Introduction

Lily is an economically important bulbous plant used for cut-flower, potted plants, and planting in the garden. In the genus Lilium, there are approximately 100 species and these plants are distributed between latitudes 10° and 60°, and more than 9,400 lily cultivars were registered (Kole 2011). These lily cultivars were divided into several commercial groups, with Oriental Hybrid lilies renowned for their large flower size and strong scent (Grassotti and Gimelli 2010). Due to this popularity, most countries producing cut lilies mainly cultivate Oriental Hybrid lilies (Kim and Kim 2015).

Geophyte plants store nutrients in their underground storage organs such as bulbs, tubers, and rhizomes. During the early stages of vernalized Lilium growth, the bulb acts as the main source of carbohydrates before the inflorescence emergence stage (Wu et al. 2012). Once the leaves are fully expanded, they become the primary source organ, transporting carbohydrates produced through photosynthesis. Meanwhile, the bulb gradually transforms into a sink organ after the flowering stage (Zeng et al. 2022). Carbohydrates, especially starch and sugar, play a significant role in bulb enlargement (Li et al. 2022). Consequently, estimating the pattern of bulb enlargement relies on carbon assimilation produced via leaf photosynthesis.

The leaf photosynthesis rate changes depending on environmental conditions such as air temperatures, light intensity, and C_i The photosynthesis model takes into account the biological system’s response to changes in light intensity, electron transport, and enzyme activity under varying environmental conditions (Farquhar et al. 2001;Kaiser et al. 2015). The biochemical C₃ photosynthesis model of Farquhar-von Caemmerer-Berry (FvCB) simply expresses photosynthesis in leaf gas exchange limited by Rubisco carboxylation (A_c) and RuBP regeneration (A_j) (Farquhar et al. 1980). The FvCB model parameters dynamically estimate carbon assimilation for plant production in conjunction with light, CO₂, and temperature. In addition, plant carbon assimilation could be estimated by expanding leaf-level photosynthesis to canopy-level photosynthesis (Sprintsin et al. 2012;Zhou et al. 2021).

Scale-up from leaf-level to canopy-level photosynthesis is sensitive to absorbed photosynthetic photon flux density (PPFD) (Kull and Kruijt 1998). The absorbed light intensity on the leaf is estimated to decrease on the ground due to leaf area, leaf angle, stem, and plant height (Liu et al. 2015;Mariscal et al. 2000;Newton and Blackman 1970). In addition, in basal leaves, the photosynthetic capacity tends to be lower than the upper leaves, likely due to their leaf age (Kitajima et al. 2002). Thus, photosynthetic rates vary at different leaf positions within a plant. It is essential to understand the photosynthetic capacity specific to each leaf position to calculate canopy photosynthesis accurately. In previous studies, plant canopy photosynthesis was estimated in sunlit and shaded parts, and then crop growth, development, and yield were simulated for various cultivation conditions (Wu et al. 2019).

In floricultural crops, there have been relatively few attempts to develop growth models based on photosynthetic carbon assimilation. To develop a growth model based on photosynthesis for floricultural crops, it is necessary to accurately estimate the photosynthetic rate under growth conditions such as CO₂ concentration and PPFD, which are known to affect the photosynthetic rate (Ma et al. 2021;Reis et al. 2009;Wang et al. 2020). This study was conducted to measure the distinct photosynthetic rate of lily (Lilium Oriental Hybrid ‘Casa Blanca’) at the upper, medium, and basal leaf positions. Additionally, we predicted A_n based on various levels of C_i and PPFD and determined FvCB model parameters at each leaf position.

