Introduction

The longevity of cut flowers is one of the most important factors contributing to flower quality and customer satisfaction (Rihn et al. 2014; Ha and In 2023). However, their vase life is often shortened due to various physiological processes within the flowers, compounded by external factors such as temperature fluctuations, relative humidity (RH) levels, and mechanical shock during harvest and transportation (Zieslin 1988; In et al. 2007). These stress sources lead to water stress, which negatively impacts the cut flowers’ ability to maintain hydration, resulting in petal wilting, incomplete flower opening, and abscission (Zieslin 1988; Woltering and Paillart, 2018) and in turn leading to diminished consumer satisfaction and a lower overall value (Doi et al. 2000; Loubaud and van Doorn 2004; Lü et al. 2010; Elhindi 2012; Fanourakis et al. 2012; van Doorn 2012; Mattos et al. 2017; Lin et al. 2019; Lin et al. 2020).

In addition to the water stress, the depletion of carbohydrates, such as soluble solids contents, accelerates the deterioration of cut flowers (Adachi et al. 2000; Nabigol et al. 2009). Carbohydrates positively influence various physiological activities and metabolic reactions in cut flowers, including blooming, petal coloration, ethylene synthesis inhibition, and water uptake regulation (Doi and Reid 1995; Ichimura 1998). After harvesting, cut flowers undergo a cessation of the existing carbohydrate supply, and acquiring carbon through photosynthesis becomes challenging owing to low indoor light intensity levels (Halevy and Mayak 1979; Ha et al. 2023a). In addition, the lack of a sucrose source leads to a more rapid decline in the quality and lifespan of cut flowers (Ho and Nichols 1977; van Doorn 1999; Ichimura et al. 2003; Ichimura et al. 2006; Ichimura et al. 2022).

Therefore, the early detection of cut flower conditions during the postharvest sorting process and the application of appropriate treatments are necessary to minimize postharvest losses. As the current sorting of cut flowers relies mainly on manual assessments of external features, such as the stem length, maturity, flower size, and leaf quality (In et al. 2009), it is challenging to reflect changes in vase lifetimes or the internal quality of cut flowers during their distribution. Precise assessments of the water content, inherent sucrose content and external appearance are necessary to reflect the postharvest quality of horticultural crops accurately.

In recent years, many studies have investigated various technologies for the non-destructive monitoring of postharvest quality changes in horticultural crops (Han et al. 2015; Kim et al. 2015; Kim et al. 2016; Kovar et al. 2019; Evelyn et al. 2020; Seehanam et al. 2022; Ha et al. 2023b). Among these methods, visible and near-infrared (Vis/NIR) spectroscopy is a valuable non-destructive technique that offers rapid analyses of the physiological and chemical characteristics of plants, such as their water content and soluble solids content (SSC) (McGlone et al. 2007; Menesatti et al. 2010; Subedi and Walsh 2011; Cortés et al. 2017; Maimaitiyiming et al. 2017; Neto et al. 2017; Evelyn et al. 2020). Reflectance changes, influenced by various factors such as plant pigments, structures, and interactions with water, can reveal stress before visible symptoms appear (Rumpf et al. 2010; Kim et al. 2011; Zahi et al. 2022). Vis/NIR is an important tool for determining the quality of agricultural products (Cortés et al. 2019; Ryckewaert et al. 2022). The use of these spectral features is thought to help identify the quality of cut flowers during the sorting stage (Kim et al. 2024). Owing to the rapid changes in quality from harvest to transportation and delivery to consumers, understanding and analyzing spectral characteristics based on changes in flower conditions are necessary for the effective utilization of Vis/NIR with cut flowers.

This study aimed to evaluate the effectiveness of Vis/NIR spectroscopy in predicting and analyzing the postharvest quality of cut roses during the sorting process. The research focused on two key aspects of cut flower quality: their physiological and phenotypic characteristics. First, we investigated physiological factors, in this case the water content, water relationships, and SSC, which are influenced by water stress, sucrose supplementation, and the transportation conditions. These factors directly impact the vase life and overall freshness of cut roses. Second, we assessed phenotypic traits, specifically flower wilting, petal abscission, and disease susceptibility. By monitoring changes in the Vis/NIR reflectance of cut roses following the treatments used here, we aimed to establish correlations between spectral data and both the physiological and phenotypic indicators of flower quality. This study highlights the potential of Vis/NIR as a non-destructive, rapid tool for assessing and improving postharvest flower sorting.

