Prediction of Forage Energy Content by Near Infrared Reflectance Spectroscopy and Summative Equations

In livestock nutrition, summative models (SM) are displacing empirical models as a preferred method to predict energy content of forages. The objective of this study was to determine the effect of using near infrared reflectance spectroscopy (NIRS) to determine nutrient subcomponents required of SM on overall ability to predict energy content of corn and legume-grass silages. Corn (n = 90) and legume-grass (n = 70) silages were collected and analyzed for CP, ADF CP, NDF, NDF CP, in vitro (IV) digestible (d) NDF, ash, and fat by standard laboratory techniques. Samples were scanned on a Model 6500 NIRS, and calibration equations were developed for each nutrient. The TDN contents of corn and legume-grass silages were then estimated using a SM, where the model nutrients were determined by laboratory or NIRS methods. The predicted TDN content of corn and legume-grass silages was compared to IV d OM to assess overall utility. The NIRS calibrations were adequate (R2 > 0.90) for CP and NDF for both corn and legume-grass silages with standard errors of calibration (SEC) <0.55 1To whom correspondence should be addressed: [email protected] for CP and <1.09 for NDF. Near infrared calibrations for ADF CP, NDF CP, fat, and ash were less accurate in both corn and legume-grass silages with R2 < 0.75. Calibrations for IV d NDF in corn and legume-grass silages had R2 = 0.87 and 0.79, respectively, but possible co-dependency with NDF is speculated. The relationship between corn and legumegrass silage SM TDN and IV d OM was excellent when model nutrients were determined by laboratory procedures. The TDN estimates when NIRS was used to determine all SM nutrients were superior to older empirical models, but SM TDN estimates using NIRS-determined nutrients were less accurate as compared with SM TDN prediction when model nutrients were determined by laboratory procedures. In particular, using NIRS to predict IV d NDF and ash for use in SM lead to the greatest challenge in TDN prediction in both corn and legume-grass silages. (