Using portable visible and near-infrared spectroscopy to authenticate beef from grass, barley, and corn-fed cattle
Meat product labels including information on livestock production systems are increasingly demanded, as consumers request total traceability of the products. The aim of this study was to explore the potential of visible and near-infrared spectroscopy (Vis-NIRS) to authenticate meat and fat from stee...
| Autores: | , , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión aceptada para publicación |
| Fecha de publicación: | 2024 |
| País: | España |
| Institución: | Universidad Pública de Navarra |
| Repositorio: | Academica-e. Repositorio Institucional de la Universidad Pública de Navarra |
| OAI Identifier: | oai:academica-e.unavarra.es:2454/52557 |
| Acceso en línea: | https://hdl.handle.net/2454/52557 |
| Access Level: | acceso abierto |
| Palabra clave: | Grass-fed Grain-fed Meat authentication Vis-NIRS Machine-learning |
| Sumario: | Meat product labels including information on livestock production systems are increasingly demanded, as consumers request total traceability of the products. The aim of this study was to explore the potential of visible and near-infrared spectroscopy (Vis-NIRS) to authenticate meat and fat from steers raised under different feeding systems (barley, corn, grass-fed). In total, spectra from 45 steers were collected (380-2,500 nm) on the subcutaneous fat and intact longissimus thoracis (LT) at 72 h postmortem and, after fabrication, on the frozen-thawed ground longissimus lumborum (LL). In subcutaneous fat samples, excellent results were obtained using partial least squares-discriminant analysis (PLS-DA) with the 100 % of the samples in external Test correctly classified (Vis, NIR or Vis-NIR regions); whereas linear-support vector machine (L-SVM) discriminated 75-100 % in Test (Vis-NIR range). In intact meat samples, PLS-DA segregated 100 % of the samples in Test (Vis-NIR region). A slightly lower percentage of meat samples were correctly classified by L-SVM using the NIR region (75-100 % in Train and Test). For ground meat, 100 % of correctly classified samples in Test was achieved using Vis, NIR or Vis-NIR spectral regions with PLS-DA and the Vis with L-SVM. Variable importance in projection (VIP) reported the influence of fat and meat pigments as well as fat, fatty acids, protein, and moisture absorption for the discriminant analyses. From the results obtained with the animals and diets used in this study, NIRS technology stands out as a reliable and green analytical tool to authenticate fat and meat from different livestock production systems. |
|---|