Materials and Methods

Plant material

Standard roses (Rosa hybrid L. ‘Unforgettable’) at a commercial stage, characterized by slightly curved external petals and without noticeable diseases or senescence symptoms (Harkema et al. 2013), were purchased from the agricultural corporation Rosepia (Jeonju, Republic of Korea). The roses used here were obtained from a commercial greenhouse in Jeonju in the Republic of Korea. The cut roses were transported to the laboratory at Andong National University, Republic of Korea, for 4 h either in tap water (wet transport, WT) or without tap water (dry transport, DT). The environmental conditions during the transport process were 23–25°C and a RH of 50–60%. The cut roses were maintained at the laboratory in a controlled environment condition, as previously described (Ha and In 2023), for the subsequent treatments and measurements.

Experiment 1: Air exposure and sucrose treatments

The cut roses (from WT) were subjected to air exposure (AE) and sucrose (ST) treatments to evaluate the effects of water stress and the internal sucrose content on the quality and reflectance wavelength of cut flowers. For the air exposure treatment, cut roses were held in a dry condition (without water) at a temperature of 23 ± 2°C and RH level of 50±5% for one, three, and five days. Cut roses placed in distilled water were used as controls (AE0). After the AE treatment, the rose flowers were recut at the bottom of the floral stem for a length of 40 cm with three upper leaves from the peduncle and were kept in glass vases containing 400 mL of distilled water to analyze the water content, Vis/NIR, and for the vase life evaluation.

For the sucrose treatment, cut roses were recut and held in vase solutions at set sucrose concentrations of 0% (control; ST0), 1% (ST1; 10 mg L^-1 sucrose), 3% (ST3; 30 mg L^-1 sucrose), and 5% (ST5; 50 mg L^-1 sucrose). Consequently, the cut roses were moved to an environmentally controlled location identical to that described in a previous study (Ha and In 2023) to measure the SSC, water relationships, vase lifetimes, and changes in the reflectance wavelength using Vis/NIR.

Experiment 2: Transport treatments

For the transport treatments, the WT and DT roses were continuously stored under wet and dry conditions to simulate the export transportation of cut flowers. The export simulation conditions used here are described in the literature (In et al. 2016). After the export simulation, the cut flowers were re-cut and transferred to a controlled environment, as described above (see Experiment 1), for postharvest quality and vase life assessments.

Vis/NIR spectroscopy acquisition

The reflectance data of the cut roses were acquired using an IMEC SNAPSCAN camera (IMEC, Leuven, Belgium; www.imec-int.com). The camera system recorded the spectral information of the objective lens of the Vis/NIR in the range of 470–900 nm. The Vis/NIR system was set up according to a previously described method (Kim et al. 2024). Briefly, four Osram halogen lamps (OSRAM, Munich, Germany) were used as the light source in the Vis/NIR device. The halogen lamps provided 350–2500 nm light with a power level of 20 W. The cut roses were positioned 50 cm away from the lens, with the lamp and camera set at a 45° angle. The camera’s exposure time was fixed at 2 ms. Spectral measurements were conducted thrice at various locations, and the data analysis utilized IMEC HSI Snapcan software version 1.8.11 (IMEC, Leuven, Belgium) to derive the average reflectance coefficient of the cut roses for the subsequent analyses.

Flower opening, fresh weight, water uptake, and water balance measurements

To assess the effect of the treatments on the quality of the cut flowers, changes in flower opening, fresh weights (FW), and water absorption levels were evaluated daily. Flower diameters were measured and calculated based on a previously established protocol (Ha et al. 2019). The water uptake (WU), daily transpiration rate (T) and water balance (WB) calculations used here are also described in the literature (Ha and In 2023).

