Abstract
Supplemental lighting, temperature control, and CO2 enrichment can improve the productivity of greenhouse crops, but operating costs for greenhouse control systems to maintain environmental parameters at desired setpoints can be expensive. To balance operating costs with productivity, growers need to be able to predict how a crop will perform as a function of photosynthetic photon flux density (PPFD), CO2 concentration, and temperature. The objective of this study was to explore the response of net photosynthetic rate (Pn) to PPFD and CO2 concentration, for plants acclimated to different growth environment temperatures or light intensities. We measured Pn at all combinations of 14 irradiances and four CO2 concentrations of calibrachoa (Calibrachoa ×hybrida ‘Superbells Lemon Slice’), petunia (Petunia ×hybrida ‘Supertunia Mini Strawberry Pink Veined’), and verbena (Verbena ×hybrida ‘Superbena Royale Whitecap’) grown at three light intensities, and of geranium (Pelargonium ×hortorum ‘Maverick Red’), pepper (Capsicum annuum ‘California Wonder’), and sunflower (Helianthus annuus ‘Pacino Gold’) grown at three different temperatures. Sunflower, pepper, and geranium were fit to a model representing Pn as a function of PPFD, CO2 concentration, and leaf temperature. Photosynthetic light response curves, at each CO2 concentration, were fit for each species and growth environment using a nonrectangular hyperbola. These models can be used to identify multiple combinations of PPFD, CO2 concentration, and leaf temperature that would result in equivalent rates of photosynthesis, allowing the most cost-effective combination to be chosen.
Many greenhouse-grown plants are started in late winter or early spring when outdoor light intensity and air temperature are lower than ideal for plant growth (Korczynski et al., 2002). To reduce heating costs, greenhouses are typically sealed as tightly as possible, which can contribute to low CO2 concentrations inside (Mortensen, 1987). Conversely, greenhouses often require some form of cooling during summer months, typically through use of vents, fans, or evaporative cooling. The timing of CO2 enrichment should be considered, as the increased air exchange rate during periods of active ventilation, either for temperature or humidity control, minimizes its efficacy. Supplemental lighting, temperature control, and CO2 enrichment can improve the productivity of greenhouse crops, but capital costs for climate control systems and operating costs for these systems to control heaters, lights, vents, fans, and/or supplemental CO2 systems to maintain environmental parameters at desired setpoints can be expensive. To balance operating costs with productivity, growers need to be able to predict how a crop will perform as a function of PPFD, CO2 concentration, and temperature. Determining which combinations can provide similar net photosynthetic rates (Pn) can assist growers in evaluating which combination is most cost-effective for their operation. Net photosynthetic rate is a useful proxy for a plant’s instantaneous productivity and is about proportional to dry matter accumulation (Seginer et al., 1991).
Plant photosynthetic response to PPFD, or photosynthetically active radiation (PAR) intensity, has been characterized in many studies, both empirically (e.g., Acock, 1991; Johnson and Thornley, 1984; Pachepsky and Acock, 1996) and biochemically (see review by von Caemmerer et al., 2009). Additional studies have examined photosynthetic response to CO2 concentration and temperature within the context of increasing atmospheric CO2 concentration and climate change (e.g., Idso et al., 1987), although much of this research has focused on field crops (Bunce and Heichel, 1986). Research focused on measuring photosynthetic responses of greenhouse crops has typically varied only one factor (irradiance, temperature, or CO2 concentration) at a time. Boldt et al. (2014) measured leaf photosynthetic response, singly, to irradiance, CO2 concentration, and leaf temperature of 13 herbaceous ornamentals, and the models were incorporated into a decision-support tool for floriculture growers, PhotoSim [U.S. Department of Agriculture, Agricultural Research Service (USDA-ARS), 2019]. Erwin and Gesick (2017) measured leaf photosynthetic response of 10 taxa of leafy greens grown in greenhouses to PPFD and CO2 concentration. In both Boldt et al. (2014) and Erwin and Gesick (2017), CO2 concentration was maintained at 400 µmol·mol−1 for the light response curves, and PPFD was maintained at 300 µmol·m−2·s−1 for the CO2 response curves. Therefore, data gathered in these studies allowed for modeling of Pn along two intersecting lines in a domain of the variables CO2 concentration and PPFD, but not in the rest of the plane. Stanghellini and Bunce (1993) measured leaf Pn of tomato (Solanum lycopersicum) at all combinations of CO2 concentration (100 to 1000 µmol·mol−1), PPFD (50 to 2000 µmol·m−2·s−1), and temperature (18, 25, and 32 °C), providing insight into the response of Pn over the PPFD and CO2 concentration domain, as well as their interaction with temperature. This allows Pn to be estimated via interpolation at any combination of CO2 concentration, temperature, and PPFD within the domain tested, which encompasses the range of parameters likely to occur for tomato production. Because it is known [e.g., results from Boldt et al. (2014) and Erwin and Gesick (2017)] that photosynthetic responses vary significantly between species, similar data for additional plant species are warranted.
It would be useful to incorporate PPFD, CO2 concentration, temperature, and their interactions into a single equation to estimate how changing multiple greenhouse environmental parameters will affect Pn. Cannell and Thornley (1998) and Körner et al. (2009) have shown clear interactive effects between these parameters. Cannell and Thornley (1998) developed a mathematical model to calculate gross photosynthetic rate as a function of temperature, CO2 concentration, and PPFD. They did so by developing relationships for quantum yield (α) and maximum gross photosynthetic rate (Pmax) as functions of CO2 concentration and temperature. These parameters, along with a fitting parameter, θ, characterized the rate of gross photosynthesis in response to PPFD using a nonrectangular hyperbolic function. Although the model developed by Cannell and Thornley (1998), which we will refer to as the Cannell and Thornley model, provides great conceptual value, the generic parameter values included in the original publication were not meant to accurately predict photosynthetic rate for any specific plant species.
The growth environment before measurement is another important factor affecting Pn, especially PPFD and air temperature. Plants grown at higher PPFDs often, but not always, have higher photosynthetic capacities (Bunce and Heichel, 1986). The temperature optimum (Topt) of plants grown at higher temperatures is often higher than plants grown at lower temperatures (Berry and Björkman, 1980).
Therefore, the objectives of our study were to 1) measure the combined effects of light intensity and CO2 concentration on Pn of plants grown at different PPFDs; 2) measure the combined effects of PPFD, leaf temperature, and CO2 concentration on Pn of plants grown at different air temperatures; and 3) fit the data to model Pn as a function of leaf temperature, PPFD, and CO2 concentration.
Materials and Methods
Plant care and treatments
Expt. 1.
Calibrachoa (Calibrachoa ×hybrida ‘Superbells Lemon Slice’), petunia (Petunia ×hybrida ‘Supertunia Mini Strawberry Pink Veined’), and verbena (Verbena ×hybrida ‘Superbena Royale Whitecap’) cuttings (Pleasant View Gardens, Loudon, NH) in 84-cell count trays were trimmed and transplanted into 11.4-cm-diameter pots filled with soilless substrate (LC1; Sun Gro Horticulture, Bellevue, WA) and grown for 14 d in a growth chamber before the start of treatments. The PPFD was ≈270 to 300 µmol·m−2·s−1, the photoperiod was 14 h, and air temperature was 22/18 °C day/night. Plants were watered as needed and fertilized at every irrigation with 20N–4.4P–16.6K (Jack’s 20–10–20; JR Peters, Allentown, PA) at 150 mg·L−1 N.
