精準農耕水稻施肥決策系統之案例分析

Abstract

本文係針對農委會農業試驗所精準農耕研究計畫在農試所的試驗田區（共10.83公頃，分為163個網格小區），以水稻施肥決策系統，進行施肥決策分析。施肥決策所需的期作生產力圖，係以試驗期間試驗小區個別期作之最高產量構成。一期作各小區之生產力均可達第一級（> 6 , 2 Mg/ha)，二期作則大部分小區可達第二級（5.2-6.2 Mgma)，亦有部分田區可達第一級（>6.2 Mg/ha)，試驗期間試驗田區之有效磷含量維持於極高（>28.3 μg / ml P by Mehlich’s No . 3 ）與高等級（11.3-28.3 μg / ml P by Mehlich’s No.3)，而鉀含量則維持於高（> 45μg / m 1 K by Mehlich’s No.3）與中（16-48 μg / ml K by Mehlich’s No.3）等級。依據決策系統之診斷與施肥推薦，二期作應可按目前施肥量減施14％。未來藉由遙測技術之開發，提供較精準之空間資訊，將相對可獲得更精準之施肥圖。 Field experiments were conducted in the experimental farm of Taiwan Agricultural Research Institute to study a Fertilization Decision Support System for precision farming on rice in Taiwan. The experimental farm set up for the study was 10.83 ha and was divided into 163 grid subplots. The map of potential productivity of rice crop is constituted by using the maximum yields of grid subplots from the experiments, and serves as the key information for the nitrogen fertilizer recommendation. In the first crops, yields harvested from the subplots may reach the first level of productivity. In the second crops, yields from some subplots may reach the first level of productivity while some may reach only the second level of productivity. The concentration of available P in soil was maintained from high (11.3-28.3μg P/ml, by Mehlich’s No. 3) to very high level (>28.3 μg P/ml, by Mehlich’s No. 3) throughout the study. The concentration of available K in soil was maintained from medium (16-48μg K/ml, by Mehlich’s No. 3) to high level (>48 μg K/ml, by Mehlich’s No. 3). According to the diagnosis and fertilization recommendation processed from the decision support system, it indicated that 14 % of total fertilizer used in the second crops may be cut out from the current application rate. In the future, when the precised space variability information can be provided from remote sensing, it can be expected to have a more accurate fertilization map for precision farming purpose.