Species substitution is rampant in the seafood industry, but how often do our furry and feathered protein sources get swapped? Was Horsegate the exception or the rule?
In 1995, researchers from the Florida Department of Agriculture and Consumer Services analyzed over 900 meat samples collected from Florida retail markets; 806 were raw samples of meat and 96 were cooked. For their analysis they turned to proteomics (the large-scale study of proteins). One of the methods they used was ELISA (Enzyme-Linked ImmunoSorbent Assay), which is based on antibodies recognizing and binding to specific animal proteins. For this technique, antibodies are developed for specific proteins, such as a heat-tolerant protein found in the muscle of pigs. The sample of ‘all beef’ sausage is mixed with the antibody and if the pig protein is present, it will bind to the antibody. The sample is then washed to get rid of any unbound sample. A second antibody can then be added, which is linked to an enzyme. This antibody binds to any bound pig protein. Finally, a labeling reagent is added. This reagent contains a substance that will be converted into a colored product by the enzyme. This color change can then be used to simply indicate presence of the protein of interest or it can be measured against a set of standards to determine the concentration of protein present.
Of the 900 meat samples, the researchers found that 149 (16.6 percent) contained more than 1 percent of an undeclared meat. The substitution rate was higher among the cooked meats (22.9 percent) than the raw meats (15.9 percent). The undeclared species found in minced beef and veal products were sheep, pork and poultry. However, it must be stated that immunoassays will only recognize the species that they have developed and added antibodies for. In other words, unless they added antibodies for rat and dog, they wouldn’t have found them.
In 2006, a group of Turkish researchers used immunoassays to test processed meat products such as fermented sausages, salami, frankfurters, pastrami, bacon and canned goods. They found 22 percent were adulterated; 11 of the 28 sausages that were labeled as beef contained only chicken.
China has been riddled with meat substitution scandals. There have been reports of rat, mink and fox meat being transformed into mutton slices. Twenty thousand tons of meat were seized and more than 900 people were detained in association with the scandal. In early 2014, Walmart’s operations in China were recalling donkey meat because it had been adulterated with fox meat. Donkey is a very expensive meat and highly sought after for its tenderness and sweetness; fox, not so much.
Pork is swapped for beef, beef is swapped for buffalo, fat trimmings and offal (internal organs) are added to minced beef, chicken is sold as lamb, pork is sold as chicken, and beef and pork gristle and bones are injected into chicken. The list is long, and this is just substitutions between animal species. We haven’t yet mentioned the undeclared ingredients – such as added water, chickpea flour, rice flour and soy – that are added to meat to bulk it up.
Of course, there’s always the possibility that some of these undeclared species are the result of accidental cross-contamination. When a processor takes a carcass from a slaughterhouse and debones it and takes it down to smaller cuts, there are a number of leftover bits that aren’t particularly useful as a cut of meat, and these are called the trimmings. For beef, about 15 to 20 percent of the carcass will end up as trimmings, so this is a significant amount of meat that it would be shameful to waste. The trimmings are shipped to a processor that then mixes the extra fatty trimmings with the extra lean trimmings to get the desired fat-to-lean content for their customers. It can then be packaged up as mince and sold on to other manufacturers or retailers, or processed further into things like burgers. Processors work with several different types of meat and so there is a possibility that some minced pork remnants will be left in the pipeline and get pushed through when the beef goes through the machine. As a result, there is some forgiveness in levels of contamination. The European Food Safety Authority (EFSA) uses a 1 percent threshold – anything above this level of contamination is considered to be intentional adulteration.
As with any food, the more processed the meat, the more difficult it is to tell by visual inspection alone whether it’s been tampered with. By definition, mince is a mixture of meat that’s ground beyond any hints as to its animal origins. The only distinguishing feature the meat sitting in a plastic tray, bound by a thin layer of protective plastic wrapping, has is its color. We can judge its animal origins based on its shade of red – ranging from pale poultry pink to vibrant venetian red venison. The fat content can be estimated based on the relative proportion of red bits and white bits. Freshness is assessed by the saturation of the color – is it a dull grey color or bright red? It’s not a lot to go on and even these attributes can be manipulated. While it’s easy to distinguish some turkey mince from beef, things can get more difficult between the red meats such as horse and beef.