After 14 d, six plants of each species were moved to growth chambers that provided a PPFD of 90, 180, or 270 µmol·m−2·s−1 from cool-white fluorescent lamps. The photoperiod was 16 h, and daily light integrals (DLIs) were 5.2, 10.4, and 15.6 mol·m−2·d−1, respectively. Air temperature remained at 22/18 °C day/night. Flowers were removed daily to minimize shading of developing leaves. Photosynthetic measurements commenced after 20 d (verbena), 28 d (petunia), and 42 d (calibrachoa) in the lighting treatments. Each crop required a different length of time for leaves initiated after plants were moved into the lighting treatments to expand enough to cover the measuring area of the instrument cuvette (described as follows).
Expt. 2.
Geranium (Pelargonium ×hortorum ‘Maverick Red’), pepper (Capsicum annuum ‘California Wonder’), and sunflower (Helianthus annuus ‘Pacino Gold’) (Ball Seed Co., West Chicago, IL) were sown in 288-cell count trays filled with soilless substrate (LC1). Seedlings were initially watered with reverse-osmosis water until true leaves emerged, at which point they were fertigated at every irrigation with 20N–4.4P–16.6K at 75 mg·L−1 N. Initial PPFD in the growth chamber was ≈150 µmol·m−2·s−1, the photoperiod was 12 h, and air temperature was 22/18 °C day/night. Sunflower and geranium were transplanted 24 d after sowing, and pepper was transplanted 31 d after sowing into 11.4-cm-diameter pots filled with soilless substrate (LC1). After transplant, PPFD was increased to 250 µmol·m−2·s−1, photoperiod was increased to 14 h, and the nutrient solution concentration was increased to 150 mg·L−1 N.
Three (pepper) or four (sunflower and geranium) weeks after transplant, seven plants of each species were moved to growth chambers set at constant air temperatures of 14, 21, or 28 °C. The PPFD remained at 250 µmol·m−2·s−1 and the photoperiod was increased to 16 h, resulting in a DLI of 14.4 mol·m−2·d−1. Photosynthetic measurements commenced after 14 d (sunflower), 21 d (pepper), and 29 d (geranium) in these treatments. As noted previously, each crop required a different length of time for leaves initiated after plants were moved into temperature treatments to expand enough for photosynthetic measurements.
Photosynthetic measurements
In both experiments, Pn was measured once during production for each crop and treatment combination, and all plants in a crop were evaluated before proceeding to the next crop. Five plants per species per treatment (DLI or temperature) were selected in random order for measurement. A single leaf from each plant was used. Recently emerged, fully expanded, unshaded leaves were chosen whenever possible. However, recently emerged leaves on calibrachoa and petunia were often too small to cover the measurement area, and this occurred in all three DLI treatments in Expt 1. In those instances, larger, but not necessarily recently emerged, leaves were selected instead. In Expt. 2, geraniums grown at 28 °C exhibited high temperature-induced leaf yellowing and bleaching on leaves that emerged while at 28 °C. Preliminary tests on the yellowed leaves resulted in negligible Pn, therefore, older (greener) leaves of geranium grown at 28 °C were used instead.
Photosynthetic response curve measurements were collected using a portable photosynthesis system (LI-6400XT; LI-COR Biosciences, Lincoln, NE). Leaf temperature, flow rate, PPFD, and reference CO2 concentration were precisely controlled in a leaf chamber. A cuvette with a 6-cm2 area and red and blue light-emitting diodes [LEDs (6400-02B, LI-COR Biosciences)] was used for most species. A cuvette with a 2-cm2 measurement area (6400-40, LI-COR Biosciences) was used for calibrachoa and petunia because of their small leaf size. The LI-6400XT reports Pn on a leaf area basis and automatically accounts for a change in measurement area when using cuvettes with different apertures. Both LED sources emitted a red:blue light ratio of 90:10. Leaf temperature was maintained at ambient growth chamber air temperature (22 °C for calibrachoa, petunia, and verbena, and 14, 21, or 28 °C for geranium, pepper, and sunflower).
Measurements were collected using an autoprogram that cycled through a predetermined sequence of conditions, logging Pn at each step once stability parameters were attained. The leaf was placed in the measurement chamber and acclimated to initial conditions of 400 µmol·mol−1 CO2 and 2000 µmol·m−2·s−1 PPFD. Net photosynthesis was then measured at 14 light intensities, ranging from 2000 to 0 µmol·m−2·s−1, and four reference CO2 concentrations (200, 400, 600, and 800 µmol·mol−1). For each leaf, each light response curve was measured at 400 µmol·mol−1 CO2, then 200, 600, and 800 µmol·mol−1 CO2. Once the leaf was acclimated, measurements were collected in descending PPFD (2000, 1500, 1000, 800, 700, 600, 500, 400, 300, 200, 100, 50, 25, and 0 µmol·m−2·s−1). An adjustment period of at least 5 min was programmed at 2000 µmol·m−2·s−1 PPFD between each light response curve at each CO2 concentration, and in addition, two consecutive Pn measurements were taken at 2000 µmol·m−2·s−1 PPFD to verify that Pn had indeed stabilized. Measurements were logged once stability parameters were attained (minimum and maximum wait times of 45 and 90 s, respectively, were defined when using the 6-cm2 cuvette, and 60 and 90 s, respectively, when using the 2-cm2 cuvette). Stability criteria included the rate of change for Pn, reference CO2 concentration, reference H2O concentration, and flow rate.
Data analysis
Parameters in Eq. [1] (Pmax, α, Rd, and θ) were fit by minimizing the sum of squares between the data points and the values modeled for measurements taken at each growth environment [GE (DLI for Expt. 1 and temperature for Expt. 2)] and CO2 concentration combination. Data from each light response curve were fit independently. Regression was performed using the Ipopt solver [version 0.2.6 (Wächter and Biegler, 2006)] in the JuMP modeling language [version 0.17.1 (Dunning et al., 2017)] written in the Julia programming language [version 0.5.2 (Bezanson et al., 2012)].
The light saturation point (LSP) and light compensation point (LCP) were calculated for each light response curve. The LSP was defined as the light intensity at which Pn equaled 90% of the maximum net photosynthetic rate Pn,max (Pmax – Rd), as previously defined (Jurik et al., 1988; Ronquim et al., 2009). The LCP was the light intensity at which Pn equaled zero.
Analysis of variance was conducted in R (version 3.3.1; R Foundation for Statistical Computing, Vienna, Austria) on the NRH parameter values within each species and GE combination. Where significant (P < 0.05), Tukey’s honestly significant difference test (α = 0.05) was used for mean separation.