Analyses of SSC and water content outcomes

The SSC (°Brix) values and water content (WC) levels of the cut flowers were measured at two-day intervals from the beginning to the end of the vase lifetimes. After the Vis/NIR spectrum analyses, the cut roses were used for destructive SSC and WC measurements. SSC measurements were conducted based on a previously described method (Ha and In 2023).

The WC of the cut flowers was determined using the following equation:

WC(%)=FW-DWFW×100

where FW represents the fresh weight and DW represents the dry weight. The cut roses were dried in an oven at 45°C for 108 h to evaluate the DW. The roses were dried further until no additional weight change was observed after removal from the oven, and Vis/NIR measurements were taken after measuring the DW.

Vase life and senescence evaluations

The longevity of the flowers was recorded from the time the cut flowers were placed in the vase solution until they lost their ornamental value. Flower senescence symptoms were evaluated daily following the card evaluation for Rosa (VBN 2014). An evaluation was conducted if one or more of the following symptoms were observed in the cut flowers: a bent neck (stem bent at an angle exceeding 45°), petal wilting (PW, ≥50% of petals lose their turgor), petal discoloration (PD, ≥50% petals turned blue), petal abscission (PA, ≥3 petals dropping), or gray mold disease (GMD, if more than 50% of petals among all petals of the flowers were infected with Botrytis cinerea).

Experimental design and data analysis

Twenty-four roses were used for each treatment. The Vis/NIR spectrum analyses, postharvest quality evaluations, and longevity assessments were conducted with nine replicates per treatment (one flower per sub-replicate). The remaining six flowers were allocated for destructive measurements of the SSC and WC. Differences among treatments were analyzed using LSD and Duncan’s multiple range test in SPSS (version 22.0; IBM Corp., Somers, NY, USA) with a significance level set at p = 0.05. Student’s t-tests were also conducted to compare the WT and DT treatments at the same significance level. Pearson’s correlation and simple linear regression were employed to examine the relationship between the vase life and various characteristics of the cut roses.

Results

AE and ST treatments influenced the postharvest quality and longevity of cut roses

The cut roses exhibited petal dryness and discoloration after exposure to the air for three (AE3) and five (AE5) days (Fig. 1A). The AE treatment significantly decreased the FW and WC of cut roses, consequently decreasing the flowers’ longevity (Fig. 1B-1D). These results show that AE treatment decreases the quality and lifespan of cut flowers.

The ST treatment retarded senescence symptoms and maintained floral freshness in the cut roses (Fig. 1E). ST-treated flowers exhibited higher relative fresh weights during vase life than ST0 flowers (Fig. 1F). Among all sucrose concentrations, the ST3 and ST5 treatments increased the SSC in the petal tissues on day 7 of the vase period compared to that in the ST0 flowers (Fig. 1G). The ST treatment also extended the lifespan of the cut flowers (Fig. 1H).

Fig. 1.

Effects of air exposure (A-D) and sucrose (E-H) treatments on the postharvest quality and vase life of cut roses. For AE0 and ST0, cut flowers were held in distilled water; for AE1, AE3, and AE5, cut roses were held in dry conditions for one, three, and five days; and for ST1, ST3, and ST5, cut roses were held in sucrose solutions at the following concentrations: 1%, 3%, and 5%. The red triangular markers on the graphs of the flower diameters and fresh weights indicate the start date of vase life measurements after the treatments were completed. Different letters (a-d) on the vertical bars present a significant difference among the treatments at p = 0.05 based on Duncan’s multiple range tests. Vertical bars represent the standard error (SE) of the mean (n = 9).

AE and ST treatments influenced the reflectance wavelength of cut flowers

Changes in the reflectance wavelength (RW) after AE and ST treatments were also analyzed in cut roses using the Vis/NIR spectra (Fig. 2). The RW at 550 nm in the AE flowers was lower than that in AE0 flowers and a decrease in the RW corresponded to an increase in the AE time (Fig. 2A). In addition, simple linear regression analyses revealed that the WC and lifespan of cut roses were highly positively correlated with reflectance at 550 nm (r² = 0.72; p = 0.05, and r² = 0.78; p = 0.05, respectively) (Fig. 2B and 2C). These results suggest that reflectance at 550 nm is closely correlated with the water status and longevity of the flowers.