Materials and Methods

Plant materials and growth conditions

The Lilium Oriental Hybrid ‘Casa Blanca’ was acquired from a commercial nursery farm (Wooriflowers Seeds Seedling CO., Ltd., Gwacheon, South Korea) and cultivated in a greenhouse at Seoul National University Farm (Suwon, Korea; 37°27’N, 126°99’E) for three months, sprouting from the bulb. Bulbs with a circumference of 18 – 20 cm size were planted in a lily bulb crate (40 × 20 × 60 cm) on September 08, 2023 with horticultural substrates (Sunshine^® Mix #4; Sun Gro Horticulture Canada Ltd., Agawam, USA). During experimental periods, the greenhouse environment conditions were illustrated in Fig. 1. Mean air temperature was controlled in 17.48 ± 4.61°C and mean relative humidity was maintained in 71.88 ± 15.23%. Daily light integral (DLI) was 7.98 ± 3.50 mol･m^-2･d^-1, depending on the weather conditions. The plants were automatically watered with a sprinkler system twice a week and fertigated at the planting date with 30 g of controlled-release fertilizer (Osmocote Plus 15N-4.8P-10.8K + 2Mg + TE; Everris International B.V., Heerlen, Netherlands) per crate.

Measurement of leaf photosynthetic rate and SPAD value

The photosynthetic rate and SPAD values of lilies were measured at each leaf position. The Leaf position was divided into the upper, medium, and basal parts. Each part was defined from the apex to the 1/4 point, 2/4 point, and 3/4 point of a plant, respectively.

The leaf A_n rates for photosynthesis model parameterization and evaluation were measured in December 2023, during the ‘Inflorescence emergence’ phenological stage of lily, using a portable photosynthesis measuring device (LI-6400 XT, LI-Cor. Inc., Lincoln, NE, USA) equipped with a 6400-02B LED light chamber. Measurements were conducted from 09:00 to 16:00. The block temperature, relative humidity, and flow rates in the LED light chamber were set at 25°C, 51±1%, and 300 μmol･s^-1, respectively. Light intensities were progressively adjusted to 1500, 1200, 900, 600, 300, 200, 100, 50, 0, and 1500 μmol･m^-2･s^-1 while maintaining a sample CO₂ concentration of 400 μmol･mol^-1. The reference CO₂ concentrations were progressively set to 400, 300, 200, 100, 60, 400, 400, 600, 900, 1200, 1500, and 1800 μmol･mol^-1, while maintaining a light intensity of 400 and 800 μmol･m^-2･s^-1. Every photosynthetic rate value was logged between 60-150s with infrared gas analyzer (IRGA) matching.

SPAD values were measured with a chlorophyll meter (SPAD-502, Konica Minolta, Inc., Tokyo, Japan). In each leaf position, SPAD values were averaged by measuring three times at the central part of a single leaf. The SPAD measurement had three replicates.

Photosynthetic rate models and statistical analysis

An empirical photosynthesis model was used followed by a previous study (Kaitala et al. 1982). In previous studies, photosynthetic rate was expressed by a rectangular hyperbola equation with two variables:

where A_n represents the leaf net photosynthetic assimilation rate (μmol･m^-2･s^-1), α denotes the photochemical efficiency (μmol･mol^-1), β signifies the carboxylation conductance (s^-1), PPFD stands for the photosynthetic photon flux density (μmol･m^-2･s^-1), C_i represents the intracellular CO₂ concentration (μmol･mol^-1), and R indicates the respiration rate (μmol･m^-2･s^-1).

In this study, the PPFD was set at 800 μmol･m^-2･s^-1 as constant. The parameters α, β, and R of the rectangular hyperbola photosynthesis model were estimated by varying C_i at each leaf position. SigmaPlot (version 10.0; Grafiti LLC, Palo Alto, CA, USA) was used for nonlinear regression analysis and fitting parameters of the rectangular hyperbola model.

The FvCB model was originally developed by Farquhar et al. (1980). This photosynthesis model was expressed by following equation:

In the equation, A_n represents the leaf net photosynthetic assimilation rate (μmol･m^-2･s^-1), while A_c and A_j denote the net photosynthetic rates limited by Rubisco and RuBP, respectively (μmol･m^-2･s^-1).

