izzo et al., 2020; Mizuno et al., 2011; Rabara et al., 2017; Son et al., 2012). However, monochromatic red light can inhibit the growth and synthesis of chlorophyll and carotenoids, decreasing their content owing to an energy imbalance between photosystems I and II (Lee et al., 2007; Tennessen et al., 1994). Moreover, it causes an imbalance between cryptochromes and phytochromes, which negatively affects plant morphology and metabolism (Gao et al., 2022; Yanagi et al., 1996). The negative effects of this monochromatic red light are referred to as ‘R light syndrome’ (Gao et al., 2022;

Fig. 3.

The seedling of balloon flower as affected by light quality at 50 days after sowing (DAS). Light qualities were red, green, blue, RG (red:green = 1:1), GB (green:blue = 1:1), RB (red:blue = 1:1), RGB (red:green:blue = 1:1:1), and white (red:green:blue = 2:5:3).

Fig. 4.

The seedling of lance asiabell seedlings at 50 days after sowing (DAS). Light qualities were red, green, blue, RG (red:green = 1:1), GB (green:blue = 1:1), RB (red:blue = 1:1), RGB (red:green:blue = 1:1:1), and white (red:green:blue = 2:5:3).

Fig. 5.

Plant height (A), internode length (B), and compactness (C) of balloon flower seedlings as affected by light quality measured at 50 days after sowing (DAS). Light qualities were red, green, blue, RG (red:green = 1:1), GB (green:blue = 1:1), RB (red:blue = 1:1), RGB (red:green:blue = 1:1:1), and white (red:green:blue = 2:5:3). Vertical bars indicate standard errors of the means (n = 6). Different letters above bars indicate significant differences by Duncan’s multiple range test at p ≤ 0.05.

McCree, 1981). Additionally, Liu et al. (2014) reported an improved growth in balloon flowers grown under blue light than those grown in vitro under monochromatic red light. Previously, various lights have been combined to address the imbalance between photosystems and phytochromes (Hogewoning et al., 2010). Compared to monochromatic red light, RB and RGB lights, which include green and blue light, increase chlorophyll content and enhance photosynthesis through the synergistic interactions between photoreceptors such as phytochrome and cryptochromes which respond various wavelength of light that regulate physiological processes (Claypool and Lieth, 2020; Usami et al., 2004; Xiaoying et al., 2012). These results validate the enhanced growth of balloon flower and lance asiabell seedlings grown under mixed light.

2. Plant height, internode length, and compactness

The plant height and internode length of balloon flower seedlings were significantly enhanced by 50 DAS under red, green, and RG light than those under other lights (Figs. 5A and B). The plant height and internode length of balloon flower seedlings treated with blue and GB lights were significantly lower than in the other light qualities. However, the number of leaves was significantly lower under GB light than under other blue-mixed lights (Table 2). Plant height and internode length were significantly longer under RGB light than under GB light but were significantly shorter than their counterparts under other mixed lights, such as RB and white light, exhibiting significantly increased compactness (Fig. 5). Therefore, RGB light was considered effective in producing shorter and more compact balloon flower seedlings. The plant height and internode length of lance asiabell seedlings measured at 50 DAS were the shortest under blue light (Figs. 6A and B). Plant height and internode length were significantly lower after blue, GB, RB, and white lights than those under other lights. The compactness of the lance asiabell seedlings was significantly higher under GB and RB lights (Fig. 6C). Therefore, GB and RB lights are considered effective for producing short and compact lance asiabell seedlings. Photons in blue light possess high energy (Lin, 2002) and induce photoprotective mechanisms in plants (Yanagi et al., 1996; Zheng et al., 2018). Continuous exposure to monochromatic blue light causes overexpression of CRY1, a photoreceptor that inhibits hypocotyl elongation, resulting in shorter hypocotyls and dwarf seedlings (Lin et al., 1995; Xiaoying et al., 2012; Zhang et al., 2020). Compactness, calculated by dividing plant height by dry weight of shoot, measures the density and sturdiness of the plant shoot parts. A higher compactness value indicates better seedling quality (Jeong et al., 2020). This study confirmed that light quality containing blue light can reduce plant height and internode length while producing compact seedlings of balloon flowers and lance asiabell.

Fig. 6.

