Search

This study proposed, verified and analyzed a potential reason (matrix composition complexity) for the subpar performance of FT-IR-based meat and bone meal (MBM) species discrimination. In order to address the aforementioned issue, the MBM matrix was modeled as a binary mixture comprising meat fragment (MF) and bone fragment (BF). BF and MF samples derived from four species were prepared, and MBM samples were subsequently formulated. The influence of matrix fragments on FT-IR-based MBM species-specific analysis was verified through the analysis of FT-IR spectral absorption peak intensities and comparison of species discrimination analysis models established using MBM, BF and MF samples. Furthermore, the coupling effect of BF and MF on FT-IR-based MBM species-specific analysis was elucidated by conducting correlation analysis between composition of matrix fragments and the prediction scores of species discrimination analysis. Additionally, FT-IR spectral variables that contributed predominantly within the coupling process were investigated in detail. • MBM matrix fragments were characterized and compared by FT-IR. • PLS-DA model using MBM performed worse than MF and BF based models. • Coupling effect of MF and BF on MBM species-specific analysis was analyzed. [ABSTRACT FROM AUTHOR]