Authenticity for the Australian Seafood Sector: A Review of Available Tools to Identify Substitution and Mislabelling Prepared by: Stephen Pahl AUGUST 2018

Authenticity for the Australian Seafood Sector: A Review of Available Tools to Identify Substitution and Mislabelling

Information current as of 07 August 2018

© Government of South Australia 2018

Disclaimer

PIRSA and its employees do not warrant or make any representation regarding the use, or results of the use, of the information contained herein as regards to its correctness, accuracy, reliability and currency or otherwise. PIRSA and its employees expressly disclaim all liability or responsibility to any person using the information or advice.

All enquiries

Dr Stephen Pahl South Australian Research and Development Institute (SARDI) division of Primary Industries and Regions SA (PIRSA) Gate 2B Hartley Grove, Urrbrae, SA 5064 GPO Box 397, Adelaide SA 5001 T 08 8429 2298 M 0477 336 181 E [email protected]

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 2 TABLE OF CONTENTS

ACKNOWLEDGEMENTS ...... 5

EXECUTIVE SUMMARY ...... 6

1. INTRODUCTION ...... 7 1.1 FOOD FRAUD – WHAT IS IT? 7 1.2 WHAT CAN BUSINESSES DO? 9 1.3 OBJECTIVE 10

2. BACKGROUND ...... 11 2.1 ORIGIN, PROVENANCE, TRACEABILITY AND AUTHENTICITY 11 2.2 SEAFOOD MISLABELLING IMPACTS AND INCIDENCES 12 2.3 REQUIREMENTS FOR SEAFOOD PRODUCT TRACEABILITY IN SEVERAL OVERSEAS JURISDICTIONS 16

3. METHODS ...... 17 3.1 LITERATURE REVIEW 17 3.2 INTERVIEWS TO ESTABLISH CASE-STUDIES 17

4. TRACK AND TRACE TECHNOLOGY ...... 18 4.1 BARCODES AND LABELS 18 4.2 SERIALISATION 21 4.3 RFID (RADIO-FREQUENCY IDENTIFICATION) 21 4.4 INKS, MICRODOT AND OTHER TAGGANT TECHNOLOGY 22 4.5 DISTRIBUTED LEDGERS 22

5. AUTHENTICITY TOOLS BASED ON PRODUCTS’ INTRINSIC PROPERTIES ...... 24 5.1 VISUAL ASSESSMENT 24 5.2 NUCLEIC ACID BASES TOOLS 25 5.3 PROTEIN/PROTEOMICS 30 5.4 FATTY ACIDS/LIPIDOMICS 33 5.5 MINERAL/ELEMENTAL ANALYSIS 33 5.5.1 Native Elements ...... 33 5.5.2 Stable Isotopes ...... 35 5.5.3 Inclusion of Rare Earth Elements ...... 36 5.6 OTHER SYSTEMS 36 5.6.1 Spectroscopic Methods ...... 36 5.6.2 High Resolution Mass Spectrometry ...... 38 5.6.3 Organoleptic and Sensor Based Technologies ...... 41

6. INDUSTRY CASE STUDIES ...... 42 6.1 NANOTAGS FOR AUSTRALIAN WILD ABALONE 42 6.2 MINERAL/ELEMENTAL ANALYSIS IN PRAWNS 44

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 3 7. FINAL REMARKS ...... 46

8. USEFUL RESOURCES ...... 49

9. REFERENCES ...... 50

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 4 ACKNOWLEDGEMENTS

This work has been made possible through the SafeFish research program and supported by the SafeFish Partnership Committee. Special thanks to Dean Lisson (Abalone Council Australia Ltd), Helen Barnard (Honey and Fox, ex Tasmanian Abalone Council), Spiro Markantonakis (Dover Ex27 Pty Ltd), Darvin Hansen (Tasmanian Seafoods), Rachel King (Australian Council of Prawn Fisheries) and Janet Howieson (Curtin University of Technology) for supplying information for the case studies. Many thanks to Dr Garry Lee for his critical review of the report.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 5 EXECUTIVE SUMMARY

Food fraud is generally considered to be a growing concern for businesses, governments and law enforcement agencies. Seafood is widely recognised as one of the four major food and food ingredient sectors that has been targeted for fraudulent activities. The seafood sector is often vulnerable due to its complex supply chains and difficulties in identifying occurrences of fraud. Fraudulent activities can involve substitution of more expensive species with less expensive or illegally fished species, mislabelling of credence values, product adulteration through blending or dilution, use of unapproved enhancements, counterfeiting. Economic losses are often the most cited impacts, although some cases fraudulent activities can implicate food safety, threat to animal conservation and/or sustainability.

The extent of food fraud associated with Australian seafood is not well documented, however recent estimates indicate that the overall cost of seafood fraud for Australian exporters was AUD$189 million (McLeod, 2017). The ability to be able to detect seafood fraud can help deter fraudulent practices. Product authentication practices may require the use of numerous features that incorporate varying levels of technological complexity and a combination of overt and covert features. Commercially available technologies are based on physical track-and-trace tools (such as attached barcodes, RFID and other labels or taggants) or interrogation of the intrinsic attributes of the product (i.e. DNA, proteins, fatty acids, minerals, elemental and other metabolomics). These can be coupled to distributed ledgers to ensure that no single entity can tamper with an entry. There is often an underlying requirement that the selected technology must be able to stand up in a court of law. Numerous agencies and research institutions are seeking to develop handheld and/or portable field based systems that can be positioned at known points of vulnerability. Furthermore, the continual improvement in advanced analytical methods coupled with chemometrics has enabled a range of food ‘fingerprints’ or ‘profiles’ to be developed with greater sensitivity and robustness.

The case studies included in this report demonstrate that Australian seafood sectors are aware of the importance of brand protection and being able to demonstrate the origin, safety and sustainability of their products. There is generally no single technology that is better than any other; with each technology having its own strengths and weaknesses. Initial consideration should be given to establishing the intended goals or outcomes, and then selecting, trialling and reviewing the technology and findings that align with your product(s) and markets. It is important to understand start-up and on-going costs, easy-of-use, reliability and accuracy of the technology and to also consider how the information obtained could be utilised.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 6 1. INTRODUCTION

Australian food is ranked the highest in the Asia-Pacific countries in terms of safety and quality; and is the fourth highest globally behind Portugal, France and the USA (The Economist Intelligence Unit, 2016). This reputation allows our products to attract a premium price in domestic and export markets. However, food fraud is increasing, costing the global food industry a reported AUD$50 billion annually with an estimated 20% of in-store and 40% of on-line food products being adulterated or counterfeited (Anonymous, 2016b; McLeod, 2017). Approximately one in three companies are victimised by fraud (PwC Australia, 2016), and seafood is recognised as one of four major foods and food ingredients susceptible to fraud (Moore et al., 2012).

In 2016 SafeFish determined that seafood fraud and authenticity was a concern for the Australian industry and should be included as an area for priority research1. This was principally due to a potential threat to the industries reputation as a provider of high quality and safe seafood. In 2017, it was estimated that the overall cost of seafood fraud for Australian exporters was AUD$189 million (McLeod, 2017). Consumer awareness of the origin, quality, safety, sustainability and production practices of the foods they consume are also generally increasing. Recent scandals such as the discovery of horse meat substituting as beef in Europe (2013) and the addition of melamine to milk formula in China (2008) have resulted in broad media attention and a potential loss of consumer confidence in the food industry. To help address food fraud, the food industry needs to strategically shift its focus to the prevention of fraudulent activities and to be able to authenticate products. The ability to be able to determine and demonstrate the authenticity of food is an important facet for quality control and food safety as it can help deter fraudulent activities from occurring, provide greater market confidence, improve public safety, increase business competitiveness and ultimately should increase the economic returns for honest fishers and traders.

1.1 FOOD FRAUD – WHAT IS IT?

Many organisations have considered that food fraud is the intentional adulteration of food for financial gain. Though a more formal definition proposed by Spink and Moyer (2011) is that “food fraud is a collective term used to encompass the deliberate and intentional substitution, addition, tampering, or misrepresentation of food, food ingredients, or food packaging; or false or misleading statements made about a product for economic gain”. More recently, Elliott (2018) suggested an alternative definition being “food fraud is any actions taken by businesses or individuals that deceive others businesses and/or individuals in terms of misrepresenting food, food ingredients or food packaging that bring about financial gain”. The latter definition by-passes the need to demonstrate intent, which is often one of the most difficult things to prove in a court of law; and would include the inadvertent misnaming of a product. Inadvertent misnaming of a product includes incorrect identification at capture or wholesale, incorrect placement of unlabelled product at the point of sale or retail staff forgetting to remove display signage. Such examples are generally considered to be a quality issue rather than fraud.

The food risk matrix, shown in Figure 1, illustrates how actions (whether intentional or unintentional) and motivations of people or consequences can affect the outcome of food. For example, a food safety risk is a public health threat that is unintentional, such as seafood contaminated with marine biotoxins or e-coli. The implementation of Hazard Analysis Critical Control Points (HACCP), Good Manufacturing Practices (GMP) and Good Hygienic Practices (GHP) can reduce the likelihood of a food safety incident. On the other hand, a food defence risk is a public health treat that is intentional, such as malicious tampering by a disgruntled individual or terrorism. Food defence systems aim to mitigate the risks and hazards of intentional contamination in food operations by introducing a selection of countermeasures to make vulnerable

1 SafeFish Prioritisation Workshop held in Adelaide, South Australia on 25 October 2016.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 7 elements more secure (Moerman, 2018). The types of intentional actions leading to food fraud can be illustrated by Figure 2.

ACTION

Unintentional Intentional

Gain: Food Food Economic Quality Fraud

/CONSEQUENCE Harm: Food Food Public health, economic or Safety Defence terror

MOTIVATION

Figure 1. The Food Risk Matrix: Effect of motivation and action on outcomes of food. Adapted from Spink and Moyer (2011).

Substitution (e.g. species, Dilution wild/farmed, (e.g. mix of catch method) Concealment species, (e.g. harvest overglazing, method, live overbreading, transport undeclared conditions, water-binding stocking density) agents, non-fish protein) FOOD Counterfeiting (e.g. replicas of FRAUD Mislabeling packaging/ (e.g. provenance) branding)

Unapproved Grey market enhancements production/ (e.g. use of theft/ diversion carbon (e.g. IUU fishing, monoxide) smuggling)

Figure 2. Types of food fraud. Adapted from Spink (2014).

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 8 1.2 WHAT CAN BUSINESSES DO?

Vulnerability to food fraud is a combination of opportunity, motivation and inadequate control measures. It has been recognised that many food and wine exporters are currently doing too little to protect their own brand or ‘Brand Australia’ in foreign marketplaces (Anonymous, 2016a). A single food safety incident can cost up to hundreds of millions of dollars, and can also permanently destroy a valuable brand, cause long- term industry-wide losses, risk market access and damage trust in public institutions (PwC, 2016).

Seafood is the most traded food commodity in the world, and its movement can involve complex and diverse supply chains. Major drivers for committing food fraud are based on opportunity (i.e. complex supply chains, largest traded food commodity, unsatisfied market demands), pressure (i.e. desire on individual or business to drive up profits, competition with other businesses) and attitude or rationalisation (i.e. being able to convince yourself that it is ok to cheat). It is often easy to do, hard to catch and even harder to prove.

Businesses or organisations should recognise and appreciate their product and any similar products, the potential complexity of their supply chain and the customs and values of end-users or consumers. The following four-step strategy outlines a process that can be followed to help protect markets and prevent fraudulent activities.

Step 1 – Conduct a documented food fraud vulnerability assessment

Vulnerability Assessment and Critical Control Point (VACCP) analysis and the implementation of subsequent mitigation plans can help guard against economically-motivated adulteration. Many service providers have established capability to assist businesses in completing VACCP. Within Australia VACCP plans are currently being requested by some of the larger food chains and may become a mandatory part of every food supplier’s food safety program.

Preventive measures should take into account the economic incentives and deceptive criminal behaviours. Proactive approaches employed by some food businesses include having better control over their supply chains from farm to supermarket shelf (PwC Australia, 2016). In addition businesses should implement awareness and prevention programmes to reassure employees, downstream stakeholders and consumers (if applicable) that they are actively preventing fraudulent practices from occurring. Step 2 – Review and determine which tool(s) are the most appropriate to protect business risk and reduce fraud opportunities

Performing in-house or external assessments can not only detect any possible fraudulent practices, but also provide important feedback on the efficacy of your chosen mitigation plans. This document should assist in identifying what tools and/or techniques are available and how they have been previously used. Many scientific publications are based on a particular technology, or group of technologies, looking for an application. It is important that when considering implementing provenance, traceability and/or authenticity tools that the desire for a solution should drive the technology choice or development, and not the other way around.

When selecting the most appropriate tool(s) and/or technique(s) be clear on your intended goal or outcome, i.e. is it to determine efficacy of mitigation plans, increased ability to be able to detect fraudulent practices, protect brand, and/or threaten or proceed with litigation? The strengths and weaknesses of each tool should also be considered to ensure its applicability to your intended use, product/s and markets. Other key considerations include the level and availability of required expertise, as well as setup and continued maintenance costs and the cost of conducting a traceback in the event of a food related incident. Care should be taken so that cost doesn’t becomes a determining factor in choice of methodology at the expense of achieving the initial objective. If the information is to be used for litigation, then a forensic approach would be required.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 9 Public education on the characteristics of various seafood species or products may also assist in consumers being able to better distinguish between two similar looking fish species. Step 3 – Implement assessment trial or strategy

Implementation will enable a range of information to be collected. This might be the ease and cost of which the technology can be implemented in the supply chain, how stakeholders interact with the technology, or being able to reliably differentiate a product based on various intrinsic attributes. It is important to consider who will do what tasks, how will these be done, when will they be achieved and who will pay. Step 4 – Review and evaluate

Monitoring and evaluation are the basis for continuous improvement and provide confidence that the strategies are on-track or need revision. They can also enable benchmarking to other enterprises if this information is available.

1.3 OBJECTIVE

The aim of this report is to review provenance and authenticity tools that can identify species substitution and mislabelling and may be suitable for uptake by the seafood industry. The objective is to provide information to the Australian seafood industry on the current ‘state of play’ and key considerations, including estimations of the development and going management costs. The report will provide examples of where such tools have been used within the seafood sector both nationally and internationally. The report will also contain two case studies where different tools have been used in Australia and why the tool was chosen, if it was successful, and any measure of success.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 10 2. BACKGROUND

2.1 ORIGIN, PROVENANCE, TRACEABILITY AND AUTHENTICITY

The origin of food is where the product comes from; for seafood this is the location of harvest. Provenance can be used to differentiate between products of various origins, often as a means to acquire additional value. Provenance branding is becoming more important as consumers become more aware of where their food comes from and how it is managed. Many markets have historically placed a premium value on Australian produce. The wholesale price of selected Australian and imported seafood is reported in Table 1. The premium can be due to Australia’s reputation for producing safe and high quality foods, however many other countries also produce safe, high quality foods. For example as illustrated in Table 1, the wholesale price of King salmon (Oncorhynchus tshawytscha) from New Zealand is greater than that of Atlantic salmon (Salmo salar) from Australia. Whilst there has been a reliance on the ‘clean and green’ mantra, it is not unique to Australia (CSIRO Futures, 2017).

Table 1. Typical wholesale price comparison of selected Australian and imported seafood. Adapted from Australian Government, Department of Industry, Innovation and Science (2017).

Approximate wholesale Approximate wholesale price (per kg) price (per kg) Percent Type of seafood – imported – Australian difference (AUD$) (AUD$) Barramundi – filleted $15.80 $27.00 +71% Flathead $16.00 (Argentinian) $30.00 +87.5% $29.50 (Chinook/King Salmon – filleted $27.50 (Atlantic salmon) -7% salmon – New Zealand) Octopus – cleaned (16/25) $8.00 $12.00 +50% Calamari – whole (6/8) $11.00 $16.00 +45.5% Prawn – cutlets (16/20) $18.00 $42.00 +133%

Food traceability programs were originally established to help provide safe food and whilst several definitions exist; it is generally considered to be the ability to track any entity (i.e. food) through all stages of production, processing and distribution. Within Australia minimum traceability systems provide a ‘one step back and one step forward’ element and can assist when corrective actions (such as product recalls) are required. Traceability systems can vary from paper-based records to electronic data systems that can include barcodes or other tags, readers/scanners and software. Full traceability systems provide all of this information at all points across the value chain. Traceability programs may also include attributes that facilitate data sharing among multiple stakeholders of the supply chain and capture critical tracking events. Traceability is also linked to consumer perception and can also support credence attributes, such as country of origin, source, and production methods. Effective traceability systems that involve best practice, as shown in Figure 3, can provide a number of benefits to businesses including reduced risk, greater market access and increased efficiencies.

Governments, businesses, and consumers have identified traceability as a valuable product attribute. Potential benefits include:  Improved consumer trust, consumers consistently receive what they are expecting  Improved safety and quality, through greater accountability and transparency  Improved market differentiation, supporting product credence claims (Tamm et al., 2016).

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 11

Business Risk Mitigation Best Practice

Market Operational Access Efficiencies

Figure 3. Key benefits to businesses when using effective traceability systems. Adapted from Sterling et al. (2015).

