We currently have an excellent set of broad-based near-infrared spectrophotometric (NIRS) prediction equations for predicting four measures of nutritive value of corn silage; NDF, IVTD, starch, and protein. IVNDFD is then obtained as a function of NDF and IVTD. These NIRS equations are now routinely used by the UW corn breeding program as well as the corn hybrid evaluation program (conducted by Dr. J. Lauer, UW Agronomy – Corn Extension). After each season of silage evaluation, our ability to predict the nutritive value of the germplasm under development is upgraded by including new NIRS spectral samples into the calibration database.
Even though our NIRS prediction equations have been proven to be reliable and accurate, knowledge of nutritional requirements has increased from the time the equations were initially developed. This will always be the case because of the need to assemble a large nutritional database encompassing a large array of different hybrids grown in different environments in order for NIRS prediction to work adequately. Assembling such a database takes a number of years to accomplish. While data are being accumulated, the nutritional concepts validating the initial effort at NIRS prediction change somewhat requiring that altered or new nutritional characteristics be considered. Kernel texture, starch digestibility, and digestion kinetics are good examples of new characteristics that should be investigated further.