AOBPreview originally published online on March 28, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Annals of Botany 91: 771-781, 2003
© 2003 Annals of Botany Company
A Coupled Model of Photosynthesis, Stomatal Conductance and Transpiration for a Rose Leaf (Rosa hybrida L.)
1 Environmental Horticulture, University of California, Davis, CA 95616, USA
* For correspondence. Fax +1 (530) 752 1819, e-mail jhlieth{at}ucdavis.edu 
Present address: Alternate Crops and Systems Laboratory,USDA-ARS, Bldg. 001 Rm 342, BARC-W, Beltsville, MD 20705, USA.
Received: 20 November 2002; Returned for revision: 6 January 2003 ; Accepted: 3 February 2003 Published electronically: 27 March 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMODEL DESCRIPTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXLITERATURE CITED |
|---|
The following three models were combined to predict simultaneously photosynthesis, stomatal conductance, transpiration and leaf temperature of a rose leaf: the biochemical model of photosynthesis of Farquhar, von Caemmerer and Berry (1980, Planta 149: 78–90), the stomatal conductance model of Ball, Woodrow and Berry (In: Biggens J, ed. Progress in photosynthesis research. The Netherlands: Martinus Nijhoff Publishers), and an energy balance model. The photosynthetic parameters: maximum carboxylation rate, potential rate of electron transport and rate of triose phosphate utilization, and their temperature dependence were determined using gas exchange data of fully expanded, young, sunlit leaves. The stomatal conductance model was calibrated independently. Prediction of net photosynthesis by the coupled model agreed well with the validation data, but the model tended to underestimate rates of stomatal conductance and transpiration. The coupled model developed in this study can be used to assist growers making environmental control decisions in glasshouse production.
Key words: Rosa hybrida L., photosynthesis, stomatal conductance, transpiration, coupled model, cut-flower, crop simulation, calibration, validation.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMODEL DESCRIPTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXLITERATURE CITED |
|---|
Crop simulation models are invaluable tools for optimization of the glasshouse microclimate and for making cultural decisions for increasing production and profit of glasshouse crops, including cut-flower roses. As one of the most important modules of such crop simulation models, a photosynthesis model should be comprehensive, encompassing all of the major variables that can be controlled and/or monitored using the environmental control system in the glasshouse. This raises the need for a mechanistic model of photosynthesis. The biochemical model of photosynthesis for C3 leaves by Farquhar, von Caemmerer and Berry (FvCB model) (Farquhar et al., 1980) has been adopted extensively in studies of ecological and physiological modelling, including studies on the effect of elevated CO2 on plant productivity (Medlyn et al., 1999). A major advantage of using the biochemical model is that it is mechanistic and, therefore, capable of describing underlying processes that might not be well described by simple empirical approaches. The disadvantage of the model has been that it requires rather extensive calibration of a number of parameters (Wullschleger, 1993; Cannell and Thornley, 1998). Application of the biochemical model in crop simulation models for agricultural or horticultural purposes has been limited due to the complex parameterization procedure. Fortunately, recent development of highly sophisticated gas-exchange systems has made the process of estimating model parameters easier. Also, the number of parameters to be fitted can be reduced by assuming that some are invariant across species of C3 plants.
Estimation of substomatal CO2 partial pressure (Ci) from given atmospheric CO2 (Ca) is critical for practical use of the model because the FvCB model operates using Ci instead of Ca. Ci may be estimated to be proportional to Ca under certain conditions. The ratio of Ci/Ca is assumed to be constant (0·7–0·8) in some studies owing to the complexity of Ci estimation. However, use of the fixed Ci/Ca ratio may not be appropriate for dynamic crop simulation models, as they should cover a variety of conditions where a fixed ratio may not be valid. To be useful in predicting gas exchange responses to varying environmental conditions, a photosynthesis model should be integrated with a model describing stomatal conductance (gs) so as to obtain realistic estimates of Ci. This allows a coupling of the supply function of diffusion through the stomata to the demand function of the CO2 fixation reaction.
A coupled approach to photosynthesis–stomatal conduct ance–transpiration modelling has been proposed (Collatz et al., 1991; Harley et al., 1992; Leuning et al., 1995; Nikolov et al., 1995) that combines the FvCB photosynthesis model with a model of stomatal conductance (Ball et al., 1987; Leuning, 1995) and an energy budget equation. This coupled-model approach can describe the photosynthetic behaviour of leaves by taking into account the biochemical limitation for CO2 (demand) as well as the stomatal limitation to supply of CO2. Sharkey (1985) included the rate of triose phosphate utilization (TPU) as one of the important biochemical limitations in photosynthesis. Harley et al. (1992) implemented the TPU limitation in their model.