Data in Expt. 2 were also fit to a model developed by Cannell and Thornley (1998), which defined α and Pmax (as defined in Eq. [1]) as functions of temperature and CO2 concentration, as shown in Eq. [2] through Eq. [9]. They modeled gross photosynthesis, but by including Rd in Eq. 1, we modeled Pn instead. In addition, we modified the reference CO2 concentration from 350 to 400 µmol·mol−1 to reflect recent increases in atmospheric CO2 concentration, and we modified the reference temperature for α (Tref,α) from 15 to 14 °C to allow our data to be modeled using a continuous function. At temperatures below Tref, Cannell and Thornley (1998) assumed the value of α was constant in relation to temperature.
To fit the general model developed by Cannell and Thornley (1998) with the addition of the parameter Rd, the parameters αref, β, θ, cT,α, Pmax,ref, KCa,Pmax, Tmax,Pmax,400, Tmax,Pmax,800, T0, and Rd were determined using least squares regression to minimize differences between data for each species and the Cannell and Thornley model. Regression was performed as previously described for individual response curves.
Results
Expt. 1.
Verbena and petunia acclimated to higher light intensities (180 and 270 µmol·m−2·s−1 PPFD) had higher rates of Pmax compared with plants acclimated to 90 µmol·m−2·s−1 PPFD (Fig. 1). In verbena, Pmax ranged from 8.6 to 25.1 µmol·m−2·s−1 CO2 for plants grown at 90 µmol·m−2·s−1 PPFD, depending on CO2 concentration, but increased to 11.2 to 33.4 µmol·m−2·s−1 CO2 and 12.2 to 33.0 µmol·m−2·s−1 CO2 for plants grown at 180 and 270 µmol·m−2·s−1 PPFD, respectively (Table 1). In petunia, Pmax ranged from 6.9 to 17.9 µmol·m−2·s−1 CO2, 10.3 to 28.5 µmol·m−2·s−1 CO2, and 9.1 to 27.1 µmol·m−2·s−1 CO2 for plants grown at 90, 180, and 270 µmol·m−2·s−1 PPFD, respectively (Table 1). Results for calibrachoa were inconsistent between treatments, with the highest values of Pmax (6.4 to 17.4 µmol·m−2·s−1 CO2) occurring in plants grown at 90 µmol·m−2·s−1 PPFD and the lowest values of Pmax (3.8 to 9.4 µmol·m−2·s−1 CO2) occurring in plants acclimated to 180 µmol·m−2·s−1 PPFD (Fig. 1, Table 1).
Net photosynthesis (Pn) of calibrachoa ‘Superbells Lemon Slice’, petunia ‘Supertunia Mini Strawberry Pink Veined’, and verbena ‘Superbena Royale Whitecap’ (top, middle, and bottom rows, respectively) grown at photosynthetic photon flux densities (PPFDs) of 90, 180, and 270 µmol·m−2·s−1 (left, center, and right columns, respectively). Plants were measured at 14 PPFDs and four CO2 concentrations. Points plotted are the means (±SD) of each treatment (n = 5) at each PPFD. Lines are the modeled values using the equation Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 − 4αθPmax·I]0.5} − Rd, with the parameter values listed in Table 1.
Citation: Journal of the American Society for Horticultural Science 147, 2; 10.21273/JASHS05115-21
Net photosynthesis (Pn) of calibrachoa ‘Superbells Lemon Slice’, petunia ‘Supertunia Mini Strawberry Pink Veined’, and verbena ‘Superbena Royale Whitecap’ (top, middle, and bottom rows, respectively) grown at photosynthetic photon flux densities (PPFDs) of 90, 180, and 270 µmol·m−2·s−1 (left, center, and right columns, respectively). Plants were measured at 14 PPFDs and four CO2 concentrations. Points plotted are the means (±SD) of each treatment (n = 5) at each PPFD. Lines are the modeled values using the equation Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 − 4αθPmax·I]0.5} − Rd, with the parameter values listed in Table 1.
Citation: Journal of the American Society for Horticultural Science 147, 2; 10.21273/JASHS05115-21
Net photosynthesis (Pn) of calibrachoa ‘Superbells Lemon Slice’, petunia ‘Supertunia Mini Strawberry Pink Veined’, and verbena ‘Superbena Royale Whitecap’ (top, middle, and bottom rows, respectively) grown at photosynthetic photon flux densities (PPFDs) of 90, 180, and 270 µmol·m−2·s−1 (left, center, and right columns, respectively). Plants were measured at 14 PPFDs and four CO2 concentrations. Points plotted are the means (±SD) of each treatment (n = 5) at each PPFD. Lines are the modeled values using the equation Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 − 4αθPmax·I]0.5} − Rd, with the parameter values listed in Table 1.
Citation: Journal of the American Society for Horticultural Science 147, 2; 10.21273/JASHS05115-21
Nonrectangular hyperbola equation parameters [maximum photosynthetic rate (Pmax), quantum yield (α), θ, and dark respiration (Rd)] for photosynthetic light response curves of calibrachoa, petunia, and verbena grown at three photosynthetic photon flux densities (90, 180, and 270 µmol·m−2·s−1 PPFD) and measured at four CO2 concentrations (200, 400, 600, and 800 µmol·mol−1). The light compensation point {LCP [PPFD at which net photosynthesis (Pn) = 0]} and light saturation point [LSP (PPFD corresponding to 90% net Pmax)] were calculated from the nonrectangular hyperbola, Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 − 4αθPmaxI]0.5} − Rd. All fitted parameters and calculated values are mean ± sd (n = 5).
In all three species, Pmax generally increased with increasing CO2 concentration, although the magnitude of increase was generally less with higher CO2 concentrations. For example, Pmax of verbena increased 144%, 156%, and 158% from 200 to 400 µmol·mol−1 CO2, at 90, 180, and 270 µmol·m−2·s−1 PPFD, respectively, but increased only 20%, 16%, and 5%, respectively, from 400 to 600 µmol·mol−1 CO2, and there was almost no change from 600 to 800 µmol·mol−1 CO2. Differences in Pmax at different CO2 concentrations were significant (P < 0.05), with the exception of calibrachoa grown at 180 µmol·m−2·s−1 PPFD (Table 1). Most common was a lower value of Pmax at 200 µmol·mol−1 CO2 compared with higher CO2 concentrations.
Across all GE and CO2 concentrations, α ranged from 0.0425 to 0.0694 µmol CO2 per micromole photons in verbena, 0.0392 to 0.0729 in petunia, and 0.0361 to 0.0727 in calibrachoa (Table 1). These values fell within the range typically observed for α (0.040 to 0.075) at ambient (350 µmol·mol−1) CO2 (Cannell and Thornley, 1998), although they were slightly below this general range at sub-ambient CO2 concentrations. Where significant, α at 200 µmol·mol−1 CO2 was lower than at 600 and 800 µmol·mol−1 CO2, and similar to or lower than at 400 µmol·mol−1 CO2.
Dark respiration ranged from 0.53 to 1.28 µmol·m−2·s−1 CO2 in verbena, 0.45 to 1.34 µmol·m−2·s−1 CO2 in petunia, and 0.29 to 0.98 µmol·m−2·s−1 CO2 in calibrachoa (Table 1). In all three species, there was a trend for Rd to generally decrease with increasing CO2 concentration; however, significant differences were not detected within any of the combinations of species and GEs.
