WITH beef commanding top dollar, the incentive to counterfeit species and cuts is stronger than ever and the ways of doing it are becoming more and more sophisticated.
While meat fraud is rare in Australia, cattle producers certainly have a strong interest in stopping it - or even accidental substitution - overseas, given three quarters of Australian beef is exported.
That's why scientists in Australia are honing in on ways to outsmart the fraudsters.
Advanced machine learning algorithms that researchers at the Queensland Alliance for Agricultural and Food Innovation are helping to develop have the power to validate the species, cut and even provenance of meat.
ALSO IN BEEF:
Chair of meat science at QAAFI, Professor Louwrens Hoffman said with population growth increasing demand for food, there was considerable economic gain in adulterating food - swapping premium products for inferior products or species.
High value products, such as meat, were especially susceptible to food fraud, he said.
While the more expensive cuts like sirloin and rump are tricky to scam, stewing beef and mince is far more homogeneous.
"But each cut has a different chemical compound structure and we can rapidly determine differences using imaging technology," Prof Hoffman said.
The machine learning models he and colleague Daniel Cozzolino are working on will be mostly used overseas in markets where, for example, large numbers of horses are also slaughtered, he said.
The end product of the scientific work will typically be used by retailers to authenticate product, he said.
It will have other implications on home soil too, however.
"We don't think Australia has much fraud in meat but it might be the consumer is not aware of practices like adding plant-based ingredients to products like sausages," Prof Hoffman said.
Processors can also utilise the science to build on their analytical testing.
"The problem is a carcse has an ID up until it is cut into steaks - and then we can't determine which animal it came from but this gives us a way to do that," Prof Hoffman said.
"It is similar to testing DNA but faster and cheaper and not invasive."
The scientists are using light-based, or spectroscopic, technology to provide data about a meat sample.
Commercially available handheld devices that emit light in the near-infrared (NIR) range can collect from a sample it's 'signature', Prof Hoffman explained.
More work on decoding is still needed, however.
The algorithms will 'solve' the puzzle of the statistical information that is the identifiers of meat traits, which are beyond the ability of human senses to detect.
This type of technology could also be used in other areas for quality control purposes, Prof Hoffman said.
"For example, in wheat for determining protein and moisture content or to detect plastic fragments in food," he said.
For all the big news in beef, sign up below to receive our Red Meat newsletter.