A model for photosynthesis of rose leaves as a function of photosynthetically active radiation (PAR), leaf temperature and leaf age was previously developed for the rose variety Cara Mia (Lieth and Pasian, 1990). The model did not include CO2 as a driving variable. The FvCB model has been used to study the photosynthetic properties of rose canopy (Gonzalez-Real and Baille, 2000). Coupled gas exchange models have rarely been developed for horticultural and/or ornamental crops such as roses.
The objective of this study was to formulate and test a coupled model of photosynthesis, stomatal conductance and transpiration for rose by combining widely accepted sub-models. The coupled model is to be used as a module in a rose crop simulation model.
|
MODEL DESCRIPTION |
|---|
|
TOPABSTRACTINTRODUCTIONMODEL DESCRIPTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXLITERATURE CITED |
|---|
Photosynthesis, stomatal conductance and energy balance sub-models
The models are summarized in the Appendix, and parameters and variables listed. The biochemical model for C3 photosynthesis by Farquhar et al. (1980) was used as modified by Harley et al. (1992) and de Pury and Farquhar (1997). The parameters of photosynthetic capacity: maximum carboxylation rate (Vcmax), potential rate of electron transport (Jmax) and rate of triose phosphate utilization (Pu) were modelled to account for the effect of leaf age, assuming that all three parameters are dependent on leaf age to the same extent [eqn (A10)].
The stomatal conductance model [eqn (A11)] proposed by Ball, Woodrow and Berry (BWB model) (Ball et al., 1987) was calibrated and tested for rose leaves. Humidity at the leaf surface (hs) was obtained by applying a quadratic equation [eqn (A16)]. The energy balance equation was used to estimate leaf temperature (TL) as a function of stomatal conductance (gs), boundary layer conductance (gb) and the environmental variables: air temperature, absorbed long-wave and short-wave radiation, and relative humidity (Campbell and Norman, 1998). The leaf temperature was determined iteratively using a linear solution of the energy budget equation [eqn (A18)] using the Newton–Raphson method. Assuming that the water vapour pressure inside the leaf is the same as saturation vapour pressure (es) at the leaf temperature, the rate of transpiration was calculated using the diffusion equation [eqn (A23)].
Coupling the models
The FvCB model uses Ci [eqns (A2) and (A3)] and TL [eqns (A7)–(A9)], among others, as driving variables. The BWB model requires the net photosynthetic rate (A) as an input [eqn (A11)], while Ci results from the interaction of A and gs [eqn (A24)]. TL is estimated iteratively from a linear solution of the energy budget equation [eqn (A18)] using air temperature (Ta), and the conductances for heat (gh) and water vapour (gv) as input variables. The diffusion equation is used to relate Ca, Cs and Ci using A, gs and gb [eqns (A13) and (A24)]. Therefore, the three sub-models (FvCB, BWB and energy balance) are interdependent. A nested iterative procedure was used to solve this relation numerically (Fig. 1). Initially, TL and Ci were assumed to be equal to Ta and 0·7Ca, respectively, so as to obtain an estimate of A, which was then used to obtain gs. Ci was estimated using the resulting A and gs [eqn (A24)]. This process was solved iteratively using the Newton–Raphson method until Ci was stable. Subsequently, TL was computed using Ta and gs [eqn (A18)] and compared with the initial TL. When the new TL agreed to within 0·001 °C with the initial TL, the iteration was assumed to have converged.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMODEL DESCRIPTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXLITERATURE CITED |
|---|
Plant material
Fifteen rose plants (Rosa hybrida L. ‘Kardinal’) grafted onto ‘Natal Brier’ rootstock and transplanted into 13 l pots in May 1997 were used for the leaf photosynthesis measurements for calibration and validation of the model. A potting mix containing sand, redwood sawdust, and peat moss (1 : 1 : 1, v/v) was used as growing medium. Tensiometer-based irrigation was used to control root-zone moisture tension, with set-points of 1·0 and 3·0 kPa (Oki et al., 2001), delivering a modified half-strength Hoagland’s solution. Plants were grown in the glasshouse at the Department of Environmental Horticulture at the University of California (Davis, USA). Air temperature set-points inside the glasshouse were 24/20 °C day/night.
