Simulation Of The Responses Of Leaf Area, Photosynthesis Rate And Dry Matter Production Of Greenhouse Cucumber To Leaf Nitrogen Content

Leaf area index (LAI) is an important parameter for modelling canopy photosynthesis and crop water status, photosynthesis and dry matter production are necessary to guarantee crop yield and quality formation. Leaf nitrogen content directly affects the crop LAI and photosynthesis rate, thus affects crop yield and dry matter production. Quantitative researches on the responses of crop LAI, leaf photosynthesis rate and dry matter production to leaf nitrogen are the important premise for the accuracy and optimization of nitrogen management in greenhouses. Cucumber is one of main greenhouse cultivated crops in China and in the world. Experiments of greenhouse cucumber with different planting dates, substrates and nitrogen application rates were conducted in a Venlo-type greenhouse and multi-span plastic greenhouse in Shanghai (121.5 (?),31.2 (?)) from 2005 to 2007 to collect data for model development and validation. Based on the data, first we established a model to simulate the responses of leaf morphological traits and leaf area index to leaf nitrogen content using the index of radiation and temperature (photothermal index, PTI); second, based on the FvCB photosynthesis model, the responses of parameters of FvCB model to leaf nitrogen content was determined; third, integrated the model of responses of LAI to leaf nitrogen content and responses of photosynthesis based on the FvCB model to leaf nitrogen content, we established a model of to simulate the response of dry matter production of greenhouse cucumber to leaf nitrogen content. The main results are:(1) Simulation of the response of leaf growth of greenhouse cucumber to leaf nitrogen content. Using the experiment data with different nitrogen application rates, seasonal time course of canopy nitrogen content and the distribution of leaf nitrogen content in canopy were quantified, then the effects of canopy nitrogen content and specific leaf nitrogen content on leaf morphological traits of greenhouse cucumber were described. On this basis, a simulation model of responses of leaf morphological traits and leaf area index to leaf nitrogen content was established. Independent experiments data were used to validate our model and SLA-based model. The coefficient of determination (r2) and the root mean squared error (RMSE) between the predicted and measured values using our method are 0.95 and 2.9 for leaf number,0.93 and 0.05 m for specific leaf length,0.87 and 0.18 m2 m-2 for LAI when using canopy nitrogen content at the start of fruit setting while using specific leaf nitrogen content (the 8th leaf counted downward) at the start of fruit setting, the r2 and RMSE are 0.93 and 3.2 for leaf number,0.92 and 0.05 m for specific leaf length and 0.8 and 0.39 m2 m-2 for LAI. For the SLA-based model, using measured SLA data as input, the r2 and RMSE are 0.74 and 0.43 m2 m-2. So, our model gives satisfactory prediction of the response of leaf growth and LAI of greenhouse cucumber to leaf nitrogen content with different nitrogen application rate in different greenhouses.(2) Simulation of the response of leaf photosynthesis rate to leaf nitrogen content using the FvCB model. Based on the FvCB model, parameters used for calculating photosynthesis rate (e.g. day respiration (Rd), conversion efficiency of linc into linear electron transport of PSII under limiting light (k2(LL)), electron transport capacity (Jmax), curvature factor (0 for the non-rectangular hyperbolic response of electron flux to Iinc, ribulose 1-5-bisphosphate carboxylase/oxygenase (Rubisco) CO2/O2 specificity (Sc/0), Rubisco carboxylation capacity (Vcmax) and mesophyll conductance (gm) were estimated, using combined measurements of photosynthesis rate and chlorophyll fluorescence, and the responses of these parameters to leaf nitrogen were determined. The results showed that K2(LL), Vcmax(To), Jmax and Rd increased linearly with leaf nitrogen content, but (?) and gm could be independent. Using estimated parameter as input, independent experimental data was used to validate and give the comparison between this model and the photosynthesis model which is based on the radiation, the maximum gross photosynthesis rate and initial light use efficiency (the asymptotic negative-exponential fucntion). The coefficient of determination (r2) and the root mean squared error (RMSE) between the predicted and measured values are 0.63 and 5.28μmol CO2 m-2 s"1 when using FvCB based model; r2 and RMSE between the predicted and measured values are 0.65 and 4.94μmol CO2 m-2 s-1 when using the the asymptotic negative-exponential fucntion. Though the prediction accuracy of the negative-exponential function is better than the FvCB model, but further analyses showed that the prediction accuracy of net photosynthesis rate is poor when the environment is differ from the model established, thus stability and adaptability is poor; the prediction accuracy of FvCB model is lower, but under different environment, the prediction accuracy is the consistent thus has stability and adaptability. The prediction accuracy can be improved when using more data to correction the parameters used in the model.(3) Integrating simulation of the responses of leaf growth and photosynthesis rate to leaf nitrogen content developed in chapter 2 and chapter 3, the responses of dry matter production to leaf nitrogen content were established. Integrating prediction of the effect of nitrogen on greenhouse cucumber leaf area developed in chapter 2 (leaf morphological traits based LAI model and SLA-based LAI model) and prediction of the effect of nitrogen on leaf photosynthesis using FvCB model and an asymptotic negative-exponential function used in Chapter 3, dry matter production of greenhouse cucumber was predicted. Independent experimental values were used to validate the model. The results show that the r2 and RMSE between the predicted and measured total dry matter based on the 1:1 line are 0.62 and 58.9 g m-2 when using leaf morphological traits based LAI model and FvCB model; 0.66 and 54.3 g m-2 when using leaf morphological traits based LAI and the asymptotic negative-exponential function; 0.71 and 44.8 g m-2 when using SLA-based LAI model and FvCB model; 0.53 and 77.2 g m-2 when using SLA-based LAI model and the asymptotic negative-exponential function, respectively. Though, using leaf morphological traits based LAI model and FvCB model did not give a better prediction as using leaf morphological traits based LAI and the asymptotic negative-exponential function and using SLA-based LAI model and FvCB model, it can avoid destructive sampling with respect to SLA-based model and prediction accuracy has a better consistency than leaf morphological leaf traits and the asymptotic negative-exponential function gives. Therefore, leaf morphological traits based LAI model and FvCB model is more applicable, stable and adaptable, the prediction accuracy can be improved through the improvement of the parameters in FvCB model to raise the prediction accuracy of photosynthesis rate.The model developed in this study could predict the number of appeared leaves per plant, leaf nitrogen content of each rank, specific leaf length, leaf area, leaf photosynthesis rate and dry matter production of greenhouse cucumber with the single-stem pruning in Shanghai using the easily gained crop and environmental parameters such as planting date, planting density, nitrogen application rate, PAR, temperature and CO2 concentration inside the greenhouse, the number of leaves on planting date and the removed leaves as input, therefore has strong mechanism and is easy to use, can be used for optimization nitrogen management for greenhouse fruit cucumber prediction.