Recommendations for businesses from a Global Food Traceability Centre (GFTC) project were to:  View traceability from a strategic perspective  Establish purpose and objectives before selecting technology  Approach traceability with big vision, small steps (Sterling et al., 2015).

Whilst traceability is one method that can be used to help combat food fraud, it is insufficient on its own as it is vulnerable to errors, omissions and/or falsification. Consequently, there is a need for authenticity. Authenticity refers to the item being genuinely as it is described, whether this is species, provenance or other credence claims including method of production (i.e. farmed/wild) or catch. Increasingly attention is turning to methods for determining authenticity in the market place to help deter against fraudulent claims. Authentication is essentially the verification necessary to underpin traceability systems. Authentication methods must be able to provide irrefutable and repeatable results.

2.2 SEAFOOD MISLABELLING IMPACTS AND INCIDENCES

Economic losses are the most commonly cited impact from product substitution, with less expensive fish being switched more expensive fish, farmed for wild, thawed for fresh, incorrect declaration of harvest method, or products labelled from different geographical zones or country of origin. In addition, the sale of endangered or vulnerable species under another name, or fishing of illegal, unreported, and unregulated (IUU) products (such as from a protected area) can result in stock depletion and may compromise or threaten resource management and fishery conservation measures. A list of several commonly substituted seafood species is reported in Table 2. In most cases, what is represented on the product labels is more expensive and/or desirable than the identity of the actual product.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 12 Table 2. Examples of substituted seafood. Adapted from the U..S Food and Drug Administration (2017). Product label Actual product identity Various snappers (Lutjanus spp.) or Rockfish Red snapper (Lutjanus campechanus) (Sebastes spp.) Mahi-mahi Yellowtail Swordfish Mako shark Orange roughy Oreo dory or John dory Cod Alaska pollock Halibut Sea bass Dover sole Arrowtooth flounder Red drum (Red fish; Southern or Gulf) Black drum Lake or Yellow perch (Great Lakes) White perch or Zander Caviar (Sturgeon species) Paddlefish and other fish roe Walleye Sauger Chum salmon Pink salmon Scallops Skate wings Walleye Alaskan pollock Salmon Steelhead trout Blue crabmeat Imported crabmeat Wild caught salmon Farm raised salmon Abalone Top shell

Aside from the financial and reputational consequences to the industry, product mislabelling or substitution can also result in food safety implications by impacting consumer health from allergens, parasites, environmental chemicals, aquaculture drug residues, natural toxins or other components (Naaum et al., 2016). Some examples include:  Mislabelling of escolar (Lepidocybium flavobrunneum) or oilfish (Ruvettus pretiosus) as rudderfish (Centrolophur niger and Tubbia species) (Yohannes et al., 2002)  Shark and King mackerel (which can contain high levels of mercury) being substituted for other species (Naaum et al., 2016). Substitution with King mackerel also increasing the chance of ciguatera poisoning (Lehane and Lewis, 2000)  Pufferfish containing tetrodotoxin labelled as monkfish (Cohen et al., 2009).

Several recent reports and/or programs have highlighted the potential extent of seafood fraud. A joint INTERPOL-Europol program has been in operation since 2011 targeting counterfeit and substandard food and beverages around the world. Some investigations have focussed on specific threats and therefore have undertaken targeted campaigns; this makes it difficult to compare results from year-to-year or from one jurisdiction to another. However, key findings with relation to fish and fishery products from the coordinated actions from each of the operations are outlined below2:

2 Data extracted from publically available reports and press releases available from https://www.europol.europa.eu/.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 13  Operation OPSON I (December 2011): 5 tonnes of substandard fish and seafood; sale of fake/substandard caviar via the internet under investigation.  Operation OPSON II (December 2012): Seafood was the most seized product category with 1.9 tonnes of fish and 112.7 tonnes of seafood seized. 89% of seafood infringements related to food safety and false labelling in order to deceive the consumer.  Operation OPSON III (December 2013): Seafood was the second most seized product category with 762.8 tonnes of fish and fish products seized.  Operation OPSON IV (December 2014 –January 2015). Seafood was the third most seized product category with nearly 819 tonnes of fish and fish products, including red and black caviar, seized. One seafood processing facility in Italy was selling previously frozen (thawed) seafood as fresh. The seafood had also been sprayed with a chemical to disguise rotting flesh.  Operation OPSON V (November 2015 – February 2016): Seafood was the third most seized product category with 906.3 tonnes of fish and fish products seized.  Operation OPSON VI (December 2016 – March 2017): Seafood was the fifth most seized product category with 480.8 tonnes of fish and fish seafood seized. The most common infringement related to seafood was deceiving consumers, especially with species substitution, followed by food safety. One seafood processing factory in Portugal was found to be operating without a license was repacking almost expired sardines.  Operation OPSON VII (December 2017 – March 2018): More than 51 tonnes of tuna was seized in Europe. Some of this tuna intended for canning was illegally treated with chemicals to alter its colour to give the misleading impression of its freshness.

In 2016, an Oceana report indicated that nearly one in five (or 20%) seafood samples tested was mislabelled. Asian catfish, hake and escolar were reported to be the most commonly substituted, with Asian catfish being sold as 18 different types of higher-value fish (Oceana, 2016a). Oceana have also compiled a list that contains the global distribution of documented seafood fraud and species substitution; this distribution is shown in Figure 4. Other studies have suggested that on average 30% of seafood products are misdescribed or mislabelled, but also highlighted that many published studies did not detail any information on the sampling design (Pardo et al., 2016). The National Fisheries Institute (2016) responded with caution to the Oceana findings as a gross exaggeration of the overall status of seafood fraud as the study was biased towards seafood products that are commonly mislabelled.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 14

Figure 4. Distribution of seafood fraud and species substitution. Reproduced from Oceana (2016b).

Seafood fraud through mislabelling and species substitution is considered to be a widespread problem in both national and international markets and can occur at every stage of the supply chain, from the point of landing through to processing, distribution, retail and catering (FAO, 2018). One of the challenges faced by the wild-catch sector is ensuring the correct identification of the fish at harvest. Inadvertent mislabelling at this point might be carried throughout the supply chain. Globally it is generally considered that the extent of fraud and/or inadvertent mislabelling is greatest at food service providers (Pardo et al., 2016). From an Australian context there have been a limited number of published studies that provide information on the level of seafood fraud or mislabelling. The following examples demonstrate that fraudulent activities have occurred within Australia.  In 2016, Australian Food News compiled a list of the most commonly mislabelled fish in Australia. This included Catfish or Basa sold as Dory, Stick fish (or South American Flathead) sold as Flathead, Asian Sea Bass sold as Barramundi, Crimson snapper or Saddletail snapper sold as Red emperor (Australian Food News, 2016).  In 2015, a study seeking to investigate labelling accuracy in seafood retailers in Tasmania reported that whilst no products (out of 38 samples that resulted in suitable DNA extraction) were deemed to have been mislabelled, there were a few cases of naming discrepancies or ambiguity with broad or obsolete fish names3 being used (Lamendin et al., 2015).  In 2003, a pilot market survey was undertaken across Australian Capital Territory, Queensland, New South Wales, Northern Territory, South Australia and Western Australia targeting fillets claimed to be Australian Barramundi (Lates calcarifer), Red emperor (Lujanus sebae) or Western Australian Dhufish (Glaucosoma hebraicum). A total of 138 samples were collected from wholesale, retail and food service institutions. 86.8% of samples sold as Barramundi (79 out of 91) were Barramundi or a species that is closely related to Barramundi, whilst only 58.8% of Red emperor (20 out of 34) and 53.8% of Dhufish were correctly labelled (FSANZ, 2003).

3 The Australia New Zealand Food Standards Code, Standard 2.2.3, does not define specific names for fish. Instead it refers to the Australian Fish Names Standard (AS 5300) which provides guidance on standard fish names to be used in Australia.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 15  In 1996, a small study seeking to determine whether species-specific profiles using random amplified polymorphic DNA (RAPD) to aid in fish species identification found that four out of six fillets purchased at retail outlets did not match the typical Barramundi profile with three consistent with those generated by Nile Perch. In addition, one out of five John Dory fillets was not a match to six authenticated samples. Whilst the study found evidence of mislabelling it was not designed to assess compliance to labelling or prevalence of mislabelling (Partis and Wells, 1996).

The Marine Stewardship Council (MSC) certification is the most widely used third party certification scheme within Australia (McLeod, 2017). In 2016, the MSC reported a 100% compliance rate from MSC labelled products from Australia (9 samples tested) and 99.6% compliance rate from MSC labelled products globally, with only one product being incorrectly labelled (256 samples tested)4 (Anderson, 2016). Whereas in 2011, a study showed that 8% (3 out of 36 samples) of fish labelled as MSC-certified Chilean sea bass or Patagonian toothfish (Disscistichis eleginoides) were actually another species (Marko et al., 2011).

2.3 REQUIREMENTS FOR SEAFOOD PRODUCT TRACEABILITY IN SEVERAL OVERSEAS JURISDICTIONS

In most countries, no single government agency is responsible for regulating fish fraud, and no single food law directly addresses all aspects of fish fraud. Responsibilities are usually shared across official food regulatory authorities, border protection agencies, customs import authorities and other specialist agencies. One of the principle challenges in tackling seafood fraud at the international level is developing an agreed list of common names that are linked to scientific nomenclature (FAO, 2018).

In 2010, the European Union introduced Catch Certificates (Regulation No. 1005/2008) to trace imported seafood back to fishing vessels. The certificate requires information about the product’s catch vessel, transport vessel, scientific name, and FAO catch area, among other data. In 2014, the European Union also introduced new seafood labelling laws (Regulation No. 1379/2013) requiring businesses to provide details on catch methods and harvesting areas to consumers. DNA barcoding is routinely used in the European meat sector and many processors and supermarkets will only purchase processed meat products that are certified and can guarantee their authenticity (FAO, 2018).

The US Congress also introduced the Safety and Fraud Enforcement for Seafood Act in 2013 and in 2014 a Presidential Task Force on IUU Fishing and Seafood Fraud recommended the creation of a risk-based traceability program to track seafood from harvest to entry into US commerce, although it failed to include a role for DNA-based verification testing (Tamm et al., 2016). The US Bioterrorism Act of 2005 required all links in the food and feed supply chain to be traceable; however in 2006 less than 1% of all imported seafood was tested (Jacquet and Pauly, 2008). In January 2017 the US National Oceanic and Atmospheric Administration (NOAA) established a program which aims to combat IUU fishing through mandating more stringent and improved seafood traceability requirements. The Seafood Import Monitoring Program (SIMP) requires reporting of fish and fish product data to verify each imported shipment from the initial harvest event to arrival at the US border (Willette and Cheng, 2017).

4 The one incorrectly labelled product was labelled as MSC certified Southern rock sole (Lepidopsetta bilineata) when it was Northern rock sole (Lepidopsetta polyxystra). Follow-up investigations concluded that it was an accidental mix- up of the two closely related species; both of which had been caught in MSC certified fisheries. In addition one sample failed to yield a successful DNA result; this was a pouched Halibut product with chilli.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 16 3. METHODS

3.1 LITERATURE REVIEW

A literature review was undertaken through key word searches and supplemented with snowball sampling of relevant sources from the internet and the University of Adelaide’s Library search engine which retrieves publications from prominent scientific journals and databases including Web of Science and Scopus. Combination of key words used in initial searches included ‘seafood’, ‘fish’, ‘traceability’, ‘mislabelling’, ‘labelling’, ‘fraud’, ‘authenticity’, ‘adulteration’, ‘substitution’, ‘provenance’ ‘method’ and ‘detection’. Results from this initial search provided the basis for the review and particular topics were then individually researched to obtain further information. Example use of different technologies have been provided, where possible, to help demonstrate the broad cross-section of application.

3.2 INTERVIEWS TO ESTABLISH CASE-STUDIES

The industry case studies aim to provide an overview of the considerations made by two different Australian seafood industries in adopting anti-fraud technologies. The selected case studies were:  Application of NanoTags for Australian Wild Abalone (AWA) Certification  Mineral/elemental analysis of Australian prawns for provenance.

Key individuals from the respective industry associations, researchers or service providers were provided with a background on the project and asked if they were willing to participate. The individuals were provided with a series of prepared questions prior to the interview. These questions aimed to elucidate the reasons why they have or are investing in this area, when it became a research priority, why they chose a particular tool, what trials have been undertaken, and if any benefits or limitations have been identified. The responses were then combined to build the case-studies.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 17 4. TRACK AND TRACE TECHNOLOGY

Track and trace tools can utilise extrinsic elements such as barcodes, labels, serialisation and tags that are physically applied to the product or packaging. Whilst the application of many of these tools can help authenticate products and make counterfeiting more difficult, their use does not make counterfeiting impossible. Many anti-counterfeiting methods rely on clonable tags produced using a deterministic process, for which the low complexity and high predictability means that these tags can be copied by counterfeiters (Arppe and Just Sørensen, 2017). A comparison of typical costs and attributes of different food traceability technologies are summarised in Table 3.

Table 3. Cost and attributes of commercial food traceability technologies. Adapted from Pigini and Conti (2017). Radio Quick Near Field Paper Frequency Response (QR) Communication documents Identification codes (NFC) (RFID) tags Cost of single tag <$0.02 $0.15 <$0.02 $0.20 ($AUD) Read distance (m) 0.5 >1 0.5 0.2 Effective read time 10 min 0.2 s 2 s 0.2s (single item) Simultaneous reads 1 100’s 1 1 Cost of reader Nil (already in Nil (already in Nil $450 ($AUD) mobile devices) mobile devices) Ease of use High Low High High Total cost of system High High Low Low Information security Low High None High

4.1 BARCODES AND LABELS

Barcodes are one of the simplest forms of product identification, are ubiquitous on packaged products and are mainly used for identifying items through the supply chain and at point of sale. The family of EAN/UPC barcodes contain a unique product identification number called a GTIN (Global Trade Item Number). The GTIN is printed below the black and white strips of a typical barcode. GS1 is a global not-for-profit organisation that has become an industry leader in creating standards that uniquely identify, capture and share information about products that help make up the GS1 System. They have developed a range of data carriers for different applications. For example, GS1 DataBar barcodes, which became a global standard in 2014, can be used at point-of-sale and can encode a greater level of information including specific product attributes such as harvest dates, harvest locations, lot numbers, quantities, weights and packaging dates and provide a safer supply chain by providing traceability at point of sale, or prevent sale of products past the use-by date. The GS1 Global Traceability Standard has been developed to help with interoperability between traceability systems.

Whilst artificial assigned identifiers have been used to identify product and for certain traceability applications, these can be subject to loss, removal or damage. Text-only ID labels are being used less frequency, but can enable integration with ‘low-tech’ supply chain members or provide a last-resort fall-back measure (Kemény and Ilie-Zudor, 2016).

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 18 Quick response (QR) codes are two-dimensional computer-generated images that can be printed on some products and/or packaging. QR codes are an extension of traditional barcodes but are generally consumer orientated. Consumers can scan the QR code with suitably enabled smart phones or similar devices which automatically directs the user to a website to access additional product information.

A typical EAN-13 barcode, a GS1 DataBar, a GS1 DataMatrix and QR code is shown in Figure 5. The GS1 DataMatrix can capture detailed product information but is not intended for point-of-sale scanning.

a) b)

c) d) Figure 5. Example of typical barcodes or labels a) EAN-13 Barcode; b) GS1 DataBar; c) GS1 DataMatrix; and d) QR Code

A large number of organisations provide industry and consumer orientated traceability services. Examples of organisations that provide consumer orientated traceability services are included in Table 4. Several of these organisations also provide enterprise resource planning software to assist businesses in production, inventory and financial analytics.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 19 Table 4. Example of organisations that provide consumer orientated traceability services.

Product Name Website Description Traceability Service Providers

MyTrace http://fishtale.co.nz/ Product information accessed through scannable QR codes. Companies can customise the information available to consumers such as where and how the product was caught. SeaTrace™ http://www.caisoft.com/ Consumers access information about the seafood they are purchasing through QR codes on product labels. Website typically maintained by the seafood processor that can control the amount of detail provided.

fTrace http://www.ftrace.com/ Allows consumers to trace a code to retrieve data on where the product comes from, when and how it was produced, and information about quality and recipes. Uses both scannable QR and traceable alphanumeric codes.

TransparenSea http://www.gulfwild.com/ Use of gill tags containing unique alphanumeric identification numbers and scannable QR code to enable customers to retrieve product details. System does not permit tag duplication.

ScoringAg http://traceback.com/ Consumers use a traceable alphanumeric and QR codes Traceback on products to trace product details online.

ShellCatch http://www.shellcatch.com/ Information on traceability and environmental practices are delivered to consumers through scannable QR and traceable alphanumeric codes.

Fish Trax™ http://marketplace.fishtrax.org/ Consumers can use a scannable QR or traceable Marketplace alphanumeric code to retrieve data on the harvesting and production of their seafood.

ThisFish http://thisfish.info/ Consumers use a website to trace the origin of their seafood and can connect to the harvester who caught it. Uses unique traceable alphanumeric codes.

