Modelling Stomatal Conductance,Transpiration And Photosynthesis Of Cut Lilium Grown In Greenhouses

Transpiration and photosynthesis are the important physiological processes with closely related to crop growth, yield and quality. As the bridge of these processes, the leaf stomatal conductance not only plays regulates transpiration and photosynthesis, but also serves as one of the most important crop parameters to estimate canopy transpiration rate and photosynthetic rate. Therefore, quantifying the effects of different conditions on stomatal conductance, transpiration and photosynthesis is useful and necessary for environment control and optimizing crop management. For this purpose, experiments of standard cut lilium(Lilium spp.’Sorbonne’) with different planting dates and water treatment levels were conducted in a solar greenhouse located in Lianyungang (34.7°N,119.5°E) and a multi-span greenhouse in Nanjing (32°N,118°E) from2008to2010. Our result showed that the model developed in this study gives satisfactory predictions of the stomatal conductance, transpiration rate and photosynthetic rate of cut lilium under different condtions, hence, can be used for environment control and optimizing management of cut lilum production. The research results are as follow:(1) Modelling stomatal conductance of cut lilium grown in greenhouses. We tested two models, i.e. the Jarvis model and a new version of BWB-type model (BWB-Leuning-Yin model), for estimating gsc of cut lilium grown in greenhouses. The models were parameterized from a subset of the experimental data. The remaining data sets for model validation were classified into two groups:group one was from experiments conducted during the similar seasons at the same sites as those for model parameterization, whereas group two was from experiments in different seasons and sites. When using data of group one, both models gave satisfactory estimations of gsc under both ample water supply and water stress conditions. The coefficient of determination (r2) and relative root mean squared error (rRMSE) between the measured and estimated gsc of lilium varied, respectively, between0.75and0.76, and between0.22and0.23for the Jarvis model. The r2and rRMSE values were between0.74and0.82, and between0.18and0.24, respectively, for the BWB-Leuning-Yin model using measured An as input. When using calculated An as input to the BWB-Leuning-Yin model, the r2was slightly decreased to values between0.70and0.72, and the rRMSE was slightly increased (between0.24and0.25). When using data of group two, the BWB-Leuning-Yin model gave better estimations of gsc than the Jarvis model did. The r2and rRMSE between the measured and estimated gsc of lilium varied, respectively, between0.61and0.62, and between0.35and0.36for the Jarvis model. The r2and rRMSE values were between0.72and0.79, and between0.23and0.25for the BWB-Leuning-Yin model using measured An as input. The r2and rRMSE values were between0.69and0.72, and between0.26and0.27for the BWB-Leuning-Yin model using calculated An as input. The results demonstrated that functions or parameters of the Jarvis model, if applied to independent environmental conditions, have to be re-derived. Compared with the three BWB-type models, the BWB model (1987) has an advantage for practical use because of its simple form and ease for its parameterization under moderate and high level of CO2concentration (250-1200μmol mol-1) and PAR (300-1500μmol m-2s-1). The BWB-Leuning-Yin model gave better estimated gsc under low light intensities (0-300μmol m-2s-1) and low CO2concentrations (50-250μmol mol’1), suggesting that the new version of BWB-type model (BWB-Leuning-Yin model) is suitable for estimation of gsc where levels of light and CO2are lower than ambient levels. The accurate estimation of stomatal conductance achieved in this study can be used for further estimation of cut lilium transpiration rate and photosynthetic rate accurately.(2) Modelling transpiration rate of cut lilium grown in greenhouses. We tested three models, i.e. big-leaf model, multilayer model and two-leaf model, for estimating transpiration rate of cut lilium grown in greenhouse. The models were parameterized from a subset experimental data. Comparison of the estimated result of daily accumulated transpiration by using three transpiration model showed that when LAI is lower than1, the estimated daily accumulated transpiration differed slightly among the three mode. But when LAI is above1, the big-leaf model overestimate daily accumulated transpiration and two-leaf model underestimate daily accumulated transpiration. Based on this result, daily averaged transpiration rate under different water treatment levels were estimated using the multilayer model, the coefficient of determination (r2) and the relative root mean square error (rRMSE) based on1:1line between the estimated and measured values are, respectively,0.85and0.23. The results demonstrated that, when LAI is lower than1, the big-leaf model and two-leaf model are suitable for estimation of transpiration rate; when LAI is above1, multilayer model is suitable for estimation of transpiration rate for cut lilium grown in greenhouses. Accurate estimation of transpiration rate achieved in this study can be used for optimizing not only the crop management but also the environment regulation inside greenhouse.(3) Modelling photosynthetic rate of cut lilium grown in greenhouses. We tested two models, i.e. the light response curve model of photosynthesis and the FvCB biochemical model, for estimating photosynthetic rate of cut lilium grown in greenhouses. The models were parameterized from a subset of the experimental data. Independent experiments data were used to validate the two models. The coefficient of determination (r2) and the relative root mean square error (rRMSE) based on1:1line between the estimated and measured An are, respectively,0.81and0.25for the light response curve model of photosynthesis;0.78and0.26for the FvCB biochemical model. But as used in our study, the FvCB biochemical model makes the model parameterization an easy task and the estimation of An possible via easily obtained information (Ca, PAR, T and VPD) and soil water potential (SWP). From the results mentioned above, the FvCB biochemical model is suitable for estimation of photosynthetic rate for cut lilium grown in greenhouses, hence, be further used for estimaung dry materr production of cut lilium accurately.The model developed in this study could be integrated to develop a model-based decision support system for greenhouse environment control and provided the theoretical basis and decision making to crop production. To improve the economic and ecological benefits of greenhouse production and greenhouse control level, and promote the development of the greenhouse industry in China have important theoretical significance and application value.