Sensores NIRS para el control nutritivo y fermentativo de ensilados de maíz en explotación

Resumen

We aimed to establish controls of nutritive and fermentative parameters on maize silages at a farmlevel. We used the sensor microPhazir™ Handheld NIR Analyzer and 97 maize silage samples todevelop quantitative prediction models that displayed determination coefficient values between0.7-0.9 for all nutritive and fermentative parameters, except for pH due to narrow variability range.Related to nutritive parameters, the best results were obtained using first derivatives as amathematical treatment and for fermentative parameters after applying a second derivative. Inboth cases it was necessary to apply a Standard Normal Variate pre-treatment. The predictionerrors, evaluated as the ratio between cross validation error and average value of population foreach parameter, were lower than 10% for nutritive quality, except for starch content (15%), whilefor fermentative parameters errors varied 19% (lactic acid) and 38% (acetic acid). The resultsdemonstrate the potential of these handheld NIRS sensors for use in Precision Livestock Farming tooptimize animal feeding practices.