Trace and Trust http://traceandtrust.com/ Closed in December 2017. Enabled fishers to market directly to retails and restaurants. Consumers could also use a website to see details of seafood deliveries to restaurant and retailer location and trace a fish’s identification tag to the boat or fisher that caught it. Proprietary Traceability Systems

Red’s Best http://www.redsbest.com/ Traceability software tracks product from boat to plate through the use of scannable QR codes.

John West http://www.john-west.co.uk/ Allows consumers to use numeric barcodes and can codes to trace the origin of their canned seafood products through a website.

Bumble Bee http://www.bumblebee.com/ Allows consumers to enter can codes to trace the origin of their canned seafood produces through a website.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 20 4.2 SERIALISATION

The healthcare and pharmaceutical industries are considered leaders in the use of serialisation. Serialisation provides a unique identifier to every product which enable product traceability and facilitate authentication. The unique identifiers can be stored in a database along with other information regarding the item or product. Downstream supply chain members, including consumers, can scan the unique code to confirm the product’s authenticity. A typical serialisation system contains four building blocks; namely:  Unique identification codes  Data capture mechanisms  Managing links across the chain  Data communication across the supply chain (Matthews, 2016).

Serialisation typically takes 18 months to fully implement (GS1, 2016) and can be applied directly to packaging or labels including gill tags. A recent Mettler-Toledo survey of 37 North America and European businesses generally responded that they were well aware of the potential added value a serialisation solution could provide for their customers, themselves and their legal distribution partners (Mettler Toledo, 2017). Whilst serialisation solutions cannot stop counterfeiting, they can serve as a powerful weapon in the fight against product and brand piracy.

4.3 RFID (RADIO-FREQUENCY IDENTIFICATION)

Radio-frequency identification (RFID) is an automatic identification method, enabling electronic storage and retrieval of tracking information without the need for physical contact or line of sight (such as with barcodes). RFID systems can operate at a variety of frequencies and are based on transponders (usually electronic tags) and a reader (or transceiver). The tags can be either passive (the tag is powered by the electromagnetic field generated by a receiver), semi-passive (tags contains an internal battery which powers its internal circuit, but relies on the reader to supply power for data transmission), or active (contains an internal battery to power its internal circuit and to send data to the reader).

For pet and livestock applications, such as with the Australian National Livestock Identification System for cattle, sheep and goats, low-frequency (134.2 kHz) passive tags are currently used. These tags are less susceptible to environmental interferences from water or metal, but can only be read over short distances and do not allow data to be written to the tag. The tag ID and relevant information are generally manually entered and stored in databases. High frequency (HF; 13.56 MHz) and ultra-high frequency (UHF; 865-868 MHz or 902-928 MHz depending on country) tags have been used in pharmaceutical, textile and food processing chains to track and help authenticate item-level products and improve inventory management. Both HF and UHF tags contain anti-collision technology which mean multiple tags can be read at the same time, but cannot be used when there is a requirement to penetrate water or moist materials. Semi-passive and active tags are more expensive than passive tags but can monitor and record attributes such as product locations and temperature.

RFID systems are generally considered to be robust against tampering, though it is costly compared to traditional approaches, and information output also depends on the availability of an electronic reader (Hofherr et al., 2016). In 2013, the Hong Kong based Fukui Shell Nucleus Factory began embedding passive UHF RFID tags as pearl nuclei in molluscs to enable each pearl to be uniquely identified, track-and-traced and authenticated (Anonymous, 2016c). For food applications, product composition and critical production environments can impede tag performance making it difficult to obtain exhaustive and reliable identification (Dabbene et al., 2016). Furthermore, some challenges remain, such as how to prevent label (RFID) cloning in offline authentication procedures and how to permanently link RFID labels to food products to impede the repeated application of the same identifier to many products (Dabbene et al., 2016).

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 21 Near Field Communication (NFC) is a subset of RFID and has been widely integrated into mobile devices such as smartphones, tablets and notebooks. NFC permits peer-to-peer communication between two devices. It is often used to help authenticate products, provide consumers with more information on the product and can provide a digital engagement platform. Pilot programs have commenced to also use NFC as a traceability technology for the food industry (Pigini and Conti, 2017).

4.4 INKS, MICRODOT AND OTHER TAGGANT TECHNOLOGY

A number of physical, spectroscopic, chemical and DNA based coding systems have been used as markers or forensic level taggants for property identification and as anti-counterfeiting measures to distinguish potential forgeries (Gooch et al., 2016). Sophisticated printing technology such as holograms, or the use of optically variable, UV sensitive, infrared sensitive, thermal, fluorescent functional, intaglio and reactive inks can be used to add security features to products or packaging. Many of these inks and printing technologies are used for financial documents (such as banknotes) or identity documents (such as birth certificates, passports, and visas). Overt features can be detected with visual analysis, whilst covert features generally require a device (e.g. microscope or hand-held reader), or may require some level of chemical analysis. Several manufacturers also produce a range of visible (overt) and invisible (covert) edible inks that enable information (such as harvest dates, logos and barcodes) to be stamped or printed directly onto products.

Microdots are tiny particles that are etched with unique identification codes or symbols and can be physically applied to a product. Properly designed and once applied they are virtually irremovable and indestructible. Microdot particles are so small that they are generally invisible to the naked eye providing covert protection, but are frequently used in combination with overt warning systems to provide greater deterrent against fraudulent activity. DataDot™ and NanoTags™ are two forms of microdots and can be used in a variety of application settings including protection of food, apparel and luxury items, pharmaceuticals, industrial goods, documents and automotive components. DataDots have been trialled in Australia to prevent the theft of oysters from leases but were considered too expensive for everyday use (McOrrie, S., personal communication). NanoTags are being used by the Australian Wild Abalone Program (see Section 6.1 for further details).

Several research groups are also aiming to develop encodable nanomaterials, known as nanobarcodes. Once commercialised these could be added into or mixed with objects of interest without being noticed. It has been reported that nanobarcodes would be highly durable, machine readable and very difficult to counterfeit (Wang et al., 2015). The Australian beef industry in conjunction with PwC are currently in the process of trialling the application of nano-scale silicon dioxide particles onto beef that provides a natural and edible fingerprint (Neales, 2018).

4.5 DISTRIBUTED LEDGERS

Distributed ledgers are synchronized databases that are stored simultaneously on multiple, often thousands of, decentralized computers at any given time. The distributed ledgers are visible to anyone within the network and the decentralised architecture ensures that there is no single point of failure. The nature of distributed ledgers is that no single entity can tamper with an entry. Blockchain is one of the most well- known digital ledger platforms as it is used for the Bitcoin cryptocurrency. However, a large number of others platforms such as Tangle and Hashgraph have been developed. These platforms can provide greater transparency in the supply chain which may lead to greater accountability and responsibility. They are a back-end technology that securely records information from a primary source.

Blockchain technology allows ownership or information regarding a product, such as product origin, batch numbers, factory and processing data, expiration dates and transport details, to be recorded through secure, distributed ledgers, without using a trusted third party intermediary. Blockchain requires all parties in the supply chain to utilise the one system and this can also rapidly speed up product recalls from days or weeks to seconds, as it overcomes many existing traceability systems that are fragmented and do not communicate

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 22 with one another. New blocks record transactions and in order to add these blocks to the ledger, the data needs to be validated by the majority of all the computers (or miners) within the network. Validation of most blockchain platforms is performed via cryptography and each validated block includes a timestamp and a link, or hash, to the previous block within the chain (Drescher, 2017). The information stored in the blocks is considered to be secure as it cannot be changed or deleted unless the subsequent blocks are also changed and the majority of the miners accept the modifications. This protects the security of the chain making is nearly impossible to change or alter any recorded data without leaving a trace. Current limitations with some blockchains include the unscalable nature, high transaction costs, low transaction rates and large computational energy needed for validation by the miners; however increase scalability and more energy- efficient consensus mechanisms are being developed (Staples et al., 2017).

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 23 5. AUTHENTICITY TOOLS BASED ON PRODUCTS’ INTRINSIC PROPERTIES

The intrinsic properties of a food product can be affected by a large number of factors including genetics, sex, feed sources/diet, life history or age and harvest location. Subsequently, the examination of morphological features or use of omic-based (i.e. genomics, proteomics, lipidomics and metabolomics) techniques can potentially prevent and/or detect fraudulent activities. These omic-based techniques can be used to characterise and quantify the various biological molecules in samples based on the respective classes (i.e. DNA/RNA, proteins/amino acids, lipids/fatty acids and other small, low molecular weight molecules). Metabolomics is often coupled with chemometric analysis5 to develop chemical or metabolic fingerprints or profiles of species or sub-populations within species. Other biological markers such as bacterial communities and parasite loading have also been used or suggested as a potential methodology to determine product origin (Martinsohn, 2011; Fiorino et al., 2018).

The types of technologies used to detect product adulteration and food fraud in published scientific literature was recently reviewed by (Hong et al., 2017). The most common analytical technologies that have been used for the detection of food fraud in seafood products were polymerase chain reaction (PCR), mass spectrometry (MS), high-performance liquid chromatography (HPLC), liquid chromatography (LC), gas chromatography (GC) and enzyme-linked immunosorbent assays (ELISA). Whilst visual identification and genetic testing have historically been the most prominent methods for fish species identification (Bremner and Snow, 2007), omics can provide information to support numerous other product claims and credence attributes.

Chemical or metabolic fingerprints are non-targeted and are usually based on generic analytical techniques, whereas chemical profiles are obtained from the use of a specific extraction, separation and detection processes (Rubert et al., 2015). Metabolomic fingerprinting has also been demonstrated to be able to distinguish between line or trawl catch methods (Black et al., 2017). Data obtained from non-targeted analytical techniques is generally very extensive and often not all the data points are required to distinguish between samples (Creydt and Fischer, 2018). However, a decision to remove any data points should be quantified to ensure that you are not losing valuable data that may subsequently create a bias during chemometric analysis.

5.1 VISUAL ASSESSMENT

Morphological identification (such as shape, colour, texture etc.) as well as microscopic examination of tissue structures are simple and low cost approaches for product identification. However, the accuracy is highly dependent on experiences and skills of the assessor and it is often difficult to distinguish between two closely related substitutes. Furthermore, identification becomes more difficult in processed products (such as surimi, fillets, breaded or battered products, or precooked products) because of the removal or destruction of the observable characteristics such as heads, tails, fins and skin. An example of several similar looking fillets are shown in Figure 6.

5 Chemometric analysis uses mathematical, statistical and other methods to provide maximum relevant chemical information by analysing chemical data (Héberger, 2008). Chemometric analysis is generally used when the raw data is too complicated to process visually.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 24 Albacore tuna Escolar

Red snapper Rock fish

Farmed Atlantic salmon Wild Atlantic salmon

Figure 6. Fillets from various fish. Adapted from Oceana (2012)

5.2 NUCLEIC ACID BASES TOOLS

Deoxyribonucleic acid (DNA) based techniques are generally considered the ‘gold standard’ for species identification. They are well established and have often been used to identify seafood mislabelling practices in commercial markets. Species differentiation using DNA is based on identifying different genetic mutations acquired by the product during evolution and several example studies which have used DNA sequencing techniques are reported in Table 5. DNA extraction is the first step for almost every method of DNA analysis, and requires lysing the cells to release the DNA from the nucleus or other DNA-containing organelles within the tissue. Following lysis, sample clean-up is often required to remove cellular debris and other constitutes including proteins and RNA (ribonucleic acid).

Gel electrophoresis has historically been used as technique to visualize the DNA from a sample whereby purified (non-amplified) DNA is loaded on a specific agar or polyacrylamide gel and an electrical current is applied. DNA fragments are separated based on size and the charge-to-mass ratio; smaller DNA fragments generally move faster across the gel, whilst larger DNA fragments move slower. A “DNA ladder” or solution of known fragment sizes is generally included as part of the assay as a comparison and verified or voucher samples as a control. In many cases the isolated DNA is amplified through a polymerase chain reaction (PCR) process which generates copies of the target gene region in order to improve sensitivity of subsequent detection technique. A range of DNA based techniques are available and have been used for species identification. The most common methods include forensically informative nucleotide sequencing (FINS), restriction fragment length polymorphism (RFLP), single-stranded conformational polymorphism (SSCP), random amplified polymorphic DNA (RAPD), amplified fragment length polymorphism (AFLP) and loop mediated isothermal amplification (LAMP). A number of rapid ID test kits are commercially available.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 25 Table 5. Examples of studies that have utilised DNA sequencing to detect seafood mislabelling on the commercial market. Adapted from Hellberg et al. (2016). Product Geographic Seafood species targeted Study Findings Type Location Clupeidae, Engraulidae, Canned cat Mislabelling rates of 40–100%, Salangidae, and Scombridae Italy food depending on the product category families Dry salted 79% of dry salted and 100% of and ready- Jellyfish (Class Scyphozoa) Italy ready-to-eat products were to-eat mislabelled products Dried, One canned sample and all sliced Abalone (Haliotidae family) canned Hong Kong (dried) samples were mislabelled abalone Fillets, 14% of samples had puffer fish, but Puffer fish (Lagocephalus canned Taiwan 28% of those were toxic varieties, and Takifugu) products and causing a potential threat fish powders Freshwater fishes 38% of commercial samples did not Whole fish Taiwan (Cyprinidae family) contain Cyprinidae 11% of samples labelled as Pacific Atlantic salmon (Salmo salmon were Atlantic salmon and salar) and Pacific salmon Fillets United States restaurant substitution was more (Oncorhynchus spp.) common than supermarket substitution 85% of salted fillets were substituted Salted fillets with Lotidae species and 100% of Gadidae family and battered Italy battered cod chunks were chunks substituted with Pollock (Pollachius virens) and Tusk (Brosme brosme) Sciaenidae family (Plagioscion Mislabelling rates of 77–100% Fillets Brazil squamosissimus and depending on labelling criteria Cynoscion leiarchus) 50% of Nile perch and Panga Tilapia (Oreochromis spp.), samples were substituted with Tra Nile perch (Lates niloticus), Fillets Egypt fish (Pangasianodon hypophthalmus) and Panga (Pangasius or and none of the tilapia was Pangasionodon spp.) mislabelled 16.8% of samples collected from Fillets United States Variety of commonly restaurants were mislabelled consumed species No samples in the study were Fillets Tasmania mislabelled European plaice 41% of Common sole samples and (Pleuronectes platessa) and Fillets Italy 35% of European plaice samples Common sole (Solea solea) were mislabelled

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 26 A comparison of the requirements, robustness, cost and accuracy of most of these methods are reported in Table 6. DNA barcoding using mitochondrial COI (cytochrome C oxidase subunit I) gene is frequently used in most studies for species identification as it is relatively heat-resistant and has a high level of recovery compared to nuclear DNA, or the 16S rRNA (Hellberg et al., 2016; FAO, 2018). A single marker or gene target is frequently used, however, this may not differentiate between closely related species and several markers might be necessary.

Sequence results are usually compared to published findings and a number of genetic-based catalogues such as FishTrace (https://fishtrace.jrc.ec.europa.eu/), FISH-BOL (http://www.fishbol.org/), NCBI GenBank (http://www.ncbi.nlm.nih.gov), Regulated Fish Encyclopedia (https://www.fda.gov/Food/FoodScienceResearch/RFE/default.htm), Global Genome Biodiversity Network (GGBN) (http://www.ggbn.org/ggbn_portal) and AquaGene (http://www.aquagene.org) are publically available. In addition, more genetic information would be housed in non-publically available databases (i.e. private businesses, universities and other research institutes). Most commercial fish species are well represented in these catalogues and a number of studies have utilised DNA sequencing to identify seafood mislabelling. However, it is also important to consider the accuracy of the information in such databases as the original samples are not always authenticated or the source materials may have originated from a distinct subpopulation. When using DNA based procedures the potential for intraspecific variations (variations within and among populations) of the target species should be understood. In addition, the similar morphological appearance of many seafood species can give rise to the possibility of misidentification and erroneous sequence information (Hellberg et al., 2016).

A number of recent benchmarking studies have also utilised citizen scientists to purchase seafood and provide samples for DNA analysis. In a recent Canadian study, SeaChoice together with Life Scanner recruited 300 citizens and tested 500 seafood samples for approximately AUD$10,000 (Shore, 2017). This value included the cost of the DNA kits and postage of the kits to and from the citizens, but excluded considerable staff time that when into planning, undertaking and reporting on the study (Foster, S., personal communication). Whilst the costs for DNA based technologies are reducing and more high-throughput facilities are becoming available, traditional DNA sequencing remains a time-consuming and expensive process making it impractical for routine use, especially in fast moving supply chains (Applewhite et al., 2016; Black et al., 2017). Real-time PCR can be cheaper and less time consuming but requires the use of specific primers for each target species. Suitable primers may not be commercially available for all species.

A number of Lab-on-a-chip technologies such as Agilent Technologies FishID solution utilising PCR-RFLP methodology can replace the need for gel electrophoresis. Recently portable testing devices suitable for field use such as the MinION (Oxford Nanopore Technologies) have become commercially available. Nano- Tracer has also recently been developed at a proof of concept stage for European perch (Perca fluviatilis) and is based on a simplified DNA barcoding analysis that allows for a rapid screening tool with interpretation by the naked eye (Valentini et al., 2017). NanoTracer can provide species confirmation results within three hours as opposed to several days. The estimated total commercial cost for the Nano-Tracer assay is AUD$15 (Pompa, P., personal communication).