Gas exchange measurements
A photosynthesis system (LI-6400; LI-COR, Lincoln, NE, USA) with a red/blue LED light source (LI6400-02B) mounted onto a 6-cm2 clamp-on leaf chamber was used to determine light and A/Ci responses under various environmental conditions. Terminal leaflets of randomly selected, fully developed, young sunlit leaves (approx. 20 d from unfolding) of the flowering shoots (developed between September and November 2000) were used. For generation of A/Ci and light response curves, an automated protocol built into LI-6400 was used. The programme was configured to advance to the next step if the sum of the three coefficients of variation (CO2, water vapour and flow rate) was less than 0·3 %, with minimum wait time of 3 min. Each leaf was equilibrated to the initial conditions by waiting at least 5 min before executing the automated protocol. The photosynthetic response to Ci of 15 individual leaves was measured at 0, 50, 100, 200, 300, 400, 600, 800, 1200 and 1500 µmol mol–1 CO2 at 1500 µmol m–2 s–1 of incident PAR at a leaf temperature of 25 °C and relative humidity (ha) of approx. 50 %. A/Ci response measurements were started at ambient conditions, decreased to nearly the compensation point, returned to ambient, and then increased to higher concentrations to ensure that the stomata stayed open throughout the measurements. The light response of nine leaves was determined at several PAR levels between 0 and 2000 µmol m–2 s–1 at 25 °C leaf temperature and 350 µmol mol–1 CO2 inside the leaf chamber. For light response curves, measurements started with a leaf equilibrated to high light and the light level was then gradually decreased.
The A/Ci response of a total of 54 leaves was investigated at various leaf temperatures (10, 15, 20, 25, 30, 35 and 40 °C) to determine the temperature dependence of the photosynthetic parameters. The photosynthesis system (LI-6400) was able to control the leaf temperatures between 20 and 30 °C under growing conditions in the glasshouse. Growth chambers were used to provide conditions resulting in leaf temperatures below 20 °C or above 30 °C. Whenever a growth chamber was used, plants were moved into it at least 2 h before measurements to allow them time to acclimate.
Using 21 leaves, the response of gs to relative humidity (0·05–0·90), PAR (>100 µmol m–2 s–1), leaf temperature (10–40 °C) and a range of CO2 levels (>50 µmol mol–1) was determined to calibrate the stomatal conductance model. Relative humidity was controlled either by using the bypass valve on the desiccant tube containing anhydrous calcium carbonate (Drierite; W.A. Hammond Drierite Company, Ltd, Xenia, OH, USA), or by adjusting the flow rate of air through the leaf chamber. Measurements used to calibrate the stomatal conductance model were collected by waiting until the rates of Ci, transpiration and CO2 assimilation had stabilized before taking readings; this wait-time ranged from 5 to 30 min depending on the leaves and the environmental conditions of the chamber.
The dependence of the model on leaf age was calibrated using a separate data set collected in 1999. Leaf gas exchange measurements were made using a CIRAS-1 photosynthesis system (PP Systems, Hitchin, UK) between May and October 1999. The date on which 130 individual leaves unfolded was noted throughout the season, and the age of the leaves used for calibration was determined in ‘days after unfolding’. Measurements for characterizing leaf age effects were made in sunlight (>900 µmol m–2 s–1) at ambient CO2 (340–370 µmol mol–1) and growth temperature (24–27 °C) conditions in the glasshouse.
Additional measurements used in the model validation were made using the LI-6400 photosynthesis system between February and May 2001. These measurements were made under conditions that differed from the calibration conditions: A/Ci response at 70 and 200 µmol m–2 s–1 PAR; temperature response at 200 µmol m–2 s–1 PAR at three different CO2 levels (200, 370 and 1200 µmol mol–1); and light responses of leaves of different ages (8, 30, 68 and 180 d after unfolding). Other measurement conditions (e.g. automated protocol criteria) remained the same.
