Source Verification using Biometric Identification - The key to future International Trade?

Associate Professor Steve Morris
Institute of Veterinary and Animal Science
Massey University

(This article quotes extensively from an address by Professor Bruce Golden, Department of Animal Science, University of Colorado) to the National Beef Science Seminar, Lethbridge, Alberta, Canada on 26-28 January 2000 [bgolden@gcp1.agsci.Colostate.EDU])

Source verification of animals refers to methods or systems that allow the source(s) of individual animals to be traced and independently verified. to satisfy various needs, such as compliance with regulations governing food safety and disease control.

Currently, the greatest push for source verification emanates from countries with serious animal disease control and contaminant issues (such as BSE in United Kingdom and the EU) and from countries that wish to export to the EU. The European Union (EU) and the United Kingdom (UK) both have mandatory animal identification systems in place (as has New Zealand). Source verification creates an accountable supply chain for consumers (and retailers/supermarket chains), but it can create a headache for producers.

Demands for source verification in the US will occur because consumers, retailers, vertically integrated producers and government expect dramatic improvements in food safety, animal welfare, environmental protection, and health management which in many cases will lead to branded products. The ability to market beef as having been produced in certain conditions requires the ability to verify the source of these products. McDonald's, for example, would like to be able to say that cattle they use have been treated humanely and food safety precautions were adhered to throughout processing.

The objective of source verification is to be able to trace the ownership and location of animals from birth to slaughter, and ultimately to the retail meat counter. Consumers want a system that is reliable. Producers want a system that is not burdensome. For different reason both want a system that is low cost.

Professor Golden argues that only a biometric method has the potential to completely satisfy the requirements of producers, consumers and regulatory agencies. Biometric identification is not a new science. Fingerprinting in law enforcement is the most commonly used biometric identification technique. There has been renewed interest into research into biometric identification led by the banking industry which intend use it as security tool for ATMs and other fraud prone transactions.

Professor Golden's research team has shown that using a digital image of the retina (located in the eye) is a low cost and reliable method for identification because retinal vascular patterns are unique, permanent, easily obtained and easily digitised (hence stored in a data base). The retinal pattern results in a simple image that is suitable for using as a key in a large database of animal images. It has the advantage of not having to attach a device to the animal (eg.electronic tags) and because digital imaging camera components have become inexpensive the cost of the system is attractive. Combine the secruity of using the innate structure of the retina with an encrypted Global Positioning System (GPS) receiver signal (patented), and it becomes virtually impossible to foil or deceive the system.

The landmarks that Professor Golden's team uses are the patterns created by the retinal blood vessels. These are highly unique, not only between animals, but also between eyes within animal. Their research to date shows that the image of the retina is completely repeatable in the normally developing bovine from birth to maturity. The United States Navy initially developed retinal scanning in the early 1970's as a secure door entry system.

The most popular solution for animal identification in source verification programs is radio frequency based ear tags. However, these are very expensive and potentially prone to falsification as is any tag based system. The use of DNA and other molecular based identification methods are currently receiving a great deal of attention as a proposed biometric identification system for livestock in source verification systems. However cost (US$ 25 per animal sample in USA) and turnaround time (currently one to six weeks) are a disadvantage for this method. This compares with a retinal image costing US$ 0.60 that can be obtained in three to five seconds. Tissue based methods (such as DNA) can be tampered while retina imaging combined with a GPS receiver signal are virtually tamper proof according to Professor Golden.

This is something the New Zealand Beef industry should seriously investigate as one shudders to imagine what might happen if a food safety scare or violation (real or imagined) does occur and closes access to our markets. To insure against this we need verifiable and tamperproof "source identification" of our beef cattle. Professor Golden and his group suggest that only a biometric method (and they advocate retinal scanning) combined with a secure location indicator (eg. GPS information) will satisfy the needs of our markets, and do it a cost-effective manner