Whilst DNA is relatively stable, the suitability of DNA based techniques can be limited when assessing some processed (i.e. canned or brined) tissues (Applewhite et al., 2016; Hellberg et al., 2016). In the case of degraded or highly processed samples approaches that investigate small fragments (or shorter base-pairs) are generally used. Tomás et al. (2017) recently used a Short-Amplicon High Resolution Melting Analysis (SA-HRMA) method to differential between processed Alaska pollock, Pacific cod and Atlantic cod.

The use of DNA techniques for population or point-of-origin identification has received less attention to date as it requires more background genetic data, more advanced statistical analysis and a broader insight into the evolution and biology of the target species (Nielsen, 2016). DNA-based analysis are often unsuitable to identify origin especially from wild fisheries, or authenticate provenance when there is a low genetic diversity amongst the population units, or that it came from a legal fishery, or that it was harvested using sustainable or ethical methods. Any lack of differentiation could be an inherent feature of the population itself or an artefact through the choice and number of markers or gene targets used.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 27

High

Cost

Medium Medium Medium Medium

Medium to high Medium

medium medium medium

Low

.

- - -

Medium Medium

Low Low Low

Potential for for Potential interspecies

variation errors variation

high high high high

- - - -

High High

database

Medium Medium Medium Medium

Potential for for Potential

construction

Rasmussen and Morrissey (2008) andRasmussen Morrissey

from

high

-

medium

-

High High High

Medium

Adapted

Low

Medium

Potential for for Potential

reproducibility

interlaboratory interlaboratory

high high high high

- - - -

medium medium

- -

Low Low

Medium Medium Medium Medium

Robustness to Robustness

DNA degradation DNA

based methods for identification of fish.of for based methods identification

-

Single Single Single Single

Multiple Multiple

analysed

Quantity of loci of loci Quantity

No No

Yes Yes Yes Yes

information

Requires prior prior Requires

DNA sequence DNA

specific specific

-

stranded stranded

-

PCR

(FINS)

(RFLP) (AFLP)

(RAPD)

Method

Comparisoncommonly various used between DNA

.

6

conformational conformational

Single

Species

polymorphic DNA DNA polymorphic

Random amplified Random

DNA sequencing + DNA

Amplified fragment fragment Amplified

Restriction fragment fragment Restriction

length polymorphism polymorphism length polymorphism length

primers and multiplex and multiplex primers

polymorphism (SSCP) polymorphism

phylogenetic mapping mapping phylogenetic

Table

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 28 The identification and selection of the most suitable target DNA fragments and approaches are critical. Common DNA based approaches for geographical origin include cleaved amplified polymorphisms (CAPS), microsatellites, species specific PCR amplification, single nucleotide polymorphisms (SNPs) and DNA sequencing. The strengths and weaknesses of these methods are reported in Table 7. CAPS is one of the most frequently used molecular approaches for point-of-origin analysis (Wilkes et al., 2016). In mixed populations the signature of an individual may not reveal the geographical origin, while a larger sample consisting of many individuals may expose the origin (Nielsen, 2016).

Future developments in advanced genetic techniques such as mini-barcoding (shorter regions of standardised gene targets) and next generation sequencing (useful when multiple species may be simultaneously present) will enable DNA sequencing to be applied to heavily processed, mixed species (Hellberg et al., 2016). In addition, microbial communities (metagenomics) present on food surfaces are related to the local environmental conditions and in some cases can be used to determine geographical origin (El Shieikha and Montet, 2016). Advanced genetic techniques have also been proposed to identify the origin or provenance, though these methods are reported to need further development prior to use in routine official food control programmes (FAO, 2018).

DNA sequence data can be readily replicated and interpreted across laboratories, which is important for forensic applications (Martinsohn, 2011). The success of DNA barcoding has been progressively recognised by government authorities (Valentini et al., 2017) and a step-by-step procedure for generating DNA barcodes for species identification of unknown fish samples has been made available by the US FDA6.

Table 7 Strength and potential weaknesses of DNA methods frequently used to assess geographical or point-of-origin. Adapted from (Wilkes et al., 2016). Method Strengths Potential weaknesses Cleaved Amplified  Simple technique  Moderate throughput Polymorphic  High specificity  Erroneous results from partial Sequences (CAPS)  High reproducibility digestion Microsatellites  High specificity  Large consumable requirement  High reproducibility  Moderate throughput  Highly informative  Limited targets  Technically challenging Species-specific  High reproducibility  Limited target availability PCR primers  Relatively simple technique  Technically challenging Single Nucleotide  Highly informative  Moderate throughput Polymorphisms  Adaptable method  Relatively expensive (SNPs)  Technically challenging DNA sequencing  Highly informative  Requires careful primer design (mitochondrial or  Reproducible  Moderate throughput nuclear genes) DNA sequencing  Highly informative  Technically challenging (metagenomics)  Adaptable  Current high cost  Reproducible  Resource intensive

6 Method available from https://www.fda.gov/Food/FoodScienceResearch/DNASeafoodIdentification/ucm237391.htm

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 29 5.3 PROTEIN/PROTEOMICS

Genetic differences between species can impact on the protein composition. Protein extraction is a key step within proteomic analysis and the recovered proteins can be separated with a variety of methods which include isoelectric focusing (IEF), capillary electrophoresis (CE), high-performance liquid chromatography (HPLC) and immunoassay (i.e. enzyme-linked immunosorbent assay (ELISA)) systems. Proteomic approaches have the potential to be used for a range of food identification or authenticity issues as highlighted in Figure 7.

Two-dimensional electrophoresis

Species authentication Food Authenticity

Tandem mass spectrometry (MS/MS) Figure 7. Proteomic approaches (outer circle) that may assist with food authenticity issues (inner circle). Adapted from Ortea et al. (2016).

IEF is a traditional gel electrophoresis based technique that can be used on fresh, refrigerated and frozen products and remains a valid official method for fish species identification by the Association of Official Analytical Chemistry7. Separated proteins are stained and comparisons made against authentic reference samples. Other detection systems such as mass-spectrometric analysis can be employed. Electrophoresis based assays can be performed 1-dimensionally or 2-dimensionally, thereby increasing the power to resolve and identify proteins (Martinsohn, 2011). Changes in single amino acids of a protein can impact the migration behaviour in electrophoresis assays or other chromatography systems. Whilst IEF can be a reliable technique, it is tedious, laborious (and therefore costly) and the interpretation of results extremely subjective (Applewhite et al., 2012). IEF require extensive databases or reference collections and is generally not suitable for mixed, cooked or marinated products. In addition, the same fish species from different geographical location can exhibit different protein profiles (Applewhite et al., 2012).

Proteomic analysis methods have also been developed to separate isozymes8 for fillet identification and have been used in courts of law (Ward, 2000). Common protocols used in the preparation of samples for proteomic analysis involve the depletion of the most abundant interfering proteins and enrichment or purification of target proteins (Ortea et al., 2016). However, proteins begin to lose their biological activity after animal death. Furthermore, many protein analysis techniques cannot be used when protein

7 AOAC International (2012). Official Method 980.16. Identification of Fish Species – Thin-layer Polyacrylamine Gel Isoelectric Focusing Method. In: Official Methods of Analysis of AOAC International. Gaithersburg, MD, USA. 8 Isozymes are different enzymes that catalyse the same reaction and are a product of gene expression.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 30 denaturation has occurred (i.e. cooked or heat treated products), and generally have low resolution when comparing between closely-related species (Sotelo and Pérez-Martín, 2002). The type of proteins present in a sample also depends on the cell type. Whilst protein analysis techniques have been widely used by control authorities for fish species identification, the US FDA is phasing them out over preference towards DNA-based systems (Martinsohn, 2011). The “Regulatory Fish Encyclopedia” developed by the US FDA contains the protein profiles of numerous species that are of importance in the United States of America. An example of an IEF protein profile from Atlantic salmon (Salmo salar) is shown in Figure 8.

Figure 8. An isoelectric focusing protein pattern from Atlantic salmon (Salmo salar). Reproduced with permission from the Regulated Fish Encyclopaedia, U.S. Food and Drug Administration (2016).

Proteomic based approaches continue to be used and/or developed to assess various seafood authenticity issues. Many of these techniques have been spurred on by the identification of species-specific peptides or biomarker patterns from sarcoplasmic proteins and improvements in analytical instruments and bioinformatics tools (Mazzeo and Siciliano, 2016). A list of recent publications utilising various proteomic based techniques for seafood authenticity purposes are reported in Table 8. These techniques have generally been used to differentiate between species, origin and/or fresh/thawed formats. Recently, matrix- assisted laser desorption/ionisation time-of-flight (MALDI-TOF) mass spectrometry was used to generate a database containing the protein fingerprint of 54 fish species (Stahl and Schrӧder, 2017).

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 31 Table 8. Proteomic approaches to assess authenticity of seafood. Year Purpose of analysis Main technique Reference published 2017 Differentiation of 54 fish species Matrix-assisted laser Stahl and Schrӧder desorption/ionization (2017) time-of-flight (MALDI– TOF) 2014 Discrimination of two scallop 2-dimensional Artigaud et al. (2014) populations electrophoresis 2013 Differentiation of 22 fish species tandem mass Wulff et al. (2013) spectrometry 2013 Differentiation of six shrimp MALDI-TOF MS protein Salla and Murray species, geographic origin, and profiling library (2013) fresh/frozen state 2013 Discrimination of two river fish 2-dimensional Barik et al. (2013) species electrophoresis, peptide mass fingerprinting and tandem mass spectrometry (MS/MS) 2012 Differentiation of four tuna species 2-dimensional Pepe et al. (2012) electrophoresis 2012 Identification of Northern shrimp, Matrix-assisted laser Pascoal et al. (2012) Pandalus borealis, biomarkers desorption/ionization time-of-flight (MALDI– TOF) and electrospray ionisation–ion trap mass spectrometry 2011 Differentiation of all hake species Selected tandem mass Carrera et al. (2011) spectrometry (MS/MS) ion monitoring 2011 Differentiation of seven shrimp Selected tandem mass Ortea et al. (2011) species spectrometry (MS/MS) ion monitoring 2010 Differentiation of 14 shrimp Isoelectric focusing Ortea et al. (2010) species 2009 Differentiation of seven shrimp Tandem mass Ortea et al. (2009b) species spectrometry (MS/MS) with data-dependent analysis and peptide fragmentation fingerprinting 2009 Differentiation of six shrimp 2-dimensional Ortea et al. (2009a) species electrophoresis and peptide mass fingerprinting

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 32 5.4 FATTY ACIDS/LIPIDOMICS

The fatty acid composition of all living organism are determined by genetic and environmental conditions. Most seafood products generally contain the same types of fatty acids; however, seafood of a given species or from a given location, are likely to have a characteristic fatty acid profile. Whilst lipidomics is relatively new for detection of fraud (Ellis et al., 2016), fatty acid profiles have been used to distinguish wild and farmed fish and for establishing seafood origin (Martinsohn, 2011; SeaFish, 2011; Lavilla et al., 2013). The European Commission’s FishPopTrace Project demonstrated that the geographical populations of Common sole (Solea solea), European hake (Merluccius merluccius), Atlantic cod (Gadus morhua) and Atlantic herring (Clupea harengus) could be differentiated from between and within the Atlantic and Mediterranean basins according to the fatty acid composition (FishPopTrace, 2011). Fatty acids composition ranges are also used to support a variety of named fish oils within the Codex Standard for Fish Oils (CAC, 2017).

5.5 MINERAL/ELEMENTAL ANALYSIS

Multi-element and stable isotope analysis, are two of the most widely used methods in traceability of aquaculture products (Li et al., 2016). The mineral content of foods can depend on the local environmental conditions and have been successfully employed for product authentication (Gonzálvev and de la Guardia, 2013). Mineral profiles in seafood can also be affected by species, sex, size, age/sexual maturity, life history and diet (Li et al., 2016). The mineral profiles of foods are generally determined through single-elemental techniques including flame atomic absorption spectroscopy (FAAS) and electrothermal atomic absorption spectrometry (ETAAS) as well as multi-elemental techniques including inductively coupled plasma optical emission spectrometry (ICP-OES) and inductively coupled plasma mass spectrometry (ICP-MS). Multi- element techniques including x-ray fluorescence (XRF) spectrometry and neutron activation analysis (NAA) have also been used. In general multi-element methods coupled with appropriate chemometric models are more commonly employed to detect fraud. Many elemental analysis processes including ICP-OES and ICP- MS, require an acid digestion or dissolution step to solubilise the components. However, XRF and NAA methods require little or no sample preparation and can measure the total concentration of a large number of elements in the sample. XRF is based on the fluorescence profile detected when a sample is exposed to x-rays. The x-rays displace electrons from the inner orbital shells of an atom, which in-turn allow higher orbiting electrons to fill the vacancies and in doing so releases energy. Whereas, NAA can identify and quantify elements present in a sample based on the detection and analysis of gamma rays emitted during radioactive decay after the sample is irradiated in a nuclear reactor.

5.5.1 Native Elements

The elemental fingerprint from otoliths, scales, spines or shells can be used as a permanent signature that is influenced by the geographical (spatial) origin (Martinsohn, 2011; Norrie et al., 2016); however, various organs or muscle tissue can also be used. The elemental fingerprint can also be influenced by diet. Otoliths can be analysed as a whole, or specific zones from within otoliths may be analysed for a specific life stage. Challenges can arise due to significant variability of the elemental composition within otoliths (Di Franco et al., 2014). The investigation of native elements has been used to examine population structures and stock tracing in a number of cases, including those reported in Table 9. Elemental fingerprinting or profiling can be less effective within aquaculture operations due to the globalisation of feed ingredients (Wang et al., 2018).

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 33 Table 9 Example applications where trace metals have been used in seafood. Species Tissue Purpose Reference

Common sole (Solea Otoliths Confirmed separation of De Pontual et al. (2000) solea) individuals from Common sole stocks originating from two estuaries 200km apart

Atlantic salmon (Salmo Scales Distinguish between wild Adey et al. (2009) salar) and farmed Atlantic salmon

Barramundi (Lates Muscle Origin of harvest D'Costa (2015) calcarifer)

Pink ling (Genypterus Muscle Origin of harvest Bremner (2009) blacodes)

Prawns (mixed Muscle and shell Origin of harvest Carter et al. (2015) species)

Common cockle Shell Geographical origin at fine Ricardo et al. (2015) (Cerastoderma edule) spatial resolution

Goose barnacles Capitular plates Geographical origin at fine Albuquerque et al. (Pollicipes pollicipes) spatial resolution (2016)

The techniques used to recover and analyse the elements present in a sample impact on the accuracy, sensitivity as well as the actual elements that can be detected. Studies or investigations should understand how the trace elements interact with each other and also the underlying geology and oceanography. Albuquerque et al. (2016) recently investigated the trace element profile of goose barnacles (Pollicipes pollicipes) from ten sites along a 300 km stretch of Portuguese coast. The study involved 100 samples and whilst the overall successful classification was 83%, 97% of individuals from an important region were successfully classified and 2.7% of specimens from outside the region were erroneously assigned to this region (Albuquerque et al., 2016). Ricardo et al. (2015) demonstrated that the geographical origin of bivalves can be distinguished at locations separated less than 1 km apart within the same estuarine system. However, the authors noted that follow-up studies are needed to determine if this level of discrimination can occur in other ecosystems.

Australian Pork Limited have invested into the development of Physi-Trace which can trace back the origin of pork products to approximately 85% of all pig producers in Australia (Lee and Watling, 2017). Since its inception in 2008 over 16,000 samples have been analysed (Lee, G., personal communication) and revised sampling and analytical plans9 have been implemented to deliver Physi-Trace to industry at a cost of AUD$0.05/pig (APL, 2015). The key benefits of Physi-Trace have included increased market confidence in the integrity and traceability of Australian pork, rapid market re-entry in the event of a food safety incident and verification of country of origin and production label claims. In 2017, it was reported that the benefit-cost ratio of the Physi-Trace R&D program used within the Australian pork industry was 15.8 (APL, 2017). The review also reported that the present day value of all costs associated with the Physi-Trace program was estimated at AUD$16.6 million, whereas the present day value of all benefits from the program was AUD $263.4 million (APL, 2017).

9 Plans entail sampling and storage of between 0.5-1.5 per cent of annual slaughter by each establishment, and five percent of these samples randomly selected on a monthly basis for analysis and including in database.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 34 5.5.2 Stable Isotopes

Isotopes are atoms with an identical number of protons, but a different number of neutrons, and therefore different atomic weights. Nearly every element has a number of naturally occurring isotopes and the abundance of each isotope in a product is influenced through physical, biochemical and chemical processes. Isotope ratios are the most widely used method for product authenticity and traceability testing and 2H/1H, 13C/12C, 18O/16O, 15N/14N, 34S/32S and 87Sr/86Sr have been used for the verification of the geographical origin of a number of commodities (Camin et al., 2017; Carrera and Gallardo, 2017). Bulk or compound-specific isotope analysis can be used and isotope reference materials that serve as a measurement standard are commercially available. A number of standard methods for wine, fruit juice, cheese, olive oil and vinegar have also been established (Camin et al., 2017).