The LCP ranged from 8.9 to 25.6 µmol·m−2·s−1 PPFD in verbena, 7.3 to 30.7 µmol·m−2·s−1 PPFD in petunia, and 4.9 to 28.9 µmol·m−2·s−1 PPFD in calibrachoa (Table 1). The LCP generally decreased with increasing CO2 concentration, and differences were significant within all species and GE combinations, except for verbena grown at 180 µmol·m−2·s−1 PPFD.
The calculated LSP, while not compared statistically across light intensities (due to lack of chamber replication), was higher in verbena and petunia acclimated to 180 and 270 µmol·m−2·s−1 PPFD, compared with those acclimated at 90 µmol·m−2·s−1 PPFD (Table 1). The LSP generally increased as CO2 concentration increased from 200 to 600 µmol·mol−1, then decreased slightly from 600 to 800 µmol·mol−1.
Expt. 2.
Pepper and sunflower showed dramatically increased Pn as growth chamber temperature increased, especially from 14 to 21 °C (Fig. 2). In pepper, Pmax at the four CO2 concentrations evaluated ranged from 1.0 to 4.2 µmol·m−2·s−1 CO2, 7.6 to 19.6 µmol·m−2·s−1 CO2, and 8.3 to 26.7 µmol·m−2·s−1 CO2 at 14, 21, and 28 °C, respectively (Table 2). In sunflower, Pmax ranged from 8.2 to 30.0 µmol·m−2·s−1 CO2, 17.7 to 53.1 µmol·m−2·s−1 CO2, and 18.2 to 65.1 µmol·m−2·s−1 CO2 at 14, 21, and 28 °C respectively (Table 2). However, in geranium, higher rates of Pn were attained at 21 °C than at 14 or 28 °C. Lower rates of Pn at 28 °C were likely due to high temperature-induced leaf yellowing and photobleaching (loss of chlorophyll). In all three species, Pmax increased with increasing CO2 concentration, although the magnitude of increase was generally less at higher CO2 concentrations. For example, in sunflower at 21 °C, Pmax increased 150% from 200 to 400 µmol·mol−1 CO2, 20% from 400 to 600 µmol·mol−1, and by less than 1% from 600 to 800 µmol·mol−1. In most instances, Pmax at 200 µmol·mol−1 CO2 was significantly lower (P < 0.05) than the other three CO2 concentrations, and Pmax at 600 and 800 µmol·mol−1 CO2 were similar (Table 2).
Net photosynthesis (Pn) of pepper ‘California Wonder’, sunflower ‘Pacino Gold’, and geranium ‘Maverick Red’ (top, middle, and bottom rows, respectively) grown at air temperatures of 14, 21, and 28 °C (left, center, and right columns, respectively). Plants were measured at 14 photosynthetic photon flux densities (PPFDs) and four CO2 concentrations. Points plotted are the means (±SD) of each treatment (n = 5) at each PPFD. Lines are the modeled values using the equation Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 − 4αθPmax·I]0.5} − Rd, with the parameter values listed in Table 2.
Citation: Journal of the American Society for Horticultural Science 147, 2; 10.21273/JASHS05115-21
Net photosynthesis (Pn) of pepper ‘California Wonder’, sunflower ‘Pacino Gold’, and geranium ‘Maverick Red’ (top, middle, and bottom rows, respectively) grown at air temperatures of 14, 21, and 28 °C (left, center, and right columns, respectively). Plants were measured at 14 photosynthetic photon flux densities (PPFDs) and four CO2 concentrations. Points plotted are the means (±SD) of each treatment (n = 5) at each PPFD. Lines are the modeled values using the equation Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 − 4αθPmax·I]0.5} − Rd, with the parameter values listed in Table 2.
Citation: Journal of the American Society for Horticultural Science 147, 2; 10.21273/JASHS05115-21
Net photosynthesis (Pn) of pepper ‘California Wonder’, sunflower ‘Pacino Gold’, and geranium ‘Maverick Red’ (top, middle, and bottom rows, respectively) grown at air temperatures of 14, 21, and 28 °C (left, center, and right columns, respectively). Plants were measured at 14 photosynthetic photon flux densities (PPFDs) and four CO2 concentrations. Points plotted are the means (±SD) of each treatment (n = 5) at each PPFD. Lines are the modeled values using the equation Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 − 4αθPmax·I]0.5} − Rd, with the parameter values listed in Table 2.
Citation: Journal of the American Society for Horticultural Science 147, 2; 10.21273/JASHS05115-21
Nonrectangular hyperbola equation parameters [maximum photosynthetic rate (Pmax), quantum yield (α), θ, and dark respiration (Rd)] for photosynthetic light response curves of pepper, sunflower, and geranium grown at three air temperatures (14, 21, and 28 °C) and measured at four CO2 concentrations (200, 400, 600, and 800 µmol·mol−1). The light compensation point [LCP; photosynthetic photon flux densities (PPFD) at which net photosynthesis (Pn) = 0] and light saturation point (LSP; PPFD corresponding to 90% net Pmax) were calculated from the nonrectangular hyperbola, Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 - 4αθPmaxI]0.5} – Rd. All fitted parameters and calculated values are mean ± sd (n = 5).
Values of α ranged from 0.0137 to 0.0765 µmol CO2 per micromole photons in pepper, 0.0398 to 0.0693 in sunflower, and 0.0213 to 0.0460 in geranium (Table 2). Significant differences occurred in eight of the nine treatments (species × GE combinations), but the relationship was inconsistent. In general, α was lowest at 200 µmol·mol−1 CO2. While not statistically compared across temperatures due to lack of replication of GE, α tended to be lower at 14 °C compared with 21 or 28 °C but only at 200 µmol·mol−1 CO2.
Dark respiration ranged from 0.01 to 1.17 µmol·m−2·s−1 CO2 in pepper, 0.18 to 1.36 µmol·m−2·s−1 CO2 in sunflower, and 0.31 to 1.19 µmol·m−2·s−1 CO2 in geranium (Table 2). Although statistical differences were detected in only four of the nine treatment combinations, Rd tended to increase with temperature and typically decreased as CO2 concentration increased.
The LCP ranged from 0.4 to 29.3 µmol·m−2·s−1 PPFD in pepper, 3.2 to 26.4 µmol·m−2·s−1 PPFD in sunflower, and 8.7 to 38.8 µmol·m−2·s−1 PPFD in geranium (Table 2). The LCP generally decreased with increasing CO2 concentration. For example, in pepper at 28 °C, the LCP was 26.3, 15.5, 12.9, and 9.7 µmol·m−2·s−1 PPFD at 200, 400, 600, and 800 µmol·mol−1 CO2, respectively.
