Meat Speciation by near Infrared Reflectance Spectroscopy on Dry Extract

Publisher Website
Google Scholar
Abstract
Professionals and consumers want to control the origin of meat, while producers can profit by mixing minced meat from low cost species into high value meat. Near infrared (NIR) spectroscopy on dry extracts was studied as a method for speciation of minced beef, pork, mutton and mechanically recovered poultry meat. This was divided into three levels: (1) classifying into “labelled species or not?”, (2) into one of four possible species or (3) predicting the degree of substitution of minced beef with other species in a model study. Minced meat from 68 carcasses from four animal species ranging from low to high fat contents were centrifuged. Extracted meat juices (0.5 mL) were dried on glass microfibre filters and the reflectance measured. Using the range 780–2500 nm and different methods of discriminant analysis by cross-validation, the 68 samples were classified 90–100% correctly according to speciation at levels 1) and 2). Level 3) included a Simplex design model study using 350 mixtures of the 68 centrifuged juices. Predicting the relative amounts from each species in samples containing 50–100% beef juice gave cross-validated prediction errors of 12% (w/w) for beef, 8% for pork and 7% for mutton and poultry. The results showed that NIR on dry extract might be used as a rapid screening method for classification and optimised to detect species adulteration.