Whilst several studies have investigated the effect of changing diets on the stable isotope ratios in seafood, principally to differentiate between wild and farmed species, few studies have been published that are specifically concerned with the geographic origin. Rooker et al. (2008) used carbon and oxygen stable isotope ratios (δ13C and δ18O) in otoliths to identify Trans-Atlantic movement and connectivity between Atlantic bluefin tuna populations. Rossier et al. (2014) used oxygen stable isotope ratios (δ18O) to detect mislabelling of fish origin in Switzerland. Kim et al. (2015) reported that carbon and nitrogen stable isotope ratios (δ13C and δ15N) could be used as a potential tool to distinguish between country of origin for three commercial fish, Chub mackerel (Scomber japonicus), Yellow croaker (Larimichthys polyactis and Larimichthys crocea) and Alaskan pollock (Theragra chalcogramma). The authors did not report the number of samples but acknowledged that more detailed sampling would be needed to confirm findings and should consider climate, seasonal, age and diet-related studies. The geographical origin of all commercial hake species, as shown in Figure 9, can also be differentiated using stable isotope ratios. Stable isotope ratios have also been used in a 2013 court case over Bulega caviar imported to Germany from the Caspian region (Camin et al., 2017).

Figure 9. Differentiation of all commercial hake species belonging to the Merucciidae family using carbon (δ13C) and nitrogen (δ15N) stable isotope ratios. Reproduced with permission from Carrera and Gallardo (2017).

Isotope mapping is generally only useful to identify different populations over large geographical origins (Lee, G., personal communication). As with many study findings, caution should be used when interpreting results especially when a relatively low number of samples are used. For example, in the study by Ortea and

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 35 Gallardo (2015), that investigated wild and farmed shrimp from different geographical localisations, increasing the total number of samples from 31 to 45 resulted in a reduction in the classification performance. It is also essential that sufficient data is available covering all the variables known to affect the particular marker. Other reviews have concluded that isotope ratios should be used in conjunction with other non- isotopic methods (SeaFish, 2015). As an example, Thomas et al. (2008) used nitrogen and oxygen stable isotope ratios (δ15N and δ18O) from lipid compounds in conjunction with the fatty acid composition to authenticate wild and farmed Atlantic salmon (Salmo salar).

5.5.3 Inclusion of Rare Earth Elements

Rare earth elements are a group of metals that despite their name are generally abundant in the environment, but are only present in very low concentrations. Whilst rare-earths are used in a number of commercial applications, the toxicological implications of these non-essential elements is poorly understood (Rim et al., 2013; Pagano et al., 2015). Several studies have investigated the inclusion of rare earth elements in formulated diets as potential markers. For example de Nanclares et al. (2016) demonstrated that a number of rare earth elements could be incorporated into the scales of Atlantic salmon (Salmo salar) during a 10- week feeding trial in a closed containment system. The rare earth elements remained detectable for at least two months after the trial. Other studies has shown that the rare earths can be detected for more than a year after treatment (Michibata, 1981; Kato, 1985), whilst another study reported that the rare earths were undetected 30 days after treatment (Zak, 1984).

5.6 OTHER SYSTEMS 5.6.1 Spectroscopic Methods

Numerous spectroscopic methods including near-infrared (NIR), mid-infrared (MIR), Fourier transform infrared (FT-IR), Raman, nuclear magnetic resonance (NMR) spectroscopy and hyperspectral imaging have been used for food authentication purposes. These techniques have some benefits over other intrinsic- based methods; this includes that they are rapid, often non-destructive, generally require no or little sample preparation and have a relatively low unit analysis cost (Esteki et al., 2018). These techniques can provide qualitative and quantitative data to help determine product quality or identify product adulteration or contamination (Cheng et al., 2013; Lohumi et al., 2015; Cozzolino, 2016). The principles, type of information obtained and the major attributes of the different spectroscopic techniques are reported in Table 10. NIR, MIR, Raman and hyperspectral images can provide complex chemical information from a sample being analysed and have been used for a variety of purposes in the agro-food, pharmaceutical and petrochemical sectors (Lohumi et al., 2015).

Infrared spectroscopy (IR) is based on the interactions between electromagnetic radiation absorbed by molecular bonds with a dipole moment of the sample. Energy absorbed in the infrared region is due to fundamental vibrations of atoms in the molecules and will generate an absorption or reflectance spectrum. NIR signals are associated with the superposition of composite, overtone and high frequency vibrations in the 780-250 nm region, whereas MIR vibrations are in the 2500-250 nm region. Fourier-transform infrared (FT-IR) spectroscopy has been used extensively in research laboratories for food and food ingredient authentication (Ellis et al., 2015). FT-IR is generally more sensitive and can detect low-level compounds within complex food matrices and subtle differences between samples.

Hyperspectral imaging incorporates spectroscopic technologies and imaging into one system and can attain both spectral and spatial (within a product) information from a sample over the ultraviolet, visible and infrared electromagnetic radiation regions. Hyperspectral systems have been widely studied for quality and safety evaluation in seafood and other muscle-based food products. Typical applications include the prediction of colour, textural properties, composition, spoilage bacteria, freshness, and fresh/thawed classification (Cheng and Sun, 2015; Kamruzzaman and Sun, 2016).

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 36 Table 10. Characteristics and major attributes of spectroscopic techniques when used in food analysis. Characteristics and Infrared Hyperspectral Raman NMR attributes Type of information Organic bond Direct Organic bonds Atomic and obtained types and identification of and physical molecular physical various structures – structures and structures – components – derived from mobility – derived derived from derived from molecular bond from change of molecular bond molecular bond vibrations when nuclear spin vibrations when vibrations when sample probed when sample sample probed sample probed over the infrared probed with over the infrared over the visible region of the radiowaves from region of the and infrared electromagnetic the electromagnetic regions of the region electromagnetic spectrum electromagnetic region spectrum Spatial information (variation within -  -  product readily determined) Spectral information (product specific    - signatures) Multicomponent information from     single analysis Suitable for use at-, on- and in-line   -  process

Raman spectroscopy is based on the inelastic scattering of light due to the interactions between the photon energy and electron cloud of the bonds associated with the individual molecules or functional groups within a sample. Raman spectroscopy is a widely adopted analytical method in the pharmaceutical industry. It has a high chemical selectivity and techniques can encompass Fourier transform Raman spectroscopy, dispersive Raman spectroscopy, spatially offset Raman spectroscopy (SORS), surface-enhanced Raman spectroscopy (SERS) and transmission Raman spectroscopy. Raman spectroscopy has several advantages over infrared spectroscopy methods. The principle advantages of Raman is a weaker interference from water which is a major component in many food products, and sharper absorption bands that result in less spectral overlapping enabling more product characteristics to be obtained in complex matrixes. In addition Raman spectroscopy can penetrate most plastic, glass and other transparent packaging materials (Ellis et al., 2015). SERS provides highly localised chemical information regarding a sample and operates at very low detection limits, whereas transmission Raman Spectroscopy can analyses the bulk composition of samples. The Raman effect is however generally weaker and the equipment more expensive that IR spectroscopy equipment (Ellis et al., 2015).

Nuclear magnetic resonance (NMR) offers high potential for the analysis of multicomponent systems, such as food matrices (Dais and Spyros, 2012). NMR is often used in the study of fats and oils. NMR is based on the difference of the nuclei spin of elemental isotopes, with 1H and 13C the most important for seafood and other organic materials. A range of NMR techniques are available. The most common are high-resolution liquid-state NMR spectroscopy, high-resolution solid-state NMR spectroscopy and Site-Specific Natural Isotope Fractionation by NMR (SNIF-NMR). NMR has been used to determine the fat content and fatty acid

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 37 profile in seafood products and provide information for other quality attributes including pH, water-holding capacity and collagen content (Xu et al., 2015). Through lipid profiling NMR has been used to distinguish the geographical origin and production method (i.e. wild versus farmed) of Atlantic salmon (Salmo salar) and Gilthead sea bream (Sparus aurato) (Aursand et al., 2009; Coco et al., 2009; Savorani et al., 2010).

The analysis of the vibrational spectra is however considered a secondary analytical technique as it doesn’t analyse the chemicals directly. NIR based technology generally lacks the required sensitivity and selectivity when compared to MIR and Raman spectroscopy and the water content in samples can make it difficult to observe minor components (Garrigues and de la Guardia, 2013; Cozzolino, 2016). Examples of published studies utilising vibrational spectroscopic methods for seafood product substitution, distinguishing geographic origin or adulteration include:  Geographical origin: Tilapia fillets were assessed using NIR from four distinctive geographical regions in China. Correct geographical classification was between 75-85% depending on source, with 1-2% of the samples incorrectly classified and the remainder not assigned (Liu et al., 2015).  Product adulteration: Substitution of surimi-based imitation crab meat to Atlantic blue (Callinectes sapidus) and Blue swimmer crab (Portunus pelagicus) meat samples based on visible and near- infrared spectroscopy (Gayo et al., 2006).  Product adulteration: Substitution of Atlantic salmon (Salmo salar) with Rainbow trout (Oncorhynchus mykiss) in seafood patties based on Fourier transform infrared (FT-IR) spectroscopy coupled with chemometric methods (Sousa et al., 2018).  Substitution of fresh with thawed fish: Differentiation between fresh and thawed Atlantic mullet (Pseudupeneus prayensis) using FT-NIR and FT-IR spectroscopy (Alamprese and Casiraghi, 2015).  Substitution of fresh with thawed fish: Combination of NIR and colour analysis was demonstrated as a potential in-line/at-line inspection of fresh and thawed West African goatfish fillets (Pseudupeneus prayensis) (Ottavian et al., 2014).  Substitution of species: FT-NIR and FT-IR spectroscopic techniques could successfully distinguish between cheaper Atlantic mullet fillets from those of the more valuable red mullet, and good results were obtained for flounder and plaice fillets (Alamprese and Casiraghi, 2015).  Wild and farmed: Differentiation between wild and farmed European sea bass (Dicentrachus labrax) using NIR (Ottavian et al., 2012).  Product substitution: Discriminated between seafood reared in fresh or brackish water and salt water through the use of Raman spectroscopy (Rašković et al., 2016).

Alamprese and Casiraghi (2015) recommended that vibrational spectroscopy methods could be used as a pre-screening technique to classify against substitution and in the case of suspected fraud, more sophisticated analyses could be carried out in order to legally assess the fraudulent claims. Aside from benchtop systems, a variety of handheld or portable Raman, IR and NMR devices are also commercially available.

5.6.2 High Resolution Mass Spectrometry

High resolution mass spectrometry is frequently used for the detection of metabolites10 and can be operated in a stand-alone mode, or is often coupled with a separation technique such as gas chromatography, liquid chromatography, or capillary electrophoresis (Rubert et al., 2015). All mass spectrometers contain an ionising module that ionises molecules as they pass through and once formed the ‘ions’ are separated by a

10 A metabolite is a low molecular weight organic compound, typically involved in a biological process as a substrate or product.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 38 magnetic field and detected based on their mass-to-charge ratio (m/z). Analytes may be ionised thermally, by electric fields, or by impacting energetic electrons, ions or photons. The ‘ions’ can be single ionised atoms, clusters, molecules or their fragments or associates (Gross, 2017).

Ambient mass spectrometry is relatively new in the field of analytical chemistry and has the potential to overcome the long and complex sample preparation and assay times typically required for most mass spectrometry systems (Black et al., 2016). Since being first identified in 1998, there are now over thirty different ambient ionisation techniques (Black et al., 2016). These techniques are outlined in Table 11.

Table 11 Ambient ionisation techniques grouped based upon their ionisation mechanism. Reproduced from (Black et al., 2016)

Electric discharge ambient Ambient gas-, heat- or laser Spray or jet ionisation ionisation assisted desorption/ionisation Atmospheric pressure glow Desorption electrospray Atmospheric solid analysis discharge desorption ionisation ionisation (DESI) probe (ASAP) (APGDDI) Desorption atmospheric Desorption atmospheric Electrospray-assisted laser pressure photo ionisation pressure chemical ionisation desorption ionisation (ELDI) (DAPPI) (DAPCI) Desorption electrospray/ Direct analysis in real time Extractive electrospray metastable induced ionisation (DART) ionisation (EESI) (DEMI)A

Desorption sonic-spray Dielectric barrier discharge High–voltage-assisted laser ionisation (DESSI) ionisation (DBDI) desorption ionisation (HALDI) Infrared laser ablation Easy ambient sonic-spray Desorption corona beam metastable-induced chemical ionisation (EASI) ionisation (DCBI) ionisation (IR-LAMICI)A Helium atmospheric pressure Electrode-assisted desorption Laser ablation electrospray glow discharge ionisation electrospray ionisation (EADSI) ionisation (LAESI) (HAPGDI) Liquid sampling atmospheric Electrostatic spray ionisation Laser desorption spray post- pressure glow discharge (ESTASI) ionisation (LDSPI) (LS-APGD) Jet desorption electrospray Low temperature plasma (LTP) Laser spray ionisation (LSI) ionisation (JeDI) Matrix assisted laser desorption Liquid extraction surface Microwave induced plasma electrospray ionisation analysis (LESA) desorption ionisation (MIPDI) (MALDESI) Plasma assisted desorption Rapid evaporative ionisation Paper spray (PS) ionisation (PADI) mass spectrometry (REIMS) Transmission mode desorption electrospray ionisation (TM-DESI) A Both DEMI and IR-LAMICI ionisation mechanisms have traits similar to that of an electric discharge ambient ionisation mechanism

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 39 Rapid evaporative ionisation mass spectrometry (REIMS) is one of the newest forms of ambient mass spectrometry and can characterise biological tissues without the need for any, or very little, sample pre- treatment. REIMS uses a laser, bipolar forceps or an electrosurgical knife to cut into the sample and the resultant ‘smoke’ or aerosol is captured and analysed through a mass spectrometer. Black et al. (2017) recently used REIMS to differentiate five white fish species (see Figure 10a) and also demonstrated that REIMS could distinguish between line and trawl catch methods within a species (see Figure 10b). REIMS has also been used to differential the species and origins of five different prawn and shrimp species, see Figure 11.

a) b) Figure 10. Models outputs developed following REIMS analysis that can differentiate between a) cod (blue), coley (red), haddock (green), pollock (blue) and whiting (black) and; b) haddock caught by trawl (red) and line (blue) methods. Reproduced with permission from Black et al. (2017).

Figure 11. Differentiation of prawn/shrimp species and geographical origin using metabolomic profiling. Reproduced from Elliott (2018).

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 40 Curie-point pyrolysis is a method whereby organic material is heated and thermochemically decomposed in the absence of oxygen to produce small molecules that are then generally separated and detected using mass spectrometry. Curie-point pyrolysis has been used in a number of applications, including the identification of fish species at the larval stage (Valcarce et al., 1991) and discriminating between spoilage bacteria in seafood (Dalgaard et al., 1997). Curie-point pyrolysis has also been used to distinguish between the country of origin of wine and fruit juice (Montanarella et al., 1995; Garcia-Wass et al., 2000).

5.6.3 Organoleptic and Sensor Based Technologies

Organoleptic based technologies are typically used to determine the freshness and other quality features of a product based on characteristic attributes (i.e. flavour, appearance, texture etc.) or volatile compounds. The use of experts or trained panels can make sensory evaluation an objective measure. Various analytical systems, such as those based on headspace sampling, solid phase microextraction and steam distillation, have also been used to determine the volatile organic components within a sample (Kim and Cadwallader, 2011), however all these systems require a separation step. A number of electronic noses (e-noses) and electronic tongues (e-tongues) have been developed to detect the aroma and flavour of samples. E-nose and e-tongue systems do not require a separation stage, and consist of an array of sensors that are either in contact with the headspace or immersed in the sample. E-noses and e-tongues have been used to assess freshness in a number of seafood species (Śliwińska et al., 2016) and detect product adulteration and product authenticity in other food commodities including honeys, oils, beverages and spices (Peris and Escuder-Gilabert, 2016). To date their use to detect species substitution and adulteration in seafood has not been reported.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 41 6. INDUSTRY CASE STUDIES

6.1 NANOTAGS FOR AUSTRALIAN WILD ABALONE

Input kindly provided from Dean Lisson (Abalone Council Australia Ltd), Helen Barnard (Honey and Fox, ex Tasmanian Abalone Council), Spiro Markantonakis (Dover Ex27 Pty Ltd) and Darvin Hansen (Tasmanian Seafoods).

In 2010 the Australian wild abalone industry initiated a market and research development program called the Australian Wild Abalone® (AWA). The program remains on-going and is working to assure AWA customers that they are purchasing a premium quality product that has been hand-harvested by experienced divers from the cool clean waters of southern Australia, and processed and packaged according to stringent Quality Assurance guidelines. The AWA program essentially consisted of five ‘overlapping’ projects with objectives to research and measure the effectiveness of different market development strategies and activities for Australian wild caught abalone in increasingly competitive Asian markets. NanoTag® labelling was a component of this program to provide assurances that any product marketed and sold with the AWA logo was genuine.

What initiated the uptake of NanoTags?

