Prediction of the chemical composition of Cenchrus clandestinus grass using Near Infrared Spectroscopy – NIRS

Forage evaluation is important as milk production depends on the availability of highly nutritious forage. The aim of this study was to develop nutrient content prediction equations for kikuyu grass (Cenchrus clandestinus), a universally used forage in Colombian specialized dairy farms, with near-infrared spectroscopy (NIRS). Kikuyu samples obtained from two hundred dairy farms in the North of Antioquia were analyzed for DM, protein, NDF and ADF. Using three-step procedure (calibration, cross validation, and prediction), equations were developed in a NIRS equipment (NIRS DS 2500 monochromator, Foss-NIRsystem, Denmark). The absorbance values (logarithm (1/R), R = reflectance) were analyzed using the software WinISI version 4.8, performing mathematical treatments to generate several equations per chemical component analyzed. The R2v for protein content was 0.96 and SEP was 0.54 indicating an appropriate prediction equation. The R2v for NDF and ADF contents were 0.89 and 0.88 respectively, however the SEP value was lower for ADF (0.69) than NDF (1.88). Chemical composition for protein, NDF and ADF in kikuyu grass can reliably predicted using equations developed in NIRS. However, it was not possible to develop a prediction equation for kikuyu DM.