Although Principal Component Analysis (PCA) has been applied as a tool for dynamic evaluation in District Suitability Classification for Agriculture*, several problem areas remain - e.g.: (i) verification of the stability of the classification scale and (ii) the method of evaluating qualitative differences of agricultural income change, etc.
In this paper, we examine the above problems through a case study of the Suwa Region, which is located in Nagano prefecture. We also discuss a classification method for the estimation of qualitative differences of agricultural income change, which is expected to be an effective method in planning processes of land use plans and regional agricultural plans.
The results of this paper are summarized as follows.
(1) PCA was applied to the data sets for three points in time, viz: 1970, 1975 and 1980. Structural changes of the first PCs, which are generally used as a classification scale, precisely reflect changes in the regional profile of agriculture.
(2) Regarding the stability analysis of the classification scales, it was verified that the influences of these structural changes of the first PC on the first PC scores are negligibly small. This fact implied that three classification scales can be unified without any damage to the validity of the classification results.
(3) Based on the above analysis, PCA was applied to the three data sets at the same time in order to unify the scales. The examination of the dynamics of both the PC scores and classification maps in the 10 years under consideration (1970-1980) made it clear that the simple quantitative method of agricultural income change might be of doubtful value in estimating the stability of regional agriculture and its future potential for development.
(4) A classification based on changing patterns of the first PC scores in the two terms under consideration (1970-1975 and 1975-1980) was excluded in order to grasp the qualitative differences of agricultural income change. Regarding the examination of the characteristics of each classification type through statistical data and field surveys, it was found that this classification method is effective in summarizing the qualitative differences of agricultural income change.
* District Suitability Classification for Agriculture is a district-based classification method of land use suitability, whose criterion is agricultural income level.
View full abstract