MATHEMATICAL MODEL OF WEED SEED DYNAMICS

研究结果包括三个部分。第一部分，将杂草生活史模型建立为一个以地表和埋地种子数为状态参数的动态系统。该模型处理除草、耕作和收获。利用该模型，我们制定了两种策略。第二部分对建立的动态模型进行了修正，以适应实际的田间栽培。状态参数的数量从2个扩展到8个。从温度模型的参数状态来看，改变每年的耕作深度是降低农田种子潜力的一种可能方法。第三部分对种子萌发、生长和播种过程进行了数学模拟。通过现场实验确定了各个过程的表达式。利用该模型，从种子产量角度对除草调度进行了优化。

The research results consists of 3 parts. On the 1st part, weed life history model was developped as a dynamical system where surface and buried seeds number are state parameters. The model deals weeding, tillage, and harvesting. With the model, 2 strategies were developed.On the 2nd part, the dynamical model developed was modified to fit actual field cultivations. The number of state prameters were expanded from 2 to 8. From parameter staties with themodefied model, changing the tillage depth in each year is a possible method to reduce a seed potensial of agricultural fields.On the third part, germination, growth, and seeding process were modelled mathematicaly. The functions express each processes were determined by field experiments. With the model, optimization of weeding scheduling in terms of seed productions was conducted.