The ST treatment affected the total RW of cut roses at 600 nm (Fig. 2D). ST flowers exhibited a higher reflectance total RW at 600 nm compared to that of ST0 flowers (Fig. 2D). The RW at 600 nm was positively correlated with the SSC of rose petals (r² = 0.53; p = 0.05) and the vase life (r² = 0.57; p = 0.05) of cut flowers (Fig. 2E and 2F). These results imply that a RW at 600 nm can be used to predict the SSC content and longevity of cut roses using Vis/NIR data.

Fig. 2.

Effects of air exposure and sucrose treatments on the reflectance wavelength of cut roses (A and D). The arrows in A and D indicate the wavelengths 550 and 600 nm, which were chosen for the linear regression analyses. Simple linear regressions between the reflectance wavelength of 550 nm and the water content and vase life outcomes (B and C), and between the reflectance wavelength 600 nm and SSC levels of the petals and the vase life (E and F) of cut roses. For AE0 and ST0, cut roses were held in distilled water; for AE1, AE3, and AE5, cut roses were held in dry conditions for one, three, and five days; and for ST1, ST3, and ST5, cut roses were held in sucrose solutions at the following concentrations: 1%, 3%, and 5%. (*) indicates p = 0.05 (n = 36).

Relationships between the AE times and the ST treatment and senescence symptoms of cut roses

Considerable levels of petal wilting (PW), GMD, and petal abscission (PA) occurred in the control flowers (AE0 and ST0) (Fig. 3A and 3B). The AE treatment decreased the incidence of GMD but accelerated the appearance of petal discoloration (PD) in cut roses (Fig. 3A). The ST reduced the incidence rates of PW but increased the susceptibility of cut flowers to GMD (Fig. 3B). The correlation analysis demonstrated that the incidence rate of PD had a significantly high positive correlation with the AE duration (r = 0.97; p = 0.01) (Fig. 3C). PW incidence rates were significantly negatively correlated with the AE time (p = 0.05) but not with the sucrose concentration (Fig. 3C and 3D). The incidence rates of GMD in cut roses were significantly negatively and positively correlated with the AE time and sucrose concentration (Fig. 3C and 3D), respectively. These results indicate that longer AE times suppress GMD growth and that higher SSC levels in the petals increase the susceptibility of cut roses to GMD.

Fig. 3.

Effects of air exposure and sucrose treatments on flower senescence symptoms (A and B), correlation coefficient values of the relationships between the senescence symptoms and air exposure time (C), and the sucrose treatment concentration (D) of cut roses. For AE0 and ST0, cut roses were held in distilled water; for AE1, AE3, and AE5, cut roses were held in dry conditions for one, three, and five days; and for ST1, ST3, and ST5, cut roses were held in sucrose solutions at the following concentrations: 1%, 3%, and 5%. GMD, gray mold disease; PA, petal abscission; PW, petal wilting; PD, petal discoloration. * and ** correspondingly denote p = 0.05 and p = 0.01 (n = 36).

Transport treatments affected the quality, longevity, and RW of cut flowers

The diameter of the WT flowers increased during vase life and was larger than those of the DT flowers on days 4–7 of the vase life (Fig. 4A). The FW of the DT flowers was lower than that of the WT flowers (Fig. 4B). The DT treatment also reduced the maintenance requirements of positive WB (PWB) and WC of cut flowers, leading to a decrease in the vase life by 1.4 d compared to that of the WT (Fig. 4C-4E). The DT treatment increased the occurrence rates of PW but reduced the incidence rates of GMD infection in the petals (Fig. 4F).