Model calibration and validation
Rather than fitting all parameters simultaneously, stepwise calibration of individual components of the model was performed so that each component of the model could be updated independently as needed. That is, the photosynthetic parameters (Vcm25, Jm25, Rd25) were first determined by fitting the biochemical model of photosynthesis (Farquhar et al., 1980) to the A/Ci response using measured Ci at controlled steady-state conditions where PAR was fixed at 1500 µmol m–2 s–1, relative humidity was around 50 %, and leaf temperature was controlled at 25 °C (Fig 2A). In addition, Vcmax and Jmax were estimated for individual leaves over a range of temperatures. Temperature dependence of Vcmax and Jmax was then determined by fitting eqns (A7) and (A8) with these estimates, respectively (Fig. 3). Temperature dependence of Pu was determined by fitting eqn (A7) with the net photosynthesis data collected between 10 and 20 °C at 1500 µmol mol–1 CO2 and 1500 µmol m–2 s–1 PAR, assuming that A is primarily governed by the rate of TPU when temperature is low while CO2 and light are not limiting (Sharkey, 1985). Pu25 was estimated by extrapolating the resulting equation (Fig. 3C). Temperature dependencies of Kc (Michaelis–Menten constant of Rubisco for CO2), Ko (Michaelis–Menten constant of Rubisco for O2), 
* (mitochondrial respiration in the light) and Rd (CO2 compensation point in the absence of Rd) were adopted from de Pury and Farquhar (1997), assuming that these parameters were invariant across species. Leaf age dependence, given by eqn (A10), was fitted with the data collected in the 1999 experiment (Fig. 2B). The parameter values of the stomatal conductance model were determined using the gas exchange data collected specifically for calibration of this sub-model. The data included a range of CO2 concentrations, relative humidity levels, PAR levels and temperatures, while those conditions where A might approach zero were excluded (Collatz et al., 1991). Non-linear regression with the Gauss–Newton method was used to estimate the parameter values of the model (Freund and Littell, 1991). No parameter values in the energy balance equation were specifically calibrated in this study. An attempt was also made to calibrate multiple parameters (Vcmax, Jmax, Pu and their temperature dependencies, and Rd) simultaneously by fitting the combined model with the pooled data. As the number of parameters increased, the attempt at fitting the model using a non-linear regression technique failed to obtain a set of converging parameters regardless of the optimization method.
|
|
Radiation absorbed by the leaf in the solar and thermal wavebands (Rabs) inside the leaf chamber was estimated using the following empirical relation (LI-COR, 1998):
Rabs = 0·5 
LEDkI0 + 
T4w(1)
where LED is the leaf absorptivity (0·84) averaged over the spectrum of the LED light source, k is an empirical conversion factor (0·19) from the incoming PAR level (I0) in the leaf chamber to total visible and near infrared energy, and Tw is the temperature of the chamber wall. Tw was assumed to be the same as air temperature (Ta) measured inside the leaf chamber.
Other photosynthetic parameters were obtained from de Pury and Farquhar (1997). The combined model was programmed with the computer programming language Pascal (Source code available from the authors upon request).
Linear regression of the model prediction on the observed values was used to evaluate the model performance. Significance of the linear regression, slope unity and zero intercept were tested. Bias and root mean square error (RMSE) were also evaluated as measures of model performance (Retta et al., 1991).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMODEL DESCRIPTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXLITERATURE CITED |
|---|
Calibration of the sub-models
A/Ci response of rose leaves, examined at a PAR of 1500 µmol m–2 s–1 and leaf temperature of 25 °C, followed typical A/Ci response patterns of C3 plants (Fig. 2A). Estimates of the photosynthetic parameters were 102·4 ± 2·04 (approximated standard error), 162·0 ± 2·04 and 1·26 ± 0·289 µmol m–2 s–1 for Vc25, Jm25 and Rd25, respectively. The photosynthesis sub-model described the photosynthetic response very well over a range of measured Ci at 25 °C. The transition from Ac to Aj occurred at Ci = 293 µbar. The limitation due to Ap did not appear to take place under this condition. Photosynthetic rates of young rose leaves increased rapidly up to 20 d after unfolding, then declined gradually as leaves aged (Fig. 2B). The model explained the temperature response of Vcmax and Jmax well (Fig. 3). The BWB model was capable of accounting for 70 % of the observed variation in measured stomatal conductance of calibration data (Fig. 4).
