在水稻根长增长模型和养分吸收动态模型基础上,将根系养分吸收CDE模型与经验模型相结合,建立了描述一个生长周期的水稻根系钾素吸收稳态模型.计算机模拟试验显示,根系钾素吸收力a和根表面水流速率V0为变量而钾素有效扩散系数De为常数是最佳参数组合,提出了a(t)和V0(t)与水稻生长时间t的关系式.用Monte Carlo法校准了水培条件下水稻一个生长周期的根系钾素吸收力常数a0和根表面水流速率常数v0以及钾素有效扩散系数De.结果表明,Monte Carlo法对α、V0和De参数的校准值与目前室内直接测定的数值有相同的数量级；稳态模型对6个供试水稻品种不同生育期吸钾量的模拟值与实测值具有高度一致性,线性回归方程的斜率达到0.978 1,吸钾量模拟误差为(10.7±3.1)％.6个品种的平行试验结果表明,模型预测吸钾量与实测值的平均误差为(9.9±6.6)％.参数敏感性分析表明,水培条件下营养液钾素浓度Cli、水稻根长L和根系钾素吸收力常数a0是钾素吸收的主要限制因子.%A steady-state model for potassium uptake on rice root of one growth period was established based on CDE model combining with rice root length increasing model and rice nutrient absorption dynamic model. Computer simulation experiment showed that the optimum parameters array was that K uptake absorbing power (a) and fluid flow velocity (Vo) on root surface were function of time and effective diffusion coefficient of K(De) was constant. The models were set up between a(t) or Vo(t) and rice growth period (t) respectively. Root K absorbing power contant (dp) and fluid velocity contant (Vo) on root surface and effective diffusion coefficient of K(De) in steady -state model for rice in a growth cycle were calibrated by Monte Carlo method under water culture condition. Results indicated that calibration value of a, Vo and De parameters had the same magnitude grades with testing mothod directly. Uptake K amount in different growth period of six observed rice varieties using the steady-state had significant coherence with testing values. The slop of linear regression equation reached by 0.978 1 and their simulation error for uptake K on average was only (10.7 ?.1)% compared with testing values. Parallel experiments in six rice varieties were showed that average error was only (9.9 ?.6)% between model prediction value and measured value. Parameters sensitivity analysis showed that K concentration in nutrient solution (Cj> and root length (L) and root K absorbing power contant M were main limited factors for K uptake.
展开▼