The need to underpin product integrity has been around for several decades. Two studies (one in 199811 and the other in 200512) concluded that the Australian wild abalone industry should develop a strategically focused market development platform including a quality assurance and integrity program underpinned by an industry endorsement brand or mark. Australian abalone are processed, packaged and exported in a variety of formats including live, frozen and canned. Trust in provenance is important and is a valuable product characteristic in key abalone export markets.

NanoTag Technology Pty Ltd initially approached an Australian abalone processor and exporter to offer their technology, services and capabilities to help authenticate abalone products in the marketplace. The authentication potential of NanoTags was subsequently raised at the abalone export association level and received a high level of interest. NanoTag Technology Pty Ltd outlined their capabilities and services and how it differed from alternative Microdot technology. Key advantages of NanoTags were that they require a high level of technology to manufacture, making them difficult to counterfeit, and are highly resistant to heat and chemicals so they cannot fade, burn, melt or degrade. Representatives from NanoTag Technology Pty Ltd also visited abalone export companies and explained the technology, how it works and how it could be integrated into their processing lines.

The combination of a commercial/association desire to authenticate Australian wild caught abalone products in order to support product differentiation and market development activities and opportunity for a technical solution aligned. Due to the range in product formats something highly versatile was needed. No other system was really considered.

What was the process?

As part of the AWA® Program, Abalone Council Australia (ACA), along with the Australian Seafood Cooperative Research Centre (SCRC), and the Australian Fisheries Research and Development Corporation (FRDC) supported a proof of concept authenticity project that worked with NanoTag Technology

11 Macarthur Agribusiness (1998) Wild Abalone Fisheries R&D Needs Review. Report prepared for FRDC Project 1998/170. 12 McKinna, D. (2005) Development of a Marketing and Market Development Strategy for the Australian Abalone Industry. Report prepared for Abalone Council Australia Limited. 209pp.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 42 Pty Ltd to develop a customised portfolio of solutions that guarantees product provenance. The NanoTags developed and used by the Australian abalone industry contain the AWA Certification Logo and could be further customised to contain the Export Establishment Number (EEN) of each processor.

What trials have been undertaken?

Approximately half of the approved AWA export companies’ trialled the NanoTag technology and approximately 50% of those trialled the technology to a significant extent. The proof of concept trials allowed the AWA export companies to assess the practicality of using NanoTag’s Brand Protection system. Tagging ‘wands’ that contained NanoTags suspended in an adhesive was used to covertly label products with the customised NanoTags, whereas, high-security stickers, labels, seals and packaging tape (each of which contained imbedded NanoTags for increased security) were used for overt protection. The trials identified that the high-security labels and packaging tape were the most cost-effective and appropriate for the products and variety of factory processes. However, the need to constantly disperse the NanoTags in the wand applicator and time taken for the adhesive to dry were excessive and an automated application system to protect brite (unmarked) cans was considered cost-prohibitive.

What has been the success?

As part of the implementation phase, the objectives of the AWA Program was explained to over 40 companies in China and Hong Kong during which they were introduced to the NanoTag technology. The companies could see the application and were unilaterally supportive of the marketing activities with no negative feedback received. A number of stakeholders in Singapore and Japan were also engaged and were supportive of the AWA Program provided that it was appropriately customised to reflect the relevant market situation within their countries.

Whilst the proof of concept phase has now been completed, no real attempt has been made to quantify success and a benefit-cost assessment has not been undertaken. Success from the proof of concept phase could be judged by the fact that abalone exporters initially embraced the technology and many are continuing to use the technology. One exporter temporarily stopped supplying AWA branded product but then experienced overseas customers requesting it. The AWA Code of Conduct stipulates that all AWA product shipments must be protected by the NanoTag Brand Protection System. Each AWA exporter can now buy the NanoTag Brand Protection System directly from NanoTag Technologies.

What are the next steps?

The ACA is currently in the process of trying to establish a levy to continue to advance its marketing strategy. If successful, other authentication technologies may be considered, though no viable alternatives have been explored.

Can NanoTags be applied to other sectors?

Absolutely, with increasing concern regarding food safety, anything that can guarantee product provenance can only be a benefit. An important consideration for other sections is to determine the objective of using any technology or tool. In this case the objective was to provide importers and customers assurances, or ability to confirm, that products marketed and sold with the AWA labelled as Australian wild caught abalone is genuine. Other important considerations are to know who the end-user is, and how they will interact with the technology.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 43 6.2 MINERAL/ELEMENTAL ANALYSIS IN PRAWNS

Input kindly provided by Rachel King (Australian Council of Prawn Fisheries) and Janet Howieson (Curtin University of Technology).

A large range of analytical tools including a variety of protein-based methods, DNA-based methods, trace metal composition, stable isotope analysis and metabolic profiling have been used to identify the origins of prawns/shrimp or survey their authenticity in the marketplace. Metabolic profiling was recently demonstrated by Elliott (2018) in its ability to differentiate prawn species and country of origin. Whereas, Li et al. (2017) used elemental profiling to distinguish between farmed Pacific white shrimp (Litopenaeus vannamei) from India, Vietnam and Thailand. Ortea and Gallardo (2015) investigated isotopic and trace metal composition to differentiate prawns based on wild/farmed production methods, geographical origins and species. The authors reported that a combination of stable isotope ratios and multi-element analysis enhanced the prediction capabilities.

In Australia, Carter et al. (2015) investigated the isotopic and trace metal composition of domestic and imported prawn meat, head and chitin, and concluded that the techniques could be used to distinguish between Australian and imported product. Gopi et al. (2018) also used stable isotopes to distinguish between farmed and wild-caught black tiger prawns originating from Queensland. However, during a recent investigation into suspected prawn substitution in Australia, results from trace element and stable isotope analysis proved inconclusive when the test results from samples of known Australian and the suspected imported Vannamei prawns both came back as more likely to be Australian (Queensland Government, 2017).

In 2016 Australian Council of Prawn Fisheries (ACPF) partnered with Australian Prawn Farmer’s Association (APFA) in a pilot project to investigate the use of trace element fingerprinting to substantiate provenance for the Australian prawn industry.

What initiated the study?

Product mislabelling or substitution is a global issue and there are often suspected cases of product substitution of prawns in the retail market. In Australia convictions for prawn substitution have been recorded against some enterprises, but the true extent of fraudulent practices are unknown as it is often difficult to detect and prosecute. ACPF and APFA recognised that there is a need to develop a rapid and robust scientific method to verify compliance and ensure product integrity. The pilot project was implemented to help act as a deterrent and will assess if trace element fingerprinting can be used to validate provenance, i.e. identify where the prawns were farmed or fished, and be able to differentiate between Australian wild and farmed prawns versus non-Australia product. If successful, the method and findings may be used to generate evidence in any future suspected prawn substitution cases.

What was the process?

The pilot project investigating the use of trace elements is currently underway and is funded through the Fisheries Research and Development Corporation (FRDC) in collaboration with Curtin University of Technology. The FRDC grant includes legal advice on methods like the use of trace element profiles and how they could be used by enforcement agencies for the collection of evidence in prosecution cases. The project was developed in consultation with supply chain members and enforcement agencies and supply chain members, and was based on the Physi-Trace model. Physi-Trace is used within the Australian pork industry and is able to trace fresh, frozen or cooked pork products back to the origin/farm.

The project aims to build on findings from a preliminary research project that determined the trace element profiles from approximately 150 prawn samples (Tan, C., MSc Thesis – The University of Western Australia (unpublished)). The analytical methodology and interpretive algorithms from this earlier project were capable

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 44 of identifying specific sites from which Australian prawns were cultivated and to distinguish these from samples of prawns imported from overseas (China, Indonesia, Malaysia, Thailand and Vietnam). The preliminary project indicated that trace element profiles could enable the Australian prawn industry and/or regulators to manage product integrity cases, such as substitution, and provide the basis for possible prosecution. However, the study was of limited scope and a more detailed survey was required to ensure that the technique is robust enough to withstand scrutiny.

Key areas that needed to be resolved at the commencement of the project were the preferred provenance determining method, the preferred supplier, and the required confidence levels of results for the level of provenance desired (site of production/harvest). Authentication using trace elements was selected over DNA-based techniques because many prawns grown in Southeast Asia are the same species or come from Australian spawning stock and are likely to have similar genetic markers. The prawn industry selected trace element profiles over stable isotope analysis as trace element analysis is cheaper and is expected to be able to better discriminate between samples harvested from different farms or over smaller geographical zones.

What trials have been undertaken?

Samples will be analysed from 17 farms and 35 fishery locations. Samples of prawns from key locations throughout Australia have been collected and the first batch are being analysed for a suite of elements from May 2018. The pilot project is expected to be completed in 2019.

What are the next steps?

The pilot project will begin to develop an Australian prawn trace element provenance database and review the results. Enforcement agencies are involved at various points within the project. The findings from the project will be disseminated through a communication strategy. ACPF and APFA will also consider future database maintenance and usage near the completion of the project.

Can trace elements be applied to other sectors?

Elemental analysis is a proven technique that can be used to differentiate origins of various commodities. When developing the project it is important to consider the major objectives, availability of technology and expertise, and how it has been used in the past.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 45 7. FINAL REMARKS

Traceability and authenticity are two distinct areas, as traceability involves record keeping in order to be able to track an entity (i.e. food) through all stages of production, harvest, processing and distribution, whereas authenticity refers to the product being genuinely as it is described. Most traceability systems are vulnerable to errors, omissions and/or falsification. Visual identification and genetic testing have been the most prominent method for species identification in Australia. The physical identification of fish for trade or stock assessments based on morphological characteristics often requires expert training. Once seafood has been processed, even after minor processing, it becomes extremely difficult to accurately identify the species because many morphological characteristics are lacking. Furthermore, the ability for industry stakeholders, regulators or consumers to robustly authenticate the geographical origin of a product, or if it was caught using sustainable harvest practices based on physical attributes is almost impossible. Whilst the benefit-cost ratio of traceability systems within the Australian seafood sector are generally unreported, in 2015 the Australian National Livestock Identification System reported that their traceability system would provide a benefit-cost ratio of 8.3 over the following 30 years13 and the Australian pork industry reported a benefit-cost ratio for the Physi-Trace R&D program of approximately 15.814.

The authentication of the traceability and origin of a species should:  Provide assurances that the product comes from a sustainable and well managed fishery and was processed and handled under appropriate quality assurance standards  Distinguish between wild and aquaculture products  Support country of origin labelling requirements and provenance declarations.

A range of techniques, as illustrated in Figure 12, are available to help identify product species or product origin. Species identification is generally based on molecular markers, such as DNA, and/or proteins, whereas geographical origins are usually established by fatty acids, trace elements and/or stable isotopes. Differentiation between wild or aquaculture/farmed products have generally been made through comparing fatty acids or protein markers. Often no single technique can be used to authenticate the identification, origin or provenance of a product and a combination of techniques have been used. As way of example, Higgins et al. (2010) used otolith microchemistry in combination with otolith shape analysis, body morphometry, microbacterial assemblages, parasite loads and microsatellite DNA to successfully classify the origin of Atlantic cod (Gadus morhua) in the northeast Atlantic. However, whilst combining results from different technologies can increase the confidence of an accurate determination, it also increases the cost.

13 Marden Jacob Associates (2015). Ex-post benefit-cost assessment of MLA’s Product Integrity Programs. Final report prepared for Meat & Livestock Australia (Project No. V-LIM.1505). 110 pp. 14 APL (2017). Industry returns on investments from APL R&D – Gilt enhancement and Physi-Trace. Retrieved from http://australianpork.com.au/wp-content/uploads/2017/12/Gilt-enhancement-Physi-trace.pdf on 13 March 2018.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 46 Species identification Species identification

Wild vs farmed

Proteins DNA-based markers

Fatty acids Elements and isotopes

Geographical origin Geographical origin

Wild vs farmed

Figure 12. Commonly used authenticity techniques and purpose of application. Reproduced with minor variation from Lavilla et al. (2013).

The FAO/IAEA15 have undertaken a number of coordinated research projects using nuclear and related techniques to trace the origin of food, check authenticity and test for contaminants. For example Project CRP D53040 titled “Field-Deployable Analytical Methods to Assess the Authenticity, Safety and Quality of Food”, is focused on the development of portable atomic and molecular spectroscopic screening technologies for front-line food fraud detection. The project aims to close the gap between instrumental capabilities found in sophisticated research laboratories and technologies that can be easily used in the field. The opportunity to accomplish this goal stems from a rapid and ongoing reduction in the cost of analytical equipment and a rapid increase in portability.

Analytical methods, especially in the wild capture sector, are generally unable to identify the owner of the seafood at various stages in the supply chain and or the processing conditions that do not directly influence the properties of the food product (FAO, 2016). To verify claims or detect fraud in these areas forensic auditing of paper-based or electronic records are needed.

There is an underlining requirement that any findings or results that are used to implicate or help deter fraudulent practices should also be sufficiently robust to stand up in a court of law. To assure the admissibility of analytical results as evidence in a court of law it is crucial that strict standards and guidelines are applied. This requires the rigorous validation of the techniques, together with appropriate sampling, evidence handling and statistical results (Martinsohn, 2011). Expert evidence and expert opinion evidence must satisfy the Daubert Criteria. Consequently, it is critical to understand and assess the likelihood and extent of inter-population variability. Some studies have demonstrated that classification rates can decrease with increasing amount of data. Therefore, results along with measurement uncertainty and variability within and amongst populations or species must be considered. It is often the aim of defence counsels to create reasonable doubt about the accuracy or validity of the findings presented. For example, if stable isotope analysis is used in a court of law then it must be demonstration to the satisfaction of the court that:  A robust databank of authentic samples exist;  An authentic sample is properly defined;  A suitable sampling design was used;  Relevant analytical and metadata have been collected;

15 FAO (Food and Agriculture Organization of the United Nations); IAEA (International Atomic Energy Agency)

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 47  All of the factors that can influence the outcome of the analysis considered;  Methods used are officially recognised or validated;  Laboratory is accredited to carry out the analysis;  Instrumentation and reference materials are suitable;  Measurement uncertainty has been calculated and taken into account;  There is confidence in all other analysis/data generated which support the stable isotope result (Camin et al., 2017).

Similar considerations would also be applicable with other methods. Full chain traceability is an essential prerequisite to better supply chain management and can support improved food security, sustainability, safety and intentional adulteration. Effective traceability systems help to reduce food waste and are essential in performing targeted product recalls. Authentic products operating with transparent supply chains provide a mechanism to help build consumer trust, which in turn can add value. Prior to implementing full chain traceability or authenticity systems, it is important to assess if the supply chain and markets are ready for more transparency and the challenges surrounding system interoperability.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 48 8. USEFUL RESOURCES

 Conservation Alliance for Seafood Solutions; available from http://solutionsforseafood.org/about/  European Union’s Knowledge Centre for Food Fraud and Quality; available from https://ec.europa. eu/jrc/en/food-fraud-and-quality  European Commission’s Joint Research Centre; available from https://ec.europa.eu/jrc/en/research-topic/genetics-traceability-fisheries-sector  FishWise’s Traceability and Counter-IUU Fishing; available from https://fishwise.org/resources/ traceability  Fiorino, G. M., Garino, C., Arlorio, M., Logrieco, A. F., Losito, I. and Monaci, L. (2018) Overview on untargeted methods to combat food frauds: a focus on fishery products. Journal of Food Quality 2018: 13.  Gonzálvev, A. and de la Guardia, M. (Eds.) (2013) Food Protected Designation of Origin: Methodologies and Applications. Amsterdam, Elsevier  Naaum, A. M. and Hanner, R. H. (Eds.) (2016) Seafood Authenticity and Traceability: A DNA-based Perspective. San Diego, USA, Academic Press  Picó, Y. (Ed.) (2012), Chemical Analysis of Food: Techniques and Applications. Boston, USA, Academic Press.  SafeFood (2017) Exploration of novel technologies for counteracting food fraud: A technology viability study; available from http://www.safefood.eu/SafeFood/media/SafeFoodLibrary/ Documents/Publications/Research%20Reports/Final-report.pdf  Verrez-Bagnis, V., Sotelo, C. G., Mendes, R. O., Silva, H. A., Kappel, K. and Schrӧder, U. (2019). Methods for Seafood Authenticity Testing in Europe. In: Mérillon, J. M. and Ramawat, K. G. (Eds.) Bioactive Molecules in Food, Springer International Publishing.  Watling, R. J., Lee, G. S., Scadding, C. J., Pilgrim, T. S., Green, R. L., Martin, A. E., May, C. D., and Valentin, J. L. (2010) The application of solution and laser ablation based ICP-MS and solution based AES for the provenance determination of selected food and drink produce. The Open Chemical and Biomedical Methods Journal 3: 177-194.  Wilkes, T., Nixon, G. and Burns, M. (2016) A review of current and emerging DNA-based methodologies for the determination of the geographical point of origin of food stuffs; available from https://www.gov.uk/government/publications/review-of-dna-based-methodologies-for-determining -origin-of-foods