The LSP ranged from 143 to >2000 µmol·m−2·s−1 in pepper, 533 to >2000 µmol·m−2·s−1 in sunflower, and 803 to >2000 µmol·m−2·s−1 in geranium (Table 2). These values were calculated as the light intensity corresponding to 90% of net Pmax. Based on our fitted equation of Pn measured at light intensities of ≤2000 µmol·m−2·s−1 PPFD, some of the calculated values exceeded our upper bound. These extrapolated values may not be biologically realistic and, therefore, are reported in Table 2 as >2000 µmol·m−2·s−1 PPFD. In general, the LSP was higher at 21 and 28 °C, compared with 14 °C. The LSP generally increased with increasing CO2 concentration from 200 to 600, then either increased or decreased from 600 to 800.
Fitting to the Cannell and Thornley model.
Values of αref, β, θ, cT,α, Pmax,ref, KCa,Pmax, Tmax,Pmax,400, Tmax,Pmax,800, T0, and Rd were determined by the fitting procedure for geranium, pepper, and sunflower (Table 3). Rates of Pmax increased with CO2 concentration, although the magnitude of increase was incrementally smaller with increasing CO2 concentration. This behavior is captured by Eq. [6], in which fCa,Pmax, the factor representing the influence of CO2 concentration on Pmax, increased asymptotically due to increasing CO2 concentration. The measured behavior of Pmax was also well-represented by Eq. [7], in which Pmax increased up to a temperature optimum, above which it decreased. The fitted temperature optima fell between 20.9 and 26.5 °C (Table 3).
Fitted parameters for the Cannell and Thornley model used to model quantum yield (α) and maximum photosynthetic rate (Pmax) as functions of temperature and CO2 concentration, as well as the root mean squared error (RMSE) of the Cannell and Thornley model using these fitted parameters (RMSECannell-Thornley) and the average RMSE of each nonrectangular hyperbola, Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 − 4αθPmaxI]0.5]} for each species (Avg. RMSENRH).
Cannell and Thornley (1998) predicted α to decrease as temperature increased above 15 °C (Eq. [4], assuming a positive value for cT), based on a review of multiple studies. However, in the present study, α increased with temperature in pepper and geranium (Table 2). As a result, when fitting our data to the Cannell and Thornley model, the parameter cT, which represents the impact of temperature on α, was negative in geranium and pepper, instead of positive as expected. The Cannell and Thornley model also assumes α increases with CO2 concentration (Eq. [3]). Although our measured values of α at 200 µmol·mol−1 were consistently lower than at the higher CO2 concentrations, α did not consistently increase with CO2 concentrations above 400 µmol·mol−1. Mean α for a given species and GE was greater (in value, but not necessarily statistically significant) at a CO2 concentration of 400 µmol·mol−1 than at 600 or 800 µmol·mol−1 for 15 of 36 species and GE combinations.
The root mean square error (RSME) of the fitted model was compared with the average RMSE from the NRH functions fit to each species. Both values are presented in Table 3. The RMSE of the Cannell and Thornley model for sunflower was 1.57 vs. an average of 1.06 when data from each temperature and CO2 concentration combination were fit individually with an NRH. Likewise, the RMSE of the Cannell and Thornley model for pepper was 1.44 vs. an average of 1.21 when data from each temperature and CO2 concentration combination were fit individually with an NRH, and 1.29 vs. an average of 1.05, respectively, for geranium.
Discussion
Previous studies have developed photosynthetic response curves for greenhouse ornamentals to light (e.g., Abreu et al., 2014; Ronquim et al., 2009), CO2 (e.g., Wullschleger, 1993), temperature (e.g., Berry and Björkman, 1980), or some combination (e.g., Boldt et al., 2014; Erwin and Gesick, 2017; Iwakiri and Inayama, 1975), but typically only along lines in the PPFD, CO2 concentration, and temperature variable space; the response to one variable was measured while holding the others constant. For example, data could be used to predict plant responses to a change in light intensity or CO2 concentration where these lines intersect, but not a simultaneous change in both. Increasing (or decreasing) light intensity and CO2 simultaneously can result in interactive effects on Pn, especially in combination with changes to leaf temperature as well. Our study is one of the first to report instantaneous Pn measurements at multiple combinations of light intensity, CO2 concentration, and leaf temperature in greenhouse ornamentals, which can be used to model the interactive effects on Pn. Stanghellini and Bunce (1993) previously evaluated these same environmental variables (light intensity, CO2, and temperature) on tomato plants grown at ambient (350 µmol·mol−1). Although general trends were similar between our results and theirs, our results show significant species-to-species variation. This means that additional studies of this type are warranted to model the responses of more species.
All species evaluated displayed the characteristic photosynthetic light response curve shape regardless of the light intensity or temperature to which they were acclimated or the CO2 concentration during measurement. The NRH parameter values, LCPs, and LSPs (Tables 1 and 2) were generally within expected ranges but also showed substantial differences between results from previous studies measuring the same species. In general, Pmax of petunia, verbena, geranium, and sunflower was higher in the present study compared with Boldt et al. (2014), which may be because of the slightly higher growth and measurement temperature (21 to 22 °C, as compared with 20 °C for PhotoSim), cultivar, GE, and/or plant age. Photosynthetic responses to irradiance in field-grown irrigated pepper ‘Quadrato d’Asti’ reached an asymptote at ≈25 µmol·m−2·s−1 CO2 at a leaf temperature of 25 °C (Delfine et al., 2001). This is higher than the Pmax of 14.0 and 17.0 µmol·m−2·s−1 CO2 observed in our study at ambient CO2 (400 µmol·mol−1) and temperatures of 21 and 28 °C, respectively, and could be attributed to cultivar, higher irradiance in the field relative to a growth chamber, or differences in production systems. We did not find previous reports of light response curves on calibrachoa.
Previously, LCPs of 24 to 52 µmol·m−2·s−1 PPFD in herbaceous ornamentals (Boldt et al., 2014) and 13 to 73 µmol·m−2·s−1 PPFD in leafy greens (Erwin and Gesick, 2017) have been reported at ambient CO2 (400 µmol·mol−1), which were higher than the calculated LCP values in our study (4.9 to 30.7 µmol·m−2·s−1 in Expt. 1 and 0.4 to 38.8 µmol·m−2·s−1 in Expt. 2). The lowest LCPs calculated in our study were typically at 600 and 800 µmol·mol−1 CO2, which may explain why some of our LCPs were lower than those reported for other ornamentals. The LCP, the irradiance intensity at which photosynthetic carbon fixation offsets carbon losses from dark respiration and photorespiration, generally decreases as acclimated light intensity decreases or CO2 concentration increases. This corresponds with what we observed. The reduction in LCP as CO2 increased indicates that growers may be able to partially offset the impacts of very low ambient irradiance intensities early in the growing season by increasing CO2; however, this strategy may be more effective at the vegetative stage, when canopy shading is minimal (low leaf area index) and plant biomass is often a primary quality metric. At the reproductive stage, however, light intensity is an important contributor not just to biomass accumulation but flower initiation and time to flower as well, and may not be able to be offset by CO2 supplementation.
