【作者】 陈杰；

【导师】 胡秉民； 杨京平；

【作者基本信息】 浙江大学 ， 生态学， 2004， 博士

【摘要】 作物生长发育的动态模拟研究始于20世纪60年代，是在作物生理生态科学中引入系统分析方法和计算机技术而兴起的。作物生长发育的动态模拟，指应用系统分析的方法，综合植物生理生态、农学、气象、土壤等相关学科的研究成果，建立计算机模型定量地描述作物的生长发育、光合生产、产量形成等生态生理过程及其与环境因子、管理技术的关系。 作物生长模拟研究在过去的40年中已取得显著的成就，但完善模型仍然需要长期的努力。水稻模型是作物模拟研究的重要内容，氮素限制下的水稻生长模拟是当前的研究热点。本研究在查阅了大量作物生长模拟领域的有关文献基础上，结合田间试验资料，组建了水稻氮素行为及施氮优化系统，并以VB6.0开发了基于Windows系统的模拟软件，简化了模型的应用。 本研究主要在以下几方面取得新进展：（1）结合作物生长模拟的有关理论和技术，建立了水稻氮素行为模拟模型和氮肥管理优化模型，行为模型用于研究水稻的生长发育、氮素动态及其影响，优化模型为水稻生产的氮肥管理提供优化的施肥方案；（2）在作物生长模型中成功地引入Price数值优化算法，解决了模型有关施肥参数的计算，进一步提高了作物生长模型指导生产实践的能力；（3）应用VB6.0完成了上述模型的模拟软件开发，软件运行于Windows环境，所有的输入输出均采用“窗口”模式，结果表达直观可视，操作过程简单方便，可以满足普通用户的使用需求。 1．施氮水平对水稻生产的影响 水稻氮肥试验于2001、2003年在杭州市余杭农科所试验站进行，设置6个氮素处理0、75、150、225、300、375kghm^-2，供试品种为武进9728、丙9363、丙9652，研究施氮水平对水稻的生长发育、氮素动态及产量性状的影响。 结果表明，随着施氮量的增加，叶面积、根茎叶生物量都持续增长，在施氮量小于225kghm^-2时，氮肥效果更明显。施氮量与植株器官氮含量之间呈线性关系，成熟期武进9728的叶、茎、根、穗的含氮量与氮肥用量的相关系数分别为0.9408、0.9649、0.9669、0.9662，达到显著水平。有效穗随着施氮水平的提高而持续增加，穗粒数、结实率、千粒重对施氮量的高低反应不同，总体上浙江大学博}论文摘要呈下降趋势。施氮量对产量的影响可用抛物线表示，根据经济学的边际收益分析原理，目前条件下当地适宜的施氮量为140一Zookghm一2。从经济、生态和效益三方面综合考虑，针对当前的施氮水平，减少30%施氮量是完全可行的。2.水稻氮素行为的模拟 在相关研究的基础上，组建了水稻氮素行为模拟模型RNDSM，包括生育期进程、叶面积发育、光合作用、呼吸作用、干物质分配、产量形成等基本生长发育过程，以及氮素的需求、供给、吸收、分配、转运等氮素动态过程。 从生育期、叶面积、生物量和植株含氮量等方面对RNDSM模型进行了验证。结果表明，生育期的模拟值仅比实际小1一2天，偏差主要发生在移栽至开花期。叶、茎、穗的生物量和总生物量的模拟结果较好，而根生物量的模拟不够理想。叶面积指数的模拟，总体效果良好，生长季节的中期略微偏低。模型对植株氮素动态的模拟好于生物量，除根以外，其他器官含氮量及总氮含量的模拟结果比较接近理想状况，误差在10%以内。3.水稻施氮优化的模拟 以ORYZAO模型为基础，组建了水稻施氮优化模拟模型RNOSM，对氮肥的施用、吸收、分配、干物质形成及土壤氮供给等过程进行了详细描述，并结合数值优化方法对施氮量、施用时间优化模拟，为水稻的氮肥管理提供服务。 验证结果表明，模型能够较好地模拟生物量的积累过程及叶氮含量、植株总氮含量的变化动态，模拟值与实测值之间的相关系数分别为0.9926、0.9130、0.9689，都达到显著水平。应用模型优化水稻的氮肥管理，首先计算理想施氮曲线，在此基础上转化为实际的施氮方案。对施氮量150kghm一的处理，模型优化的施氮时间和比例分别为移栽后10、20、30、45天和0.2、0.3、0.3、0.2，模拟产量比实际值高10%以上。4.水稻氮素模拟软件的开发 应用面向对象的编程语言VB6.O，结合VB数据库技术，开发了水稻氮素行为及施氮优化模拟软件RICEN。模拟软件在windows环境下运行，有简洁的菜单工具和命令按钮，操作过程简单，用户能够方便地编辑保存作物参数文件和气象资料，模拟结果以数据表和图形的方式直观显示，可以分别保存为纯文本文件和BMP图像文件。了浙江大学博士论文摘要5.讨论与研究展望 模型对土壤氮素动态和施氮优化方法的处理还有较大的不足，有待进一步深入研究。展望水稻生产的模拟研究，需要在完善模型自身的结构、改进模拟方法的同时，加强与现代信息技术的融合，开发基于网络环境运行的模拟软件，提高模型在生产实践中的应用能力。更多还原