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 49 9. REFERENCES

Adey, E. A., Black, K. D., Sawyer, T., Shimmield, T. M. and Trueman, C. N. (2009). Scale microchemistry as a tool to investigate the origin of wild and farmed Salmo salar. Marine Ecology Progress Series 390: 225-235. Alamprese, C. and Casiraghi, E. (2015). Application of FT-NIR and FT-IR spectroscopy to fish fillet authentication. Food Science and Technology 63: 720-725. Albuquerque, R., Queiroga, H., Swearer, S. E., Calado, R. and Leandro, S. M. (2016). Harvest locations of goose barnacles can be successfully discriminated using trace elemental signatures. Scientific Reports 6: 27787. Anderson, L. (2016). From Ocean to Plate: How DNA Testing Helps to Ensure Traceable, Sustainable Seafood. Marine Stewardship Council. 16 pp. Anonymous (2016a). Fight Food Waste & Fraud CRC: To Protect and Profit Australia's Food & Wine Industry. Anonymous (2016b). Food Waste & Fraud Prevention CRC; Round 19 Prospectus - March 2017. Retrieved from https://crca.asn.au/wp-content/uploads/2016/11/Food-Waste-Fraud-Prevention-CRC- Prospectus.pdf on 07 December 2017. Anonymous (2016c). Pearl Identification Technology. Retrieved from http://www.fukuishell.com/index.php/en/component/spsimpleportfolio/item/5-metakaku on 28 March 2018. AOAC International (2012). Official Method 980.16: Identification of Fish Species - Thin-layer Polyacrylamine Gel Isoelectric Focusing Method. In: Official Methods of Analysis of AOAC International. Gaithersburg, MD, USA. APL (2015). Annual Report 2014-2015. Australian Pork Limited. ACT, Australia. APL (2017). Industry returns on investment from APL R&D - Gilt enhancement and Physi-Trace. Retrieved from http://australianpork.com.au/wp-content/uploads/2017/12/Gilt-enhancement-Physi-trace.pdf on 13 March 2018. Applewhite, L., Larkin, P. and Naaum, A. M. (2016). Species identification using other tools. In: Naaum, A. M. and Hanner, R. H. (Eds.) Seafood Authenticity and Traceability: A DNA-based Perspective. San Diego, USA, Academic Press: 133-148. Applewhite, L., Rasmussen, R. and Morrissey, M. (2012). Species identification of seafood. In: Granata, L. A., Flick, G. J. and Martin, R. E. (Eds.) Seafood Industry: Species, Products, Processing, and Safety, Second Edition, Blackwell Publishing Ltd: 193-219. Arppe, R. and Just Sørensen, T. (2017). Physical unclonable functions generated through chemical methods for anti-counterfeiting. Nature Reviews Chemistry 1: 0031. Artigaud, S., Lavaud, R., Thébault, J., Jean, F., Strand, Ø., Strohmeier, T., Milan, M. and Pichereau, V. (2014). Proteomic-based comparison between populations of the Great Scallop, Pecten maximus. Journal of Proteomics 105: 164-173. Aursand, M., Standal, I. B., Praël, A., McEvoy, L., Irvine, J. and Axelson, D. E. (2009). 13C NMR pattern recognition techniques for the classification of Atlantic salmon (Salmo salar L.) according to their wild, farmed, and geographical origin. Journal of Agricultural and Food Chemistry 57: 3444-3451. Australian Food News (2016). Fish fraud concerns are mounting. Retrieved from http://www.ausfoodnews.com.au/2016/06/06/fish-fraud-concerns-are-mounting.html on 11 December 2017. Australian Government, Department of Industry, Innovation and Science (2017). Addendum to Seafood Origin Information Working Group Paper. Retrieved from https://industry.gov.au/industry/IndustrySectors/FoodManufacturingIndustry/Documents/Seafood- Origin-Labelling-Working-Group-Paper-Addendum.pdf on 23 April 2018. Barik, S. K., Banerjee, S., Bhattacharjee, S., Das Gupta, S. K., Mohanty, S. and Mohanty, B. P. (2013). Proteomic analysis of sarcoplasmic peptides of two related fish species for food authentication. Applied Biochemistry and Biotechnology 171: 1011-1021. Black, C., Chevallier, O. P. and Elliott, C. T. (2016). The current and potential applications of ambient mass spectrometry in detecting food fraud. Trends in Analytical Chemistry 82: 268-278. Black, C., Chevallier, O. P., Haughey, S. A., Balog, J., Stead, S., Pringle, S. D., Riina, M. V., Martucci, F., Acutis, P. L., Morris, M., Nikolopoulos, D. S., Takats, Z. and Elliott, C. T. (2017). A real time

