Estimation of mineral content in forages by near-infrared reflectance spectroscopy

In 100 dried samples of forages, harvested three or four times during 1997 and 1999 in different morphological stages the macro- (Ca, P, Mg, K and Na) and trace- (Zn, Mn and Se) mineral contents were determined by chemical methods and by near infrared reflectance spectroscopy (NIRS). The accuracy c...

Full description

Bibliographic Details
Main Authors: Andrej LAVRENČIČ, Bruno STEFANON, Andrej OREŠNIK
Format: Article
Language:English
Published: University of Ljubljana Press (Založba Univerze v Ljubljani) 2002-08-01
Series:Acta Agriculturae Slovenica
Subjects:
Online Access:https://journals.uni-lj.si/aas/article/view/15352
Description
Summary:In 100 dried samples of forages, harvested three or four times during 1997 and 1999 in different morphological stages the macro- (Ca, P, Mg, K and Na) and trace- (Zn, Mn and Se) mineral contents were determined by chemical methods and by near infrared reflectance spectroscopy (NIRS). The accuracy calibration equations of NIRS for macrominerals Ca, P, Mg and K showed a high degree of accuracy, with an R2 of 0.963, 0.884, 0.892 and 0.939, respectively. These results suggest that these macrominerals are probably closely associated with the organic components, such as plant cell wall (Ca and Mg), chlorophyll (Mg), phytates, phospholipids, phosphoproteins and nucleic acids (P) and organic acids, such as malate (K). On the contrary, NIRS did not estimate Na content very accurately (R2 = 0.52; coefficient of variability of cross validation (CVSECV) = 35.5%), suggesting that Na is not closely associated with organic functional groups or that the amount of the mineral is below the level of NIRS detection. Similar conlusions are also valid for trace minerals, where the R2 was 0.642 (SECV = 4.4) for Zn, 0.427 (SECV = 55.6) for Mg and only 0.186 (SECV = 6.5) for Se. The CVSECV for trace minerals always exceeded 20%. Despite the insensibility of NIRS to accurately determine Na and certain trace minerals, the NIRS technique may provide an acceptable prediction of macro mineral contents.
ISSN:1854-1941