|
Prediction of A by the combined model
Prediction of A by the combined model is represented graphically against calibration data (Fig. 5). At 10 °C, A was insensitive to high CO2 levels. The combined model simulated this observed pattern well; that is, a linear increase up to the ambient level of CO2, followed by a flat line as CO2 increased further (Fig. 5A). The model predicted a flat response at high CO2 levels as a result of a limitation due to Ap. At 10 °C, the model predicted a nearly direct transition from the Ac- (linear increase) to the Ap-limited (flat response) region, with a brief period of Aj limitation (170–230 µbar of Ci) between the two regions. At 20 °C, the model behaved such that the transition from Ac to Aj occurred around 250 µbar and the transition between Aj and Ap took place around 700 µbar (Fig. 5A). At 30 °C, the limitation due to Ap was not realized at high CO2 concentrations (up to 1500 µbar), while the transition from Ac to Aj occurred at 320 µbar (Fig. 5B). At 40 °C, the model predicted that A was solely limited by Aj throughout all CO2 levels (Fig. 5B). The model successfully reproduced the observed pattern of A/Ci responses at all four leaf temperatures.
|
The photosynthetic response to leaf temperature was simulated fairly well over the entire range of temperatures at various ambient CO2 concentrations (Fig. 5C). At Ca = 1200 µbar, the optimal leaf temperature that yielded the maximal net photosynthetic rate of 37·0 µmol m–2 s–1 was around 32 °C. The optimal temperature decreased with decreasing ambient CO2 concentrations. Despite the fact that parameters related to the light response were not determined experimentally, the model response to PAR was also simulated fairly well, exhibiting saturation at a net photosynthetic rate of 21·8 µmol m–2 s–1 when incident PAR was above 1100 µmol m–2 s–1 and Ac started limiting A (Fig. 5D).
Model validation
The combined model predicted the observed pattern of A/Ci response quite well at both light flux densities (70 and 200 µmol m–2 s–1) (Fig. 6A). The model tended to underestimate A slightly at high CO2 concentrations. Light responses at high CO2 (Ca =1000 µbar) were also investigated (Fig. 6B), and the results from the model agreed with the observed light response pattern. The model was capable of predicting the observed pattern of temperature responses at three ambient CO2 concentrations with PAR at 200 µmol m–2 s–1 (Fig. 6C). The model also predicted the photosynthetic response to PAR at various leaf ages reasonably well (Fig. 6D). For a leaf age of 180 d, the model slightly overestimated A compared with observed values.
|
Performance of the combined model was evaluated by comparing observed values with predicted values. The regression F-value was significant for all variables listed in Table 1. The regression line slope deviated significantly from unity for all variables but A. The combined model successfully reproduced the observed response in A (r2 = 0·956). The model tended to underestimate both gs and E as indicated by negative bias values (Table 1).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMODEL DESCRIPTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXLITERATURE CITED |
|---|
Values for Vcmax of 102·4 and Jmax of 162·0 µmol m–2 s–1 were obtained for young rose leaves (‘Kardinal’) at 25 °C. Using young leaves of rose ‘Sonia’, Gonzalez-Real and Baille (2000) reported values of 66 µmol m–2 s–1 for Vcmax and 155·8 µmol m–2 s–1 for Jmax. Their Jmax value agrees reasonably well with present results, whereas their estimate of Vcmax is lower than that found here.
Various functions have been used to describe the temperature dependence of Vcmax and Jmax. For example, Harley et al. (1992) and Leuning (1995) employed a compound function with an optimum for both Vcmax and Jmax, whereas de Pury and Farquhar (1997) used an exponential growth function (the Arrhenius function) for Vcmax. In our parameterization of Vcmax, we observed a slight decline between 35 and 40 °C. However, we opted to use the Arrhenius function to describe the temperature dependence of Vcmax because it resulted in better overall performance for typical growing temperatures. In a recent survey, Leuning (2002) reported that functions describing the temperature response of the photosynthetic parameters Vcmax and Jmax at 25 °C show little variation between different species at leaf temperature <30 °C, while above this temperature variation is large and species-dependent. Bernacchi et al. (2001) published a set of modified temperature response functions of Rubisco-related parameters that improved predictions of Rubisco-limited A over the temperature range 10–40 °C.
The A/Ci response at low temperature (10 °C) shows that short-term increases of CO2 did not result in increased photosynthesis (Fig. 5A). We hypothesized that this response was the result of limitation because of TPU, and implemented this in the model following Harley et al. (1992). Inclusion of the TPU limitation greatly improved the model prediction of response to high CO2 at low temperature. Implementation of the TPU limitation in the model could be critical, especially for glasshouse crops, including cut-flower roses, because many commercial glasshouses practice CO2 enrichment during winter. The model could help growers decide whether or not to provide plants with additional CO2 under unusual temperature and light conditions. Some modellers have introduced smoothing factors for the transition between the limitations, as it appears, in reality, to be more gradual than that predicted by eqn (A1) (Collatz et al., 1991; Nikolov et al., 1995).