Products such as sausages are among the most prone to adulteration. While there is the possibility of cross-contamination as we just mentioned, the more cynical (and one could argue realistic) viewpoint is that cheaper substitutions are easier to hide in a processed product. In 1991, researchers from the University of New South Wales, Australia went out and bought samples from butchers and supermarkets of the most commonly consumed sausages – thick beef, thin beef and thick pork. The researchers were interested in nutritional quality, but as there had been an article in the media about adulteration, they decided to test for other species in the samples using the ELISA method as well. Cow, sheep and pig meat were detected in all of the ‘all beef’ sausages, thick and thin. Of the 10 pork sausages tested, three contained only pork as labeled, three contained undeclared cow meat and the remaining four samples contained cow and sheep in the ‘pork’ sausage. Of the 30 sausages tested, only the three pork sausages were labeled correctly.
In 2012, researchers in South Africa examined a total of 139 processed meat products – from minced meat to deli meat – to look at what ingredients were not being declared on the label.8 They used the ELISA method to detect undeclared plant proteins, but also used DNA-based methods to look for a total of 14 animal species. They found undeclared plant and/or animal species in 95 (68 percent) of the samples. The highest rates of adulteration were in sausages; nearly half of the sausages contained undeclared pork. All together, the meat products tested contained undeclared soy, gluten, beef, water buffalo, sheep, goat, donkey and chicken. The majority of the products were not complying with labeling laws.
Adding or substituting meats and using vegetable fillers to bulk up the end product in sausages and other highly processed meats isn’t difficult. It’s a matter of adding another ingredient into the giant mixer as it blends together the meats and spices. Doner meat is similar – it’s mince and spices mixed together – which is probably why 70 percent of lamb kebabs from British takeaways tested in 2013 contained cheaper, undeclared meats.
Substitutions aren’t limited to highly processed meat
The techniques of the fraudsters are now sophisticated enough that substitutions can happen beyond the minced and processed meats. Let’s return to the example from China of fake mutton. Thinly shaved mutton slices are a popular hotpot ingredient. There have now been several scams unveiled in China that have involved the sale of fake mutton; one operation, raided in January 2013, had 40 tons of fake mutton and another 540 tons of materials to make more. Allegedly, rat, mink, fox and duck meat have all been used as the base for this fake product. These meats are apparently soaked in a cellulose gum (sodium carboxymethyl cellulose), which is commonly used in food manufacturing to extend shelf life, improve freeze/thaw stability and help bind water. In the making of fake mutton, this process allows the meat to take on more water and therefore increase its apparent weight. Food coloring is used to provide the ideal shade of mutton, and food adhesives (more on meat glue later) are used to bind the fake meat with real mutton fat. The end product is a passable, but not indistinguishable, version of mutton. What sets the fake apart is that the fat is not marbled throughout the meat as would be the case naturally. The fat and meat are quite separate and when it is thawed or cooked, the adhesive fails and one is left with fragments of fat and fragments of meat.
Despite what seems like an arduous process, making the fake mutton is worth the effort. The fraudsters can sell it wholesale to restaurants for about £2.12/kg (US$1.45/lb) less than the real thing, allowing them to undercut competitors selling real mutton. Forty tons of fake mutton would turn a profit of about £128,000 (US$192,000) – nearly 23 times the average annual salary in China for 2014.
In September 2013, police confiscated 20,000kg (44,000lb) of pork masquerading as beef from a factory in north-west China. Not mince, not sausages, not even thinly sliced ‘mutton’, but whole cuts of pork that had been made to look like beef. One wouldn’t think it was possible. The pork is mixed with beef extract and a glazing agent and left to sit for ninety minutes. When cooked, the meat takes on a dark beef-like appearance rather than the characteristic white pork color. The beef extracts even give it the beefy aroma one would expect. Though it may be more difficult to swindle people over whole cuts of meat, this shows it’s not impossible.
If one end of the spectrum is to transform a pork chop into a steak, the other end is to change things at a microscopic scale. In 2001, the UK FSA released results of an investigation that was carried out jointly with 22 local authorities. They tested 68 samples of chicken breasts that were being sold to the catering trade and found that more than half of them were mislabeled, including some that contained undeclared hydrolysed protein. Hydrolysed protein is protein that’s been broken down into smaller segments known as peptides, usually using an enzyme This can be a very useful process as it can remove the allergenic properties of proteins and make them more easily digestible. Baby formula, for example, contains hydrolysed milk proteins (casein or whey). Collagen, which is the main structural protein derived from bone, connective tissue, skin and hide, forms the ideal water-retaining agent when it is hydrolysed – gelatin – and it was this that was being added to the chicken breasts.
