The Procedure of Arable Land Classification on the Basis of Landscape Ecological Factors

Japanese Journal of Crop Science

Yoichi TORIGOE¹ 

关键词: Arable land classification;    耕地分類;    Land use;    耕地利用;    Landscape ecology;    多変量解析;    Mesh data;    地形;    Multiple regression analysis;    土壌;   

DOI  :  10.1626/jcs.58.305

学科分类：农业科学（综合）

来源: Nihon Sakumotsu Gakkai / Crop Science Society of Japan

【 摘 要 】

In order to develop the procedure of arable land classification, we forced to adopt the viewpoint of landscape ecological classification depending on the analogy of topographic and edaphic factors, natural vegetation and microclimate. The region of our case study, Yasato-cho, Ibaraki, covers approximately 138 square kilometers. Data processing system based on the personal computer was developed to use the mesh data of topographic and edaphic factors, soil types and agricultural census. The dimension of a mesh was about 500 m×500 m. The first step was to demarcate agricultural domains. Twentyseven agricultural domains were demarcated according to the resemblance of paddy and upland soils of each mesh into individual watersheds. The second step was to choose the factors showing the regional diversity between the agricultural domains. Seven factors were detected and categorized at six classes. The principal component analysis was applied to the categorized data of individual agricultural domains. As a result, we concluded that the regional veriation of these factors could be interpreted by three principal components. The third step was to apply the furthest neighbor method of cluster analysis to three component scores obtained by the previous analysis. 27 agricultural domains were divided into five clusters which could characterize the regional difference of landscape ecology and current agriculture in the given region.

【 授权许可】

Unknown