Triose phosphate utilization (TPU) limitation, which typically does not affect photosynthetic rates under common growth conditions and can influence photosynthesis model parameters, was excluded from consideration in this study (Lamour et al. 2023;Rogers et al. 2020).

where V_c represents the carboxylation capacity at a certain light intensity (μmol･m^-2･s^-1), Γ^* denotes the CO₂ compensation point when mitochondrial respiration is zero (μmol･mol^-1), K_m, K_c, and K_o are the Michaelis-Menten constants for Rubisco, carboxylation, and oxygenation, respectively (μmol･mol^-1). V_cmax is the maximum rate of Rubisco activity (μmol･m^-2･s^-1), Q stands for the incident photosynthetic active photon flux density (μmol･m^-2･s^-1), J represents the electron transport rate at specific light intensity (μmol･m^-2･s^-1), J_max is maximum electron transport rate, α signifies the quantum yield of electron transport (mol e^-･mol^-1 photon) and θ denotes the curvature of the light-response curve (dimensionless).

The calculation of FvCB model parameters, V_cmax, J_max, R, α, θ, K_m, and Γ^*, and the plotting of graphs were performed using the “fitacis” function of the “plantecophys” package (Duursma 2015) in R (version 4.3.2; R foundation for statistical computing, Vienna, Austria). α and θ were set to 0.24 mol e^-･mol^-1 photon and 0.85, respectively. The temperature-dependent parameters K_m and Γ^* were calculated by Arrhenius function (Medlyn et al. 2002). Statistical analysis of SPAD values and FvCB model parameters were performed using the “stats” and “agricolae” packages (de Mendiburu 2023) in R.

Results

Photosynthetic capacity in different leaf positions

SPAD values of Lilium Oriental Hybrid ‘Casa Blanca’ were significantly different in each leaf position (Fig. 1). Mean values of upper, medium, and basal leaves were 66.57, 54.53, and 43.27, respectively. Also, A_n in A - C_i curves saturated at higher in the upper leaf position compared with the medium and basal leaf position (Fig. 2). At 800 μmol･m^-2･s^-1 PPFD, calculated saturation values of A_n were 31.88, 20.41, and 17.56 in upper, medium, and basal leaves, respectively, and this tendency was consistent with SPAD values.

Parameter estimation of FvCB model

In the FvCB model parameterization using measured A_n values, the temperature-dependent parameters K_m and Γ^* were calculated to be 701.32 and 42.75 μmol･mol^-1, respectively, in the meausrement block temperature. Saturated A_n in the upper leaf position was more higher than the medium and basal leaf positions (Fig. 3A–C). The parameters calculated from each curve were averaged by leaf position (Table 1). V_cmax and J_max were the highest in upper leaves, with 113.239 and 161.273, respectively. However, there were no significant differences between the medium and basal leaves. The difference in J_max across leaf positions were larger than that of V_cmax . Additionally, R was not significantly different across the leaf position.

Estimating photosynthetic rates by FvCB model compared to rectangular hyperbola model

Parameters α, β, and R in the rectangular hyperbola photosynthesis model (Eq. 1) were calculated for each leaf position using A_n values in Fig. 2. The photochemical efficiency, carboxylation conductance, and respiration rate were the highest in the upper leaf, followed by the medium leaf, and were the lowest in the basal leaf (Table 2). The respiration rates by the rectangular hyperbola model were higher than that of the FvCB model for each leaf position.

To compare the performance of FvCB and rectangular hyperbola photosynthesis model, A_n was measured under the same conditions for each leaf position (Fig. 4). The estimated A_n values for both the FvCB and rectangular hyperbola models were calculated. The evaluation results showed little difference between the two models across leaf positions. Although both models exhibited lower accuracy at the basal leaf compared with the upper or medium leaf positions, the coefficient of determination (R²) values remained higher than 0.86 for all leaf positions, and the root mean square error (RMSE) was lower than 2.39.

When A_n values were estimated using the FvCB and rectangular hyperbola photosynthesis models in the range of 0 to 1500 μmol･m^-2･s^-1 PPFD and 0 to 1500 μmol･mol^-1C_i (Fig. 5), the estimated A_n values at various levels of PPFD and C_i for each model exhibit distinct shapes. In similar environments to the A_n measurements, both models demonstrated high accuracy. However, the estimated A_n values by the rectangular hyperbola model tended to overestimate under high C_i and PPFD conditions and underestimate under low C_i and PPFD conditions.