The correlation analysis also revealed that WC, PWB maintenance, and FW had strong positive correlations with the vase life of cut roses (p < 0.01), whereas WB exhibited a significant negative correlation with the vase life (r = ‒0.62; p < 0.01) (Fig. 4G). Based on these results, the correlation between RW at 550 nm and WC and the flowers’ longevity was analyzed (Fig. 4H and 4I). We found that RW at 550 nm was positively correlated with both the WC (r² = 0.77; p < 0.05) and vase life (r² = 0.68; p < 0.05) of cut roses (Fig. 4H and 4I). In this study, there were positive correlations between cut flower quality and reflectance of WC and the vase life. Better flower quality (lower incidence rates of flower wilting, petal abscission, disease, and discoloration) corresponded to higher WC levels, longer vase lifetimes, and higher reflectance (data not shown). These findings indicate that the water relationships are closely related to the vase life of the cut flowers and suggest that Vis/NIR spectroscopy can be utilized to predict the WC of cut roses.

Fig. 4.

Effects of the transport treatment on the flower diameter (A), fresh weight (B), maintenance of the water balance (C), water content (D), vase life (E), incidence rates of senescence symptoms (F), correlation coefficient values of the relationships between the vase life and physiological characteristics of cut roses (G), and a simple linear regression analysis between RW 550 nm and water content (H) and vase lifetimes (I) of cut roses. WT, water transport; DT, dry transport; GMD, gray mold disease; PA, petal abscission; PW, petal wilting; PD, petal discoloration; FW, fresh weight; WC, water content; WB, water balance; and WU, water uptake. Vertical bars represent the SE of the mean (n = 9 for A-E). * and ** correspondingly denote p = 0.05 and p = 0.01 (n = 18 for G-I).

Discussion

The postharvest quality and longevity of cut roses are often negatively influenced by unsuitable postharvest conditions, such as air exposure or dry transport (Ha et al. 2019; Harkema et al. 2013; Ha et al. 2021; Kim et al. 2024). Dry handling strategies for cut roses after harvest, such as DT or AE, to control GMD growth or maximize the quality of cut roses, can trigger dehydration in water-stressed rose cultivars and reduce their vase lifetimes (Macnish et al. 2009; Harkema et al. 2013; Ha et al. 2019; Ha et al. 2021; Fanourakis et al. 2022; Kim et al. 2024). These results align with our findings showing that AE and DT treatments significantly reduced GMD infection rate in petals but negatively influenced the water status and postharvest quality of cut roses. This suggests that, although disease control is critical, it comes at the cost of reducing water uptake, which is essential for maintaining flower turgor and overall vase life. This trade-off between disease prevention and WC monitoring in cut flowers highlights the need for a more balanced approach that considers both aspects of postharvest flower quality.

In the present study, ST effectively prolonged the cut flowers’ lifespans by enhancing their FW, WB, and SSC outcomes. These results are consistent with previous findings showing that external carbohydrate treatments significantly extended the lifespans of cut flowers (Ichimura and Hiraya 1999; Elgimabi 2011; van Doorn and Han 2011; Arrom and Munné-Bosch 2012; Elhindi 2012; Rabiza-Świder et al. 2020; Ichimura et al. 2022). Despite the positive effects, the ST treatment also increased the flowers’ susceptibility to GMD. Higher sucrose concentrations led to severe GMD infection rates and necrotic symptoms in rose petals. Similarly, a previous study reported a significant correlation between the SSC of strawberry fruits and their susceptibility to GMD (Mancini et al. 2023). Therefore, while sucrose treatments can effectively improve cut flower quality, careful management of SSC concentrations is necessary to mitigate GMD infection.

