Search

A statistical regression model for rapid prediction of moisture content based on measurements of dielectric capacitance and test weight was developed for Eastern Canadian corn (Zea mays L.). For 336 samples of the 1986 crop, dielectric readings were determined with a Model 919 grain moisture meter, test weight values with an Ohaus half-litre measure and moisture content values by a single-stage air-oven procedure. The regression model, which incorporates linear terms for dielectric reading and test weight plus an interaction term which is a product of the two, is an excellent predictor of corn moisture as indicated by analysis of the residuals and by the high value of the coefficient of determination (R2 = 0.95) and low value of the standard error of estimate (SEE = 0.85). Although the relationship between moisture content and dielectric reading for Ontario samples differed from that for Quebec samples, the proposed regression model helped to compensate for the difference. This model was also effective in predicting moisture content for 365 samples of 1987-crop Eastern Canadian corn. As well, it yielded a better fit to 1986-87 crop data than did the dielectric-based regression model used in CGC Corn Moisture Conversion Table No. 9.

A statistical regression model for rapid prediction of moisture content based on measurements of dielectric capacitance and test weight was developed for Eastern Canadian corn (Zea mays L.). For 336 samples of the 1986 crop, dielectric readings were determined with a Model 919 grain moisture meter, test weight values with an Ohaus half-litre measure and moisture content values by a single-stage air-oven procedure. The regression model, which incorporates linear terms for dielectric reading and test weight plus an interaction term which is a product of the two, is an excellent predictor of corn moisture as indicated by analysis of the residuals and by the high value of the coefficient of determination (R2 = 0.95) and low value of the standard error of estimate (SEE = 0.85). Although the relationship between moisture content and dielectric reading for Ontario samples differed from that for Quebec samples, the proposed regression model helped to compensate for the difference. This model was also effective in predicting moisture content for 365 samples of 1987-crop Eastern Canadian corn. As well, it yielded a better fit to 1986-87 crop data than did the dielectric-based regression model used in CGC Corn Moisture Conversion Table No. 9.