Plant height (A), internode length (B), and compactness (C) of lance asiabell seedlings as affected by light quality measured at 50 days after sowing (DAS). Light qualities were red, green, blue, RG (red:green = 1:1), GB (green:blue = 1:1), RB (red:blue = 1:1), RGB (red:green:blue = 1:1:1), and white (red:green:blue = 2:5:3). Vertical bars indicate standard errors of the means (n = 6). Different letters above bars indicate significant differences by Duncan’s multiple range test at p ≤ 0.05.

3. Root characteristics

Among the monochromatic lights, the total root length, root surface area, average root diameter, and root volume of balloon flower seedlings under blue light were significantly higher than in other monochromatic lights, and there was no significant difference between green and red light (Table 4 and Fig. 7). Hence, blue light is considered a major light source that positively induces root development in balloon flower seedlings. Under RG light, the total root length, root surface area, average root diameter, and root volume of balloon flower seedlings, which did not include blue light, were significantly lower than in other light qualities. The effects did not significantly differ among GB, RB, RGB, and white lights, which included blue light. Among the monochromatic lights, blue light induced the highest root length, root surface area, average root diameter, and root volume of lance asiabell seedlings (Table 5 and Fig. 8). Root growth of lance asiabell seedlings under GB and RB lights tended to be superior to that induced by RGB light. GB and RB lights were considered effective for the development of lance asiabell roots at the seedling stage. Monochromatic red light primarily increases stem biomass, leading to smaller root systems (Izzo et al., 2020). Contrastingly, monochromatic blue light activates phototropin-1, promoting root growth and significantly increasing root biomass and dry weight compared to the counterparts grown under red and green lights (Galen et al., 2007; Xiaoying et al., 2012; Kim et al., 2019). Johkan et al. (2010) reported an increased dry weight of lettuce roots grown using mixed blue light than that treated with monochromatic blue and red lights. Consistently, in the present study, balloon flowers and lance asiabells showed enhanced root growth under monochromatic blue light used alone or in combinations.

Fig. 7.

The root of balloon flower as affected by light quality measured at 50 days after sowing (DAS). Light qualities were red, green, blue, RG (red:green = 1:1), GB (green:blue = 1:1), RB (red:blue = 1:1), RGB (red:green:blue = 1:1:1), and white (red:green:blue = 2:5:3).

Fig. 8.

The root of lance asiabell as affected by light quality measured at 50 days after sowing (DAS). Light qualities were red, green, blue, RG (red:green = 1:1), GB (green:blue = 1:1), RB (red:blue = 1:1), RGB (red:green:blue = 1:1:1), and white (red:green:blue = 2:5:3).

4. Predicted leaf area by multispectral camera

The R² of the leaf area regression analysis showed the tendency to decrease when the DAS of balloon flower was increased (Table 6). The R² value for the lance asiabell was lower than that of the balloon flower. Leaf overlap is a factor that interferes with the estimation of leaf area using spectral images. During the leaf area estimation of Chinese cabbage seedlings on a plug-tray basis, a tendency of increasing overlaps and decreasing R² were detected with an increasing number of plugs (Ban et al., 2023). This study also revealed a gradual increase in the overlaps of leaves with the increasing period for balloon flower cultivation, which led to a decrease in the coefficient of determination for leaf area estimation using multispectral imaging. Lance asiabell is a vine plant; during spectral image capture, the leaves do not spread flatly. Therefore, the overlap in many areas is considered to cause a lower coefficient of determination in lance asiabell than in the balloon flower seedlings. Contrastingly, the R² of the balloon flower showed high values (> 0.76) across all DAS, indicating that leaf area estimation using multispectral imaging is effective.

Conclusion

This study was conducted to select the appropriate light quality suitable for cultivating balloon flowers and lance asiabell seedlings in a CPPS. RGB light treatment induced short plant height and internodes in balloon flower plug seedlings. GB and RB lights supporting excellent shoot growth with short internodes and vigorous root development are considered suitable for producing lance asiabell plug seedlings in a CPPS. These results can potentially facilitate setting the light environment for the effective production of plug seedlings of balloon flowers and lance asiabells. Multispectral imaging for leaf area estimation is particularly applicable to balloon flowers, as it provides valuable research data that can be utilized for data-driven cultivation studies.