In general, the LSP increased with increasing CO2 concentration. At elevated CO2 concentration (>400 µmol·mol−1), some LSP values were quite high, above 2000 µmol·m−2·s−1 PPFD, and involved extrapolating outside the tested domain. Therefore, they are only reported as >2000 µmol·m−2·s−1 in Tables 1 and 2. In addition, values >2000 µmol·m−2·s−1 are unlikely in greenhouses or other controlled environments. The lower values of LSP calculated likely have more practical meaning to growers. For instance, in pepper at 14 °C and 200 µmol·mol−1 CO2, the measured LSP was only 214 µmol·m−2·s−1 PPFD. Such information can be useful for deciding on light, temperature, and CO2 set points during winter, especially at higher latitudes. At 14 °C and 200 µmol·mol−1 CO2, adding additional light above the LSP would likely be an unnecessary expense and not contribute to increased growth, unless the CO2 concentration was also concurrently increased or unless the crop canopy was quite dense, in which case the additional light would penetrate deeper into the canopy. Rather, increasing the ambient temperature would be more beneficial. It is worth noting that there are limitations in extrapolating individual leaf responses to whole-plant or canopy measures of Pn (Erwin and Gesick, 2017), e.g., plant age or leaf canopy denseness and depth (and subsequently, the degree of leaf shading). However, modeling these instantaneous measurements, along with knowledge of the crop age, growth stage, and the current environmental conditions they are acclimated to, can provide a starting framework to adjust temperature, irradiance, and CO2 to optimize growth and yield or maximize biomass production per cost of inputs.
In Expt. 1, we wanted to see if the light intensity of the growth chamber in which plants were grown before measurement impacted photosynthetic response. The increased Pmax observed in verbena and petunia acclimated to higher light intensities agreed with previous findings that the higher the light intensity during growth, the higher the photosynthetic capacity of the leaf will be (Bunce and Heichel, 1986). However, this was not the case with calibrachoa. It is possible that the inconsistent results between GE were due to selection bias. As noted in the methods, many of the recently mature calibrachoa leaves of plants grown at 180 and 270 µmol·m−2·s−1 PPFD were too small to cover even the smaller (2 cm2) cuvette used to measure Pn of calibrachoa and petunia; therefore, larger, older leaves (that unfolded and matured in the lighting environment) and did fully cover the small cuvette were measured in those instances. Studies (e.g., Iwakiri and Inayama, 1975) have shown that older leaves can have lower rates of Pn.
We observed that all three species (calibrachoa, petunia, and verbena) had larger, but fewer leaves at the lowest light intensity and smaller, but more tightly packed leaves at the higher light intensities. These differences can be attributed to morphological responses to light intensity. First, leaves acclimated to lower light intensities tend to be larger and thinner, compared with leaves acclimated to higher light intensities, as a way to maximize light capture (Poorter et al., 2019). Second, internode length often increases as light intensity decreases, leading to a less “tightly packed” appearance (Garcia and Lopez, 2020; Poorter et al., 2019).
In Expt. 2, Pmax increased with increasing GE temperature in sunflower and pepper, but not in geranium. At 14 °C, Pn in pepper was especially low. This shows that adding supplemental CO2 and/or light to pepper (a cold-sensitive plant) would increase Pn somewhat at 14 °C, but not nearly as much as increasing the temperature. In geranium, Pmax increased from 14 to 21 °C, then declined at 28 °C. Geraniums grown at 28 °C exhibited photobleaching on leaves that developed after placement in the temperature treatments, a physiological response sometimes observed in geranium in response to high temperature (Owen and Lindberg, 2017). Several measurements on the yellowed leaves resulted in very low rates of Pn. Therefore, older (but greener) leaves were selected for measurement. Also, the lower Pmax in geranium at 28 vs. 21 °C could also be attributed to its calculated temperature optimum of 22.5 °C at 400 µmol·mol−1 (Table 3), which is close to the 21 °C environment. The temperature optimum for plants in Expt. 2 was determined to be between 20.9 and 26.5 °C, based on regression performed for fitting the Cannell and Thornley model. This fits within the temperature optimum range of 11.1 to 30.8 °C that Boldt et al. (2014) reported for 13 herbaceous ornamentals.
This study may be the first, to our knowledge, that fits NRH parameter values to empirical data to model Pn as a function of light intensity, CO2 concentration, and leaf temperature, allowing for the estimation of Pn at any combination of these three variables within the domain tested. The Cannell and Thornley model was a good starting point, although we adjusted the reference values of temperature (14 °C instead of 15 °C) and CO2 (400 µmol·mol−1 instead of 350 µmol·mol−1) and included Rd to better model our scenario, as described in the materials and methods.
Contour plots for sunflower (Fig. 3) show a comparison of Pn response surfaces interpolated for light intensity and CO2 by the NRH functions (Fig. 3A) and the fitted Cannell and Thornley model (Fig. 3B). The similarities indicate either one could be used as a decision-support tool. However, the NRH surface plots are limited to the three temperatures evaluated, whereas the fitted Cannell and Thornley model could produce a plot at any temperature between 14 and 28 °C. The quality of fit of the Cannell and Thornley model to our experimental data (RMSE of 1.3 to 1.6, depending on species) may be adequate for a number of applications, including decision support for greenhouse climate control. The Cannell and Thornley model fits all the experimental data with a single model, as compared with fitting separate NRHs for each temperature and CO2 concentration, hence the greater RMSE compared with individual NRH models (comparison of RMSE in Table 3). However, the small tradeoff in accuracy may be worthwhile for applications that require predicting Pn at conditions that do not fall along one of the NRHs, as it would be experimentally impractical to measure and develop Pn response curves for irradiance and CO2 at each degree or couple of degrees Celsius for multiple species and cultivars.
Comparison of net photosynthesis (Pn) of sunflower ‘Pacino Gold’ modeled by fitting the nonrectangular hyperbola (NRH) equation to data at 14, 21, and 28 °C and each CO2 concentration (A) and by fitting the Cannell and Thornley model to the same data (B). The NRH equation: Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 − 4αθPmaxI]0.5} − Rd. The Cannell and Thornley model includes equations for α and Pmax as functions of temperature and CO2 concentration.
Citation: Journal of the American Society for Horticultural Science 147, 2; 10.21273/JASHS05115-21
Comparison of net photosynthesis (Pn) of sunflower ‘Pacino Gold’ modeled by fitting the nonrectangular hyperbola (NRH) equation to data at 14, 21, and 28 °C and each CO2 concentration (A) and by fitting the Cannell and Thornley model to the same data (B). The NRH equation: Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 − 4αθPmaxI]0.5} − Rd. The Cannell and Thornley model includes equations for α and Pmax as functions of temperature and CO2 concentration.
Citation: Journal of the American Society for Horticultural Science 147, 2; 10.21273/JASHS05115-21
Comparison of net photosynthesis (Pn) of sunflower ‘Pacino Gold’ modeled by fitting the nonrectangular hyperbola (NRH) equation to data at 14, 21, and 28 °C and each CO2 concentration (A) and by fitting the Cannell and Thornley model to the same data (B). The NRH equation: Pn = 1/2θ {αI + Pmax – [(αI + Pmax)2 − 4αθPmaxI]0.5} − Rd. The Cannell and Thornley model includes equations for α and Pmax as functions of temperature and CO2 concentration.