In our model, solutions for Ci and TL were obtained by numerical methods. In most cases, the number of iterations was kept below ten for both Ci and TL determinations when the model was tested for a range of environmental situations. A drawback of applying numerical solutions is that they are time- and resource-consuming. In addition, Baldocchi (1994) reported that iterative solutions for A become unstable when gb and * exceed critical values. While some modellers (Collatz et al., 1991; Harley et al., 1992) have used iterative solutions to couple photo synthesis–stomatal conductance models, others have successfully yielded analytical solutions for coupling processes so as to reduce the load of the iterations (Baldocchi, 1994; Nikolov et al., 1995; Wang and Leuning, 1998). Linking the processes through analytical solutions could speed up the model execution when used in extensive simulation tasks (e.g. simulation of monthly or annual canopy productivity for various environmental scenarios).
While the combined model yielded very good estimation of net photosynthesis, it failed to achieve high accuracy in predicting gs and E. There was considerable variation in estimation of gs that was not explained by the coupled model (Table 1; r2 = 0·491). The light and A/Ci response measurements in this study were made using the automated protocol built into the photosynthesis system (LI-6400). Despite the fact that fairly conservative criteria were used to advance to the next step after leaves had equilibrated to the initial condition, it is possible that second or later readings in the measurement series were taken before the stomata were fully acclimated to the new environment. Therefore, the performance of the model was also tested using only the first readings from each automated measurement series, since these readings were taken when leaves were equilibrated to the initial conditions. The test revealed that the model performed slightly better against the first readings than against all readings for predicting gs and E. That is, the slope of linear regression for E (= 0·881) was not different from unity. The r2-values for both gs and E were also improved. However, the slope for gs (= 0·651) was still significantly different from unity at P < 0·01 (Table 2). This indicates that some measurements might have been made before leaves entered steady state, and this could account, in part, for the discrepancy in performance of the model for gs and E when tested against all readings of validation data. This suggests that when data are to be used for calibration and validation of stomatal conductance models, gas exchange measurements using an automated protocol built into a photosynthesis system should be made with prolonged equilibration times. Leaves in the canopy are more likely to be exposed to dynamic changes of multiple environmental factors than to stay in steady state. A dynamic model rather than a steady-state model would be of more use for predicting stomatal conductance in dynamically changing environments. Models for dynamic stomatal responses to changing light regimes (Pearcy et al., 1997) and humidity (Jarvis et al., 1999) have been introduced.
|
In modelling stomatal conductance there is no consensus such as for the FvCB model in photosynthesis, and available models are basically empirical. Two main approaches have commonly been used in modelling stomatal conductance. The first, proposed by Jarvis (1976), is based on empirical stomatal responses to environmental conditions, such as radiation, vapour pressure deficit (VPD), temperature, soil water potential and CO2 concentration. The second commonly used approach is based on the observed link between stomatal conductance and photosynthesis (Wong et al., 1979). Ball et al. (1987) implemented this relation to model stomatal conductance as a function of CO2 concentration, relative humidity at the leaf surface (hs) and net photosynthesis. Lack of a mechanistic basis for using hs in the stomatal conductance model was criticized, and it was suggested that hs be replaced by VPD (Aphalo and Jarvis, 1991; Lloyd, 1991). Leuning (1995) modified the BWB model by replacing hs with VPD at the leaf surface (Ds) to allow for low intercellular CO2 concentrations by using (Cs –
) in the denominator so that the data when A 
0 could be included. We also tested a coupled model incorporating the Leuning model for rose leaves and found that it performed similarly to the BWB model (data not shown; see Kim, 2001 for details). BWB-type models (including the Leuning model) have been widely used because of their simplicity and plasticity in linking leaf photosynthesis to stomatal conductance (Collatz et al., 1991; Harley et al., 1992; Nikolov et al., 1995). The benefit of including the BWB model in a coupled model is that its variables can be determined from mechanistic photosynthesis and energy balance models (Baldocchi, 1994). It is also advantageous because stomatal responses to various factors can be realized indirectly through their effects on photosynthesis since the model operates as a function of A. The present model is to be extended to include water- and nutrient-related variables of both the root-zone and plant. The effect of soil and plant water status might be implemented through regulating the slope coefficient (m) in the stomatal model as a function of leaf water potential. Van Wijk et al. (2000) examined the applicability of the commonly used stomatal conductance models when a soil water stress function was incorporated. They reported that the slope coefficient of the BWB model was related to both air temperature and soil water content, while the Leuning- and Jarvis-type models showed a relationship only with soil water content. Thus, Van Wijk et al. (2000) concluded that the models incorporating VPD other than hs would be better suited for their study. Leaf nitrogen content can also be linked to photosynthetic parameters (Leuning et al., 1995; Gonzalez-Real and Baille, 2000). To date, information on the effect of other macro- and micronutrients on the gas exchange characteristics of leaves is not as complete as that for nitrogen.
