Measurement of nutrients in Saskatchewan hog manures using near-infrared spectroscopy

摘要

The objective of this study was to use near-infrared spectroscopy (NIRS) to predict nutrient concentrations and physical parameters of hog manure from 13 Saskatchewan farms and producer submissions to a commercial laboratory. Nutrients predicted weretotal Kjeldhal nitrogen (TKN), total phosphorus (P), and potassium (K). Total solids (TS) were also measured. Multiple linear regression (MLR) was used to develop calibration models using a calibration set of 63 samples. Performance of the calibration models was evaluated using a validation set of 25 samples. The performance of the models was quantified by determining the coefficient of determination (r~2), the ratio of the standard deviation of the reference chemistry values to the standard error of prediction (RPD), and the ratio of the range of reference chemistry values to the standard error of prediction (RER). Total Kjeldhal nitrogen (r~2 =0 89, RPD = 2.29, and RER = 10.22) and TS (r~2 = 0.92, RPD = 3.63, and RER = 17.04) were predicted successfully, while models predicting P (r~2 = 0.79, RPD = 1.92, and RER = 6.33) and K (r~2 = 0.68, RPD = 1.14, and RER = 6.00) were less accurate kw:near-infrared reflectance spectroscopy; NIRS; manure management; multiple linear regression; MLR; spectroscopie aux infrarouges rapprochés; SIRR; gestion du lisier; régression linéaire multiple; RLM