Citation: Journal of the American Society for Horticultural Science 147, 2; 10.21273/JASHS05115-21
Interestingly, we observed an increase in α with increased temperature in pepper and geranium, instead of the expected decrease. It is possible that differences in plant development during growth at different temperatures accounts for this. Peri et al. (2005) found that α decreased by ≈40% when leaf nitrogen (N) decreased from 4% to 2%, which corresponded to a decrease in total chlorophyll, from ≈0.6 to 0.1 g·m−2. Although we did not measure N directly, previous research [e.g., review by Sinclair and Horie (1989)] has shown a high correlation between Pn per unit leaf area and leaf N per unit leaf area. As such, it is possible that leaf N may have been lower at 14 °C for pepper and geranium, compared with 21 or 28 °C, and this relationship may be responsible for the increased values of α at the higher temperatures. Johnson et al. (2010) built on the Cannell and Thornley model using the results from Peri et al. (2005) by incorporating a factor for protein concentration into the expressions for α and Pmax. If leaf N had been measured, the incorporation of this factor may have improved the fit of the model.
One application of either the NRHs or the fitted Cannell and Thornley model is that they allow growers to assess various combinations of environmental conditions simultaneously, rather than individually. This provides them with the opportunity to see which sets of conditions would result in similar rates of Pn. In Fig. 3, a similar rate of Pn could be achieved at any point located on a contour within the plot. For instance, in sunflower grown at 21 °C (Fig. 3), a Pn of 20 µmol·m−2·s−1 CO2 is achievable at ≈600 µmol·mol−1 CO2 and 500 µmol·m−2·s−1 PPFD or at 300 µmol·mol−1 CO2 and 1000 µmol·m−2·s−1 PPFD. At a temperature of 14 °C, much higher levels of light intensity and CO2 concentration would be needed to attain the same Pn. Based on the species being grown, geographic location, time of year, greenhouse light transmission, electricity and CO2 costs, and other associated factors, a grower could use these plots, or an associated decision support tool that incorporates them, to assess which combination of light intensity, CO2 concentration, and temperature is most feasible and economically sound for them to achieve the desired Pn at a minimum cost. In the northern United States, low ambient irradiance is a limiting factor during winter and early spring greenhouse production. Supplemental lighting, historically from high-pressure sodium lamps but now also including LEDs, can be quite expensive to install and operate. If CO2 supplementation could partially or substantially reduce the need for supplemental lighting, even for a few hours per day when venting is not actively occurring or imminent, without reducing Pn, cost savings may be realized. This is likely to be more advantageous to growers producing leafy crops, ornamental or edible, or during the vegetative stage of crop growth, where the impact of DLI on flower initiation and development is not critical to crop timing.
In lieu of CO2 supplementation from liquid CO2 or CO2 generators, growers could temporally increase the CO2 concentration of the greenhouse, from sub-ambient to near ambient, by opening vents periodically and using fans to mix the air around the plant canopy to reduce the boundary layer. These strategies would be most feasible when venting is already scheduled for humidity or temperature control; if not, the tradeoff for increasing the CO2 concentration would be the loss of pre-heated air from the greenhouse and higher heating costs. A cost calculator, based on greenhouse location, time of year, and desired crop growth rates would benefit growers.
In addition, plant acclimation to supplemental CO2 should be considered. Although most crops exhibit an overall positive response to CO2 enrichment (Kimball, 1983), they will begin acclimating after a few weeks (McKinney and Craver, 2021; Yelle et al., 1990). Therefore, CO2 supplementation may only be early in the cropping cycle when plants are most responsive. For plants with short crop cycles, such as annual bedding plants, CO2 supplementation may be appropriate throughout production, especially if the greenhouse contains plants at multiple growth stages. Due to acclimation, photosynthetic and growth models developed from instantaneous changes in Pn to CO2 concentration will likely overestimate plant response as crop duration increases.
In summary, our data are one of the first studies to report instantaneous Pn measurements for ornamental species at multiple combinations of light intensity, CO2 concentration, leaf temperature, and GE, which can be used to model their interactive effects on Pn. Incorporation of these models into the existing PhotoSim decision-support tool (USDA-ARS, 2019) is planned, to provide growers the opportunity to use these models. The interactive effect of light intensity, CO2 concentration, and leaf temperature indicates that growers should consider lighting, temperature control, and CO2 decisions jointly. If ambient CO2 concentrations in a greenhouse or controlled environment are sub-ambient, and they often are (Erwin and Gesick, 2017; Mortensen, 1987), the expected increase in Pn with supplemental lighting will be tempered by limitations in CO2 fixation due to lack of available substrate. Similarly, low enzymatic activity at low ambient temperature can result in very low rates of Pn despite high CO2 concentration and PPFD. Therefore, as sole-source or supplemental lighting guidelines are developed for crops, simultaneous consideration of CO2 concentration and temperature will be critical to achieve the ideal rates of Pn, either based on maximizing photosynthetic rate or optimizing the return on investment of inputs.
Future work measuring and modeling Pn as a function of light intensity, CO2 concentration, and temperature would benefit from taking data points of these variables in combination, especially in relation to CO2 and temperature. Algorithm development and/or the use of surface plots could improve this type of modeling, as it would not require collection of data points at all possible combinations of variables and levels. In addition, including a sub-model to represent Rd as a function of temperature and CO2 concentration might improve the fit of the Cannell and Thornley model.