In conclusion, the present coupled gas exchange model for a rose leaf is capable of predicting photosynthesis, intercellular CO2 concentration and leaf temperature as a function of radiation, air temperature, ambient CO2, leaf age and relative humidity, but predictions of stomatal conductance and transpiration are less satisfactory. The model has simple input and output structures and can be used as a module in a crop simulation model. As a stand-alone application the model can assist rose growers making glasshouse environmental control decisions.
|
ACKNOWLEDGEMENTS |
|---|
This research was supported in part by a grant from the Joseph Hill Foundation. We thank Dr M. Parrella for use of the photosynthesis system, and Drs L. Oki, J. Baker and G. McMaster for helpful comments on the manuscript.
|
APPENDIX |
|---|
|
TOPABSTRACTINTRODUCTIONMODEL DESCRIPTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXLITERATURE CITED |
|---|
|
|
|
|
LITERATURE CITED |
|---|
|
TOPABSTRACTINTRODUCTIONMODEL DESCRIPTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXLITERATURE CITED |
|---|
-
Aphalo PJ, Jarvis PG. 1991. Do stomata respond to relative humidity? Plant, Cell and Environment 14: 127–132.
Baldocchi D. 1994. An analytical solution for coupled leaf photosynthesis and stomatal conductance models. Tree Physiology 14: 1069–1079.[Web of Science][Medline]
Ball JT, Woodrow IE, Berry JA. 1987. A model predicting stomatal conductance and its contribution to the control of photosynthesis under different environmental conditions. In: Biggens J, ed. Progress in photosynthesis research. The Netherlands: Martinus Nijhoff Publishers.
Bernacchi CJ, Singsaas EL, Pimentel C, Portis ARJ, Long SP. 2001. Improved temperature response functions for models of Rubisco-limited photosynthesis. Plant, Cell and Environment 24: 253–259.[CrossRef]
Campbell GS, Norman JM. 1998. Introduction to environmental biophysics. 2nd edn. New York: Springer.
Cannell MGR, Thornley JHM. 1998. Temperature and CO2 responses of leaf and canopy photosynthesis: a clarification using the non-rectangular hyperbola model of photosynthesis. Annals of Botany 82: 883–892.
Collatz GJ, Ball JT, Grivet C, Berry JA. 1991. Physiological and environmental regulation of stomatal conductance, photosynthesis and transpiration: a model that includes a laminar boundary layer. Agricultural and Forest Meteorology 54: 107–136.[CrossRef][Web of Science]
de Pury DGG, Farquhar GD. 1997. Simple scaling of photosynthesis from leaves to canopies without the errors of big-leaf models. Plant, Cell and Environment 20: 537–557.[CrossRef]
Farquhar GD, von Caemmerer S, Berry JA. 1980. A biochemical model of photosynthetic CO2 assimilation in leaves of C3 species. Planta 149: 78–90.[CrossRef][Web of Science]
Freund RJ, Littell RC. 1991. SAS system for regression. Cary, NC: SAS Institute Inc.
Gonzalez-Real MM, Baille A. 2000. Changes in leaf photosynthetic parameters with leaf position and nitrogen content within a rose plant canopy (Rosa hybrida). Plant, Cell and Environment 23: 351–363.[CrossRef]
Harley PC, Thomas RB, Reynolds JF, Strain BR. 1992. Modelling photosynthesis of cotton grown in elevated CO2. Plant, Cell and Environment 15: 271–282.[CrossRef]
Jarvis AJ, Young PC, Taylor CJ, Davies WJ. 1999. An analysis of the dynamic response of stomatal conductance to a reduction in humidity over leaves of Cedrella odorata. Plant, Cell and Environment 22: 913–924.[CrossRef]
Kim S-H. 2001. Photosynthesis models and canopy management optimization in cut-flower roses. PhD Thesis. University of California, Davis, USA.
