当前作物肥效模型可分为结构模型和经验模型两类。结构模型复杂和参数多，应用尚有困难，经验模型是国内外实现定量施肥的主要途径之一。实践证明，经验模型存在不符合生产实际的缺陷：单位养分增产量与施肥量之间的函数采用线性模型，最高施肥量之前施肥的增产“速率”与最高施肥量之后施肥的减产“速率”是对称关系，结果导致应用上出现大量非典型式，严重制约了测土配方施肥计量精确性。针对经验模型存在的缺陷，本项目借鉴种群生态学建模理论，构建“半机理半经验”的非结构肥效模型。以氮磷钾盆栽和田间试验结果为依据，研究水稻在不同土壤肥力水平下，单位养分增产量与施肥量之间的函数关系，进而建立一元非结构肥效模型；根据植物营养元素功能不能相互替代而具有交互作用的原理，构建二元和三元非结构肥效模型；探讨新模型参数估计方法、推荐施肥技术及其可靠性。旨在扩大肥效模型的适用性，为水稻肥效模型建模和施肥专家系统研发提供理论依据。

Fertilizer response model for crops can be divided into structure model and empirical model. Structure model is complicated and has many parameters, which is difficult for application. Therefore, empirical model is the main way to realize quantitative fertilization at home and abroad. Practice has proved that empirical model has two defects with the actual practices: 1) Linear model is used in the function between the increased in production of unit nutrient and the fertilization rate. 2) Model existed a symmetric relationship between the rate of production increase before the highest fertilization rate and the rate of production decrease afterwards. These detects in model have brought about plenty of no-typical forms in practices and thus severely limited the practicability of empirical model and measurement accuracy in soil testing and formula fertilization. Aiming at the scientific and application problems in empirical model, this project intends to build a "semi-mechanism and semi-experience" unstructured fertilizer response model by using the modeling theory of population ecology. Based on the results from pot experiments and field trials for N, P and K, the exact functional relationship between the unit production increase and the fertilization rate under different soil fertility levels will be studied and a single-element fertilizer response model will be set up. According to the principle that nutrients can’t be substituted by each other and the synergistic effects between them, the binary and ternary unstructured fertilizers response model will be set up as well. The parameter estimation methods, the recommendation fertilization technique and its scientificity of the new model will be explored. The goal is to expand the model applicability and further provide theoretical basis for the research and development of nutrient management expert system for paddy rice.