【Abstract】 Simulation studies on crop growth and development began in 1960’s with the application in crop physiological ecology of system analysis methodology and computer technology. Crop simulation is referred to set up computer model to analyze quantitatively the ecophysiological processes including crop growth and development, photosynthesis, yield formation, etc., and the effects of environmental factors, management measures, by applying the methods of system analysis and integrating the related disciplines, such as plant physiological ecology, agronomy, meteorology and soil science.Outstanding achievements have been made for crop simulation studies in past 40 years, but it still needs more efforts to perfect the simulation models. Rice model is an important research field, especially for simulation of rice growth limited by nitrogen. Based on related literatures and field experiments, a simulation model for nitrogen dynamics and fertilizer-N optimization in rice was set up, and the corresponding software was also developed to simplify model’s application. The software running for Windows98 or above was programmed with Visual Basic 6.0. Innovated progress in present study includes the following aspects. (1) By combining related theory and technology of crop growth simulation, two models were ret up to simulate nitrogen dynamics and fertilizer-N management in rice growth and production, which can be used to study growth, development, nitrogen dynamics and optimizing plans for fertilizer-N application. (2) Price’s numerical optimization algorithm was integrated in rice nitrogen model to realize the computation of some fertilizing parameters and improve the model’s ability in production. (3) Programmed with Visual Basic 6.0, simulation software for above models was developed. The software is designed on the basis of windows interface, so it is easy to operate and can meet common users’ applying demand.1. Effects of fertilizer-N application levels on rice productionField experiments were carried out to study the effects of fertilizer-N application levels on rice production in Hangzhou city in 2001 and 2003. There were six treatments with nitrogen application of 0, 75, 150, 225, 300, 375 kghm"2. Three varieties of Wujin-9728, Bing-9363 and Bing-9652 were used as materials.The results showed that the fertilizer-N application levels obviously affected on leaf area index (LAI), biomass, yield formation and nitrogen content of crop. LAI, dry matter of leaves, stems and roots increased with the rise of fertilizer-N continuously, and the effects were more distinctly at low or medium levels of fertilizer-N. There existed a linear relationship between nitrogen content of plant organs and fertilizer-N input. For yield components, panicles increased with the rise of fertilizer-N all along, while number of grain per panicle, fertility rate and 1000-grain weight had different response to low or high fertilizer-N, though a decreasing tendency could be found as a whole. The effects of fertilizer-N on yield could be expressed by quadratic parabola. Fertilizer-N application of 140-200 kghm"2 was considered suitable for present local conditions, according to the parabola and the principles of marginal benefits of economics. By integrating related factors, it is feasible to cut 30% of fertilizer-N from current application levels.2. Simulation of rice nitrogen dynamicsRNDSM, a simulation model for nitrogen dynamics in rice crop, was built to study rice production under conditions limited by nitrogen, based on previous models. The model is made up of basic processes including phenological development, leaf area development, photosynthesis, expiration, dry matter partitioning, yield format;on, etc., and nitrogen related processes, such as the demand, supply, uptake, partitioning, transfer of plant nitrogen.RNDSM was validated by the field experimental data. From the results of validation, the model could simulate the phenological development accurately, only one or two days’ differences between the simulated and observed growth du更多还原