Leuning R. 1995. A critical appraisal of a combined stomatal-photosynthesis model for C3 plants. Plant, Cell and Environment 18: 339–355.[CrossRef]
Leuning R. 2002. Temperature dependence of two parameters in a photosynthesis model. Plant, Cell and Environment 25: 1205–1210.[CrossRef]
Leuning R, Kelliher FM, de Pury DGG, Schulze E-D. 1995. Leaf nitrogen, photosynthesis, conductance and transpiration: scaling from leaves to canopies. Plant, Cell and Environment 18: 1183–1200.[CrossRef]
LI-COR. 1998. Using the LI-6400: Portable photosynthesis system. 2nd edn. Lincoln, NE: LI-COR, Inc.
Lieth JH, Pasian CC. 1990. A model for net photosynthesis of rose leaves as a function of photosynthetically active radiation, leaf temperature, and leaf age. Journal of the American Society for Horticutural Science 115: 486–491.
Lloyd J. 1991. Modelling stomatal responses to environment in Macadamia integrifolia. Australian Journal of Plant Physiology 18: 649–660.[Web of Science]
Medlyn BE, Badeck FW, de Pury DGG, Barton CVM, Broadmeadow M, Ceulemans R, De Angelis P, Forstreuter M, Jach ME, Kellomaki S et al. 1999. Effects of elevated CO2 on photosynthesis in European forest species: a meta-analysis of model parameters. Plant, Cell and Environment 22: 1475–1495.[CrossRef]
Nikolov NT, Massman WJ, Shoettle AW. 1995. Coupling biochemical and biophysical processes at the leaf level: an equilibrium photosynthesis model for leaves of C3 plants. Ecological Modelling 80: 205–235.[CrossRef]
Oki LR, Lieth JH, Tjosvold SA. 2001. Irrigation of Rosa hybrida L. ‘Kardinal’ based on soil moisture tension increases productivity and flower quality. Acta Horticulturae 547: 213–219.
Pearcy RW, Gross LJ, He D. 1997. An improved dynamic model of photosynthesis for estimation of carbon gain in sunfleck light regimes. Plant, Cell and Environment 20: 411–424.[CrossRef]
Retta A, Vanderlip RL, Higgins RA, Moshier LJ, Feyerherm AM. 1991. Suitability of corn growth models for incorporation of weed and insect stresses. Agronomy Journal 83: 757–765.
Sharkey TD. 1985. Photosynthesis in intact leaves of C3 plants: physics, physiology and rate limitations. Botanical Review 51: 53–105.[CrossRef]
Van Wijk MT, Dekker SC, Bouten W, Bosveld FC, Kohsiek W, Kramer K, Mohren GMJ. 2000. Modeling daily gas exchange of a Douglas-fir forest: comparison of three stomatal conductance models with and without a soil water stress function. Tree Physiology 20: 115–122.[Abstract]
Wang YP, Leuning R. 1998. A two-leaf model for canopy conductance, photosynthesis and partitioning of available energy. I: Model description and comparison with a multi-layered model. Agricultural and Forest Meteorology 91: 89–111.[CrossRef][Web of Science]
Wong SC, Cowan IR, Farquhar GD. 1979. Leaf conductance in relation to rate of CO2 assimilation. Plant Physiology 78: 821–825.
Wullschleger SD. 1993. Biochemical limitation to carbon assimilation in C3 plants - A retrospective analysis of the A/Ci curves from 109 species. Journal of Experimental Botany 44: 907–920.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. F. Reynolds, W. L. Bauerle, and Y. Wang Simulating carbon dioxide exchange rates of deciduous tree species: evidence for a general pattern in biochemical changes and water stress response Ann. Bot., September 1, 2009; 104(4): 775 - 784. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. H. Fleisher, R. M. Shillito, D. J. Timlin, S.-H. Kim, and V. R. Reddy Approaches to Modeling Potato Leaf Appearance Rate Agron. J., April 11, 2006; 98(3): 522 - 528. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||