Discussion

In plant photosynthesis, physiological factors influence the capacity of carbon assimilation rate. In particular, chlorophyll, the green pigment responsible for absorbing light energy and converting it into chemical energy, acts as a photocatalyst (Rabinowitch 1965). Leaf chlorophyll content significantly correlates with the photosynthetic rate (Takai et al. 2010). In this study, SPAD values in Lilium Oriental Hybrid are significantly different according to leaf positions (Fig. 1). However, there was little difference in the photosynthetic rate between the basal and medium leaves than medium and upper leaves (Fig. 2). Likewise, FvCB model photosynthetic parameters (V_cmax and J_max) have no significant differences in basal and medium leaves (Table 1). The differences in the tendency of SPAD values and photosynthetic rates by leaf position can be attributed to the effects of other photosynthetic pigments. Since a SPAD meter only measures the chlorophyll content, the photosynthetic rate would not be correlated with the SPAD value directly. Carotenoids absorb the 450-550 nm region of the solar spectrum and transfer light energy to enhance photosynthesis (Hashimoto et al. 2016). For example, higher leaf photochemical efficiency, which was determined by carotenoids content along with chlorophyll content, led to increased CO₂ uptake in evergreen chaparral species (Stylinski et al. 2002).

Leaf respiration, defined as non-photorespiratory O₂ consumption, is essential for calculating photosynthetic rates (Tcherkez et al. 2017). Respiration rates of C₃ plants, including lilies, are similar to those calculated from the FvCB model than the rectangular hyperbola model (Gandin et al. 2014;Romanowska et al. 2002). The difference in J_max values among leaf positions showed a wider range compared to V_cmax values. This implied that the variation in photosynthetic capacity among leaf positions influenced by light intensities can contribute more than CO₂ concentrations. In this respect, it is necessary to consider the light interception to calculate the whole plant photosynthetic rate.

In natural environments, light interception occurs by leaf area, leaf angle, and plant canopy (Trouwborst et al. 2010). Considering the consistent FvCB model parameters observed between medium and basal leaves, along with comprehensive light interception, the photosynthetic rate of the entire plant can be simply divided into two components: sunlit and shaded to simplify the prediction of crop carbon assimilation (Wu et al. 2019). In this study, although the photosynthetic rate for each leaf position was measured at the same light intensity, the photosynthetic model presented can estimate photosynthetic rates at varying light intensities. Furthermore, if the ratio of sunlit to shaded areas in lilies can be quantified, it will be possible to estimate whole plant photosynthesis. However, the estimated photosynthetic rates can vary depending on the photosynthetic model used.

The performance of both the rectangular hyperbola and the FvCB models showed no significant difference (Fig. 4). This would be attributed to the fact that the rectangulalr hyperbola model fit the best parameters only mathematically in the measured conditions. However, estimating photosynthetic rates by the rectangular hyperbola model showed over 40 μmol･m^-2･s^-1A_n in high C_i and PPFD conditions (Fig. 5A). Likewise, Jung et al. (2021) reported that the rectangular hyperbola model overestimated photosynthetic rate in high light intensity and CO₂ concentrations in paprika. In this study, the FvCB model demonstrated more appropriate values across various environments. The FvCB model represents the biochemical process of plant mechanism related to photosynthetic CO₂ uptake as mathematically (Yin et al. 2021). Given the simplicity and accuracy of the FvCB model, using this model to estimate carbon assimilation rates would allow for the model’s application under diverse growth conditions.

The present study verified the leaf photosynthetic rates at the upper, medium, and basal positions in Lilium Oriental Hybrid ‘Casa Blanca’. Based on our findings, ‘Casa Blanca’ showed higher photosynthetic rates and FvCB model parameters in upper leaves compared with medium and basal leaves. When only the C_i changed, both the rectangular hyperbola model and FvCB model have a similar estimation. However, when estimating photosynthetic rates under various and PPFD conditions, the FvCB model that considers the plant’s biochemical responses provided more reasonable values. With this information, whole plant photosynthesis of lily can be estimated in the dynamic environment.