Agri-systems variations determined through Principal Component Analysis

The present study tested the capability of Principal Component Analysis (PCA) in determining variations of the agri-systems landscape using ten qualitative characters viz. levels of nitrogen, phosphorus, potassium and pH, as well topography, vegetation type, presence of rocks, and characters of color such as hue, chroma and value. The 60 hectare farm in Manresa was used as model where assessment of variations using PCA was done. Exactly 103 random samples were collected from 11 sites within the model farm, and soils were analyzed and characterized using numerical coding technique. There were 1030 coded data generated and PCA was implemented on these data using PAST (Paleontological Statistics) software version 1.78. The similarity of sites was tested using cluster analysis and validation of results was made using Discriminant Function Analysis using SPSS ver. 17. Here, we found that PCA effectively deciphered variation existing in the agri-systems landscape, and dominant factors contributing to such variations identified. Practical applications of the method in agriculture is discussed.