The Use of Cluster Analysis in Distinguishing Farmland Prone to Residential Development: A Case Study of Sterling, Massachusetts.

Abstract

/ Residential development of farmland is one of the primary driving forces of land degradation in both rural and urban fringe areas throughout the world. The loss of prime agricultural land is of great concern to planning offices and organizations seeking to preserve open space. The objective of this study is to demonstrate the use of clusteranalysis as a possible tool for the identification of farms prone to residential development. Eighty-four farms in Sterling, Massachusetts, were separated into two groups by k-means nonhierarchical cluster analysis using farm size, slope, and distance to the nearest city center and highway as surrogates of farmland conversion. Discriminant analysis showed that the two groups derived from the cluster analysis were 98.8% accurate (P < 0.0000). Results from the statistical analysis may serve as a starting point for the identification of individual farms prone to residential development. To explain the driving forces of farmland conversion to residential uses, interviews should be conducted with farmers, landowners, and land buyers. The use of multivariate statistical techniques to identify farms in jeopardy of residential development, in conjunction with qualitative assessments that explain the probability of development of individual farms, may prove a useful strategy to understand and predict farmland conversion.