Literature Cited
Abreu, P.P., Souza, M.M., de Almeida, A.A.F., Santos, E.A., de Oliveira, J.C. & Figueiredo, A.L. 2014 Photosynthetic responses of ornamental passion flower hybrids to varying light intensities Acta Physiol. Plant. 36 1993 2004 https://doi.org/10.1007/s11738-014-1574-0
Acock, B 1991 Modeling canopy photosynthetic response to carbon dioxide, light interception, temperature and leaf traits 41 55 Boote, K.J. & Loomis, R.S. Modeling crop photosynthesis–From biochemistry to canopy Crop Sci. Soc. Amer., Special Publ. No. 19 Amer. Soc. Agron. Madison, WI https://doi.org/10.2135/cssaspecpub19.c3
Bezanson, J., Karpinski, S., Shah, V.B. & Edelman, A. 2012 Julia: A fast dynamic language for technical computing 26 July 2021. <https://arxiv.org/abs/1209.5145>
Berry, J. & Björkman, O. 1980 Photosynthetic response and adaptation to temperature in higher plants Annu. Rev. Plant Physiol. 31 491 543 https://doi.org/10.1146/annurev.pp.31.060180.002423
Boldt, J.K., Fisher, P.R. & Erwin, J.E. 2014 Development of photosynthetic response curves and their integration into a decision-support tool for floriculture growers HortScience 49 9 S389 (abstr.), https://doi.org/10.21273/HORTSCI.49.9S.S1
Bunce, J.A. & Heichel, G.H. 1986 Measurements and modeling of photosynthesis in field crops Crit. Rev. Plant Sci. 4 47 77 https://doi.org/10.1080/07352688609382218
Cannell, M.G.R. & Thornley, J.H.M. 1998 Temperature and CO2 responses of leaf and canopy photosynthesis: A clarification using the non-rectangular hyperbola model of photosynthesis Ann. Bot. 82 883 892 https://doi.org/10.1006/anbo.1998.0777
Delfine, S., Loreto, F. & Alvino, A. 2001 Drought-stress effects on physiology, growth and biomass production of rainfed and irrigated bell pepper plants in the Mediterranean region J. Amer. Soc. Hort. Sci. 126 297 304 https://doi.org/10.21273/JASHS.126.3.297
Dunning, I., Huchette, J. & Lubin, M. 2017 JuMP: A modeling language for mathematical optimization Soc. Ind. Appl. Math. Rev. 59 295 320 https://doi.org/10.1137/15M1020575
Erwin, J. & Gesick, E. 2017 Photosynthetic responses of swiss chard, kale, and spinach cultivars to irradiance and carbon dioxide concentration HortScience 52 706 712 https://doi.org/10.21273/HORTSCI11799-17
Garcia, C. & Lopez, R.G. 2020 Supplemental radiation quality influences cucumber, tomato, and pepper transplant growth and development HortScience 55 804 811 https://doi.org/10.21273/HORTSCI14820-20
Idso, S.B., Kimball, B.A., Anderson, M.G. & Mauney, J.R. 1987 Effects of atmospheric CO2 enrichment on plant growth: The interactive role of air temperature Agr. Ecosyst. Environ. 20 1 10 https://doi.org/10.1016/0167-8809(87)90023-5
Iwakiri, S. & Inayama, M. 1975 Studies on the canopy photosynthesis of the horticultural crops in controlled environment J. Agr. Meteorol. 30 161 166 https://doi.org/10.2480/agrmet.30.161
Johnson, I.R. & Thornley, J.H.M. 1984 A model of instantaneous and daily canopy photosynthesis J. Theor. Biol. 107 531 545 https://doi.org/10.1016/S0022-5193(84)80131-9
Johnson, I.R., Thornley, J.H.M., Frantz, J.M. & Bugbee, B. 2010 A model of canopy photosynthesis incorporating protein distribution through the canopy and its acclimation to light, temperature and CO2 Ann. Bot. 106 735 749 https://doi.org/10.1093/aob/mcq183
Jurik, T.W., Weber, J.A. & Gates, D.M. 1988 Effects of temperature and light on photosynthesis of dominant species of a northern hardwood forest Bot. Gaz. 149 203 208 https://doi.org/10.1086/337708
Kimball, B.A 1983 Carbon dioxide and agricultural yield: An assemblage and analysis of 430 prior observations Agron. J. 75 779 788 https://doi.org/10.2134/agronj1983.00021962007500050014x
Korczynski, P.C., Logan, J. & Faust, J.E. 2002 Mapping monthly distribution of daily light integrals across the contiguous United States HortTechnology 12 12 16 https://doi.org/10.21273/HORTTECH.12.1.12
Körner, O., Heuvelink, E. & Niu, Q. 2009 Quantification of temperature, CO2, and light effects on crop photosynthesis as a basis for model-based greenhouse climate control J. Hort. Sci. Biotechnol. 84 233 239 https://doi.org/10.1080/14620316.2009.11512510
McKinney, D. & Craver, J. 2021 The effect of CO2 enrichment timing and duration on pansy and petunia seedling quality for indoor production HortScience 56 9 S267 (abstr.), https://doi.org/10.21273/HORTSCI.56.9S.S1
Mortensen, L.M 1987 CO2 enrichment in greenhouses. Crop responses Scientia Hort. 33 1 25 https://doi.org/10.1016/0304-4238(87)90028-8
Owen, W.G. & Lindberg, H. 2017 Heat stress causes foliar bleaching and chlorosis of zonal geranium e-GRO Alert 6(30):1–4. 26 July 2021. <http://www.e-gro.org/pdf/2017_630.pdf>
Pachepsky, L.B. & Acock, B. 1996 An adequate model of photosynthesis—II. Dependence of parameters on environmental factors Agr. Syst. 50 227 238 https://doi.org/10.1016/0308-521X(94)00052-S
Peri, P.L., Moot, D.J. & McNeil, D.L. 2005 Modelling photosynthetic efficiency (α) for the light-response curve of cocksfoot leaves grown under temperate field conditions Eur. J. Agron. 22 277 292 https://doi.org/10.1016/j.eja.2004.04.002
Poorter, H., Niinemets, Ü., Ntagkas, N., Siebenkäs, A., Mäenpää, M., Matsubara, S. & Pons, T.L. 2019 A meta-analysis of plant responses to light intensity for 70 traits ranging from molecules to whole plant performance New Phytol. 223 1073 1105 https://doi.org/10.1111/nph.15754
Ronquim, C.C., Prado, C.H.B.A. & de Souza, J. P. 2009 Growth, photosynthesis and leaf water potential in young plants of Copaifera langsdorffii Desf. (Caesalpiniaceae) under contrasting irradiances Braz. J. Plant Physiol. 21 197 208 https://doi.org/10.1590/S1677-04202009000300004
Seginer, I., Shina, G., Albright, L.D. & Marsh, L.S. 1991 Optimal temperature setpoints for greenhouse lettuce J. Agr. Eng. Res. 49 209 226 https://doi.org/10.1016/0021-8634(91)80040-L
Sinclair, T.R. & Horie, T. 1989 Leaf nitrogen, photosynthesis, and crop radiation use efficiency: A review Crop Sci. 29 90 98 https://doi.org/10.2135/cropsci1989.0011183X002900010023x
Stanghellini, C. & Bunce, J.A. 1993 Response of photosynthesis and conductance to light, CO2, temperature and humidity in tomato plants acclimated to ambient and elevated CO2 Photosynthetica 29 487 497
U.S. Department of Agriculture, Agricultural Research Service 2019 PhotoSim 26 July 2021. <https://data.nal.usda.gov/dataset/photosim>
von Caemmerer, S., Farquhar, G. & Berry, J. 2009 Biochemical model of C3 photosynthesis 209 230 Laisk, A., Nedbal, L. & Govindjee Photosynthesis in silico. Springer Dordrecht, Netherlands https://doi.org/10.1007/978-1-4020-9237-4_9
Wächter, A. & Biegler, L.T. 2006 On the implementation of an interior-point filter line-search algorithm for large-scale nonlinear programming Math. Program. 106 25 57 https://doi.org/10.1007/s10107-004-0559-y
Wullschleger, S.D 1993 Biochemical limitations to carbon assimilation in C3 plants—A retrospective analysis of the A/Ci curves from 109 species J. Expt. Bot. 44 907 920 https://doi.org/10.1093/jxb/44.5.907
Yelle, S., Beeson, R.C. Jr., Trudel, M.J. & Gosselin, A. 1990 Duration of CO2 enrichment influences growth, yield, and gas exchange of two tomato species J. Amer. Soc. Hort. Sci. 115 52 57 https://doi.org/10.21273/JASHS.115.1.52