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 50 metabolomic profiling approach to detecting fish fraud using rapid evaporative ionisation mass spectrometry. Metabolomics 13: 153. Bremner, A. and Snow, A. (2007). A scoping study to provide FRDC with advice on future investment options in species identification. Final Report prepared for the FRDC. Project No. 2006/246. 86 pp. Bremner, G. (2009). Marine fish of disputed origin: can trace elements provide a forensic solution? - A case study from southern New Zealand. Ministry of Fisheries, Dunedin, New Zealand. CAC (2017). CODEX STAN 329-2017: Standard for Fish Oils. Camin, F., Boner, M., Bontempo, L., Fauhl-Hassek, C., Kelly, S. D., Riedl, J. and Rossmann, A. (2017). Stable isotope techniques for verifying the declared geographical origin food in legal cases. Trends in Food Science & Technology 61: 176-187. Carrera, M., Cañas, B., López-Ferrer, D., Piñeiro, C., Vázquez, J. and Gallardo, J. M. (2011). Fast monitoring of species-specific peptide biomarkers using high-intensity-focused-ultrasound- assisted tryptic digestion and selected MS/MS ion monitoring. Analytical Chemistry 83: 5688-5695. Carrera, M. and Gallardo, J. M. (2017). Determination of the geographical origin of all commercial hake species by stable isotope ratio (SIR) analysis. Journal of Agricultural and Food Chemistry 65: 1070- 1077. Carter, J. F., Tinggi, U., Yang, X. and Fry, B. (2015). Stable isotope and trace metal compositions of Australian prawns as a guide to authenticity and wholesomeness. Food Chemistry 170: 241-248. Cheng, J.-H., Dai, Q., Sun, D.-W., Zeng, X.-A., Liu, D. and Pu, H.-B. (2013). Applications of non-destructive spectroscopic techniques for fish quality and safety evaluation and inspection. Trends in Food Science & Technology 34: 18-31. Cheng, J.-H. and Sun, D.-W. (2015). Recent applications of spectroscopic and hyperspectral imaging techniques with chemometric analysis for rapid inspection of microbial spoilage in muscle foods. Comprehensive Reviews in Food Science and Food Safety 14: 478-490. Coco, L. D., Papadia, P., Pascali, S. A. D., Bressani, G., Storelli, C., Zonno, V. and Fanizzi, F. P. (2009). Comparison among different Gilthead sea bream (Sparus aurata) farming systems: activity of intestinal and hepatic enzymes and 13C-NMR analysis of lipids. Nutrients 1: 291-301. Cohen, N. J., Deeds, J. R., Wong, E. S., Hanner, R. H., Yancy, H. F., White, K. D. and Gerber, S. I. (2009). Public health response to puffer fish (Tetrodotoxin) poisoning from mislabeled product. Journal of Food Protection 72: 810-817. Cozzolino, D. (2016). Near infrared spectroscopy and food authenticity. In: Espiñeira, M. and Santaclara, F. J. (Eds.) Advances in Food Traceability Techniques and Technologies: Improving Quality Throughout the Food Chain, Woodhead Publishing: 119-136. Creydt, M. and Fischer, M. (2018). Omics approaches for food authentication. Electrophoresis: 1-13. CSIRO Futures (2017). Food and Agribusiness - A Roadmap for unlocking value-adding growth opportunities for Australia. CSIRO. Australia. 80 pp. D'Costa, J. (2015). Preliminary study to determine the provenance of Barramundi (Lates calcaeifer) based on its trace element composition. Masters Thesis, The University of Western Australia. Dabbene, F., Gay, P. and Tortia, C. (2016). Radio-frequency identification usage in food traceability. In: Espiñeira, M. and Santaclara, F. J. (Eds.) Advances in Food Traceability Techniques and Technologies: Improving Quality Throughout the Food Chain, Woodhead Publishing: 67-89. Dais, P. and Spyros, A. (2012). Nuclear Magnetic Resonance. In: Picó, Y. (Ed.) Chemical Analysis of Food: Techniques and Applications. Boston, Academic Press: 91-115. Dalgaard, P., Manfio, G. P. and Goodfellow, M. (1997). Classification of photobacteria associated with spoilage of fish products by numerical taxonomy and pyrolysis mass spectrometry. Zentralblatt für Bakteriologie 285: 157-168. de Nanclares, M. P., Dessen, J. E., Rørvik, K. A., Thomassen, Y. and Thomassen, M. S. (2016). Feasibility of using rare earth elements (REEs) to mark and identify escaped farmed Atlantic salmon Salmo salar L. Aquaculture Research 47: 1885-1898. De Pontual, H., Lagardére, F., Troadec, H., Batel, A., Désaunay, Y. and Koutsikopoulos, C. (2000). Otoliths imprinting of sole (Solea solea) from the Bay of Biscay: a tool to discriminate individuals from nursery origins? Oceanologica Acta 23: 497-513. Di Franco, A., Bulleri, F., Pennetta, A., De Benedetto, G., Clarke, K. R. and Guidetti, P. (2014). Within-otolith variability in chemical fingerprints: implications for sampling designs and possible environmental interpretation. PLoS ONE 9: 1-11.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 51 Drescher, D. (2017). Blockchain Basics: a Non-Technical Introduction in 25 Steps. New York, NY, Springer. 250 pp. El Shieikha, A. F. and Montet, D. (2016). How to determine the geographical origin of seafood? Critical Reviews in Food Science and Nutrition 56: 306-317. Elliott, C. T. (2018). The Sins of Seafood. Food Fraud 2018 Conference, London, UK. Ellis, D. I., Muhamadali, H., Allen, D. P., Elliott, C. T. and Goodacre, R. (2016). A flavour of omics approaches for the detection of food fraud. Current Opinion in Food Science 10: 7–15. Ellis, D. I., Muhamadali, H., Haughey, S. A., Elliott, C. T. and Goodacre, R. (2015). Point-and-shoot: rapid quantitative detection methods for on-site food fraud analysis - moving out of the laboratory and into the food supply chain. Analytical Methods 7: 9401-9414. Esteki, M., Shahsavari, Z. and Simal-Gandara, J. (2018). Use of spectroscopic methods in combination with linear discriminant analysis for authentication of food products. Food Control 91: 100-112. FAO (2016). Seafood Traceabiltiy Systems: GAP Analysis of Inconsistencies in Standards and Norms. Fisheries and Aquaculture Circular No. 1123. Rome, Italy. FAO (2018). Overview of Food Fraud in the Fisheries Sector. Fisheries and Aquaculture Circular No. 1165. Rome, Italy. Fiorino, G. M., Garino, C., Arlorio, M., Logrieco, A. F., Losito, I. and Monaci, L. (2018). Overview on untargeted methods to combat food frauds: a focus on fishery products. Journal of Food Quality 2018: 13. FishPopTrace (2011). Traceability of Fish Populations and Fish Products: Advances and Contributions to Sustainable Fisheries. 40 pp. FSANZ (2003). A pilot survey on the identity of fish species as sold through food outliets in Australia. Retrieved from http://www.foodstandards.gov.au/publications/documents/FishSpeciationReportfinal.pdf on 08 December 2017. Garcia-Wass, F., Hammond, D., Mottram, D. S. and Gutteridge, C. S. (2000). Detection of fruit juice authenticity using pyrolysis mass spectroscopy. Food Chemistry 69: 215-220. Garrigues, S. and de la Guardia, M. (2013). Vibrational spectroscopy. In: Gonzálvev, A. and de la Guardia, M. (Eds.) Food Protected Designation of Origin: Methodologies and Applications. Amsterdam, Elsevier: 101-122. Gayo, J., Hale, S. A. and Blanchard, S. M. (2006). Quantitative analysis and detection of adulteration in crab meat using visible and near-infrared spectroscopy. Journal of Agricultural and Food Chemistry 54: 1130-1136. Gonzálvev, A. and de la Guardia, M. (2013). Mineral profile. In: Gonzálvev, A. and de la Guardia, M. (Eds.) Food Protected Designation of Origin: Methodologies and Applications. Amsterdam, Elsevier: 51- 76. Gooch, J., Daniel, B., Abbate, V. and Frascione, N. (2016). Taggant materials in forensic science: A review. Trends in Analytical Chemistry 83: 49-54. Gopi, K., Mazumder, D., Saintilan, N. and Sammut, J. (2018). Distinguishing between farmed and wild- caught black tiger prawns, Penaeus monodon, using stable isotopes. Journal of Aquaculture & Marine Biology 7: 00176. Gross, J. H. (2017). Introduction. In: Mass Spectrometry: A Textbook (Third Edition). Switzerland, Springer International Publishing: 1-28. GS1 (2016). Pharmaceutical Traceability for Manufacturers and Wholesalers. GS1 Healthcare Conference. 25 October 2016 in Beijing. Héberger, K. (2008). Chemoinformatics—multivariate mathematical–statistical methods for data evaluation. In: Vékey, K., Telekes, A. and Vertes, A. (Eds.) Medical Applications of Mass Spectrometry. Amsterdam, Elsevier: 141-169. Hellberg, R. S., Pollack, S. J. and Hanner, R. H. (2016). Seafood species identification using DNA sequencing. In: Naaum, A. M. and Hanner, R. H. (Eds.) Seafood Authenticity and Traceability: A DNA-based Perspective. San Diego, USA, Academic Press: 113-132. Higgins, R. M., Danilowicz, B. S., Balbuena, J. A., Daníelsdóttir, A. K., Geffen, A. J., Meijer, W. G., Modin, J., Montero, F. E., Pampoulie, C., Perdiguero-Alonso, D., Schreiber, A., Stefánsson, M. Ö. and Wilson, B. (2010). Multi-disciplinary fingerprints reveal the harvest location of cod Gadus morhua in the northeast Atlantic. Marine Ecology Progress Series 404: 197-206.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 52 Hofherr, J., Martinsohn, J., Cawthorn, D., Rasco, B. and Naaum, A. M. (2016). Regulatory frameworks for seafood authenticity and traceability. In: Naaum, A. M. and Hanner, R. H. (Eds.) Seafood Authenticity and Traceability: A DNA-based Perspective. San Diego, USA, Academic Press: 47- 82. Hong, E., Lee, S. Y., Jeong, J. Y., Park, J. M., Kim, B. H., Kwon, K. and Chun, H. S. (2017). Modern analytical methods for the detection of food fraud and adulteration by food category. Journal of the Science of Food and Agriculture 97: 3877-3896. Jacquet, J. L. and Pauly, D. (2008). Trade secrets: Renaming and mislabeling of seafood. Marine Policy 32: 309-318. Kamruzzaman, M. and Sun, D. W. (2016). Introduction to Hyperspectral Imaging Technology. In: Sun, D.- W. (Ed.) Computer Vision Technology for Food Quality Evaluation (Second Edition). San Diego, Academic Press: 111-139. Kato, M. (1985). Recent information on europium marking techniques for chum salmon. In: Singermann, C. J. (Ed.) Vol 27 NOAA (National Oceanic and Atmospheric Administration) Technical Report NMFS (National Marine Fisheries Service): 67-73. Kemény, Z. and Ilie-Zudor, E. (2016). Alphanumerical and optical coding systems for food traceability. In: Espiñeira, M. and Santaclara, F. J. (Eds.) Advances in Food Traceability Techniques and Technologies: Improving Quality Throughout the Food Chain, Woodhead Publishing: 49-65. Kim, H. and Cadwallader, K. R. (2011). Instrumental analysis of seafood flavour. In: Alasalvar, C., Shahidi, F., Miyashita, K. and Wanasudara, U. (Eds.) Handbook of Seafood Quality, Safety and Health Applications. Oxford, UK, Wiley-Blackwell: 50-67. Kim, H., Kumar, K. S. and Shin, K. H. (2015). Applicability of stable C and N isotope analysis in inferring the geographical origin and authentication of commercial fish (Mackerel, Yellow Croaker and Pollock). Food Chemistry 172: 523-527. Lamendin, R., Miller, K. and Ward, R. D. (2015). Labelling accuracy in Tasmanian seafood: an investigation using DNA barcoding. Food Control 47: 436-443. Lavilla, I., Costas-Ridríguez, M. and Bendicho, C. (2013). Authenticity of Fishery Products. In: de la Guardia, M. and Gonzálvev, A. (Eds.) Food Protected Designation of Origin: Methodologies and Applications. Amsterdam, Elsevier: 657-717. Lee, G. and Watling, J. (2017). Physi-Trace: Rapid origin traceability of Australian pork products. FoodIntegrity 2017 Conference: Assuring the integrity of the food chain: Turning science into solutions, Parma, Italy. Lehane, L. and Lewis, R. J. (2000). Ciguatera: recent advances but the risk remains. International Journal of Food Microbiology 61: 91-125. Li, L., Boyd, C. and Sun, Z. (2016). Authentication of fishery and aquaculture products by multi-element and stable isotope analysis. Food Chemistry 194: 1238-1244. Li, L., Boyd, C. E., Racine, P., McNevin, A. A., Somridhivej, B., Minh, H. N., Tinh, H. Q. and Godumala, R. (2017). Assessment of elemental profiling for distinguishing geographic origin of aquacultured shrimp from India, Thailand and Vietnam. Food Control 80: 162-169. Liu, Y., Ma, D.-h., Wang, X.-c., Liu, L.-p., Fan, Y.-x. and Cao, J.-x. (2015). Prediction of chemical composition and geographical origin traceability of Chinese export tilapia fillets products by near infrared reflectance spectroscopy. Food Science and Technology 60: 1214-1218. Lohumi, S., Lee, S., Lee, H. and Cho, B.-K. (2015). A review of vibrational spectroscopic techniques for the detection of food authenticity and adulteration. Trends in Food Science & Technology 46: 85-98. Marko, P. B., Nance, H. A. and Guynn, K. D. (2011). Genetic detection of mislabeled fish from a certified sustainable fishery. Current Biology 21: R621-R622. Martinsohn, J. T. (2011). Deterring illegal activities in the fisheries sector: genetics, genomics, chemistry and forensics to fight IUU fishing and in support of fish product traceability. European Commission Joint Research Centre. Belgium. 76 pp. Matthews (2016). Ultimate guide to serialisation in the food and beverage industry. Retrieved from https://www.matthews.com.au/resource-library/documents/whitepapers/wp- confirmations/ultimate-guide-to-serialisation-ebook on 09 March 2018. Mazzeo, M. F. and Siciliano, R. A. (2016). Proteomics for the authentication of fish species. Journal of Proteomics 147: 119-124.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 53 McLeod, R. (2017). Counting the cost: lost Australian food and wine export sales due to fraud. Food Innovation Australia Limited (FIAL). 48 pp. Mettler Toledo (2017). Serialization and falsification protection: what potential do companies see? Retrieved from https://www.mt.com/dam/product_organizations/pi/KnowHow/PCE/Serialization-and- Falsification/Serialization-&-Falsification-EN.pdf on 13 March 2018. Michibata, H. (1981). Labeling fish with an activable element through their diet. Canadian Journal of Fisheries and Aquatic Sciences 38: 1281-1282. Moerman, F. (2018). Food Defense. In: Holban, A. M. and Grumezescu, A. M. (Eds.) Food Control and Biosecurity. London, UK, Academic Press: 135-223. Montanarella, L., Bassani, M. R. and Bréas, O. (1995). Chemometric classification of some european wines using pyrolysis mass spectrometry. Rapid Communications in Mass Spectrometry 9: 1589-1593. Moore, J. C., Spink, J. and Lipp, M. (2012). Development and application of a database of food ingredient fraud and economically motivated adulteration from 1980 to 2010. Journal of Food Science 77: R118-R126. Naaum, A. M., Warner, K., Mariani, S., Hanner, R. H. and Carolin, C. D. (2016). Seafood Mislabeling Incidence and Impacts. In: Naaum, A. M. and Hanner, R. H. (Eds.) Seafood Authenticity and Traceability: A DNA-based Perspective. San Diego, USA, Academic Press: 3-26. National Fisheries Institute (2016). Oceana on fish fraud: make new laws, don’t just enforce ones we have? Retrieved from https://www.aboutseafood.com/oceanas-fish-fraud-report-favors-bureaucracy- enforcement on 13 January 2017. Neales, S. (2018). Aussie beef exporters declare war on Chinese counterfeiters. The Australian. Published on 14 May 2018. Retreived from https://www.theaustralian.com.au/business/aussie-beef- exporterd-declare-war-on-chinese-counterfeiters/news- story/1ace93c183e7604880e8a4abbde11920 on 22 May 2018. Nielsen, E. E. (2016). Population or point-of-origin identification. In: Naaum, A. M. and Hanner, R. H. (Eds.) Seafood Authenticity and Traceability: A DNA-based Perspective. San Diego, USA, Academic Press: 149-169. Norrie, C. R., Dunphy, B. J., Baker, J. A. and Lundquist, C. J. (2016). Local-scale variation in trace elemental fingerprints of the estuarine bivalve Austrovenus stutchburyi within and between estuaries. Marine Ecology Progress Series 559: 89-102. Oceana (2012). Name that Fish. Retrieved from https://oceana.org on 25 July 2018. Oceana (2016a). Deceptive Dishes: Seafood Swaps Found Worldwide. Oceana. 24 pp. Oceana (2016b). The global reach of seafood fraud. Retrieved from http://usa.oceana.org/seafood- fraud/2016-global-reach-seafood-fraud on 11 December 2017. Ortea, I., Cañas, B., Calo-Mata, P., Barros-Velázquez, J. and Gallardo, J. M. (2009a). Arginine kinase peptide mass fingerprinting as a proteomic approach for species identification and taxonomic analysis of commercially relevant shrimp species. Journal of Agricultural and Food Chemistry 57: 5665-5672. Ortea, I., Cañas, B., Calo-Mata, P., Barros-Velázquez, J. and Gallardo, J. M. (2010). Identification of commercial prawn and shrimp species of food interest by native isoelectric focusing. Food Chemistry 121: 569-574. Ortea, I., Cañas, B. and Gallardo, J. M. (2009b). Mass spectrometry characterization of species-specific peptides from arginine kinase for the identification of commercially relevant shrimp species. Journal of Proteome Research 8: 5356-5362. Ortea, I., Cañas, B. and Gallardo, J. M. (2011). Selected tandem mass spectrometry ion monitoring for the fast identification of seafood species. Journal of Chromatography A 1218: 4445-4451. Ortea, I. and Gallardo, J. M. (2015). Investigation of production method, geographical origin and species authentication in commercially relevant shrimps using stable isotope ratio and/or multi-element analyses combined with chemometrics: an exploratory analysis. Food Chemistry 170: 145-153. Ortea, I., O'Connor, G. and Maquet, A. (2016). Review on proteomics for food authentication. Journal of Proteomics 147: 212-225. Ottavian, M., Facco, P., Fasolato, L., Novelli, E., Mirisola, M., Perini, M. and Barolo, M. (2012). Use of near- infrared spectroscopy for fast fraud detection in seafood: application to the authentication of wild European sea bass (Dicentrarchus labrax). Journal of Agricultural and Food Chemistry 60: 639- 648.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 54 Ottavian, M., Fasolato, L., Serva, L., Facco, P. and Barolo, M. (2014). Data fusion for food authentication: fresh/frozen–thawed discrimination in West African Goatfish (Pseudupeneus prayensis) fillets. Food and Bioprocess Technology 7: 1025-1036. Pagano, G., Aliberti, F., Guida, M., Oral, R., Siciliano, A., Trifuoggi, M. and Tommasi, F. (2015). Rare earth elements in human and animal health: state of art and research priorities. Environmental Research 142: 215-220. Pardo, M. Á., Jiménez, E. and Pérez-Villarreal, B. (2016). Misdescription incidents in seafood sector. Food Control 62: 277-283. Partis, L. and Wells, R. J. (1996). Identification of fish species using random amplified polymorphic DNA (RAPD). Molecular and Cellular Probes 10: 435-441. Pascoal, A., Ortea, I., Gallardo, J. M., Cañas, B., Barros-Velázquez, J. and Calo-Mata, P. (2012). Species identification of the Northern shrimp (Pandalus borealis) by polymerase chain reaction–restriction fragment length polymorphism and proteomic analysis. Analytical Biochemistry 421: 56-67. Pepe, T., Ceruso, M., Carpentieri, A., Ventrone, I., Amoresano, A., Anastasio, A. and Cortesi, M. L. (2012). Differentiation of four tuna species by two-dimensional electrophoresis and mass spectrometric analysis. In: Heazlewood, J. L. and Petzold, C. J. (Eds.) Proteomic Applications in Biology. Rijeka, Croaia, InTech: 191-208. Peris, M. and Escuder-Gilabert, L. (2016). Electronic noses and tongues to assess food authenticity and adulteration. Trends in Food Science & Technology 58: 40-54. Pigini, D. and Conti, M. (2017). NFC-based traceability in the food chain. Sustainability 9: 1910. PwC (2016). Food fraud vulnerability assessment and mitigation: are you doing enough to prevent food fraud? Retrieved from https://www.pwc.com/gx/en/services/food-supply-integrity- services/assets/pwc-food-fraud-vulnerability-assessment-and-mitigation-november.pdf on 27/09/2017. PwC Australia (2016). Fighting $40bn food fraud to protect food supply. Retrieved from https://www.pwc.com.au/press-room/2016/food-fraud-jan16.html on 06 December 2017. Queensland Government, Gold Coast Health (2017). An investigation into prawn substitution on the Gold Coast. Retrieved from https://www.ehaqld.org.au/documents/item/963 on 23 April 2018. Rašković, B., Heinke, R., Rösch, P. and Popp, J. (2016). The potential of Raman spectroscopy for the classification of fish fillets. Food Analytical Methods 9: 1301-1306. Rasmussen, R. S. and Morrissey, M. T. (2008). DNA-based methods for the identification of commercial fish and seafood species. Comprehensive Reviews in Food Science and Food Safety 7: 280-295. Ricardo, F., Génio, L., Costa Leal, M., Albuquerque, R., Queiroga, H., Rosa, R. and Calado, R. (2015). Trace element fingerprinting of cockle (Cerastoderma edule) shells can reveal harvesting location in adjacent areas. Scientific Reports 5: 11932. Rim, K. T., Koo, K. H. and Park, J. S. (2013). Toxicological evaluations of rare earths and their health impacts to workers: a literature review. Safety and Health at Work 4: 12-26. Rooker, J. R., Secor, D. H., De Metrio, G., Schloesser, R., Block, B. A. and D., N. J. (2008). Natal homing and connectivity in Atlantic bluefin tuna populations. Science 322: 742-744. Rossier, J. S., Maury, V., de Voogd, B. and Pfammatter, E. (2014). Use of isotope ratio mass spectrometry (IRMS) determination (18O/16O) to assess the local origin of fish and asparagus in Western Switzerland. International Journal for Chemistry 68: 696-700. Rubert, J., Zachariasova, M. and Hajslova, J. (2015). Advances in high-resolution mass spectrometry based on metabolomics studies for food – a review. Food Additives & Contaminants. Part A: Chemistry, Analysis, Control, Exposure & Risk Assessment 32: 1685-1708. Salla, V. and Murray, K. K. (2013). Matrix-assisted laser desorption ionization mass spectrometry for identification of shrimp. Analytica Chimica Acta 794: 55-59. Savorani, F., Picone, G., Badiani, A., Fagioli, P., Capozzi, F. and Engelsen, S. B. (2010). Metabolic profiling and aquaculture differentiation of gilthead sea bream by 1H NMR metabonomics. Food Chemistry 120: 907-914. SeaFish (2011). Recent advances in DNA and other techniques for identifying seafood provenance in the supply chain. SeaFish. Grimsby, UK. 10 pp. SeaFish (2015). Literature review: isotope ratios in seafood. SeaFish. Grimsby, UK. 8 pp.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 55 Shore, R. (2017). Seafood fraud waning, labels lack sustainability info, SeaChoice study finds. Vancouver Sun. Published on 03 October 2017. Retreived from http://vancouversun.com/news/local- news/seafood-labels-lack-sustainability-information-seachoice-study-finds on 09 October 2017. Śliwińska, M., Wiśniewska, P., Dymerski, T., Wardencki, W. and Namieśnik, J. (2016). Advances in Electronic Noses and Tongues for Food Authenticity Testing. In: Downey, G. (Ed.) Advances in Food Authenticity Testing, Woodhead Publishing: 201-225. Sotelo, C. G. and Pérez-Martín, R. I. (2002). Species identification in processed seafoods. In: Bremner, H. A. (Ed.) Safety and Quality Issues in Fish Processing, Woodhead Publishing: 450-474. Sousa, N., Moreira, M., Saraiva, C. and de Almeida, J. (2018). Applying Fourier transform mid infrared spectroscopy to detect the adulteration of Salmo salar with Oncorhynchus mykiss. Foods 7: 55. Spink, J. (2014). GFSI Direction on Food Fraud and Vulnerability Assessment (VACCP). Retrieved from http://foodfraud.msu.edu/2014/05/08/gfsi-direction-on-food-fraud-and-vulnerability-assessment- vaccp/ on 10 October 2017. Spink, J. and Moyer, D. C. (2011). Defining the public health threat of food fraud. Journal of Food Science 76: R157-R163. Stahl, A. and Schrӧder, U. (2017). Development of a MALDI-TOF MS-based protein fingerprint database of common food fish allowing fast and reliable identification of fraud and substitution. Journal of Agricultural and Food Chemistry 65: 7519-7527. Staples, M., Chen, S., Falamaki, S., Ponomarev, A., Rimba, P., Tran, A. B., Weber, I., Xiu, X. and Zhu, J. (2017). Risks and opportunities for sysyems using blockchain and smart contracts. Data61 (CSIRO). Sydney, Australia. Sterling, B., Gooch, M., Dent, B., Marenick, N., Miller, A. and Sylvia, G. (2015). Assessing the value and role of seafood traceability from an entire value-chain perspective. Comprehensive Reviews in Food Science and Food Safety 14: 205-268. Tamm, E. E., Schiller, L. and Hanner, R. H. (2016). Seafood Traceability and Consumer Choice. In: Naaum, A. M. and Hanner, R. H. (Eds.) Seafood Authenticity and Traceability: A DNA-based Perspective. San Diego, USA, Academic Press: 27-45. The Economist Intelligence Unit (2016). Global Food Security Index 2016: An annual measure of the state of global food security. London, UK. Thomas, F., Jamin, E., Wietzerbin, K., Guerin, R., Lees, M., Morvan, E., Billault, I., Derrien, S., Moreno Rojas, J. M., Serra, F., Guillou, C., Aursand, M., McEvoy, L., Prael, A. and Robins, R. J. (2008). Determination of origin of Atlantic salmon (Salmo salar): the use of multiprobe and multielement isotopic analyses in combination with fatty acid composition to assess wild or farmed origin. Journal of Agricultural and Food Chemistry 56: 989-997. Tomás, C., Ferreira, I. M. P. L. V. O. and Faria, M. A. (2017). Codfish authentication by a fast Short Amplicon High Resolution Melting Analysis (SA-HRMA) method. Food Control 71: 255-263. U..S Food and Drug Administration (2017). Seafood Species Substitution and Economic Fraud. Retrieved from https://www.fda.gov/Food/FoodScienceResearch/RFE/ucm071528.htm on 12 December 2017. U.S. Food and Drug Administration (2016). Regulated Fish Encyclopaedia, Atlantic Salmon PAGplate IEF. Retrieved from https://wayback.archive- it.org/7993/20170113200741/http://www.fda.gov/ucm/groups/fdagov- public/documents/image/ucm059541.jpg on 15 December 2017. Valcarce, R., Stevenson, D., Smith, G. G. and Sigler, J. W. (1991). Discriminant separation of fishes: a preliminary study of fish eggs by Curie‐Point Pyrolysis–Mass Spectrometry–chemometric analysis. Transactions of the American Fisheries Society 120: 796-802. Valentini, P., Galimberti, A., Mezzasalma, V., De Mattia, F., Casiraghi, M., Labra, M. and Pompa, P. P. (2017). DNA barcoding meets nanotechnology: development of a universal colorimetric test for food authentication. Angewandte Chemie International Edition 56: 8094-8098. Wang, M., Duong, B., Fenniri, H. and Su, M. (2015). Nanomaterial-based barcodes. Nanoscale 7: 11240- 11247. Wang, Y. V., Wan, A. H. L., Lock, E.-J., Andersen, N., Winter-Schuh, C. and Larsen, T. (2018). Know your fish: a novel compound-specific isotope approach for tracing wild and farmed salmon. Food Chemistry 256: 380-389.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 56 Ward, R. D. (2000). Genetics in fisheries management. Hydrobiolgia 420: 191-201. Wilkes, T., Nixon, G. and Burns, M. (2016). A review of current and emerging DNA-based methodologies for the determination of the geographical point of origin of food stuffs. Retrieved from https://www.gov.uk/government/publications/review-of-dna-based-methodologies-for- determining-origin-of-foods on 15 June 2018. Willette, D. A. and Cheng, S. H. (2017). Delivering on seafood traceability under the new U.S. import monitoring program. Ambio: 1-6. Wulff, T., Nielsen, M. E., Deelder, A. M., Jessen, F. and Palmblad, M. (2013). Authentication of fish products by large-scale comparison of tandem mass spectra. Journal of Proteome Research 12: 5253-5259. Xu, J. L., Riccioli, C. and Sun, D. W. (2015). An overview on nondestructive spectroscopic techniques for lipid and lipid oxidation analysis in fish and fish products. Comprehensive Reviews in Food Science and Food Safety 14: 466-477. Yohannes, K., Dalton, C. B., Halliday, L., Unicomb, L. E. and Kirk, M. (2002). An outbreak of gastrointestinal illness associated with the consumption of escolar fish. Communicable Diseases Intelligence 26: 441-448. Zak, M. A. (1984). Mass-marking American shad and Atlantic salmon with the rare eath element, samarium. Master's Thesis, Pennsylvania State University.

AUGUST 2018 SEAFOOD FRAUD & AUTHENTICITY: A REVIEW PAGE 57