Prediction of paddy productivity using apparent soil electrical conductivity model

Inexpensive, rapid and accurate methods for spatially measuring within-field soil differences associated with crop productivity are needed to improve planning for site-specific management and enhance crop yield. This study was conducted at MARDI Seberang Perai between 2002 and 2004 to investigate the application of apparent soil electrical conductivity (soil ECa) model derived from log-normal peak function using boundary line analysis. Grain yield data were obtained from crop cutting test and ECa measurements by using Veris 3100 cart equipped with data logger and a differential global positioning system. Significant relationships between potential yield and ECa were found in all three crop-season studies in the form of log-normal peak function. When comparing between predicted potential yields (Ypo) and observed yields (Yob) farm areas can be delineated into four different management zones and allows for discriminate management practices. The ECa model could correctly indicate that significant higher yield can be obtained from areas with high predicted potential yield through more intensive management practices. Accurate yield prediction couple with site-specific practices could result in less wastage of applied inputs, less pollution, lower input costs and most important, higher return.