The Vis/NIR spectra have been reported in various studies to respond to changes in water stress and sugar content levels in plants, and the RW, related to water and sugar contents, differs among plant species and plant organs (McGlone et al. 2007; Maimaitiyiming et al. 2017; Neto et al. 2017). Plants subjected to water stress exhibit alterations in their spectral reflectance, particularly at the green (500–570 nm) and red (610–700 nm) wavelengths (Katsoulas et al. 2016). Dehydration-induced changes in the optical properties and WC effects have been observed in the 550–580 nm and 630–750 nm ranges for soybean leaves (Kovar et al. 2019). The Vis/NIR spectroscopy analysis in this study provided valuable insights into the physiological changes in cut roses under different postharvest treatments. Our results revealed a strong correlation between the spectral reflectance at 550 nm and both the WC and the vase life of cut ‘Unforgettable’ roses, with reduced WC leading to lower reflectance. These results are consistent with previous studies indicating that water stress caused a decline in plant reflectance, particularly in the visible spectrum (Jackson and Ezra 1985; Kim et al. 2024). This suggests that Vis/NIR spectra, especially in the 550 nm range, could serve as reliable indicators of water stress in cut roses. Similar patterns of reflectance associated with water stress have been observed in sunflower leaves (Neto et al. 2017). RW at 500–1050 nm was observed under slight, moderate, and severe water stress levels. Leaves subjected to slight water stress exhibited a higher RW at 550 nm than those subjected to moderate and high levels of water stress (Neto et al. 2017). A previous study also revealed that the NIR region is correlated with the WC of plants, depending on the water level stored in the leaf tissues (Govender et al. 2009). These results underscore the broad applicability of the Vis/NIR technique for monitoring water status in various plants.

Changes in the spectral reflectance at 670, 798, 830, and 888 nm in potato tubers have been linked to the sugar content in these plants (Chen et al. 2010), whereas sugar content estimations in pears using the 710–930 nm wavelength band exhibited high accuracy (Choi et al. 2017). In apples, the spectral bands at 530–570 nm and 700–720 nm are also associated with the sugar content (Zhang et al. 2015). A wavelength range of 700–1000 nm has been used for sugar content measurements in sugarcane plants (Phuphaphud et al. 2020). The RW at 680 and 750 nm has been successfully used to assess the SSC levels of Korla fragrant pears (Yang et al. 2022). In the present study, the RW at 600 nm was related to SSC levels in the petals of cut roses (Fig. 2E). Our findings differ from those of previous studies, presumably due to the low SSC levels in the rose petals. It is very challenging to evaluate SSC levels in flowers with a non-contact method by measuring the spectral reflectance because the flower surface is complicated and double-curved compared to the surfaces of fruits. Although this wavelength differed from the ranges reported in other crops, our results highlight the potential of the Vis/NIR spectra for evaluating SSC levels in cut roses. However, it is important to note that evaluating SSC levels in flowers presents unique challenges compared to doing so in fruits due to the complex and irregular surfaces of petals. The low SSC levels in rose petals and the varying sucrose concentrations also complicate the use of Vis/NIR for sugar estimations. This suggests that further refinement of Vis/NIR techniques, possibly through the use of advanced calibration models and tailored algorithms, is necessary to assess SSC levels accurately in floral tissues.

Overall, our results demonstrate the potential of Vis/NIR spectroscopy as a non-destructive tool for evaluating main postharvest quality indicators, such as the WC and SSC, in cut roses. However, to optimize its application in the flower industry fully, additional research is required to refine the spectral models, account for flower-specific variables, and incorporate diverse analytical methods. Further studies should explore the influence of factors such as cultivar variations, storage conditions, and flower maturity on the spectral reflectance. Additionally, combining Vis/NIR with other techniques, such as imaging or machine-learning algorithms, may enhance its predictive accuracy and allow for more precise sorting of cut roses based on internal quality parameters. Such advancements would not only improve postharvest handling but also help minimize losses during export and distribution, ultimately extending the vase life and maintaining the market value of cut roses.

Conclusions

We explored the feasibility of using Vis/NIR spectra for water status and SSC evaluations of cut roses under various postharvest conditions. Our results uncovered a positive correlation between the longevity of cut roses and both the WC and SSC values of the petals. Specifically, spectral reflectance at 550 nm showed a strong association with WC, while reflectance at 600 nm was correlated with SSC levels in cut roses. These findings underscore the potential of Vis/NIR spectroscopy to serve as a valuable tool for assessing postharvest quality and predicting the longevity of cut roses. Thus, we conclude that the Vis/NIR method can be applied widely in the floriculture postharvest industry as a rapid technique to classify flower quality.