Provisional assessment of quality components in lucerne (Medicago sativa) and white clover (Trifolium repens) using a near infrared reflectance spectrophotometer

Abstract

A near infrared reflectance spectrophotometer (NIRS) with tilting filters was calibrated for the determination of the forage quality components, crude protein (CP), acid detergent fibre (ADF), neutral detergent fibre (NDF), in vitro organic matter digestibility (IVOMD), dry matter (DM) and the elements Ca, K, P and Mg. Calibrations were carried out separately for lucerne and white clover. Apart from validating the equations with the help of part of the calibration sample population, lucerne samples from a different growing season and clover samples from subsequent harvests as well as a different location were evaluated for certain components. Standard errors of calibration (SEC) for lucerne were 1,01% for CP; 1,6% (ADF); 1,7% (NDF); 2,4% (IVOMD); 0,4% (DM); 0,24% (Ca); 0,25% (K); 0,05% (P) and 0,03% (Mg). The calibration of white clover produced SECs of 0,71% for CP; 1,3% (ADF); 2,0% (NDF); 1,9% (IVOMD); 0,7% (DM); 0,08% (Ca); 0,34% (K); 0,03% (P) and 0,02% (Mg). The results of the validation test showed that standard errors (SEs) were within 1,5 times that of the SECs for all the components in both lucerne and white clover. The additional validation set for lucerne also showed that for Mg the standard error of prediction (SEP) was 1,5 times greater than the standard error of calibration. For white clover the additional validation set based on samples from a different locality produced SEPs that were 1,5 times larger than the SEC of the calibration set in respect of DM, Ca, K and Mg. It was concluded that a NIRS with tilting filters can successfully be used to simultaneously analyse CP, ADF, NDF, IVOMD and DM in lucerne and CP, ADF, NDF and DM in white clover samples provided a monitoring system is used.