The protein powder is purchased by processors and made up into a brine solution. This solution is then directly injected into the breasts using needles, or the chicken breasts are tumbled with the solution in a machine like a cement mixer. Either way, the breast meat takes up this solution and the hydrolysed protein helps retain water, even while cooking. The result can be a product that actually contains as little as 55 percent chicken; the rest is additives, including water. This is a perfectly legal process, but it must be labeled correctly as ‘chicken breast fillets with added hydrolysed chicken protein.’
The technique was developed by Dutch processors to introduce protein and water into salted chicken that they were importing from Brazil and Thailand. The processors were taking advantage of an EU tax loophole, as salted meat is subject to much lower import tariffs. By adding water to the chicken, they were making it more palatable but also effectively selling water for the price of chicken. Of the 68 samples taken by the FSA in 2001, 20 percent contained undeclared hydrolysed protein.
Shortly thereafter, it was revealed that some Dutch manufacturers were not only adding undeclared hydrolysed protein, but also the protein was being extracted from other animals. The FSA conducted DNA testing on 25 samples and found that almost half of them contained traces of DNA from pigs, though all but one of those samples were labeled as halal (meat that adheres to Islamic law and certainly would not include pork).They suspected that beef protein was also being used, but their DNA-based methods weren’t picking up any beef DNA. The hydrolysed protein powders are extremely processed, making any DNA, if present at all, very difficult to detect – particularly when looking for a very small amount of beef or pork DNA in a lot of chicken.The proteins are also fragmented through processing, which eliminated the use of immunoassays, such as ELISA. The FSA needed a new test.
The FSA collaborated with researchers from the University of York who had developed new procedures to identify species of ancient bone fragments dug up in archaeological sites. Archaeological work suffers the same challenges faced by food forensics in that the proteins have decayed – though in the case of archaeology it is through time rather than processing. The researchers had discovered that the collagen protein found in bone has enough variation between species to be useful in fingerprinting collagen-based tissues (for example, bone, cartilage, skin, tendon, blood vessels). Luckily, the hydrolysed protein in the chicken breasts had been extracted from these types of tissues.
The technique the York researchers have developed is called ZooMS, short for ZooArchaeology by Mass Spectrometry. For the analysis, proteins in the tissue sample are cut up into peptide fragments using the enzyme trypsin. The mass of each peptide is then determined using time-of-flight mass spectrometry, which essentially shoots the peptides out using an electric field and uses a detector to see how fast they fly a particular distance. This provides a unique mass-to-charge ratio for the peptides. Certain peptides (fragments of the collagen protein) are species-specific and can be identified by comparing them with a library of collagen proteins developed for different species.
As the chicken processors don’t make the hydrolysed protein powders themselves, the FSA conducted an investigation of the powders directly. They obtained five sample powders, four of which were made in the UK and all of which were labeled as containing only poultry-derived protein. They ran them through numerous tests, including the methods developed atYork.They used real-time PCR to look for chicken DNA in three of the powders and found that two powders tested positive for chicken DNA only and one tested positive for chicken and pork DNA. Had the tests ended there, one might have suspected that only one of the powders had used species other than chicken. Luckily they didn’t stop. Analysis of the collagen protein showed that none of the protein in all five powders had been derived from chicken. All of them contained bovine collagen-specific peptides and two of them contained bovine and porcine-specific peptides. Interestingly, two of the powders also contained unidentified non-food animal peptides.The powders had tested positive for chicken DNA probably because a small amount of chicken blood had been added, which would mask any pork or beef DNA that was likely to be highly degraded. It was the analysis of the collagen proteins themselves that revealed the true sources of the hydrolysed protein in the powders. Mislabeled powder means that some chicken processors may be unaware that they’re injecting protein from other species. Some processors have shifted to the use of plant-based protein powders as a result.Yet it does not eliminate the fact that the process is introducing a lot of water that consumers are paying for, which is fine as long as it’s labeled as an added ingredient. Consumers can then make their own decisions about whether they want to pay for water. But when it’s undeclared ... that’s fraud.
Excerpted from "Sorting the Beef From the Bull: The Science of Food Fraud Forensics" by
Richard Evershed and Nicola Temple. Published by Bloomsbury Sigma, a division of Bloomsbury Publishing. Copyright © 2016 Richard Evershed & Nicola Temple. Reprinted with permission of the publisher. All rights reserved