Land suitability assessment for agricultural crops in Enrekang, Indonesia: combination of principal component analysis and fuzzy methods

Abstract

Land suitability assessment is essential for the efficient use of diminishing fertile agricultural land. Assessment parameters include soil texture, pH, the sum of basic cations, base saturation, cation exchange capacity, organic carbon, soil depth, slope, and mean annual temperature and precipitation data. Results showed that 76.28% and 23.26% of the total area were optimally and moderately suitable for coffee growth, respectively; 9.6% and 90% were optimally and moderately suitable for cocoa growth, respectively; 1.98%, 78.74%, and 19.26% were optimally, moderately, and marginally suitable for clove growth, respectively; and 6.68%, 86.89%, and 6.41% was optimally, moderately, and marginally suitable for pepper growth, respectively. The final land suitability index (LSI) was strongly influenced by the threshold values used by the researcher and the quality of the land indicator itself. Plant threshold values differed due to variations in plant recruitment. The main limiting factors were mean annual temperature <26°C, acidic soil pH, and low CEC. This study showed that the fuzzy method is ideal for converting the numerical data of various magnitudes into membership function values and representing land suitability. The principal component analysis is an effective method to determine the weights of multiple factors in a systematic and objective manner. The linearity test found a correlation between LSI and production with f = 0.00, indicating that the applied model can predict agricultural production and is applicable to other agricultural land management.