2021 Volume 67 Issue 1 Pages 1-7
Corn silage is an excellent forage. In this study, 141 silages ensiled in laboratory silo were prepared by cutting four varieties of whole crop maize harvested from the milk-ripe to the full-ripe stage. The conventional partial least squares regression (PLSR) method, which uses the total wavelength information of the target, and the iterative stepwise elimination PLSR (ISE-PLSR), which eliminates wavelength information that does not contribute to target component estimation or selectively includes useful wavelength information, were employed, along with near-infrared spectra measurement in the wet uncrushed state after opening. A calibration curve was constructed for the easy and rapid evaluation of feed components and fermentation quality, and the estimation accuracies of both methods were compared. Based on the findings from the verification experiment, the calibration curve achieved greater accuracy using the ISE-PLSR method than by using the PLSR method for all items. Further evaluation is necessary to present a versatile calibration curve; however, the various possibilities for the rapid evaluation of pH, DM, butyric acid, valeric acid content, and V-SCORE were presented with significant accuracy.
